Sustainability of Constructions - Volume 2 - Integrated Approach To Life-Time Structural Engineering

COST Action C25
Sustainability of Constructions
Integrated Approach to Life-time Structural Engineering

Volume 2
Summary Report of the Cooperative Activities
Integrated Approach to Life-time Structural Engineering

Editors:
Lus Bragana, Heli Koukkari, Raffaele Landolfo,
Viorel Ungureanu, Erkki Vesikari, Oliver Hechler

COST is supported by the EU RTD Framework Programme and ESF provides the COST Office through an EC contract
COST - the acronym for European COoperation in Science and Technology - is the oldest and
widest European intergovernmental network for cooperation in research. Established by the
Ministerial Conference in November 1971, COST is presently used by the scientific
communities of 36 European countries to cooperate in common research projects supported by
national funds.

The funds provided by COST - less than 1% of the total value of the projects - support the COST
cooperation networks, COST Actions, through which, with only around 20 million per year,
more than 30.000 European scientists are involved in research having a total value which
exceeds 2 billion per year. This is the financial worth of the European added value which
COST achieves.

A bottom up approach (the initiative of launching a COST Action comes from the European
scientists themselves), la carte participation (only countries interested in the Action
participate), equality of access (participation is open also to the scientific communities of
countries not belonging to the European Union) and flexible structure (easy implementation
and light management of the research initiatives) are the main characteristics of COST.

As precursor of advanced multidisciplinary research COST has a very important role for the
realisation of the European Research Area (ERA) anticipating and complementing the activities
of the Framework Programmes, constituting a bridge towards the scientific communities of
emerging countries, increasing the mobility of researchers across Europe and fostering the
establishment of Networks of Excellence in many key scientific domains such as: Biomedicine
and Molecular Biosciences; Food and Agriculture; Forests, their Products and Services;
Materials, Physics and Nanosciences; Chemistry and Molecular Sciences and Technologies;
Earth System Science and Environmental Management; Information and Communication
Technologies; Transport and Urban Development; Individuals, Societies, Cultures and Health. It
covers basic and more applied research and also addresses issues of pre-normative nature or of
societal importance.

Summary Report of the Cooperative Activities of the COST Action C25
Sustainability of Constructions - Integrated Approach to Life-time Structural Engineering
Volume 2
Sustainability of Constructions - Integrated Approach to Life-time Structural Engineering
Editors: Lus Bragana, Heli Koukkari, Raffaele Landolfo, Viorel Ungureanu, Erkki Vesikari,
Oliver Hechler
Cover Design: Bogdan Oprescu
Cover illustration is based on Happy Street, th
Architect: John Kormeling; Structural Designer Rijk Blok (C25 Member and Chair of WG1)

The production of this publication was supported by COST: www.cost.eu

ISBN:

2011 The authors and the Editors
All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form
or by any means, without prior written permission from the publisher.

LEGAL NOTICE
The Editors, the Authors and the publisher are not responsible for the use which might be made of the following
information.
978-99957-816-2-0
February 2011
e Dutch Pavilion for World Expo 2010 in Shanghai;
Printed by Gutenberg Press Ltd, Malta.
Published by: Department of Civil & Structural Engineering,
Faculty for the Built Environment,
University of Malta, Malta.

Life-time structural engineering:
Design for durability, life-cycle
performance, maintenance and
deconstruction

i
Contents

Foreword.............................................................................................. ......................................... v
L. Bragana

Preface......................................................................................................................................... vii
R. Landolfo

Chapter 1. State-of-the-Art Report on Service Life Prediction and Design
Methodologies

Introduction............................................................................................................................. 1
R. Landolfo & E. Vesikari
Part 1. Current Methodologies for Service Life Prediction
Service Life Prediction Methods Referred to by Eurocode 0 ................................................. 2
E. Vesikari & R. Landolfo
Standardization of Service Life Methodologies.................................................................... 10
W. Trinius & C. Sjstrm
Part 2. Service Life Design Methods of Constructions
Service Life Design Methods for Civil Engineering Structures ........................................... 17
Risk Based Approach to Service Life Assessment of Building Structures........................... 28
Sz. Wolinski

Chapter 2. State-of-the-Art Report on Deterministic and Probabilistic Degradation
Models

Introduction........................................................................................................................... 39
Part 1. Degradation Models for Service Life Design
Degradation Models of Concrete Structures ......................................................................... 40
E. Vesikari, Ch. Giarma & J. Bleiziffer
Modelling of Corrosion Damage for Metal Structures ......................................................... 51
R. Landolfo, L. Cascini & F. Portioli
Fatigue of Steel and Composite Bridges............................................................................... 61
U. Kuhlmann, H.-P. Gnther, J. Raichle, & M. Euler
Degradation modes and models for masonry structures ....................................................... 69
Z. Lj. Bozinovski, J. Lahdensivu & E. Vesikari
Durability and service life of wood structures and components -State of the art ................. 81
H.A. Viitanen, T. Toratti, R. Peuhkuri, T. Ojanen & L. Makkonen
Part 2. Application of Degradation Models to Service Life Design
Service life design of reinforced concrete structures - Alternative probabilistic
approaches............................................................................................................................. 94
Sz. Wolinski
Service Life Design of Metal Structures Based on Lifetime Safety Factor Method........... 106
L. Cascini, F. Portioli, R. Landolfo

ii
Chapter 3. Survey and condition assessment of structures

Introduction......................................................................................................................... 117
G. Hauf & O. Hechler
Part 1. Survey and condition assessment of structures by monitoring
Monitoring an introduction.............................................................................................. 120
Part 2. Damage monitoring and control of materials and structures
Monitoring of structural behaviour of buildings................................................................. 130
G. Fabbrocino & C. Rainieri
Wireless sensor networks: A new tool for structural monitoring........................................ 142
G. Feltrin, R. Bischoff, J. Meyer & O. Saukh
Structural Damage and Vulnerability Assessment for Service Life Estimation
through MEDEA tool.......................................................................................................... 154
G. Zuccaro & M. F. Leone
Part 3. Condition assessment
Survey and Condition assessment....................................................................................... 169
B. Khn & O. Hechler
Concrete deterioration: Specification and assessment ........................................................ 179
A. Taushanov & Ch. Giarma
Condition assessment and service-life prediction of reinforced concrete structures:
description and application examples of a method developed for aged
and new buildings ............................................................................................................... 192
T. Teruzzi
Part 4. Examples
MEDEA: a multimedia and didactic handbook for structural damage and
vulnerability assessment LAquila Case Study................................................................ 201
G. Zuccaro, F. Cacace & M. Rauci
Conclusions......................................................................................................................... 214

Chapter 4. Maintenance, repair and rehabilitation techniques and planning

Introduction......................................................................................................................... 221
O. Hechler
Part 1. General
Management systems for maintenance, repair and rehabilitation ....................................... 223
E. Vesikari, A. Borrmann & K. Lukas
Part 2. Buildings
Maintenance, repair and rehabilitation of buildings ........................................................... 235
V. Ungureanu, A. Dogariu, D. Dubina, A. Ciutina, R. Landolfo, F.M. Mazzolani
& F. Portioli
Swiss codes for existing structures Principles and challenges......................................... 282
E. Brhwiler, T. Vogel, T. Lang & P. Lchinger
Part 3. Bridges
Bridge Maintenance, Repair and Rehabilitation in a Life-Cycle Context .......................... 288
D.M. Frangopol & N.M. Okasha
iii
Strengthening of civil structures ......................................................................................... 296
J. Andersson, A. Geler, J. Naumes & O. Hechler
Conclusion .......................................................................................................................... 310
O. Hechler

Chapter 5. Demolition and deconstruction

Introduction......................................................................................................................... 315
F. Portioli & O. Hechler
Part 1. General
Sustainable refurbishment, deconstruction and demolition: Applications of MCDM
methods ............................................................................................................................... 317
F. Portioli, L. Cascini, R. Landolfo & V. Ungureanu
Assessment of the structural materials at high stain-rates in tension.................................. 327
E. Cadoni
Design for Deconstruction .................................................................................................. 339
O. Hechler, O. P. Larsen & S. Nielsen
Demolition and recycling of demolition rubble after deconstruction ................................. 355
P. Kamrath & O. Hechler
Part 2. Buildings
Deconstruction of buildings: Masses and types .................................................................. 369
P. Kamrath
Part 3. Bridges
Demolition and deconstruction of bridges .......................................................................... 377
F. Portioli, M. DAniello, E. Cadoni, R.P. Borg & O. Hechler
Conclusions......................................................................................................................... 390
F. Portioli & O. Hechler

List of all contributions from COST C25393

Author Index397

iv

v
Foreword
This publication is the Final Outcome of the COST Action C25 Sustainability of Constructions:
Integrated Approach to Life-time Structural Engineering which includes research reports, datasheets
and guidelines in two Volumes. The Action has actively involved more than one hundred researchers,
engineers and architects from 28 countries during four years since December 2006.

The Action has summarized its research field by the wording Sustainability of Constructions that
refers to creative combination of methods of structural engineering with those of sustainable
construction. The Action Members have collaboratively examined theories, methods and tools for
cross-boundary and trans-disciplinary knowledge production, management and communication. These
include approaches to assess environmental, social and economic impacts of construction activities;
methods to analyse and verify eco-efficiency of materials, components, buildings and infrastructures;
and methods of structural design that incorporate holistic understanding of safety, eco-efficiency and
sustainability.

The Action has been a scientific, social and cultural journey that has given opportunities not only to
co-operate over the borders but also to learn about various traditions. Ten Management Committee
meetings have been organised in nine countries, from Norway to Turkey and from Portugal to
Romania. Most of them have been large gatherings with invited experts and merged working group
and team meetings. The events with true milestone results are the launching meeting in Eindhoven in
December 2006, the First Workshop in Lisbon in 2007, the Mid-term Conference in Dresden in 2008,
the Open Seminar in Naples in 2009 and the Second Workshop in Timisoara in 2009. The Action has
also organised Training Schools in Eindhoven in 2008, in Thessaloniki in 2009 and in Valletta in
2010. A Student Competition was arranged in 2008-2009.

The Action has successfully contributed to the scientific advancement of the methods of life-time
structural engineering and to the implementation of sustainable construction approaches. The ideas
and knowledge have matured throughout the Action, from the initial brainstorming of the proposal in
2005 until the production of this final outcome and organisation of the Final Conference in Innsbruck
in February 2011. The achievements are mainly published in four Proceedings and Training School
Books in relation to the Action events. Especially the following achievements are worth highlighting:
- Sustainability assessment guidelines for both bridges and buildings
- A methodology to assess sustainability of bridges
- Integrated methodology for lifetime engineering including risk analysis and maintenance
scenarios
- Special Issue of Sustainability Journal with the Action Chair as a Guest Editor and several
papers from the Members.

The Action was organised in three Working Groups in accordance with the main research areas
identified as necessary for the objectives:
WG1 Criteria for Sustainable Constructions (topics like global methodologies, assessment
methods, global models and databases);
WG2 Eco-efficiency (topics of eco-efficient use of resources in construction materials, products
and processes);
WG3 Life-time structural engineering (topics like design for durability, life-cycle performance,
including maintenance and deconstruction).

The Action activities were taken care in several work packages in order to ascertain a good coverage
of the most important topics as follows:
WP1 State-of-the-art on LCA and LCCA methodologies as applied in participating countries
WP2 Collection of information on databases of LCI and LCC for construction materials,
construction products and processes and assessment of existing data and criteria
WP3 Life-cycle performance: deterministic and stochastic simulation models
WP4 Implementation of global methodologies for Sustainable Design and Building - case-study
WP5 Identification and evaluation of existing and new functional materials, construction
products and processes to comply with decrease of material use, decrease of waste, decrease of
emissions and energy saving goals

vi
WP6 Improvement of environmental performance of constructions (civil engineering structures,
building structures and building envelopes). Improvement of the comfort in buildings (thermal,
acoustic, lighting and quality of air), energy performance and the integration of innovative systems
in buildings (mechanical, electrical and automation)
WP7 Analysis of functional materials and applications and new technologies - case-study
WP8 Life-cycle performance: verification methods for durability of constructions (degradation
models and service design life)
WP9 Monitoring of life-cycle performance (life-cycle safety, functionality, quality, demolition
and deconstruction)
WP10 Sustainable construction assessment and classification systems

The Action piloted a project-like organisation of a scientific network by application of cross-cutting
case-studies which aimed to support a coordinated outcome. The case-studies were continuously
reassessed and their complexity was increasing from essentially bare structures such as a bridge to
completed buildings including structural parts and non-structural parts to allow a clear identification of
all relevant aspects. By the aid of case-studies, the current technologies and methodologies were
compared and initiatives for further developments were established.

The Action and the Working Group Chairs want to warmly thank all the colleagues who have
contributed to the accomplishments by writing, reading, debating and, not the least, by putting up so
much energy, mind and heart in hosting and organizing the C25 events.

Special thanks to Luis Silva for managing the Action website where all the Action publications are
available and will be kept in coming years.

A special gratitude is also addressed to Dr. Thierry Goger and Ms. Carmencita Malimban from COST
Office and ESF (European Science Foundation) for their support and assistance in administrative
matters.

The Action Chairs hope that the achievements of the Action will inspire European researchers and
practitioners to new efforts and innovations in the field of sustainable construction.

December, 2010.

The Chairs and the Working Group Chairs of COST Action C25,

Action Chair Action Vice-chair
Lus Bragana
University of Minho
Portugal
Heli Koukkari
VTT Technical Research Centre of Finland
Finland

WG1 Chair WG1 Vice-chair
Rijk Blok
University of Technology Eindhoven
The Netherlands
Helena Gervsio
University of Coimbra
Portugal

Milan Veljkovic
Lule University of Technology
Sweden
Zbigniew Plewako
Rzeszw University of Technology
Poland

Raffaele Landolfo
University of Naples Federico II
Italy
Viorel Ungureanu
Politehnica University of Timisoara
Romania

COST Action C25 Webpage: www.cmm.pt/costc25/
vii
Preface
This book reports the outcomes of four years of activity carried out within the framework of the
COST Action C25 by the members of the Working Group 3 (WG3): Life-time structural
engineering (design for durability, life-cycle performance, including maintenance and
deconstruction).
The main objective of WG3 was the collection of methods and practices in the field of Life-
time structural engineering, which deals with the analysis of structures according to the basic
assumption that the performance is a time-dependant variable.
The design of structures based on the evaluation of life cycle performance represents a key
issue for the sustainable development of built-up areas, having the costs of construction sector a
significant social and environmental impact on the world-wide economy.
The assessment of life cycle performance of structures involves different questions related
both to design and management stages, including durability against the effects of deterioration
phenomena, monitoring and planning of maintenance operations.
As a consequence, the development of an integrated approach to life-time structural
engineering entails the definition of design procedures based on performance levels considering
mechanical safety as well as durability and obsolescence of structures.
In recent years, a large amount of research was carried out on these topics, but the
development of a holistic design approach, based on time-dependant performance and
maintenance interventions, is still missing and far from application in current practice. For this
reason, the collection of design methods and practices presented in this book aims at
representing a step forward in the development of user-oriented design approaches and in the
dissemination of life-cycle analysis of structures.
To cover the different aspects of lifetime structural engineering, WG3 has been organized in
two Working Packages, namely WP8 and WP9.
The main aim of WP8 was to analyze the different procedures for durability design based on
semi-probabilistic and probabilistic approaches. In particular, its main tasks were the survey of
the state-of-the-art in verification methods for durability of constructions and the analysis of
degradation models for prediction of service life of structures. According to performance based
durability design, the safety of constructions is satisfied if the failure event corresponding to the
maximum allowable value of degradation occurs after the design service life has expired, with a
proper safety margin. In particular, the calculation of life-time of constructions requires the
identification of the environmental loads affecting durability that will likely act on the
structures. On the basis of the identification of environmental loads, the degradation factors and
mechanisms have to be evaluated. Once deterioration mechanisms that could act on the
structures during the life cycle have been identified, corresponding damage models have to be
selected which provide degradation as a function of time on the basis of statistical formulations.
The probabilistic characterization of the damage development over time is necessary since it
influences the value of the safety factor to be considered in design phase. The definition of the
maximum allowable degradation values for different limit states represents the final step of
development of a durability design procedure. Both serviceability and ultimate limit states
related to basic requirements such as functionality in use and structural safety need to be
defined. With respect to mechanical limit states, more specific performance levels can be
classified for durability design. In particular, serviceability limit states can be related to changes
in functionality or aesthetics on the basis of maintainability, economy and environmental
impacts. Finally, ultimate limit states have to be defined for the damage of structural material
which compromises the mechanical safety of constructions.
WP9 was concerned with the monitoring of life-cycle performance and with maintenance and
deconstruction of structures, which are other important issues for the assessment of residual
service life and for the definition of end-of-life scenarios. Its main tasks were: the analysis of
maintenance, repair and rehabilitation techniques and planning; the survey and condition
assessment of structures in terms of safety and functionality; the classification and planning of
demolition and deconstruction systems. Different techniques and numerical methods for damage
monitoring of materials and evaluation of actual structural safety are available in the literature.
The aim of monitoring analysis is to define the actions to avoid or defer obsolescence that is the
viii
inability to satisfy the changing of functional and structural requirements. Generally these
actions, such as maintenance, demolition and repairing, have the purpose of minimising the
impacts of obsolescence before costs become substantial. In particular, the final objective is to
reduce demolishing of facilities that have not reached their mechanical or durability limit states,
and thus promote the sustainable development. In a performance based approach, the factors
and causes of obsolescence should be defined and identified, both in quantitative and qualitative
terms, selecting different limit states for functionality of different structural typologies. The
different methods for decision making such as life cycle costing method and risk analysis have
to be used to select the proper intervention strategy.
In this context, the present book collects a selection of the contributions presented by WG3
members and invited experts at the meetings and workshops. The contributions have been
revised on the basis of the other papers presented during the Action and dealing with similar
topics. The list of these reference papers is reported at the end of the book.
The book is organized in five Chapters, each corresponding to one of the expected WG3
output as it is reported in the Memorandum of Understanding. In particular, the first two
Chapters, corresponding to the expected Deliverable 8.1 and 8.2, represent the outcome of the
WP8, and the remaining three Chapters, matching the Deliverables D9.1, D9.2 and D9.3,
summarise the activities carried out within WP9.
The Deliverables were edited and revised with the contribution of E. Vesikari and O. Hechler,
coordinators of WP8 and WP9, to whom I would like to express my gratitude for the excellent
work they have done in the management of the research activity during these years.
Finally, I would like to thank V. Ungureanu, vice-chair of WG3, and all the people that
contributed actively to the achievement of the results presented in this book.

Raffaele Landolfo
Chairman of WG3

Chapter 1. State-of-the-Art Report on Service Life Prediction
and Design Methodologies

Chapter 1. State-of-the-Art Report on Service Life Prediction and Design Methodologies:
Introduction
1
R. Landolfo
University of Naples Federico II, Italy
E. Vesikari
Technical Research Centre of Finland
Introduction
This chapter reports the state-of-the-art on service life design methods and degradation models
of building materials. It consists of 4 papers, first presented by the authors at seminars arranged
by COST Action 25 Sustainability of Constructions. Integrated Approach to Lifetime
Structural Engineering and later on reviewed and expanded. The chapter is organized in two
parts.
Part 1 is focused on Current Methodologies for Service Life Prediction and consists of two
state-of-the-art papers. In the first paper the principles and methods based on experimental
investigation, experience from earlier constructions, theoretical calculations or a combination of
these all, as referred to by Eurocode 0, are presented. In the second paper an introduction into
the concepts related to service life planning based on international standardization, is given.
Especially the methods used in ISO 15686 series are addressed.
Part 2 is devoted to Service Life Design Methods of Constructions and is also made up of two
state-of-the-art papers. The first paper deals with service life design methods especially with
reference to bridges and civil engineering structures. The basic durability requirements reported
in the structural codes, as well as full probabilistic procedures that take into account all basic
variables with their statistical properties are considered. In particular, the limiting values
method, the factor method (ISO Series 15686), the semi-probabilistic dimensioning methods
such as the lifetime safety factor method and the alternative approaches proposed in the Fib
Model Code on Service Life Design are introduced. The chapter is closed with a paper dealing
with some aspects of risk based assessment of service life of building structures.

Life-time structural engineering: Design for durability, life-cycle performance, maintenance and deconstruction
2
E. Vesikari
R. Landolfo
University of Naples Federico II, Naples, Italy
Service Life Prediction Methods Referred to by Eurocode 0
1 INTRODUCTION
Eurocode 0, Basis of structural design, emphasizes the importance of durability and service life
in todays structural design. Eurocode 0 pushes the development towards service life design by
stating that the design working life should be specified and by imposing indicative service life
requirements for different categories of structures. For the design methods of service life
Eurocode 0 does not take a position but it states that the degree of any degradation may be
estimated on the basis of calculations, experimental investigation, experience from earlier
constructions, or a combination of these considerations.
Whatever service life design methods are used in practical design degradation and service life
data is needed. In this paper the methods of provision for reliable service life data are discussed.
They may be categorized as follows:
1. Long-term aging tests;
2. Tests performed in laboratories as combined with long-term aging tests and;
3. Theoretical/analytic methods as combined with long-term aging tests and laboratory tests.
2 LONG-TERM AGING TESTS
Long-term aging tests generate data on long-term ageing under in-use conditions. Long-term
aging tests are necessary for obtaining reliable degradation and service life data in an objective
way. Long-term aging tests can be divided into the following categories (ISO 15686-1):
1. Field exposure tests. This is the traditional way of exposing material specimens or
components on racks and stands in specific environmental conditions. Large series of
materials can be tested in an exposure field.
2. Inspection of old buildings. The idea is to use the existing building stock to acquire
reliable service life data. However, the usefulness of the inspection data from old
buildings may be limited by the difficulty of obtaining data on the history of the building
and on the difficulty in describing the initial material characteristics and the
environmental conditions of the items observed.
3. Experimental buildings. Specially for testing designed buildings which can expose
specific materials and components.
4. In-use exposure tests. Specimens can be exposed in buildings that have not been
specifically designed for experiment. This approach is necessary where degradation is
directly related to user actions and behavior.
To get more reliable data on degradation and service life of concrete structures new field tests
have been started in Nordic Countries in recent years. The purpose of these tests has been
provision for real data on the chloride penetration rate into concrete, corrosion on reinforcement

Current Methodologies for Service Life Prediction
3
and on frost scaling both in marine and motorway conditions. Other field tests in non-chloride
environments have been conducted for providing data on carbonation, corrosion of
reinforcement and frost attack without salt.

Figure 1. Field tests for concrete specimens beside a motorway near Kotka, Finland.
One of the most extensive long-term exposure programme has been carried out within the
International Cooperative Programme on Effects on Materials, including Historic and Cultural
Monuments in the framework of the UN ECE convention on long range transboundary air
pollution (Kucera 2004). The programme aimed at evaluating quantitatively the effect of air
born acidifying pollutants on the atmospheric corrosion of several materials (metallic and stone
materials, paint coatings, electric contact materials, glass materials and polymeric materials.)
The programme involves field exposure at 39 sites in 12 European countries and in the United
States and in Canada (ISO/FDIS 15686-2, Annex A).
Long-term aging tests are a prerequisite for reliable degradation and service life data as
eventually only the data obtained by field tests can be considered as firm ground. As such the
importance of the field tests is self-evident. However, there may be serious problems in
extending the application of these data to other materials and to different environmental
conditions than those applied in the actual tests. To extend the application of the field test data
beyond the limits of the actual test may require supplementary tests either in laboratory or in
field.
There are also serious problems with field tests resulting from the fact that field tests are
long-lasting and expensive. That is why the need of service life data can never be satisfied by
field tests only. Other methods should be developed to generate reliable service life data more
quickly. Such more rapid methods may be accelerated tests performed in laboratories and
application of theoretical/analytic modeling methods. However, both of them must be verified
and calibrated with observations from the field.
3 TESTS PERFORMED IN LABORATORIES AS COMBINED WITH LONG-TERM
AGING TESTS
To get data on the durability of materials and structures accelerated test methods have been
widely applied in different countries. Accordingly, many kinds of accelerated tests for material

4
testing are available. However, only a few of them are applicable to service life evaluation as
the relationship between the degradation rate in the test and the real in-use degradation rate is
not known. That is why when developing laboratory tests for service life evaluation the
degradation rate of the material or component in in-use exposure conditions should be studied
and related to the degradation rate in the test. The procedure which originally was proposed by
RILEM TC 71-PSL is presented in ISO 15686-1 and ISO 15686-2 as slightly modified. The
methodology presented in ISO 15686-2 is shown in Figure 2.
The ISO 15686 methodology contains the following phases:
1. Definition
2. Preparation
3. Pretesting
4. Aging exposure testing
5. Analysis and interpretation.
Definition of the problem includes definition of the materials or structures to be studied, the
range of the study to be established, and identification or specification of essential data.
Essential data may be user needs, performance requirements, building context, exposure classes
and materials characterization.
Preparation phase includes identification of possible degradation mechanisms and agents
(exposure conditions). Also postulations are made on which performance tests are used in
observing the degradation and on how the degradation can be accelerated without causing
undesired effects. When using accelerated short-term exposure care shall be taken to ensure that
extreme acceleration does not result in degradation mechanisms that would not be experienced
in service.
Pretesting is based upon the postulates made above. By pretesting a confirmation to the
degradation mechanisms, agents, selected test procedures and performance characteristics tests
is provided.
Testing of degradation is performed by observing and measuring
1. long-term degradation under in-use conditions (long-term-exposure) and
2. degradation under exposure conditions designed to accelerate the degradation (short-
term-exposure).
The ways to generate data on long-term aging under in-use conditions are the methods based
on experience (ref. chapter 2). Short-term exposure is carried out over a time period
considerably shorter than the anticipated service-life. A short-term test may include operating at
higher intensities of accelerating exposure or only be based on a shortened exposure period but
with controlled exposure conditions. Accelerated exposure is usually designed from the
information obtained from the pretests.
In the Analysis and Interpretation phases the results of the long-term-exposure tests are
compared and related with the results of short-term tests. Consistencies and inconsistencies are
used to increase understanding. If there is little or no correlation between the results, the data
should be re-examined and the short-term tests may need to be altered. In case of high
consistency between the results performance-over-time functions or dose-response functions
can be formulated by experts so that the results of the short-term tests can be used in prediction
of service life.
Unfortunately very seldom accelerated tests are qualified for prediction of service life. There
are many reasons for this (Pommersheim & Clifton 1985):
1. In a laboratory the specimens are usually tested in conditions substantially different from
real outdoor conditions. As a result the relationship between the time in the test and
service life is obscured.
2. Factors affecting service life may be complex and not well known leading to biased test
configurations in laboratory.
3. Acceleration of the test may change the degradation mechanism of a material from that
found in in-service conditions.

5

Figure 2. Systematic methodology for service life prediction of building components (ISO 15686-2).
In normal outdoor weather the temperatures and moisture conditions vary within a day and
within a year very much. That is why it is almost impossible to arrange conditions in a
laboratory that would completely correspond to the outdoor weather conditions.
This is especially true for degradation types that are based on phase transitions such as frost
damage. In natural weather conditions the possible frost damage of a structure is highly
dependent on occasional periods of rain and sunshine leading to a complex variation of
temperature and moisture content inside the structure. The same also applies very much to
chemical reactions such as carbonation and to electrochemical reactions such as corrosion.
That is why most laboratory tests can not give a firm basis for service life prediction.
However, laboratory tests may be useful as they generate data on relative degradation rates
between materials. Laboratory tests can also be used in studying the relative degradation rates at
different temperatures and moisture contents. In air conditioned test rooms of a laboratory we
can study e.g. corrosion rate in constant temperature and moisture conditions, or carbonation
rate on concrete in constant temperature, moisture content and CO2 content etc. This data can
be utilised in the degradation modelling based on a theoretical/analytic framework.

6
4 THEORETICAL/ANALYTIC METHODS AS COMBINED WITH LABORATORY
TESTS AND LONG-TERM AGING TESTS
The theoretical/analytic methods of durability and service life can be grouped as follows:
1. Theoretical/analytic modeling
2. Computer simulation.
Theoretical/analytic degradation modeling is based on general theories of diffusion, dissolution,
convection, erosion, abrasion, chemical reaction, electrochemical corrosion, phase transition etc.
Parameters to the theoretical framework are usually searched by experimental research both in
laboratory and in field. For a practically satisfactory model the parameters in the model should
be those commonly used in the design of structures.
Computer simulation refers to:
1. theoretical emulation of in-use climatic conditions,
2. determination of the temperature and moisture variations in a cross-section of a concrete
structure, and
3. application of temperature and moisture sensitive degradation models so that the
degradation over time and the service life can be predicted.
4.1 Theoretical/analytic modeling
As an example of theoretical/analytic modeling of a simple degradation model carbonation of
concrete is chosen. The fact that the depth of carbonation is approximately proportional to the
square root of time can be theoretically derived by applying the diffusion theory with a moving
boundary. In this theory the carbon dioxide is diffused through the already carbonized layer of
concrete and reacts with the non-carbonated calcium minerals at the moving boundary, that is
at the distance of X
c
(depth of carbonation) from the surface of a structure. The carbon dioxide
content between the surface and the moving boundary is assumed to be linear. Then the flux of
carbon dioxide towards the moving boundary can be evaluated as:
carb
S
X
C
D J
A
=

(1)
where J is flux of carbon dioxide, kg/m
2
s,
D the diffusion coefficient of concrete with respect to CO
2
, m
2
/s,
X
carb
distance of the moving carbonation boundary from the surface of the structure,
m,
AC
S
= C
s
- C
xc
, kgCO
2
/m
3
,
C
S
CO
2
content of air at the surface of concrete, kgCO
2
/m
3
, and
C
xc
CO
2
content of air at the moving boundary, kgCO
2
/m
3
.
The carbon dioxide flux into concrete must be in balance with the chemical reaction rate in
concrete at the boundary. The consumption rate of CO
2
depends on the amount of non-
carbonated calcium in concrete:
dt
dX
a
dt
dQ
carb carb
=
(2)
where Q
carb
is the amount of bound carbon dioxide in concrete, kgCO
2
/m
2

a CO
2
-binding capacity of concrete, kgCO
2
/m
3
and
t time, s.
By combining Equations 1 and 2 and integrating over time (X
Carb
= 0 when t = 0), the
following solution is obtained:
a
t C D
X
S
Carb
A
=
2

(3)

7
From Equation 3 it is seen that at constant conditions the depth of carbonation is proportional
to the square root of time. If AC
s
is constant in Equation 3, the equation can be presented in a
more simplified form as:
t k X
Carb Carb
= (4)
where k
Carb
is coefficient of carbonation, mm/a
0.5
.
The coefficient of carbonation can be studied experimentally in the laboratory at constant
moisture conditions and at constant temperature and for different mix designs of concrete. The
real carbonation in in-use conditions can be studied by field tests.
Another example of a typical degradation model is that developed for atmospheric corrosion
in metal structures (Klinesmith, McCuen R. & Albrecht 2007).
) ( 2
0
1 1 ) (
T T J
H F D
B
e
G
Cl
E
SO
C
TOW
t A t y
+
|
.
|
\
|
+ |
.
|
\
|
+
|
.
|
\
|
=

(5)
where y is corrosion loss, m
t exposure time, years
TOW time of wetness, h/year
SO
2
sulfur dioxide concentration, g/m
3
Cl chloride deposition rate, mg/m
2
/day
T air temperature, C.
A, B, C, D, E, F, G, H, J and T
0
are empirical parameters calibrated on the basis of the result
of long term exposure tests. The power function of t can be theoretically developed. The rate of
corrosion is declining (B < 1) because of the increasing oxide formation on the steel with the
result of retarding diffusion through the layer. Further details on degradation models for metal
structures are discussed elsewhere (Cascini, Portioli & Landolfo 2007).
Stochastic models are usually extensions of the deterministic models. As the deterministic
models show the average or the expected value of degradation the stochastic models express
also the standard deviation or coefficient of variation by which the variation of degradation can
be evaluated. If the variation of the material, structural and environmental parameters of the
degradation model are known the variation of degradation can also be evaluated by calculation.
The service life and/or the distribution of service life can be determined if the maximum
allowable degradation is given.
The reliability of the simple theoretical/analytic models has suffered due to poor
understanding on the effects of exposure which usually has been dealt with only by some
exposure classification and "environmental coefficients". The exposure classification can,
however, be considered an oversimplification as it cannot reflect the true effect of the local
weather. Another problem is the interaction between degradation types which is normally not
dealt with simple theoretical/analytic models. Still other problems are the changes of material
properties, such as the diffusion constant, with time and the effect of protective measures, such
as coatings, that normally are not considered in simple models. The answer to all these problems
might be computer simulation.
4.2 Computer simulation
Computer simulation offers a possibility to better manage the time related changes in materials,
structures or environmental stresses. Computer simulation utilizes theoretical/analytic
degradation models as exemplified above, however, as their differential approximations.
For example by combining Equations 1 and 2 the carbonation depth could be solved by
numeric simulation as follows:
t
) t ( X
C
a
D
) t t ( X
X X
Carb
s
Carb
Carb Carb
A
A
= A + A
A =

(6)
where AX
carb
is increment of carbonation depth during At
At increment of time.

8
The numeric solution as presented in Equation 6 does not bring any benefit as compared to
the analytic solution presented in Equation 3. However, much benefit can be gained if
parameters of Equation 6 are assumed to be time dependent. For example, if it is assumed that
the diffusion coefficient decreases with time (as a result of hydration), the diffusion coefficient
is affected by the surrounding temperature and moisture conditions, and/or the diffusion
coefficient is increased by simultaneous internal frost attack (as a result of internal cracking)
computer simulation based on Equations 6 can give the solution. Or if the concrete cover is
reduced by simultaneous frost-salt scaling (or other type of degradation which causes loss of
concrete on the surface of the structure) the numeric solution provides an answer. Then the
Equations 6 can be expanded as presented in Equations 7:
t
t X t X
C
t k t k t k
t
t
a
D
t t X
X X
FrSc Carb
s
IntFrost RH T
n
Carb
Carb Carb
A
A
|
.
|
\
|
= A + A
A =

) ( ) (
) ( ) ( ) ( ) (
0 0

( )
where D
0
is diffusion coefficient determined at age t
0,

t
0
age of concrete when the diffusion coefficient was determined
k
T
coefficient taking into account the effect of the momentary temperature on
the diffusion coefficient degradation,
k
RH
coefficient taking into account the momentary relative humidity on the
diffusion coefficient,
n exponent of time,
k
IntFrost
coefficient taking into account the effect of internal frost attack on the
diffusion coefficient, and
X
FrSc
depth of scaled concrete as a result of frost-salt attack.
The consequence of interaction between two degradation modes depends on the manifestation
of the degradation in the structure. The manifestation of degradation may be internal (cracking,
weakening, loosing material internally) or external (surface scaling, eroding, material loss from
the surface). In the former case the material parameters, such as the diffusion coefficient, are
changed as a function of the degree of internal degradation. Instead of frost attack some
chemical attack could be the reason for changing the diffusion properties and instead of frost-
salt scaling some other physical action, such as abrasion or erosion, could be the reason for
external loss of concrete.
By coefficient k
T
and k
RH
the effect of the real environmental conditions to the rate of
carbonation is taken into account. By computer simulation the thermal and moisture movements
in structures can be emulated closely as they occur in natural weather conditions. If the
increment of time is one hour the simulation program calculates hour by hour the temperature
and moisture content in various points of a cross-section having the real in-use conditions as
initial data on the surfaces of the structure. The weather is usually modeled based on long-term
statistics of weather parameters such as temperature, relative humidity, wind, rain and solar
radiation. Alternatively the measured weather data of a typical year are used as such.
By this means the effects of time dependent changes of material properties, temperature and
moisture variations and other degradation modes during the time interval At are considered. The
total degradation is obtained by totaling the differential increments of degradation during all
time intervals At. Service life is determined as the point of time when the allowable limit of
degradation is exceeded.
The effects of coatings on the degradation rate can also be evaluated by computer simulation.
The coatings effect on the degradation of the substructure both directly and indirectly. The
direct influence results from the diffusion barrier of the coating which retards the chemical
and/or electrochemical processes in the substructure. The indirect effect is the result of the
different moisture environment caused by the coating. The indirect effect may retard or promote
the degradation depending on the case. The total effect when considering both the direct and
indirect effects may be difficult to evaluate without computer simulation. By computer
simulation it is also possible to take into account the degradation of the coating itself. The
degradation impairs the protective properties of the coating by time (Vesikari 2002).
7

9
5 CONCLUSIONS
It is evident that all possible methods are needed to provide reliable durability and service life
data and to develop models for service life design. Usually reliable service life models can be
developed only by combining field tests to some laboratory tests and theoretical/analytic
modeling. Computer simulation might be a powerful tool in service life modeling when the
variations in environmental conditions, interaction between different degradation types and the
effects of protective measures such as coatings need to be considered.
REFERENCES
ISO 15686-1: 2000. Buildings and constructed assets - Service life planning - Part 1: General principles.
41 p.
ISO/FDIS 15686-2: 2000. Buildings and constructed assets - Service life planning - Part 2: Service life
prediction procedures. 25 p.
CIB W80/RILEM 71-PSL 1987. Prediction of service life of building materials and components.
Materials and Structures 20(1987)115. Pp. 55 - 72.
Klinesmith, D.E., McCuen R. & Albrecht P. 2007. Effect of environmental condition on corrosion rate.
Journal of Materials in Civil Engineering.
Kucera V. 2004. Mapping effects on materials in Manual mapping critical load. ICP Materials
Coordination Centre, Sweden.
Pommersheim, J. & Clifton, J. 1985. Prediction of concrete service-life. Materiaux et Constructions
18(1985)103. Pp. 21 - 30.
Cascini, L, Portioli, F. & Landolfo, R. 2007. Durability design of metal structures based on lifetime
safety factor method. Proc. Int conf on Sustainable Construction, Materials and Practices. SB07.
Lisbon, Portugal 2007. Pp 522 - 529.
Vesikari, E. 1998, Prediction of service life of concrete structures by computer simulation. Helsinki
University of Technology. Faculty of Civil and Environmental Engineering. Licentiates thesis. 131 p.
(in Finnish).
Vesikari E. 2002, The Effect of Coatings on the Service Life of Concrete Facades. Proc. 9th International
Conference on Durability of Building Materials and Components. Brisbane, Australia, 1721 March
2002. 10 p.

10
W. Trinius
Centre for Built Environment, University of Gvle, Gvle, Sweden
C. Sjstrm
Centre for Built Environment, University of Gvle, Gvle, Sweden
Standardization of Service Life Methodologies
1 INTRODUCTION
During the last 25 years, the subject field of service life prediction, service life planning and
service life declaration has gained intensive attention of the international research and
development as well as the international standardization community. In the European Union,
this interest manifested in the formulation of the Construction Products Directive (CPD), in
which six essential requirements to be fulfilled by construction products are established. These
requirements are to be fulfilled throughout the working life of the construction works.
Essentially, this necessitates that a service life planning exercise is being conducted.
The topic was in parallel being developed through activities in CIB, and gained the attention
of international standardization, first through CEN activities which due to the international
attention to the subject rapidly resulted in ISO activities. At the brink of completion, the ISO
standard series 15686 on "service life planning" provides standards on methodologies for
service life planning, including items such as life cycle costing, the link between LCA and
service life planning, data demands, feedback from practice, and methods to transfer and adapt
information between different scenarios.
The standards of the 15686 series are applied within the emerging ISO and CEN standards
concerning sustainability of construction works. The concept of service life is at the very core of
any "life cycle" consideration; additionally, it relates to the concept of performance-based
building, which itself is another field of growing international attention.
This paper gives a brief introduction to the concepts of
- service life planning,
- service life declaration,
- factor method
- failure mode and effects analysis
- deterministic methods
- feedback from practice (experience based information)
- CE marking
- CPD
- current ISO and CEN standards relating to service life.
2 SERVICE LIFE PLANNING
The International Standardization Organization (ISO) standard series on service life planning
(SLP) for buildings and constructed assets (ISO 15686) developed by ISO/TC59/SC14, as well
as the European Union thematic network on performance-based building (PeBBu), respond at
least partly to the European Union Construction Product Directive (CPD) (CEC, 1988).

11
Particularly, they respond to a horizontal requirement of the CPD, necessitating that buildings
and constructed assets be the subject of planning activities that reflect their performance
throughout their operating life, or in other words, require SLP.
The concept of SLP aims to enable planners to design buildings that meet or exceed identified
project-specific targets and performance requirements. Requirements may concern a variety of
issues, stretching from lifecycle costs and technical performance to environmental
considerations. To enable this, planners also have to assess information available for materials
or components in the light of the requirements for the building in its entirety - including aspects
of usage. In other words, they must consider and merge performance requirements and
performance properties at different scales. Planners may handle the task intuitively. However,
due to the increasing demands imposed by clear documentation of planning activities and the
need for verification (especially when tendering is performance based), a clear relationship
between requirements and performances on various levels needs to be established. (Sjstrm et
al) argue, that one difficulty is that the object on a higher system level (e.g. the building) is
comprised of objects from lower system levels (e.g. components and materials), but the
application in this higher system level decides their ultimate functionality. Requirements
originate from the highest system level, where fitness for purpose must be sustained. The
performance and functionality on the higher system level is substantially different from the sum
of the functionalities on the lower system levels.
The procedure of service life planning, involving the analysis of the preconditions, through
the identification and the setting of requirements to the assessment of design options in relation
to these requirements is indicated in Figure 1.

Figure 1. Service Life Planning according to ISO 15686-6.
3 SERVICE LIFE DECLARATION
The concept of service life planning relies on the availability of product and component related
service life information. Together with the application context of a product or a component, the
ability of these products and components to withstand deterioration agents and to maintain
performance forms the information basis on which the routine of service life planning can be
performed. A declaration of the reference service life, a service life declaration, needs to
indicate the development of performance over time, for a set of well-identified reference
conditions.
A general limitation to the application of service life planning is the fact, that despite the
increasing awareness of product manufacturers that information on life performance of their
products may be a persuasive argument in the market, such information is not typically
disclosed. Additionally, the majority of product standards do not include declarations of product
life performance.

12
If service life information, or information on life performance pf products and components is
provided, such information typically may have two origins. It can be based on experience
gained from product application, or "feedback from practice", or it can be based on qualified
indications, possibly based on testing, and provided by the manufacturer or qualified third
parties. In either situation, the conditions for which the provided information is valid necessarily
needs to be communicated alongside the service life information, as otherwise no scrutinization
of the information and no analysis as to the relevance to a given application situation could be
performed.
4 FACTOR METHOD
The existing empirical models are usually described for some given (reference) case e.g.
regarding material quality, maintenance and environment; and the same holds true for service
life information in general. At the time being, the design team is therefore generally required to
use the information from a reference in-use condition (e.g. declared service life from producer).
The process of service life estimation aims to transfer reference service life information to the
current design case, and to enable the designer or other decision maker to estimate service life
information relevant to his specific application situation, reflecting the current in use conditions,
see Figure 2.

Figure 2. The route from Reference Service Life to Estimated Service Life.
Where the conditions for which a reference service life has been established significantly differ
from the intended application situation, the information will need to be modified in order to
better reflect the given context. Such modulation of the information has then to take into
account the differences between the reference situation and the intended design situation, see
ISO/DIS 15686-8. One available procedure for this, called the Factor method, is described in
ISO 15686. The Factor method has been criticized for being too simple and that the results will
not be reliable enough. Due to its character and its potential applicability even as a simple
checklist, the method though is practical for designers.
The factor method addresses a set of factor classes, each representing a typical set of aspects
having an influence on the service life or products and components, when being integrated into
a building or other construction works. These factor classes are:
A. Inherent Performance Level
B. Design Level
C. Work Execution Level
D. Indoor Environment
E. outdoor Environment
F. In-Use Conditions
G. Maintenance Level.

13
5 FAILURE MODES AND EFFECTS ANALYSIS
The methodology of failure modes and effects analysis (FMEA) has recently been adapted to an
application to building products and components. The methodology is based on the
apprehension of degradations and failures of building products. An identification of degradation
scenarios and an assessment of their relevance allows the generation of a multi-performance
profile. A system analysis establishes a functional model of a building product, and the FMEA
allows an analysis of various relevant failure route scenario. For each scenario, information can
be gathered, but as this information does not originate from harmonized information sources,
fusion of data is another essential step of the methodology. A temporal quantification of the
failure scenarios ultimately leads to the establishment of the desired service life information,
reflecting the modelled failure modes and the chosen performance aspects and performance
requirements. (Talon et al 2005)
6 PREDICTION MODELS AND DETERMINISTIC METHODS
Chemical or physical deterioration or corrosion mostly causes the performance loss of building
products considered here. Other important reasons for, sudden, performance failures may be
fundamental abuse of products, fire, or explosion, e t c. Degradation can normally be expressed
as mathematical models. With a limiting value - a minimum performance requirement - the
service life can be calculated. Figure 3 shows generically some performance over time
functions.

Figure 3. A spread of performance measure with its mean value, printed relative to a requirement on the
performance measure, resulting in a lowest, average and highest expected service life. The performance
curve and requirement curve are simplified to illustrate the concept. (Trinius et al 2005).
Performance over Time (P O T) models are often called Damage Models. P o T or Damage
Models are characterised by the inclusion of a minimum performance requirement which
enables determining the service life. The establishment of P o T models requires knowledge and
data on the degradation environment and the essential degradation factors, the properties of the
material under study, and a established Performance Requirement relevant to the studied in-use
situation.

14
Important sets of models are Dose-Response functions, which are normally used to describe
corrosion processes. These can be expresses as a power function of degradation factors and
elapsed time:
M = a t
b

where M = corrosion rate at time t
a = a rate constant, e.g. expressed by the deposition of pollutants or other
degradation factors to a material surface
b = a power exponent governed by e.g. diffusion processes.
Dose-Response functions can only be used for service life prediction when an agreed
performance requirement is applied. P o T and Dose-Response models are established through
field and laboratory experiments in interaction with analyses of feedback from practice studies.
Degradation processes have a stochastic character, most easily understood by considering the
stochastic behaviour of e.g. climate and weather, and ideally P o T models should cover
stochastic variations. Most degradation models do, but it is also true that in engineering
applications they are mostly used in a deterministic way, i.e. with fixed mean values of
parameters introduced in the calculation process.
7 FEEDBACK FROM PRACTICE
ISO 15686 provides a separate part (part 7) of the standard series aiming at the topic "feedback
of service life data from practice". Acknowledging that significant experience with numerous
materials, products and components has been gained, and will continuously be made available,
this immense source of information is intended to be made applicable for service life planning.
As for the application of reference service life information from manufacturers, the user of
information must be aware that information based on experience from a singular application
necessarily includes the strengths as well as the weaknesses of that case from which the
information has been drawn. Information representing the conditions of product application
need to be communicated together with the service life information itself, otherwise the value of
the information is unclear. With an increasing number of well-documented cases however, it can
be assumed, that information sets for numerous applications of common products can be made
available.
A significant restriction remains the general unavailability of feedback information related to
innovative products, or for very specific applications. While it is not intended to stifle
innovation nor to stifle creativity, while sustaining the requirement that innovative solutions
need to meet the same requirements as established solutions, is must however be recognized,
that the non-availability of feedback information related to innovative products may be an
obstacle difficult to overcome.
8 CE-MARKING AND THE CPD
Within the European market for construction products, also innovative products are to meet the
generic requirements stated by European directives and resulting codes and regulations. The
basis for CE-marking of innovative building products is, as for all products, compliance with
the CPD.
The goal of the CPD is the removal of technical barriers in the construction products sector,
through:
- compliance of works with the six Essential Requirements
- transformation of the six Essential Requirements to product requirements by means of
Interpretative documents
- determination of performances for products used in the works by technical specifications
- attestation of conformity of the product.
The six essential requirements established in the CPD are to be fulfilled throughout the
"working life" - consequently a resulting horizontal requirement is the ability to handle service
life and life performance information.

15
Products are "fit for their intended use" if they comply with a Harmonised European Standard
(hEN), a European Technical Approval (ETA) or a non-harmonized technical specification
recognized at EC level.
9 CURRENT ISO AND CEN STANDARDS RELATING TO SERVICE LIFE
According to ISO 15392 Sustainability in Building Construction - General Principles,
"sustainable" buildings bring about the required performance and provide value while
minimizing adverse impacts. The understanding of "value" includes non-monetary value
considerations, and the "adverse impacts" include impacts on economic, environmental as well
as social aspects. The life cycle of a building or other construction works, including the
processes required to obtain and to maintain a building, plus any legacy, is the basis for the
quantification of impacts and for the consideration of sustainability. Addressing the life cycle
and the performance provided throughout the life cycle defines the prominent link of
sustainability considerations to the concepts of performance-based building and to service life
planning. Other ISO standards developed under ISO TC59 as well as the standards under
development within CEN TC350 do apply the same conceptual view, setting assessments in
relation to performance, including the consideration of life performance. The conceptual inter-
linkage between performance based building, service life planning and declaration and
assessment of sustainability aspects is illustrated in Figure 4.

Figure 4. Performance based building, service life planning and declaration and assessment in their
common context.
10 CONCLUSIONS
Service Life Planning as well as performance-based building can be considered to be pre-
requisites for the move towards a more sustainable development of the built environment and its
elements. The international standards on sustainability in building construction very well

16
demonstrate this connection. The information demand to be fed in order to enable service life
planning to be performed is immense and requires highly transparent and comprehensive
information, communicated in a way that enables designers and other building-sector actors to
work with that information. A good understanding of life performance aspects of construction
works and their parts is a necessary requirement when striving to increase performance, while
minimizing undesired impacts.
REFERENCES
CEC 1988, Council of the European Communities (CEC), Council Directive of 21 December 1988 on the
Approximation of Laws, Regulations and Administrative Provisions of the Member States Relating to
construction Products. 89/106/EEC, CEC, Brussels
ISO 15686, ISO 15686 standard series Buildings and Constructed Assets - Service Life Planning,
International Organization of Standardization, Geneva - consisting of parts
1 General Principles (under review)
2 Service Life Prediction Principles (published)
3 Performance Audits and Reviews (published)
4 Data Requirements (under development)
5 Maintenance and Life Cycle Costing (close to publication)
6 Procedures for considering Environmental Impacts (published)
7 Performance Evaluation and Feedback of service life data from practice (published)
8 Reference Service Life (final editing)
9 Guide on the inclusion of requirements of service life assessment and service life declaration in
product standards (final editing)
10 Performance standards in building (under development).
Talon et al 2005, Temporal quantification method of degradation scenarios based on FMEA, 10DBMC
International Conference on Durability of Building Materials and Components, Lyon, France, 2005.
Trinius et al 2005, Service Life Planning and Performance Requirements, Building Research and
Information (March-April 2005) 33(2), 173 - 181.

Service Life Design Methods of Constructions
17
R. Landolfo
E. Vesikari
VTT Technical Research Centre of Finland, Espoo, Finland
Service Life Design Methods for Civil Engineering Structures
1 INTRODUCTION
The optimization of construction performances over the time is one of the most important goal
to achieve for the sustainable management of the built up environment. Nowadays, it is a
common knowledge that the building sector is one of major consumer of non renewable raw
materials and energy resources and the costs of management, maintenance, repair and
refurbishment account for a large part in the economy of a country (CIB 2004a). Indeed, service
life prediction methods and durability of construction have been deeply investigated in the last
years. In the Agenda21 on Sustainable Constructions (CIB 1999) the development of models for
service life predictions was considered as one of the challenges toward a sustainable
construction, while in the Construction Product Directive (CPD, European Union 1988) it is
clearly stated that durability is an essential requirement for construction works (Member States
have provisions, including requirements, not only to building safety but also to health,
durability, energy economy, protection of the environment, and other aspects important in the
public interest).
During the 90s, the interest in research on durability of buildings and components has greatly
increased. Several reports, bulletins, papers and standards have been produced by renowned
international organizations, such as the International Association for Building Materials and
Structures (RILEM), the International Council for Research and Innovation in Building
Construction (CIB), the Fdration Internationale du Bton (fib), the International Organization
for Standardization (ISO). In order to give a short overview of the documents produced over the
years, a selection of references is presented in Table 1.
While the structural requirements together with the design and verification methods related to
mechanical safety are fully codified, the main European structural design codes give only
qualitative and common provisions for durability. No mandatory design procedure is
recommended for the evaluation of the durability performances over time under different
deterioration mechanisms. On the other hand, several design methods have been developed in
the framework of international European research projects to cope with this lack of codification.
As a matter of fact, several studies have been developed during the years but currently there is a
gap between the outcomes of the research and the day-to-day engineering practice. As a
consequence, there is a urgent need for the definition, at code level, of a standardized approach
to durability design and verifications of structures.

18
Table 1. Selected references for durability research and service life of constructions.
TITLE EDITOR(S) YEAR
Construction Products Directive European Commission 1988
Guide to Durability of Buildings and Building Elements,
products and components
British Standards Institution 1992
Guide for Service Life Planning of Buildings (English
version)
Architectural Institute of Japan 1993
Durability design of concrete structure
RILEM 1994
Guideline on Durability in Buildings
Canadian Standards Association 1995
Buildings and constructed assets. Service life planning.
ISO 2000
Performance Based Methods for Service Life Prediction,
CIB Report Publication 294.
CIB - RILEM 2004
Guide and Bibliography to Service Life and Durability
Research for Building Materials and Components, CIB
Report Publication 295.
CIB - RILEM 2004
Guidance Paper F: Durability and the Construction Products
Directive
European Commission 2004
Model code for service life design, bulletin 34
fib 2006
General principles on the design of structures for durability
ISO 2008

With this regard, the International ISO Standard on the design of structure for durability (ISO
13823 2008) represents a first step towards the codification of service life methods. Even if the
Standard does not address a specific design procedure for different construction materials (no
specific models for the degradation neither numerical or reference values for the partial safety
factors to be applied are provided), it represents a first and important attempt to create a general
framework and a unification role in the field of verification of durability. In the following, the
approach developed in the ISO 13823, the basic requirements on durability reported in the
Eurocodes and the probabilistic approach developed by RILEM and fib are discussed.
2 SERVICE LIFE DESIGN METHODS
Each construction, during its life cycle, will face with deterioration depending on several factors
such as the environmental condition, the natural aging, the quality of the material, the execution
of the works, the planned maintenance.
A deterioration could lead to a decrease of performance to such an extent that a structure
could be not able to satisfy the basic serviceability and safety requirements before the design
life has expired. In order to prevent the premature failure of a construction, the structural codes
provide several practical principles and application rules such as the use of protective systems
for material exposed in aggressive environment, the construction detailing aimed at avoiding the
initiation of degradation, the maintenance actions to be regularly performed etc.
Beside that, several design procedures based on the prediction of the deterioration that will
likely act on the structure have been developed in the framework of the international research,
especially for reinforced concrete structures (Sarja & Vesikari 1996, Sarja 2004, fib 2006). The
service life design methods developed so far are aimed at estimating the period of time during
which a structure or any component is able to achieve the performance requirements defined at
the design stage with an adequate degree of reliability. Such methods can be considered as
mathematical interpretation of the service life of a structure as a function of different design
parameters (Sarja & Vesikari 1996). On the basis of the quality of input information (mainly
concerning with the available degradation models), it is possible to distinguish among
deterministic methods, usually based on building science principles, expert judgment and past

19
experience, which provide a simple estimation of the service life, and probabilistic methods. As
for traditional structural design, different approaches can be used to evaluate the reliability
against durability, namely:
Prescriptive approach (instructions and design rules are provided)
Deterministic approach (level 0)
Semi-probabilistic approach (partial safety factors are defined, level I methods)
Probabilistic approach (reliability structural analysis is performed, level III and II)
According to Sarja & Vesikari (1996) "deterministic durability models are used in
deterministic durability design where the scatter of degradation (or performance or service life)
is not taken into account. With known values of parameters the model yields only one value (of
degradation or performance or service life) which is often the mean value. In some cases,
deterministic models are formulated to give an upper or lower fractile value instead of the
mean. In many cases the information yielded by deterministic models is insufficient to evaluate
the risk of not reaching the target service life. Especially in the mechanical design of structures,
stochastic design methods are considered essential as the scatter due to degradation is normally
wide and the degree of risk may be great".
The basic durability requirement can be expressed by means of service life t
S
versus design
life t
D
, according to equation (1). Service life is intended as the actual period of time during
which a structure fulfills the design performance requirement, while the design life is a specified
period of time during which the functioning of the element (the structure and/or a single
component) is ensured without major repairs. The numerical value of the design life is usually
specified in the structural codes (e.g. design working life in EN 1990:2002, indicative categories
are given such as 50 years for building structures and other common structures, 100 years for
monumental building structures, bridges and other civil engineering structures)
S D
t t > (1)
In the current practice, structural design is performed according to semi probabilistic and full
probabilistic approach considering different limit states while for durability only general
principles and application rules are provided. Indeed, the ISO Standard 13823 provides a
general framework to perform the verification of durability at different design levels. In this
section the approach developed in the ISO Standard 13823:2008 for the design of structures for
durability is presented. The Standard has been published in May 2008 resulting in a very useful
and comprehensive guide for the design of durability. Structures subjected to foreseeable
environmental actions leading to a decrease of performances over time are covered. Mechanical
actions producing deterioration are included, fatigue is neglected. Five informative annexes are
included to provide illustrative examples on the procedures defined in the Standard (limit state
method, the environmental actions causing degradation, the transfer mechanism from the macro-
environment to the structure level, the environmental actions for structural materials and their
control and procedures for ensuring durability). For the prediction of service life, three different
options are available. In particular, durability assessment shall be carried out according to past
experience, modeling and testing. As for modeling, conceptual and mathematical modeling are
defined.
Durability is defined as the capability of a structure or any component to satisfy, with
planned maintenance, the design performance requirements over a specified period under the
influence of the environmental actions, or as a result of a self-aging process and for the
verification of durability the use of limit state method is recommended. The limit state method
consists mainly in two steps: the definition and characterization of relevant degradation
mechanism that will likely act onto the structure and the verification of durability according to
formula 1.
The first step involves the characterization of the structure environment, the identification of
the transfer mechanism and the definition of environmental actions. In particular, the
environmental action that could lead to a loss of performance (damage, loss of resistance,
displacements, aesthetic losses, etc.) is the result of the global environmental condition that,
through a transfer mechanism, transforms the atmospheric influences in a local attack. The
structure environment is defined as the macro environment which the structure is located in and
it comprises outdoor or indoor atmosphere, the ground or the water. It could be characterized by

20
different parameters such as the humidity level, rainfall intensity, pollutants and so on. Gravity,
capillarity, condensation, ventilation are some of the transfer mechanisms causing
environmental actions.
In Table 2 an example of the characterization of relevant deterioration mechanisms according
to ISO 13823 is reported.
Table 2. Example of characterization of relevant deterioration mechanisms according to ISO 13823.
MATERIAL
STRUCTURE
ENVIRONMENT
TRANSFER
MECHANISM
ENVIRONMENTAL
ACTION
ACTION
EFFECT
Steel Outdoor atmosphere
Condensation; no
drainage
Atmospheric
corrosion
Thickness loss,
aesthetic loss, rust
expansion
Reinforced
concrete
Outdoor/indoor
atmosphere
Diffusion Chloride attack
Loss of bond,
failure of
reinforcement
Wood Ground Direct exposure
Subterranean
termites
Loss of material,
strength

Regardless of the adopted level of the analysis, the verification of durability requirements can
be performed either by service-life format or limit state format.
Service life format consists in defining t
D
and calculate the predicted service life.
In a full probabilistic approach the verification of durability consists in evaluate the
probability of failure and verify that it falls short of a target probability P
target
(formula 2). The
semi-probabilistic approach involve the definition of a characteristic value for the service life t
sk

and a partial safety factor
S
according to formula 3, while a deterministic verification of
durability consists in evaluate the numerical value of the predicted service life t
SP
and check the
basic requirements through formula 4.
arg
( )
S D t et
P t t P s s (2)
sk
D
S
t
t
> (3)
S SP D
t t t = > (4)
As for limit state format, three limit states can be defined, namely ultimate limit state (ULS)
if the resistance R is equal or falls short of the internal mechanical force, serviceability limit
state (SLS) when local damage and/or relative displacement affect the functionality and/or the
appearance of the construction and initiation limit state (ILS) corresponding to the initiation of
significant deterioration.
Also in this case, the verification shall be performed at different levels by means of
probability of failure and/or partial safety factors. The action effect S and the resistance R are
regarded as time dependant variables and, in a full probabilistic approach, they are both
modeled as random variables. The basic requirement for ULS is defined as:
( ) ( ) R t S t > (5)
and the verification of durability can be formulated a s followings:
| |
arg ,
( ) ( ) ( ) 0
f t et ULS
P t P R t S t P = s (6)
Serviceability limit states refers to the functionality in use, an upper bound S
lim
for the action
effect can be defined. The basic requirement and the verification according to a full probabilistic
approach are expressed by Formula 7, 8:
lim
( ) S S t > (7)

21
| |
lim arg ,
( ) ( ) 0
f t et SLS
P t P S S t P = s (8)
In the current practice structural design is performed according to Level I methods (limit state
analysis) while for durability only general principles and application rules are provided.
Nevertheless advanced methods have been developed. In the following sections, after a short
review of the principles and application rules given in the structural codes, some examples of
advanced procedures are presented.
3 LITERATURE REVIEW OF SERVICE LIFE DESIGN METHODS
3.1 Durability in structural design codes
In this section the approach of the structural codes used in practice is reported. In general terms
it is possible to say that durability is defined as one of the basic requirement to be achieved
during working life but usually only general provisions are recommended such as (JCCS 2001):
choice of suitable materials that will not deteriorate during the design working life (if the
maintenance is performed regularly).
accounting for deterioration allowance in the dimensioning of structural elements.
selecting a shorter working life for the elements that cannot be inspected and/or replaced.
performing regular inspection and maintenance.
The European Standard EN 1990:2002 Basis of structural design (EC0) addresses the general
principles and requirements for the design and verification of structures according to a limit-
state analysis. As for durability, Section 2 states that the structure shall be designed such that
deterioration over its design working life does not impair the performance of the structure
below that intended, having due regard to its environment and the anticipated level of
maintenance. Although each structure shall be designed to have adequate durability, only
general provisions are recommended to meet these requirements (i.e. the use of preventative and
protective measures, par. 2.2.5 b, the proper choice of the design working life, par.2.2.5 e).
Calculations, experimental investigation, experience from earlier constructions, or a
combination of these considerations are recommended to estimate the degree of any
degradation.
In case of steel structures, in EN 1993-1-1 (2004) only few common principles are stated for
durability of metal structures and in particular for preventing steel buildings from possible
causes of corrosion damage. The code refers to EC0 for durability in general and gives some
recommendations such as the opportunity of providing corrosion protection measures by means
of surface protection systems, improving the use of weathering or stainless steel and resorting to
structural redundancy. No analytical models and/or safety factors to be applied in order to
achieve these requirements are specified, thus the general provision to be applied are left to
practitioner experience and expert judgment.
Further more, it can also be noted that according to EC0, the mechanical design and
verification of structures shall be performed by limit state analysis using partial safety factors. A
design is considered to be sufficient if the relevant limit states are not reached when the design
values of the variables are introduced into the analysis.
The design values are derived from representative values (characteristic or nominal values)
through the use of partial safety factors. As for the resistance R the design value R
d
can be
derived from characteristic value R
k
through the partial safety factor
M
which take into account
the uncertainties related to material properties, geometrical deviation from nominal values and
model uncertainties, the effect of deterioration on the resistance is neglected. In addition, it is
worth to emphasize that either the actions on the structure and the resistance itself are time-
dependant variables but the time is usually neglected in ordinary design thanks to specific
assumptions. By the resistance side, the basic assumption that allows to neglect the variable
time is that the maintenance action will ensure the adequate resistance level as at in-built
condition time. Thus, if planned maintenance actions are inadequate or not performed at all, the
reliability of the structure over time could fall short of the target reliability level defined at the
design stage.

22
Only fatigue is regarded as time-dependant phenomena but also in this case the degradation
of resistance over time is neglected. In case of steel structures it is assumed that the fatigue
strength refers to structures operating under normal atmospheric conditions and with sufficient
corrosion protection and regular maintenance (Eurocode 1993-1-9:2005, par. 1.1.7).
In the European Standard EN 206-1:2002 requirements for concrete to withstand the
environmental actions are given either in terms of limiting values for concrete compositions and
established properties or requirements may be derived from performance-related design
methods. The requirements shall take into account the intended working life of the concrete
structure.
The performance-related design methods according to the following:
a) The refinement of the limiting values method based on long-term experience of local
materials and practices, and on detailed knowledge of the local environment.
b) Methods based on approved and proven tests that are representative of actual conditions
and have approved performance criteria.
c) Methods based on analytical models that have been calibrated against test data
representative of actual conditions in practice.
3.2 The Limiting values method
The limiting values method refers to limiting values for the composition of materials and proven
properties of materials and structural measures in various classes of exposure. By applying the
limiting values a structural designer can prove the structure to fulfil the requirements of the
intended working life.
An example of the limiting values method is given in EN 206-1:2002. A recommendation for
the choice of limiting values for concrete composition and properties is given in Table 3 when
using CEM I cement. The values in Table 3 are based on the assumption of an intended working
life of the structure of 50 years.
Table 3. Example of a limiting values method. Recommended limiting values for composition and
properties of concrete according to EN 206-1:2002.

The abbreviations XC, XS, XD, XF and XA refer to exposure classes which are defined in
the same standard. The limiting values of Table 3 are informative, meaning that they can be
revised on national bases. The minimum values for concrete cover are given elsewhere.
3.3 The Factor Method
The factor method (ISO 15686-1 2000) belongs to level 0 design procedure, the assessment and
evaluation of service life is deterministic and it is based on the identification of the main factors
influencing service life.

23
It is clearly specified that the method does not provide a guarantee on service life, but it
simply gives an estimation on the long lasting of components under certain condition.
The expected numerical value of service life (ESL) is obtained by adjusting the value of the
reference service life (RSL), through series of factors according to Formula 9:
ESL RSL A B C D E F G = (9)
where: A. Quality of components as supplied to the project;
B. Design level of a component or assemblys installation;
C. Work execution level or skill level of the installers;
D. Indoor environment;
E. Outdoor environment;
F. In-use conditions;
G. Maintenance level.
For the evaluation of the ESL, the RSL as well as the factors A - G are required as input. The
RSL shall be a documented period of time in years based on data provided by tests, building
codes and expert judgments. As for the adjustment factors, the standard does not provide
explicit values. The designer has to establish their numerical values (CIB 2004b).
3.4 The Life time safety factor method
The lifetime safety factor method in durability design was first time presented in the RILEM
Report on concrete structures (Sarja & Vesikari 1996) and then developed within the framework
of the EU Project LIFECON (Sarja 2004). This method represents a first attempt to apply
simplified probabilistic method to everyday engineering design procedures.
The method is used for the calculation of design life of constructions and it is based on
probabilistic degradation models, which consider the decrease of structural resistance caused by
different classes of environmental loads. The durability design based on lifetime safety factor
method is analogous with the static limit state design. In particular, it is related to control the
failure probability by considering the effects on R of the environmental loads acting during the
entire life time cycle, while static limit state design is devoted to control the structural reliability
of constructions under external mechanical loading. In durability limit state design the
resistance R is considered as a time dependant variable, contrary to static limit state design,
where the effects of time are usually neglected for R (Figure 1).
In particular, a deterioration function D can be formulated on the basis of the time dependant
resistance R according to formula 10:
( ) (0) ( ) D t R R t = (10)
where D(t) is the deterioration at time t; R(0), R(t) are respectively the resistance at t = 0 and at
the generic time t of the life cycle.

Figure 1. Representation of the life time safety factor design approach by R, S, t and D, t variables.

24
Because of the different sources of uncertainties that are involved in the definition of the
variation of capacity with time, the values in the previous equation are usually taken as mean or
characteristic values. The load S is usually adopted to be constant with time, and its design
value is also taken as a mean or characteristic value, multiplied by the relevant safety factors.
The failure event corresponds to time t
max
when the capacity R is equal to load S. The difference
R(t)-S represents the reduction with time of the safety margin:
max
( ) R t S = (11)
On the basis of previous considerations, a durability design procedure organized into
different steps can be formulated, as specified in the following.
First, the target service life t
g
has to be defined for the considered constructions. The target
service life corresponds to the design life defined in section 2. The reference value of the target
service life has to be selected according to mechanical design procedure.
Once the reference period has been stated, it is possible to identify the environmental loads S
that will likely act onto structure. Each environmental load is analysed and quantified, where
relevant, in a statistic way. The analysis of the environmental condition has to be performed in
order to define the project background. On the basis of the identification of environmental
loads, the degradation factors and mechanisms should be evaluated.
Once deterioration mechanisms that could act onto structures during the life cycle have been
identified, corresponding damage curves should be considered as a function of time in the form:
( )
n
m
D t t o = (12)
where D
m
(t) is the mean value of degradation
is a constant coefficient
t time
n degradation mode coefficient.
Substituting t
max
in eq. (12), we have:
max max
( )
m
D t D = (13)
where D
max
= R(0) - R(t
max
) = R(0) - S

represents the maximum allowable value of degradation
(e.g. the maximum allowable mass loss and/or corrosion depth in the cross section of a beam).
Durability requirements are fulfilled if the failure event (13) occurs after the design service
life had expired, with a proper safety margin. That could be expressed according to formula 14:
max
0
d g
t
t
t t
= > (14)
where t
d
is the (design) service life
t
max
is the calculated mean value of the service life corresponding to D(t
max
) = D
max

t0
is the central lifetime safety factor
t
g
is the target service life.
Assuming that degradation is normally distributed and the standard deviation of D is
proportional to the mean degradation, the coefficient of variation V
D
being constant, it can be
shown that the central lifetime safety factor of the design life depends only on safety reliability
index, the coefficient of variation of D and the exponent n, according to the formula 15:
( )
1
0
1
n
t D
V | = + (15)
On the basis of such assumptions, the safety reliability index depends on both the maximum
allowable failure probability for the selected limit state and the degradation function. In
particular, it is a function of the mean values and standard deviations of R and S, as follows
(formula 16):

25
2 2
R S
R S

|
o o
=
+
(16)
3.5 The fib model code for service life
The fib was founded in 1998 from the unification of the Euro-International Concrete Committee
(CEB, Comit Euro-International du Bton) and the International Federation for Prestressing
(FIP, Fdration Internationale de la Prcontrainte). Among different tasks related to structural
design of concrete structures, durability concerns have been deeply investigated since 1978
(CEB Bulletin 148 Durability of concrete structures 1982, Bulletin 182 Durable concrete
structures 1989 and Bulletin 238 New approach to durability design 1997).
The latest report Model Code for Service life Design (MC-SLD, fib 2006) was prepared
within the fib Task Group 5.6 with the aim to provide a basis and a general framework for
standardization of durability based design approach. It is worth to point out that the code refers
only to concrete structures, other construction materials are not considered thus far.
The aim of MC-SLD is to provide a design approach for durability similar to mechanical
design. To this end the environmental actions and the resistance of concrete against this action
are modeled quantitatively. The design approach could be summarized in four steps. The first
step consists in the definition of the deterioration mechanism affecting the durability of the
structure in terms of quantitative models. Thus the durability performances have to be defined
by means of limit states, the likelihood of attaining the defined limit states has to be evaluated
and finally the classification of the type of limit state by means of ultimate and/or serviceability
limit states shall be provided. According to MC-SLD, full probabilistic, partial factor, deemed
to satisfy design and avoidance of deterioration design criteria are established (Figure 2).

Figure 2. MC-SLD, an extract of the flow chart for service life design.
Full probabilistic approach is suggested for exceptional structures and in order to apply a full
probabilistic procedure, validated probabilistic model should be available and the model
parameters as well as the related uncertainties shall be quantifiable. In the partial factor
approach the probabilistic nature of the variables is taken into account by means of partial
factors. Deemed to satisfy methods consists in giving a set of rules based on statistical
evaluation of experimental data and field observation and/or on a long term experience for the
dimensioning, the selection of material and product and execution procedures. Specific
requirements to the workmanship, concrete composition, cover thickness etc. are recommended.
The avoidance of deterioration method consists in adopting solution and material so that
deterioration will not take place (i.e. the adoption of membranes to separate the structure from
the environment, the use of non reactive material etc.). Further details of the procedure
developed in the MC-SLD can be found elsewhere (fib 2006).

26
4 CONCLUDING REMARKS
The problem of durability can be described by two basic variables, the capacity R of the
structure to withstand the deterioration that will likely occur during service life and the loads S
acting onto the structure during service life. It shall be verified that during service life the
capacity meets or exceed the demand onto the structure. In general both the capacity R and the
load S are time dependant variables. In the mechanical design usually the time is neglected
thanks to specific assumptions, but in durability design the time is explicitly considered on the
resistance side and the decrease of resistance over time is calculated. As for traditional structural
design, different approaches can be used to evaluate the reliability against durability, namely:
Prescriptive approach (instructions and design rules are provided)
Deterministic approach (level 0)
Semi-probabilistic approach (partial safety factors are defined, level I methods)
Probabilistic approach (reliability structural analysis is performed, level III and II)
In the common practice, structural design is performed according level I methods while for
durability only design rules and general provisions are established. Indeed, several design
approaches have been developed in the framework of international research. The publication of
the ISO Standard 13823:2008 represents a first step towards the codification of durability
design. The Standard does not address a specific design procedure for durability as no specific
model for the degradation of construction material neither numerical or reference values for the
partial safety factors to be applied are provided but it represents a first and important attempt to
create a general framework and a unification role in the field of verification of durability.
Currently there is a gap between researchers and practitioners in the field of durability and
additional efforts have to be provided to introduce advanced service life design methods into
day to day engineering practice.
REFERENCES
Architectural Institute of Japan 1993. The English Edition of Principal Guide for Service Life Planning of
Buildings, Architectural Institute of Japan, Japan.
British Standards Institution 1992. BS 7543:1992 Guide to Durability of Buildings and Building
Elements, Products and Components. British Standards Institution, London, UK.
Canadian Standards Association 1995. CSA S478-1995 Guideline on durability in buildings. Ottawa,
Canada.
CIB-Council for Research and Innovation in Building and Construction 1999. Agenda21 on Sustainable
Construction. CIB Report Publication 237. Rotterdam.
CIB-Council for Research and Innovation in Building and Construction 2004a. Performance Based
Methods for Service Life Prediction State of the Art Reports. Part A, Part B. CIB Report Publication
294.
CIB-Council for Research and Innovation in Building and Construction 2004b. Guide and Bibliography
to Service Life and Durability Research for Building Materials and Components. CIB Report
Publication 295.
CEN 2001. EN 1990. Eurocode. Basis of structural design.
CEN 2004. EN 1993-1-1. Eurocode3: Design of steel structures. Part 1-1.
CEN 2002. EN 206-1. Concrete - Part 1: Specification, performance, production and conformity.
European Union 1988: The Construction products directive. Council Directive 89/106/EEC of 21
December 1988 on the approximation of laws, regulations and administrative provisions of the
Member States relating to construction products, European Union, Brussels, Belgium, December.
Fib-Fdration Internationale du Bton 2006. Model Code for service life design Bulletin 34. Stuttgard.
ISO-International Organization for Standardization 2000. ISO 15686-1:2000, Building and constructed
assets- Service life planning- Part 1: General principles.
ISO-International Organization for Standardization 2008. ISO 13823:2008(E). General principles on the
design of structures for durability.
European Commission 1999: Durability and the Construction Products Directive, Guidance Paper F,
European Commission, DG III, Brussels, Belgium.
JCCS-Joint Committee on Structural Safety 2001. Probabilistic model code Part I. available on
www.jcss.ethz.ch.

27
Sarja, A. 2004. Lifecon deliverable D2.1. Reliability based methodology for lifetime management of
structures. Finland.
Sarja, A. & Vesikari, E. 1996. Durability design of concrete structures. RILEM Report of TC 130-
CSL.RILEM Report Series 14.

28
Sz. Wolinski
Rzeszow University of Technology, Rzeszow, Poland
Risk Based Approach to Service Life Assessment of Building
Structures
1 INTRODUCTION
Building structures should be designed in such a way that they will sustain all actions likely to
occur during execution and intended service life, with appropriate degree of reliability, in
economical way and taking into consideration environmental aspects. The reliability required
for structures can be achieved by design in accordance with current standards, appropriate
execution and quality management measures (EN 1990, 2002, ISO 2394, 1998). The difference
between design of a new structure and assessment of an existing structure consists in the
available or collectable information about the structure related to materials, actions and the
structural performance characteristics. Assessing of existing structures can be understood as a
process of model formulation, consequence evaluation and model updating by introducing new
information (Faber at al., 2006, Wolinski, 2006).
An existing structure needs an assessment in situations where: unexpected or excessive
degradation has been observed, the structure has been subjected to non-foreseen events or
increased loading, modification in use or increased service life is planned, the structure has not
been inspected for a long period of time, etc. A structure can be measured, inspected, tested and
proof-loaded. All information relevant for assessing the performance of a structure should be
collected, but at a cost.
Current design codes are based on the limit states concept and the reliability based methods
are recommended to provide a minimum safety level of structures. The measure of reliability is
identified with the probability of survival that depends on safety of people and balance between
the costs of failure and safety measures. Calibration of reliability levels is generally based on the
past experience and target values of the survival probability depends on various design
situations, failure modes, reference periods and so on, are generally notional numbers. Recently
there is generally recognition that the structural design and assessment should involve the
considerations of reliability, safety, economic and environmental aspects and that these aspects
should be addressed explicitly and fully quantitatively. Acceptance criteria in structural
engineering should take into considerations also different consequences of failure, especially:
loss of human life or serious injuries, economical and environmental losses. They can be
derived from social indicators, which are statistics that express some aspects of quality of life in
society. The reliability based criteria that take into consideration only the probability of failure
and loss of functionality does not meet these requirements.
The present paper deals with recent developments of service life assessment of building
structures based on the general methods for assessing risk of systems constructed by standard
components. The approach presented utilizes quantitative definitions of exposure, robustness
and vulnerability of building structures and taking into consideration both, direct and indirect

29
consequences of their failure and loss of functionality. The presented approach is illustrated
through an example application.
2 RISK INDICATORS
2.1 Hazard and risk in structural engineering
Hazard is defined as an attribute of activities which may cause harm to persons and assets or as
a set of circumstance with the potential for causing events with undesirable consequences
(Vrouwenvelder et al., 2001). However, in the Eurocode hazard is defined as an unusual and
severe event. The two major categories of hazard involved in the building process can be
distinguished on the basis of their nature, namely natural hazards and man-made hazards.
Natural hazards resulting from variations of structural materials and products properties, actions
applied to a structure, geometrical data, which are modeled as stochastic variables with
corresponding statistical distributions and parameters. Man-made hazards include uncertainties
due to unintentional and intentional departure from the accepted practice and verified
procedures (human errors), new materials and types of structures, innovations in design and
construction, new technologies and methods of execution, new or modified models for structural
analysis and dimensioning, fires, explosions and other severe events caused by man who are not
involved in the building process but also the pressure on the designers due to the shortage of
time, money and the political climate.
Negative consequences are defined as a possible outcome of desired or undesired events that
may be expressed quantitatively or qualitatively in terms of personal injury, death, monetary
loss, environmental and social damage.
Risk may be referred to as a measure of the danger or hazard that undesired events represents
for people, economy and environment, and is defined as a combination (usually a product) of
the probability of occurrence and the consequence of a specified hazardous or undesired event
(Steward & Melchers, 1997, Vrouwenvelder et al., 2001).
For a set of hazardous design situation
i
H the total risk R can be calculated as follows (Faber
et al., 2006):

= = =
=
H D S
n
i
n
j
n
k
k j k i j i
S C D S p H D p H p R
1 1 1
) ( ) ( ) ( ) ( (1)
where the structure is subjected to
H
n different hazards that may damage the structure in
D
n
different ways and the performance of the damage structure can be discretised into
S
n adverse
states S
k
with corresponding consequences C(S
k
), and p(H
i
) is the probability of occurrence of
the i-th hazard
i
H , ) (
i j
H D p is the conditional probability of the j-th damage state of the
structure given in the i-th hazard and ) (
j k
D S p is the conditional probability of the k-th
adverse overall structural performance S given in the i-th damage state.
2.2 Exposure and consequences
Methods for assessing risk of systems can be efficiently applied for assessing service life of
building structures. Generally, a building structure can be defined as a bounded group of
interrelated, interdependent or interacting elements forming an entity that achieves a defined
objective in its environment through interaction of its parts. A building structure considered as
the system involves structural system formed from a physical subsystem, environmental, social
and human subsystems.
An exposure is considered as any event or set of events with the potential to cause damage,
loss of functionality or failure to a system. A description of a series of events in time and space,
and inter-relationship among the events, provided the occurrence of a hazard is called a
scenario. Exposure events could come from different types of direct and indirect actions or
environmental processes, errors and other disturbances. In the risk based approach conventional
action or load combination models established in reliability-based design methods should be
replaced with event scenarios, for example in the form of event tree. The event tree

30
representation is a tool for evaluating event scenarios that could occur to the structure, and it
also incorporates the associated probabilities of occurrence (Fig. 1).
Two significant types of consequences, both immediate and arise after a certain time,
associated with the exposure to a structural system may be distinguished:
- direct consequences induced by damage to the individual constituents of the system,
- indirect consequences beyond the direct consequences, induced by changes of the system.
C

DIRECT RISK
F EX
k

C

D INDIRECT RISK

F INDIRECT RISK

SD
i
EX
j

D NO RISK

Figure 1. Representation of the exposure in form of an event tree. Clarification: SD
i
- the i
th
version of
structure design, EX
j
- the j
th
scenario of exposure, D - damage, D - no damage, F - failure, F - no
failure, EX
k
- the k
th
extraordinary event, C - collapse, C - no collapse.
2.3 Direct and indirect risk
The risk R
D
associated with all direct consequences due to exposure events may be assessed as
follows (Faber et al., 2006):
=
=
EX CD
n
k
n
l
k l D k l D
EX p C c EX C p R
1
) ( ) ( ) ( (2)
where n
EX
is a number of exposure events, n
CD
is a number of possible different states of all
constituents of the element C
l
, ) (
k l
EX C p is the conditional probability of the l-th damage
state of the element C
l
on the exposure event EX
k
with probabilistic characterization ) (
k
EX p
and ) (
l D
C c is the direct consequence associated with the l-th of n
CD
possible state of damage of
all constituents of the element C
l
.
The risk R
ID
due to all indirect consequences of exposure events may be calculated using the
formula:
= = =
=
EX CD ST
n
k
n
l
n
m
k k l k l m l D m ID ID
EX p EX C p EX C S p C c S c R
1 1 1
) ( ) ( ) , ( )) ( , ( (3)
where
ST
n is a number of possible different structure states S
m
associated with indirect
consequences )) ( , (
l D m ID
C c S c and ) , (
k l m
EX C S p is the conditional probability of indirect
consequences on a given state of the constituents C
l
and the exposure EX
k
.
Following (Faber et al., 2006) the direct risk R
D
is the measure of system vulnerability and
the ratio of direct risk to the total risk (R = R
D
+ R
ID
) is the measure of system robustness.
For a given set of the N different scenarios of actions, the risk associated with each scenario
can be computed and the corresponding index of robustness can be calculated using the
generalization of equation (Faber, 2007a, Faber et al., 2007b):

31

= =
=
+
=
N
i
i ID
N
i
i D
N
i
i D
R
R R
R
I
1
,
1
,
1
,
(4)
The robustness index takes values 1 0 s s
R
I depending on the source of risk. If all risk is
due to direct consequences, the system is completely robust and 1 =
R
I . If all risk is due to
indirect consequences, the structural system has no robustness and 0 =
R
I .
3 RISK ACCEPTANCE
Acceptance criteria in structural engineering should take into considerations different
consequences of failure, especially: loss of human life or serious injuries, economical, social
and environmental losses. They can be derived from social indicators, which are statistics that
express some aspects of quality of life in society. When discussing the problem of acceptable
risk levels it is necessary to take into account the fact that individuals may have a different
viewpoint to what is acceptable risk as compared to the viewpoint of the society and decision
makers. Considering activities related to civil engineering it is obvious that the preferences of
individuals may be in contradiction with the preferences of society. Therefore, a normative
approach to modelling of preferences and criteria for risk acceptance are essential. Commonly
used formats for risk acceptance are the so-called FN-diagram or FN- matrixes.

Figure 2. FN-diagram according to ISO 2394: 1998.
In Figure 2, the FN-diagram for acceptable risk to human life according to the formula:
o
< N A year f P ) / ( , from the ISO 2394 is presented. On the x-axis the consequences in
terms of the number of fatalities are given, and on the y-axis the probability of occurrence of the
collapse events are given. The plots in Figure 2 show domains of unacceptable, tolerable and
negligible risk to human life characterized by lines corresponding to constants: A = 0.1, 2 = o
and A = 0.01, 2 = o (Wolinski, 2008).
In many cases relative costs of safety measures and consequences of structural failures are
defined by means of linguistic variables in natural language. For instance, in the ISO 2394:
1998 standard the target probabilities of failure are referred to the relative costs of safety
measures defined as: high, moderate and low, and consequences of structural failure as: great,
moderate, some and small.
In Figure 3 consequences of failure expressed by means of the fuzzy target probabilities of
failure and the relative costs of failure defined in terms of the membership functions, are
presented together with the corresponding frequency-consequences diagram corresponding to
recommendations given in ISO 2394: 1998 and EN 1990: 2002.

32

Figure 3. Fuzzy frequency-consequences FN-diagram.
Complex social indicators are commonly used for optimization of the tolerable risk.
Unfortunately, sufficiently reliable economic and demographic data as well as actual failure
rates for different types of building and engineering structures are insufficient, time-dependent
and generally questionable. Moreover, life quality has more dimensions than consumption, long
life and leisure time which are usually considered. Ethical, esthetic, cultural and religious values
can not be measured by the standard life quality indicators, that usually incorporate questionable
or unreliable data related to the quality of life: cost of averting to fatality, life expectancy,
personal well-being, time for leisure, healthy ecological environment, cultural heritage, etc.
4 RISK BASED ASSESSMENT OF SERVICE LIFE
The difference between design of a new structure and assessment of an existing structure
consists in the available or collectable information about the structure related to exposure,
vulnerability, robustness and costs of improvements of the structural performance
characteristics. The assessment of service life of existing structures may be understood as a
process of model formulation, consequence evaluation and model updating by introducing new
information. The following main steps in the risk-based framework of the assessment process
may be distinguished:
- Select a list of doubts about a structure, an exposure, relevant limit states (deterioration,
ultimate and serviceability limit states), reference period, etc.
- Confirm doubts by means of the preliminary inspection (study of documentation, site
visit, simple check, etc.).
- If doubts are confirmed, then further inspection, investigations and analysis should be
undertaken which include steps as follows:
- formulation of prior uncertainty models referred to actions, deterioration mechanisms,
behaviour of elements and the structure and models for the service life prediction,

33
- formulation of time-dependent performance and limit state functions, establishing
posterior probabilistic models (using different frameworks: conventional or Bayesian
or fuzzy-statistical),
- calculating direct and indirect risk, an index of robustness and remaining service life.
- Check compliance with codes and regulations (vulnerability), the level of robustness and
remaining service life of the structure.
- If the results of checking are positive, update the maintenance strategy.
- If the results of checking are negative, choices of repair and strengthening or demolition
of the structure is necessary, taking into account:
- assumed period of the remaining service life,
- refined limit states analysis,
- direct and indirect risk after repairs and strengthening,
- economical decision analysis.
5 ILLUSTRATIVE EXAMPLE
An existing open-area support structure for warehouse crane consists of n = 215 = 30
prefabricated reinforced concrete beams simply supported on columns fixed in isolated pad
foundations. The structure is aged 22 years. Due to exposure conditions (i.e. large variation of
moisture and temperature as well as freeze and thaw) deterioration such as rebar corrosion,
cracking and spalling or delamination of concrete and the premature loss of functionality or
collapse of beams can occur.
- The preliminary inspection has shown the appearance of moisture and rust stains, few
steep cracks and poor quality of concrete.
- Considered beams have been designed using the limit states method and a partial factors
format. The design values load bearing capacity for bending is M
d
= 224 kNm.
Corresponding value of the notional failure probability of one beam is equal
6
10 5

=
f
p . The structural system consists of 30 beams and forms the series system.
For series system with perfectly correlated elements 1 =
ij
,
6
,
10 5

= =
f i f
p p while
for series system with uncorrelated elements 0 =
ij
, . 10 5 . 1 ) 1 ( 1
4
,

= =
n
i f f
p p
When elements of the considered series system are equally correlated and the coefficient
of correlation is 80 . 0 =
ij
, the probability of failure equals . 10 1 . 2
5
~
f
p
Unfortunately, values of correlation coefficients between the resistance of different beams
are extremely uncertain.
- Five stages in development of reinforcement corrosion in concrete are commonly
distinguished (Bertolini et al., 2004, fib, 2006): depassivation of concrete cover, initiation
of reinforcement corrosion, cracking of concrete, spalling of concrete cover, end of
service life or collapse.
- Using simplified formulas based on the Ficks law of diffusion and assuming that the
increase in the volume of the crack is equal to the volume of the corrosion products
produces when the diameter of the reinforcement bar is reduced, the corrosion initiation
time t
cor
, the cracking initiation time t
crc,
the spalling initiation time t
spl
and the time to
collapse t
col
were calculated (Thoft-Christensen, 2000, fib Bulletin, 2006, Liu & Weyers,
1998). The Monte Carlo Method was used to perform the probabilistic calculations.
Results of these calculations are presented in Figure 4.

34

Figure 4. Density functions for: (a) corrosion initiation time t
cor
, (b) cracking initiation time t
crc
, (c)
spalling initiation time t
spl
, (d) time to collapse t
col
; where t is time in years,
(...)
t is mean value,
01 . 0 )
~
(
~
= s
col col
t t p t

- Diagrams of theoretical and updated relationships between the age and the relative value
of direct consequences are presented in Fig. 5. According to the results of inspection after
22 years of service the cost of necessary repairs and protection of damaged beams was
estimated at 5% of all beams replacement, instead of calculated theoretically 2%. In order
to correct the theoretical diagram its second part was shifted up in a parallel direction and
the updated mean values of corrosion and cracking initiation time were calculated:
cor
t =
14.4 years and
crc
t = 23 years. Then next two parts of the theoretical diagram were
shifted up in the same way and the updated mean value of spalling time
spl
t = 38.8 years
and time to collapse
col
t = 73.2 years were assessed.

Figure 5. Diagrams of theoretical and updated relationships between the age and direct consequences.

35
- Service life of the structure may be defined in two ways: when the relative value of direct
risk R
D
achieves the target value (economic criterion) or when the relative value of
indirect risk R
ID
due to collapse caused by corrosion of tension reinforcement achieves the
target value (human safety criterion, for instance according to ISO 2394: 1998).
Replacement cost of all 30 beams makes up 100% of direct consequences and the
conditional probability of a person being killed when at least one beam collapses makes
up 100% of indirect consequences.
- The updated t C C
rep D D

,
/ diagram in Fig. 4 can be used to assess the service life of the
structure. For example, assuming that the target value of relative direct consequences is
0.50, the mean value of service life equals 52.4 years (R
D
= 0.50.5 = 0.25) and
corresponding probability of human death 0005 . 0 ) 4 . 52 ( = s t p . For uncorrelated
resistance of beams indirect risk is R
ID
= 1[1 - (1 - 0.0005)
30
] = 0.015 and belongs to the
domain of tolerable risk shown in Figure 1.
6 CONCLUSIONS
Risk in the structural engineering can be analyzed and evaluated by means of quantitative
criteria for identified hazard scenarios, probabilities of considered events and their
consequences, information obtained from inspections and assumed risk acceptance criteria.
The assessment of service life of existing structures may be understood as a process of model
formulation, evaluation of consequence and model updating by introducing new information.
The approach presented in the paper utilizes quantitative definitions of exposure, robustness
and vulnerability of building structures. Both direct and indirect consequences associated with
loss of functionality, failure or collapse should be taken into consideration.
The tolerable risk should be optimal with regard to benefits and costs and should be based on
human safety and socio-economic values.
The remaining service life of a structure may be defined in different ways, but it is strongly
recommended to assume the socially allowable indirect risk related to human safety and to
minimize direct risk connected with all types of costs.
Hazard identification, modeling of hazard scenarios, estimation of probabilities and
consequences compels the designer to careful examination of the whole building process and its
interactions with safety, economy, social and natural environment. The presented approach
provides a helpful supplement to conventional assessment of building structures.
REFERENCES
Bertolini, L., et al. 2004. Corrosion of steel in concrete: prevention, diagnosis, repair. Wiley-VCH
Verlag GmbH & Co. KgaA, Weinheim, ISBN 3-527-30800-8: 392.
EN 1990:2002 E (2002). Eurocode - Basis of structural design. Brussels: CEN.
Faber. M., et al. 2006. On the qualification of robustness of structures. Proceedings of OMAE 2006, 25
th

Offshore Mechanics and Arctic Engineering Conference, Hambur, Germany, June 4-9, 2006,
(OMAEE 2006-92095).
Faber, M.H. 2007a. Framework for risk assessment of structural systems. Proceedings o f Workshop
COST C26: Urban habitat constructions under catastrophic events. Prague 30-31 March, 2007.
Czech TU in Prague: 359 - 367.
Faber M., et al. 2007b. Principles of risk assessment of engineered systems. 10
th
International
Conference on Applications of Statistics and Probability in Civil Engineering. The University of
Tokyo, Japan. fib Bulletin No. 34. 2006. Model Code for Service Life Design. EPFL Lausanne.
ISO 2394: 1998(E). General principles on reliability for structures. Geneve: International Standard ISO.
Liu, Y., Weyers, R. 1998. Modeling of the time to corrosion cracking in chloride contaminated reinforced
concrete structures. ACI Journal, Vol. 95: 675 - 681.
Steward, M.G. & Melchers, R. E. 1997. Probabilistic risk assessment of engineering systems. London:
Chapman Hall.

36
Thoft - Christensen, P. 2000. Modeling of the deterioration of reinforced structures. Proc. 9
th
IFIP WG
7.5 Working Conference on Reliability and Optimization of Structural Systems. The University of
Michigan, Ann Arbor: 15 - 26.
Wolinski, Sz. 2006. Risk reliability-based design. Proceedings of the XI
th
International Conference on
Metal Structures: Progress in steel, composite and aluminium structures, Rzeszow, Poland, 21-23
June 2006, Taylor&Francis Group: London, pp. 358 - 3599 (full text on CD- ROM).
Wolinski, S., 2008. Risk-based design and assessment of concrete structures. Conference Proceeding of
6
th
International Conference on Analytical Models and New Concepts in Concrete and Masonry
Structures. AMCM 2008. Lodz, Poland, 9-11 June, 2008, pp. 345 - 346 (full text on CD-ROM).
Vrouwenvelder, T., et al. 2001. Risk Assessment and Ris k Communication I n Civil Engineering CIB
Report: Publication 259, Rotterdam: CIB General Secretariat.

Chapter 2. State-of-the-Art Report on Deterministic and
Probabilistic Degradation Models

Chapter 2. State-of-the-Art Report on Deterministic and Probabilistic Degradation Models
Introduction
39
E. Vesikari
R. Landolfo
Introduction
This chapter is focused on Degradations models and is organized in two parts.
Part 1 deals with degradation modes and models for different construction materials. The
state-of-the-art related to concrete, metal, masonry and wood structures is given in four papers.
The fifth paper deals with fatigue in steel and composite structures.
Part 2 addresses application of degradation models to durability design. It consists of two
state-of-the-art papers. The first paper gives an overview of more important developments and
some general remarks on service life design of reinforced concrete structures with emphasis on
probabilistic approach. Three levels of probabilistic methods are taken into consideration: semi-
probabilistic, simplified probabilistic and full probabilistic. The second report provides a first
proposal for a general approach to the durability design of metal structures based on lifetime
safety factor method. All the phases relevant to the design procedure have been discussed and
defined in detail, according to a performance based approach.
In modern design of structures the consideration of service life is necessary. The degradation
models presented in this chapter can be used in practical service life prediction and service life
design of structures. Other sectors of design where these models can be applied are the
ecological and economical life cycle analyses and time variant risk analyses.

40
E. Vesikari
C. Giarma
Aristotle University of Thessaloniki, Thessaloniki, Greece
J. Bleiziffer
University of Zagreb, Faculty of Civil Engineering, Zagreb, Croatia
Degradation Models of Concrete Structures
1 INTRODUCTION
Degradation of concrete structures is a normal consequence of the ageing process. Both physical
and chemical changes occur in concrete structures with time. However, a number of parameters
like the quality of concrete, the climatic conditions, or the lack of maintenance, can greatly
influence this process.
The most important degradation mechanism in concrete structures is corrosion of
reinforcement which starts as a result of chemical changes in concrete. The chemical changes
are usually a result of carbonation and chloride penetration in concrete. Other forms of
degradation in concrete are freeze-thaw deterioration and deterioration due to aggressive
chemical exposure.
The levels of moisture on the surface and inside a concrete component are of decisive
importance for its life duration. Indeed, the water that is accumulated inside a concrete
component and the periodic variations of its water content are strongly related to numerous
degradation mechanisms. Water inside or on the surface of a construction element can act either
directly, being the main aggressive agent, or indirectly, facilitating the action of other
degradation mechanisms and factors and accelerating the progress of numerous deterioration
processes (Fagerlund 1996). The rate of the chloride ion diffusion, the time needed for the
carbonation of the concrete cover, as well as the propagation of corrosion are all influenced
from the moisture in the component (Rendell et al. 2002, Neville 1995, Lea & Hewlett 1998).
2 CHEMICAL CHANGES IN CONCRETE IN CONCRETE AND CORROSION
REINFORCEMENT
2.1 General
Corrosion of reinforcement in concrete is the primary degradation type of concrete structures in
general. Corrosion may be initiated either by carbonation or chlorides that penetrate into
concrete and depassivate steel. In structures exposed to sea water both chloride penetration and
carbonation take place as also in bridge structures that are exposed to deicing chlorides. In
buildings no chloride attack can normally take place but carbonation is always possible in
structures exposed to air. In cold regions frost attack may occur concurrently with carbonation
and chloride penetration thus accelerating the degradation process.

Chapter 2. State-of-the-Art Report on Deterministic and Probabilistic Degradation Models:
Degradation Models for Service Life Design
41
In the following the European models for carbonation, chloride penetration and corrosion
of reinforcement are presented. These models were developed during many joint European
projects such as DuraCrete, DARTS, LIFECON etc. The advantage of these models is the large
acceptance on them although full agreement for a European standardization has not been
attained. In this paper the first four reports in the Reference list have been the main sources for
the presented models. The notation of the parameters may be different, however.
Corrosion of reinforcement can be divided into two subtypes:
- General corrosion which takes place on the reinforcement as a result of general
depassivation of steel by carbonation or chloride penetration.
- Corrosion of reinforcement at cracks of concrete.
Different degradation models have to be applied for these two types of corrosion. Also the
structural consequences of general corrosion and crack corrosion are different which warrants a
different treatment. If the number of cracks is small the cracks do not usually become a
determinative factor for the service life of the structure because of self-healing and the
possibility of filling cracks with epoxy. However, if the number of cracks is great corrosion at
cracks can be determinative for service life.
2.2 Process of reinforcement corrosion
General corrosion of reinforcement can only start as a result of physical and/or chemical
changes in concrete around the reinforcement. Physical changes are cracking and spalling of
concrete. Chemical changes in concrete are carbonation and penetration of chlorides. Corrosion
starts when either carbonation or the front of the critical chloride content reaches the
reinforcement.
During the propagation phase corrosion on reinforcement reduces the cross sectional area of
the steel profile thus weakening the structure. The corrosion products also have a tendency to
cause cracking and spalling in the protective concrete cover thus jeopardizing the cooperation
between steel and concrete and the performance of the structure. The rate of corrosion depends
on the moisture content and the temperature of concrete surrounding the steel.

Corrosion
0 1
2 3
Initiation Propagation of corrosion
Depassivation
Cracking
Spalling
Collapse
t
0
At
1
At
2

Figure 1. Illustration of corrosion process as a function of time (SB 2006).
The service life of a structure is determined as the sum of the initiation time and the
propagation time. Depending on the limit state of corrosion the service life is determined as
follows

42
2 1 0
) (
1 0
) (
0
) (
t t t
L
t c
t t
L
t b
t
L
t a
A + A + =
A + =
=
(1)
where t
L
is service life, yrs
t
0
depassivation time of corrosion, and
At
1
additional service life to account for the propagation time until the limit state of
p
1
, yrs.
At
2
additional service life to account for the propagation time until the limit state p
2
,
yrs.
In Case (a) no corrosion is allowed and the service life is determined only based on the
depassivation time. In Case (b) corrosion is allowed only until the first cracks appear as a result
of corrosion and the service life is determined as the sum of depassivation time and additional
service life At
1
. In Case (c) corrosion is allowed until the spalling of concrete cover and the
service life is determined as the sum of the depassivation time and the additional service lives
At
1
and At
2
.
2.3 Carbonation
The protection provided by concrete against steel corrosion is based on the alkalinity of
concrete. In an alkaline environment a tight oxide layer is produced around the steel, preventing
further corrosion. The pH of pore water in concrete containing solid calcium hydroxide is
always at least 12.5, which is adequate to passivate steel. However, normally the pH of pore
water in alkali hydroxides containing cement pastes is even higher, about 13 - 14. When
concrete carbonates its pH is reduced to about 9. Then the passive layer around the steel is
destroyed and corrosion is initiated.
The process of concrete carbonation is complicated. In fact, there are many physicochemical
processes involved in it (chemical reactions from which carbonatable materials are produced,
diffusion of atmospheric CO
2
in the gaseous phase of the concrete pores, dissolution of Ca(OH)
2

in the aqueous phase of the concrete pores and the diffusion of dissolved Ca(OH)
2
in this phase,
dissolution of CO
2
in the aqueous phase of the pores and its reaction with dissolved Ca(OH)
2
,
reaction of CO
2
with other solid constituents of cement paste, etc.) (Papadakis et al.1991). A
mathematical model for concrete carbonation, based on the consideration of all these processes,
was proposed by Papadakis et al. (Papadakis et al. 1991). This model consists of a set of partial
differential equations, each one representing the mass balance of the substances involved in the
carbonation process (CO
2
, Ca(OH)
2
, calcium silicate hydrate and unhydrated silicates).
The mathematical modeling of carbonation has been dealt with, by some researchers, in the
context of the development of models for the analysis of the simultaneous transfer of heat,
moisture and carbon dioxide through and inside concrete components. A well-known model
formulated on the basis of this principle is the one introduced by Saetta et al. (Saetta et al. 1993,
Saetta et al. 1995), consisting of three differential equations, one accounting for the moisture
mass balance, one expressing the energy conservation principle (heat storage and transfer) and
one accounting for CO
2
mass balance. This set of equations is accompanied by a fourth
differential equation, which is formulated to express the rate of carbonation reaction. The
simultaneous transfer of heat, moisture and carbon dioxide, as well as chloride penetration,
inside concrete components is calculated in the model proposed by Steffens for the concrete
structures corrosion risk prediction (Steffens 2000, Steffens et al. 2002).
The solution of such mathematical systems is acquired with the discretisation of the
calculation domains (time and space) and with the use of numerical methods. It is evident that
these calculations cannot be conducted manually. A simpler approach for the prediction of the
carbonation depth is the use of relationships, according to which, the carbonation depth is
approximately proportional to the square root of time. The complex mathematical models of the
form that was previously described can be reduced to such simple relationships with the use of

43
some assumptions and simplifications. Furthermore, this proportionality has been detected early
in the study of the phenomenon, and several empirical relationships of this form have been
proposed. Of course, there is a theoretical background to some equations of such form, as
explained in the following.
The fact that the depth of carbonation is approximately proportional to the square root of time
can be theoretically derived by applying the diffusion theory with a moving boundary. In this
theory the carbon dioxide is diffused into concrete and reacts with the non-carbonated calcium
minerals at the moving boundary, that is at the distance of x
c
(depth of carbonation) from the
surface of a structure. The carbon dioxide content between the surface and the moving boundary
is assumed to be linear. Then the flux of carbon dioxide towards the moving boundary can be
evaluated as (Lay & Schiessl, 2003):
ca
x
c
D J
A
= (2)
where J is flux of carbon dioxide, g/m
2
s,
D the diffusion coefficient of concrete with respect to CO
2
, m
2
/s,
x
ca
distance of the moving carbonation boundary from the surface of the structure,
m,
Ac = c
s
-c
x
, kgCO
2
/m
3
,
c
s
CO
2
content of air at the surface of concrete, kgCO
2
/m
3
, and
c
x
CO
2
content of air at the moving boundary, kgCO
2
/m
3
.
The carbon dioxide flux into concrete must be in balance with the rate of reaction occurring
at the boundary. The rate of reaction depends on the amount of non-carbonated calcium in
concrete:
dt
dx
a J
ca
R
= (3)
where J
R
is reaction rate of carbon dioxide, g/m
2
s,
a CO
2
-binding capacity of concrete, kgCO
2
/m
3
and
t time, s.
By combining the above Equations 1 and 2 (J = J
R
) and integrating over time (x
ca
= 0 when
t = 0), the following solution is obtained:
a
t c D
x
ca
A
=
2
(4)
From Equation 4 it is seen that at constant conditions the depth of carbonation is proportional
to the square root of time. If Ac is constant in Equation 4, the equation can be presented in a
more simplified form as:
t k x
ca ca
= (5)
where k
ca
is coefficient of carbonation, mm/a
0.5
.
The coefficient of carbonation depends on the permeability of concrete, quality of cement,
possible cement replacements (blast furnace slag, silica fume etc.) and the environmental
conditions. A less permeable concrete will yield a slower carbonation rate. In wet concrete,
carbonation is much slower than in only slightly moist concrete.
To provide better coherence with field observations the Equation of carbonation (5) has been
developed further in joined European projects. The following model considers better the time
dependent properties of concrete during the carbonation process (Bamforth 1997):
ca
n
ca
S c e
ca
t
t
R
t c k k
x |
.
|
\
|

=
0
2
(6)
where R
ca
is carbonation resistance of concrete determined by a carbonation test (= a/D),
kg/m
3
/mm,

44
k
e
environmental factor
k
c
curing factor
t
0
the age of concrete during the test, d and
t age of concrete, d.
2.4 Penetration of chlorides
Another chemical change in concrete that can depassivate steel and initiate corrosion is the
result of chloride penetration into concrete. Chlorides in sufficiently rich concentrations can
break the passive film around the reinforcement, even in very alkaline solutions.
The chloride penetration into concrete can be assumed to obey Fick's second law of diffusion.
Accordingly in a semi-infinite wall the following solution can be derived for chloride content in
concrete:
|
|
.
|
\
|
t D
x
erf
c c
c c
a
cl
S
2
1
0
0
(7)
where c
0
is the initial chloride content of concrete pore water, mol/m
3
, and
c
S
chloride content at the surface of concrete, mol/m
3
,
x
cl
distance from surface, m,
D
a
apparent diffusion coefficient determined by formula, m
2
/s, and
t time, s.
In the project DuraCrete the following equation was developed for the effective chloride
diffusion coefficient (Bamforth 1997):
cl
n
c e RCM t e
t
t
k k D k D |
.
|
\
|
=
0
(8)
where D
e
is effective diffusion coefficient of concrete with respect to chloride
penetration, m
2
/s,
D
RCM
chloride diffusion coefficient of concrete determined by a rapid chloride test
(Tang 2005) at the age of t
0
, m
2
/s,
k
e
environmental factor
k
c
curing factor
t
0
age of concrete during the test, d
t age of concrete, d, and
n
cl
parameter.
According to Bazant the error-function solution (Equation 7) can be closely approximated by
the following parabola function (Bazant 1979):
2
3 2
1
|
|
.
|
\
|

=
t D
x
c c
a
Cl
S
(9)
When Equation 9 is solved for x
Cl
and the chloride content c is replaced by the critical
chloride content c
crit
the following function is obtained for the depth of critical chloride content:
t D
c
c
x
a
S
crit
Cl

|
|
.
|
\
|
= 3 2 1 (10)
where x
Cl
is depth of critical chloride content at moment t, m.
In this form it is obvious that the depth of the critical chloride content approximately
complies with the "square-root-of-time"-law in the same way as the depth of carbonation:
t k x
cl Cl
= (11)
where

45
|
|
.
|
\
|
=
s
crit
a cl
c
c
D k 1 3 2
*
(12)
where k
cl
coefficient of chloride penetration, mm/\year and
t time, year.
In principle determination of coefficients of carbonation and chloride penetration call for
testing for carbonation resistance (R
ca
) and the chloride migration coefficient (D
RCM
). That is not
always possible in the design of structures. To avoid testing some data on R
ca
and D
RCM
as a
function of normal concrete design parameters, such as water-cement ratio and the cement type,
are available in literature (fib Model Code 2006, SB 2006, Lifecon D3.2). They contain also
other parameter data that are required in the model equations.
2.5 Corrosion of Reinforcement
The process of corrosion is electrochemical which means that electrons together with chemical
compounds take part in the reactions. Two reactions are necessary to build up an
electrochemical reaction: (1) anodic reaction, which generates electrons, and (2) cathodic
reaction, which consumes electrons. In concrete the anode and cathode areas of the
reinforcement form an electric circuit with concrete as the electrolyte.
In corrosion reactions oxygen together with iron ions are chemically bound to ferrous or
ferric hydroxides. The volume of corrosion products is much greater than the volume of the
original metal. As a result of the expanding layer of corrosion products, pressure is produced in
concrete at the anodic areas of rusting steel. If corrosion is allowed to continue the pressure will
grow so great that it exceeds the tensile strength of concrete. As a result the concrete cover
cracks. With continued corrosion and an increasing layer of corrosion products, concrete will
spall from the surface of the structure. During this phase concrete cover completely loses its
ability to protect the steel. In addition, the capacity of the structure may become questionable as
a result of the reduced bond between steel and concrete. Furthermore the aesthetic appearance
suffers from cracking and corrosion stains. So for many reasons the phases of corrosion
cracking and spalling can be regarded as the limit states for service life of a concrete structure.
Applying Faradays law, the corrosion rate can be used from direct electrical measurement:
corr corr
i V = 6 . 11 (13)
where V
corr
is corrosion velocity, m/yr and
i
corr
mean value of the corrosion current, A/cm.
The corrosion velocity can be determined from the following eguation (Lifecon D3.2 2003).
2
0
O Galv cl corr
F F F
m
V =
(14)
where m
0
is constant (= 882) m.Ohm/yr,
resistivity of the corrosion circuit, Ohm.m.
F
cl
factor accounting for chlorides in concrete,
F
galv
factor accounting for galvanic effects,
F
O2
factor accounting for availability of oxygen.
The resistivity can be determined from the following equation:
res cl res RH res T res c
n
k k k k
t
t
res
, , , ,
0
0

|
|
.
|
\
|
= (15)
where
0
is chloride diffusion coefficient of concrete determined by a test (Lifecon D3.2
2003) at the age of t
0
, m
2
/s,
k
T,res
factor accounting for temperature
k
RH,res
factor accounting for relative humidity
k
c,res
curing factor
k
cl,res
chloride factor

46
t
0
age of concrete during the resistance test, d
t age of concrete, d, and
n
res
parameter.
The active corrosion time which can be added to the depassivation time to provide the
evaluated service life depends on the maximum allowable depth of corrosion on reinforcement.
For the maximum allowable depth of corrosion several limit states may be applied:
p
max
= 0
p
max
= p
1
i.e. limit state of first cracking in concrete cover
p
max
= p
2
i.e. limit state of spalling of concrete cover.
In some cases, as in the case of prestressing steel, no corrosion can be allowed (p
max
= 0). It
means that there is no active corrosion time allowed. Then the service life is determined based
only on the depassivation time (t
0
) which depends on the carbonation and the chloride
penetration processes. In general the service life can be determined as follows:
2 1 0
t t t
L
t A + A + =
(16)
where t
L
is service life, yrs
t
0
depassivation time of corrosion, and
At
1
additional life to account for active corrosion time until the limit state of p
1
, yrs.
At
2
additional service life to account for the active corrosion between the limit state
p
2
and p
1
, yrs.
The extensions of service life At
1
and At
2
are determined from the following formulae:
= A
corr
V
p
t
1
1
(17)
= A
corr
V
p p
t
1 2
2
(18)
Model functions for corrosion depths p
1
and p
2
have been presented in the Swedish instructions
(SB 2006).
2.6 Corrosion at cracks
The following theoretical dependency between the carbonation depth at a crack and the
carbonation depth at uncracked areas has been presented (Schiessl, 1979):
) ( 2 ) ( t x w
D
D
t y
B
R
= (19)
where y is carbonation depth at crack, mm,
x carbonation depth at uncracked surface, mm,
D
R
diffusion coefficient at crack,
D
B
diffusion coefficient at uncracked surface, and
w width of crack, mm.
According to Equation 5 and 11 the carbonation depth at an uncracked surface can be
assumed proportional to the square root of time. So for carbonation depth at a crack we get
t k w
D
D
t y
dep
B
R
= 2 ) ( (20)
where k
dep
is carbonation coefficient, mm/year
0.5
.
By solving Equation for t we get:

47
2
4
2
) (
|
.
|
\
|

=
dep
cr
k w
D
D
y
y t (21)
At the time of depassivation at the crack y is the same as concrete cover, C
c
. We can also note
that the depassivation time at an uncracked surface is:
2
2
0
dep
c
k
C
t = (22)
So for the depassication time at the crack we get:
0
2
2
0
0
t
D
D
w
C
k t
cr
c crack crack
|
|
.
|
\
|
|
.
|
\
|
= (23)
where t
0
crack
is depassivation time at a crack, and, yr,
t
0
depassivation time at uncracked surface, and
k
0
crack
crack coefficient.
A similar equation can be derived also for chloride initiated corrosion at a crack. However,
the value of the crack coefficient is different for carbonation initiated and chloride initiated
corrosion (Equations 24) (Vesikari, 2009):
( )
2
0
0
1 12
7
25 0
|
|
|
.
|
\
|

=
=
s
crit
c
c
crack
cl ;
crack
ca ;
k
. k
(24)
where k
0;ca
crack
is crack coefficient for carbonation initiated corrosion, and,
k
0;cl
crack
crack coefficient for chloride initiated corrosion.
Very little is known about corrosion rate at a crack, However, it is reasonable to assume that
the corrosion rate at a very small crack (< 0.05 mm) is about the same as that in uncracked
concrete. By increasing the crack width the corrosion rate is also increased. So we can write:
V k V
crack crack
=
1
(25)
where V
crack
is corrosion rate at a crack, m/yr,
k
1
crack
crack factor for corrosion rate, and
V corrosion rate in uncracked concrete, m/yr.
The following crack factor for corrosion rate is proposed (Vesikari, 2008):
w
w
N
k
crack
crack
1
1
1
1
+
+
= (26)
where N
crack
is a number (> 1).
According to Alexejev (Alexejev 1993) the corrosion rate is approximately doubled when the
crack width was increased from 0.1 mm to 0.7 mm. So, the value of N
crack
would be around 3. In
Table 1 the values of the k
1
crack
are presented as a function of crack width (N
crack
= 3):

48
Table 1. Crack factor of corrosion as a function of crack width.
Crack width (w), mm k
1
crack

0
0.05
0.1
0.2
0.5
1

1
1.1
1.18
1.33
1.67
2
3

The maximum depth of corrosion is determined based on structural demands. The active
corrosion time is defined as follows:
= A
crack
crack
crack
V
p
t
max
(27)
where p
max
crack
is maximum depth of corrosion. The service life based on crack corrosion is:
crack crack
t t
crack L
t A + =
0
,
(28)
where t
L,crack
is service life with respect to corrosion at cracks, yrs
t
0
crack
depassivation time of corrosion at a crack, and
At
crack
additional life to account for active corrosion time until the limit state of
p
max
, yrs.
3 OTHER DEGRADATION MECHANISMS
3.1 Freeze-thaw deterioration
Freeze-thaw deterioration is due to repeated cycles of freezing and thawing of water contained
within the concrete matrix. Damage is due to expansion that occurs when water is converted
into ice, as this will cause forces exceeding the tensile strength when concrete is critically
saturated. Additionally, smaller pores will be more or less freeze-dried by the ice in the larger
capillary pores, and consequently the ice will grow. If temperatures subsequently rise, the ice
will expand and due to the five times larger temperature expansion coefficient of ice than
concrete, it will exert force (Bijen 2000). Surface deterioration will occur in form of spalling,
scaling and cracking, leading to loss of structural strength and integrity. Freeze-thaw damage is
aggravated by the use of de-icing salts. Salts lower the freezing point and due to hygroscopic
character increase the water content of the pores.
In fib Model Code for Service Life Design 2006 a probabilistic method for service life design
with regard to internal frost damage is presented. The method is based on the requirement that
the probability of the critical degree of saturation is smaller than the actual degree of saturation
within the design service life is smaller than the preset failure probability. Mathematically the
limit state equation is presented as follows:
0 /
} 0 ) ( { {} p t t S S p P p
SL ACT CR damage thaw freeze
< < < = = (29)
where S
CR
is critical degree of saturation,
t
SL
design service life (years),
S
ACT
(t) actual degree of saturation at the time t,
t
SL
design service life (years),
p0 target failure probability, and
p{} probability that freeze/thaw damage occurs.
The variables S
CR
and S
ACT
(t) need to be quantified in a full probabilistic approach. As the
actual degree of saturation depends on the environmental conditions (rain, snow, drying, etc.)

49
the full use of Equation 29 may require a detailed hygrothermal analysis of the structure through
its service life. The service life of a structure may be strongly dependent on small variations in
the environmental moisture conditions.
Measures to prevent freeze-thaw deterioration include air-entrainment, use of frost-resistant
aggregates, low water/cement ratio and surface treatment with hydrophobic agents.
3.2 Deterioration due to aggressive chemical exposure
Chemical attacks may be categorized in four types: (1) acid, (2) sulphate, (3) alkali and (4)
other.
Acid attack is due to reaction of acid with cement paste. The consequence of this reaction is
that the layer by layer conversion of the hardened cement and destruction of microstructure. If
not removed, converted layer is more permeable than sound concrete. Reaction products are
removed by dissolution or abrasion. Once the cement paste is removed, exposed or loose
aggregate remains, and eventually total disintegration will occur. Acid attack weakens the
structure, and can thus contribute to other deterioration mechanisms. For instance, the removal
of concrete cover can lead to steel reinforcement corrosion in concrete. Acid attack may be
caused by atmospheric pollutants (i.e. acid rain), or gaseous acids (i.e. sour gas in a sewer pipe).
Sulphate attack (Skalny J. et al. 2002) is due to sulphates reacting with calcium and aluminate
ions. The reaction causes expansion and formation of cracks, thus speeding up the deterioration
process, as cracks allow further sulphate penetration. As with acid attack, the sulphate attack
can also lead to other deterioration mechanisms, as it damages the concrete cover.
Alkali attack (Swamy 1992) is due to alkalis reacting with silica aggregates. The reaction
causes expansion and leads to surface map cracking.
The permeability of concrete is a critical measure to prevent the deterioration of concrete due
to chemical exposure. Mix design should be appropriate to exposure of the concrete element.
Remediation generally comprises the repair and removal of damaged concrete and application
of a protective surface treatment.
The degradation models for chemical attack of concrete are still rare or missing.
4 DISCUSSION AND CONCLUSION
The models for general corrosion in concrete structures, presented in this paper, are widely
accepted and they can be used for predicting the service life of a concrete structure with
reference to carbonation, chloride penetration and corrosion of reinforcement. The models can
be applied to concrete structures exposed to seawater or deicing chlorides or just to air.
A deficiency in these European models is that they do not include any models for corrosion
at cracks. Tentatively, a method for evaluating the rate of carbonation, chloride penetration and
corrosion cracks of concrete is presented in this report.
Another deficiency is that the frost attack model is given without practical instructions. So,
also the interaction between frost attack and corrosion processes (including carbonation and
chloride penetration) is not considered in these models.
A practical difficulty in the application of these models is that they call for testing of
concrete. The model of carbonation requires testing of the carbonation resistance, the model of
chloride penetration calls for testing of the chloride migration coefficient and the model of frost
attack requires testing of both critical degree of saturation and capillary degree of saturation.
However, there are available test data that can be used for rough evaluation of these material
properties of concrete based on normal concrete design parameters, such as the w/c ratio, the air
content and the cement type. Also other parameter data for the application of the presented
models can be found in the literature (fib Model Code 2006, Lifecon D 3.2, SB 2006).

50
REFERENCES
fib Model Code for Service Life Design. 2006. Federation internationale du beton. Bulletin 34. 110 p.
Lifecon D3.2 2003 Service life models. Instructions on methodology and application of models for
prediction of residual service life for classified environmental loads and types of structures in Europe.
EC, FP5: Growth. RDT Project: Life cycle management of concrete infrastructures for improved
sustainability: LIFECON Deliverable D3.2. 169 p. http://lifecon.vtt.fi/
SB 2006. Vgldning fr livslngdsdimensionering av betongkonstruktioner. (Swedish directions for
demensioning of service life of concrete structures). 2006. Svenska Betongfreningen. Betongrapport
nr 12.
Bamforth, Ph. 1997, Probabilistic Performance Based Durability design of Concrete Structures. Proc.
Int. Sem. Management of Concrete Structures for longterm serviceability. University of Sheffield. pp.
32 - 44.
Bazant, Zd. P. 1979. Physical model for steel corrosion in concrete sea structures - application. Journal
of the Structural Division ASCE 1979: June, pp. 1155 - 1165.
Schiessl, P. 1979. Zur Frage der zulssigen Rissbreite und der erforderlichen Betondeckung in
Stahlbetonbau unter besonderer Bercksichtigung der Karbonatisierung des Betons. Berlin, Deutscher
Ausschuss fr Stahlbeton, Heft 255. 175 p.
Alexejev, S.N. 1993 Corrosion of Steel Reinforcement. In Mallick, S.K. Durability of Reinforced
Concrete in Aggressive Media. Russian Translation Series 96. Rotterdam: A.A. Balkema, Brookfield.
S. 310 - 324.
Vesikari, E. 2008. Degradation and service life of concrete structures in nuclear power plants. Technical
Research Centre of Finland. Research Report VTT-R-02323-08. 40 p.
Vesikari, E. 2009. Carbonation and chloride penetration in concrete with special objective of service life
modeling by the factor approach. VTT Technical research Centre of Finland. Research Report VTT-
R-04771-09. 32 p.
NT Build 492. 1999. Concrete, Mortar and Cement Based Repair Materials. Chloride migradation
coefficient from non-steady-state migration experiments. Nordtest method.
Bijen, J. 2000. Durability of engineering structures, Design, repair and maintenance. Cambridge:
Woodhead Publishing Limited.
Fagerlund, G. 1996. Moisture mechanics as a tool for service life prediction. In: C. Sjstrm (editor),
Durability of building materials and components; Proc. of seventh intern. conf. on Durability of
Building Materials and Components, Stockholm, 19-23 May 1996. London: E & FN Spon.
Skalny, J., Marchand, J. & Odler, I. 2002. Sulphate attack on concrete. London: Spon Press.
Swamy, R.N. (ed.). 1992. The Alkali-silica reaction in concrete. Glasgow&London: Blackie and Son Ltd.
Tang L., Final evaluation of test methods, WP 5 report, EU Project ChlorTest: Resistance of concrete
to chloride ingress - From laboratory tests to in-field performance. G6RD-CT-2002-0085, 2005.
Papadakis, V. G., Vayenas, C. G. and Fardis, M. N. 1991. Fundamental modelling and Experimental
Investigation of Concrete Carbonation. ACI Materials Journal. 88: 363 - 373.
Saetta, A.V., Screfler, B.A. and Vitaliani R.V. 1995. 2-D model for carbonation and moisture/heat flow
in porous materials. Cement and Concrete Research. 25: 1703 - 1712.
Saetta, A.V., Screfler, B.A. and Vitaliani R.V. 1993. The carbonation of concrete and the mechanism of
moisture, heat and carbon dioxide flow through porous materials. Cement and Concrete Research. 23:
761 - 772.
Steffens, A., Dinkler, D. and Ahrens, H. 2002. Modelling carbonation for corrosion risk prediction of
concrete structures. Cement and Concrete Research. 32: 935 - 941.
Steffens, A. 2000. Modellierung von Karbonatisierung und Chloridbindung zur numerischen Analyse
der Korrosionsgefhrdung der Betonbewehrung, PhD thesis, Braunschweig.

51
R. Landolfo, L. Cascini, F. Portioli
Modelling of Corrosion Damage for Metal Structures
1 INTRODUCTION
The growing attention that the scientific community has paid in the last decades to the corrosion
phenomena is related to the huge economic, social and environmental losses consequent upon
spread damage on several metal constructions.
During the 1970s and 80s, several studies on the cost of corrosion were carried out. It was
showed that the corrosion impact on the domestic economy of various countries accounted for
3% of the GDP (C.C. Technologies Laboratories & NACE International 2001). When the
damage has greatly enlarged, expensive maintenance and replacement are required for the safe
and reliable use of constructions and, even if the serviceability of the structure is not
compromised, the deterioration of appearance produces reduced value of constructions.
From the structural point of view, the thickness loss of the cross section due to corrosion
attack leads to a smaller resistant area, producing a decrease of the structural performances in
terms of strength, stiffness and ductility. In some cases, the local failure of a member or joint
could affect the stability of the whole structure. In addition, in case of cyclic loads, the corrosion
phenomenon can produce a significant reduction in the fatigue strength, mainly in zones with
high stress concentrations (Landolfo et al. 2005).
In order to prevent failures due to corrosion, a durability design of metal structures is
necessary, both for ultimate and serviceability limit states.
With this regard, the main European structural design codes provide only general protective
measures, such as the use of weathering or stainless steel, of surface protection systems and of
structural redundancy. It should be also noted that no predictive models, which are able to
estimate the corrosion depth of structural material, are provided but only general
recommendations concerning the minimum thickness to be used for structural members are
reported.
Nevertheless, several models concerning the evaluation of the damage produced by corrosion
are available in the literature and standards, even if they refer to different parameters and
conditions.
The object of this study is to present a report on the predictive models relevant to atmospheric
corrosion which are available in the literature in order to evaluate their suitability with respect to
use in a durability design procedure based on lifetime safety factor method (Landolfo et al.
2007). The procedure is composed of different phases which are the analysis of environmental
loads, selection of influencing parameters and the evaluation of thickness loss by means of
corrosion models and related safety factors.
With this aim, corrosion rates and classification methods of environmental corrosivity
considered in standards are reported in the following. Moreover, several models are shown
which take into account the different factors affecting the evolution of damage, such as the
presence of chlorides, sulphates and pollution in the atmosphere.

52
Finally, a comparison among considered dose response curves corresponding to selected
corrosion classes and metals is presented and main differences are described with respect to
degradation rates reported by ISO standards.
2 CORROSION OF METALS AND ALLOYS
The corrosion phenomena of metals and alloys involve mainly two elements: the material and
its environment. In particular corrosion is defined as the deterioration of a material, usually a
metal, that results from a reaction with its environment (NACE 2002), causing the degradation
of both.
Corrosion attacks could manifest in different forms and several classification systems are
provided by the international scientific literature.
A first classification is made on the basis of the reaction that takes place on the metal surface.
In this case, it is possible to distinguish between chemical and electrochemical corrosion, the
latter involving at least one cathodic-anodic reaction.
Another traditional classification system (NACE 2002, EN ISO 8044, NPL et al.)
characterizes corrosion phenomena according to their appearance and the basic forms of
corrosion, which could be uniform or localised (Figure 1), are defined as follows:
- general corrosion: corrosion that is diffused more or less homogeneously over the surface
of a material;
- uniform corrosion: it is a kind of general corrosion that proceeds at the same rate onto the
metal surface;
- pitting corrosion: localized corrosion that is restricted to a small area, taking the form of
pits;
- crevice corrosion: localized corrosion of a metal surface at or close to an area that is
protected by another material;
- erosion corrosion: a conjoint action involving a corrosive flowing which leads to
accelerated loss of material;
- cavitation corrosion: development and rapid collapse of cavities or bubbles;
- galvanic corrosion: corrosion of a metal due to a contact with a more noble metal in a
corrosive electrolyte;
- fatigue corrosion: cracking of metal caused by repeated stresses in a corrosive
environment;
- stress corrosion: brittle cracking caused by the parallel occurrence of an applied stress
and a specific environment.
Finally, on the basis of the corrosive environment, degradation mechanisms are categorized
in microbial and bacterial corrosion, gaseous corrosion, marine corrosion, underground
corrosion and atmospheric corrosion. The firs one is usually associated with the presence of
micro-organism and/or bacteria on the metal surface that occurs in natural and sea water as well
as in soils. Gaseous corrosion occurs when a dry gas is the main corrosive agent and there is not
a liquid layer on the surface. Marine corrosion refers to sea water acting in an immersion and/or
a splash zone, while underground corrosion is related to corrosion in soils.
This paper focuses on the atmospheric corrosion of metal structures in outdoor atmospheres,
that represents one of the most severe form of corrosion, resulting in huge economic and
structural performance losses.
The atmospheric corrosion phenomena comprise three different forms of corrosion,
particularly dry, damp and wet corrosion. Damp and wet corrosion usually occur in indoor and
outdoor atmosphere, as a consequence of wet and dry cycles induced by rainfall and
condensation, which produce thin water films at the metal surface. In damp corrosion, the thin
film of electrolyte is formed on metal surface by adsorption of water molecules when a critical
humidity level is reached, while wet films are associated with water flowing (Roberdge 2000).
Atmospheric corrosion is mainly an electrochemical process that occurs in the presence of
thin film electrolytes formed on the metal surface. The attack proceeds by balancing anodic
oxidation reaction, which involves the dissolution of the metal in the electrolytic film, and

53
cathodic reactions, involving the oxygen reduction reaction. The anodic process controls the
overall rate of atmospheric corrosion and it is represented by equation (1):
M M
n+
+ ne
-
(1)

Figure 1. Main forms of corrosion (Roberdge 2000).

The electrolyte films tend to form only when a certain critical humidity level is reached. For
example, in the case of iron a threshold value for humidity of about 60% in a pure atmosphere
can be defined. Corrosion rate can be neglected for lower values of RH while for higher ones it
rapidly increases.
The overall rate of the metal dissolution process is strongly influenced by the formation of
corrosion products, their solubility in the water film and the formation of passive coatings as
well as by the corrosiveness of the environment. These phenomena processes are influenced by
several factors, that could be divided into two different classes: the endogenous factors, which
are related to the metal itself (the effective electrode potential of a metal in a solution, the
composition of the metal, the chemical and physical homogeneity of the surface, etc.), and the
exogenous factors that are connected to the atmospheric composition (Landolfo, Di Lorenzo,
Guerrieri 2005).
Among these factors, important practical environmental variables are the time of wetness
(TOW) and the concentration of contaminants, such as sulphur dioxide and chlorides.
The time of wetness represents the average period of time during which the electrolyte is on
the corroding surface (Roberdge 2000), for practical purpose it could also be defined as the
number of hours during which the relative humidity is greater than 80% and the average
temperature T > 0 (ISO 9223, 1992). The negative effect of the humidity is exacerbated by the
concentration of the atmospheric pollutants, mainly sulphurs and chlorides.
Chlorides in the atmosphere can definitely increase the corrosion rate, such as in the case of
marine environment, where the critical humidity level is about 30 - 40%. The main effect of
chlorides is to prevent formation of protective oxide films on the metal surface, thus increasing
the overall corrosion rate.
As far as sulphurs are concerned, small concentrations could dissolve in the thin electrolyte,
the relative humidity being over 60 - 70%, and the acidic electrolyte that is formed seems to
stimulate both the anodic and the cathodic reactions (Roberdge 2000). It has been shown that
there is an increase in rate at about 60% relative humidity even in pure air, and a further sharp
increase at 80% in the presence of 0.01% of sulphur dioxide (NPL).
3 STANDARDS AND CODES
With respect to corrosion phenomena, the European structural design codes provide only
general recommendations and basic principles that are mainly concerned with the use of coating
protective systems, the choice of corrosion resistant materials and structural redundancy. Such

54
specifications aim to prevent the initiation and propagation of the corrosion attacks by means of
design detailing, surface protective systems and proper inspection and maintenance
programmes, but no references are reported to models able to predict the damage effects during
service life.
The EN 1993-1-1 states few common principles, such as the opportunity of providing
corrosion protection measures by means of surface protection systems, improving the use of
weathering and stainless steel and by structural redundancy. It should be noted that in this case
no references are made to models able to estimate the corrosion depth.
Nevertheless, in the new Italian structural code (DM 2005) an important innovation has been
introduced concerning corrosion, which is expressly included among the different loads acting
on constructions. Corrosion is classified as a type of entropic load, which comprises
deteriorating actions, caused by natural degradation mechanism of materials and environmental
loads, which affects the structural integrity. With respect to durability, it is recommended to
adopt a 2mm extra thickness in aggressive environment if maintenance and inspection can not
be performed for design life up to 100 years.
The European Standard EN 12500-2000, edited by the European Committee for
Standardization (CEN), defines the procedure to be adopted for the classification, determination
and estimation of corrosiveness of atmospheric environments by assessing the mass loss of
standard samples, after 1 year exposure. The procedure is described in detail the next section.
Important references to the corrosion of metals and alloys are reported in the almost 40
standards compiled by the ISO Technical Committee TC 156. The TC working group 4 worked
on categorization of the corrosivity of environment in terms of practical environmental
variables, atmospheric corrosion testing and classification of corrosivity of atmosphere.
In particular, EN ISO 9223 provides the classification of the corrosivity of atmospheres on
the basis of three key factors: the TOW (time of wetness), the deposition rate of chlorides and
sulphur dioxide concentration. It defines five corrosivity classes, ranging from C1 to C5.
EN ISO 9224 (1992) provides guiding values for the corrosivity categories of atmospheres.
EN ISO 9225 (1992) is concerned with the methods of measurement of pollution data, while EN
ISO 9226 (1992) gives the corrosivity rates of standard specimens for the evaluation of
corrosivity of atmospheres. In EN ISO 14713 (1999), specific recommendations are provided
for each corrosivity class with respect to different coating typologies. In particular, the life
duration of zinc and aluminium coatings is related both to thickness loss and corrosiveness of
environment.
Other references can be found in international standards, but a design procedure has still to be
codified for predicting and preventing the potential damage that a specific environment could
lead to both coatings and structural materials, during the entire service life.
4 CORROSIVITY OF ATMOSPHERIC ENVIRONMENTS
As far as the corrosivity of atmospheres is concerned, EN 12500 (2000) defines five outdoor
environments on the basis of the presence of corrosive agents in the air, namely:
- Rural atmosphere: countryside and small towns, minor corrosive agents contamination
(carbon dioxide, chlorides, artificial fertilizers).
- Urban atmosphere: densely populated areas, few industrial activities, medium corrosive
agents contamination (sulphur dioxides);
- Industrial atmosphere: intensive industrial activities, high corrosive agents contamination
(sulphur dioxides);
- Marine atmosphere: areas close to the sea, or internal zones strongly affected by airborne
salinity. Corrosion effect are influenced by topographic conditions, prevailing wind
direction.
- Marine Industrial atmosphere: complex environment, areas close to both the sea and
industrial districts, or internal zones located in the prevalent wind direction. Medium and
/or high corrosive agents contamination (sulphur dioxides, chlorides).
The methodology for the quantitative evaluation of the corrosivity of a specific environment
standard is provided in EN 12500 on the basis of guiding values of the mass loss of standard flat

55
specimens (rectangular shape, 50x100 mm) of four materials (carbon steel, copper, zinc,
aluminium), after one year of exposure. The corrosivity classification is defined according to
Table 1.
Although it is strongly recommended to follow the previous procedure, a qualitative
classification of the corrosiveness of atmosphere has been established whether field test data are
not available (Table 2).
Table 1. Mass loss (g/m
2
) for one year field test exposure in the five corrosivity classes C1-C5, the order
being from the least to the most corrosive. (EN 12500/2000).
MASS LOSS g/m
2

Carbon Steel Zinc Copper Aluminium
Corrosiveness category
10 0.7 0.9 Negligible Very low C1
10 - 200 0.7 - 5 0.9 - 5 0.6 Low C2
200 - 400 5 - 15 5 - 12 0.6 - 2 Medium C3
400 - 650 15 - 30 12 - 25 2 - 5 High C4
650 - 1500 30 - 60 25 - 50 5 - 10 Very high C5

Table 2. Corrosivity qualitative classification. An extract of EN 12500.
Corrosiveness category Typical outdoor atmospheric environments
C1 Very low
Dry or cold zones; very low pollutants contamination; time of
wetness very low. E.g.: Desert, Antarctic zone.
C2 Low
Temperate zone; low pollution (SO
2
[g/m
3
] < 12). E.g.: rural
areas and small towns.
Dry or cold zones; short damp periods. E.g.: desert, sub-artic
zones.
C3 Medium
Temperate zones; medium pollutant contamination (12 < SO
2
[g/m
3
] < 40); low chloride influences. E.g.: urban areas, coastal
area characterized by low chlorides deposition rate.
Tropical zones with low pollution.
C4 High
Temperate zones; high pollution levels (40 < SO
2
[g/m
3
] < 80);
important chloride influences. E.g.: polluted urban areas,
industrial areas, costal areas (no splashing zones), deicing salt
influence.
Tropical zones with medium pollution level.
C5 Very high
Temperate zones; very high pollution levels (80 <SO
2
[g/m
3
] <
250); strong chlorides deposition rates. E.g.: industrial zones,
coastal and sea areas (no splashing zones).
Tropical zones with high pollution levels and/or strong chlorides
influences.

As an alternative to this qualitative classification, it is possible to refer to ISO 9223
corrosivity chart where each class is defined by the guiding values of TOW, sulphur dioxide
concentration and the chloride deposition rate. The time of wetness (hours/year) is defined as
the number of hours per year during which RH is greater than 80% and the average temperature
exceed 0 C. Five TOW classes are identified (T1T5), ranging from the indoor environment
with climate control (TOW 10 hours/year) to the most corrosive damp climate (TOW 5500
hours/year). Finally, four sulphur dioxide (P0P3) and chloride (S0S1) deposition rates classes
have been stated for the classification of environmental corrosivity. An application of ISO 9223
is presented in Table 3.

56
Table 3. An application of the ISO 9223 chart to carbon steel for TOW = T3.
TOW = T
3

P3 C2 C2 C3 C4
P2 C1-2 C1-2 C2-3 C3-4
P1 C1 C1 C2 C3-4
Sulphur dioxide
concentration
[g/m
3
]
or deposition rate
[mg/m
2
day]
P0 C1 C1 C2 C3-4
S0 S1 S2 S3

Chloride deposition rate [mg/m
2
day]
5 CORROSION MODELLING
Several models concerning the evaluation of the damage produced by atmospheric corrosion are
available in the literature. They are formulated according to different approaches which depend
on the objectives of the model itself. Such models can be classified as 1
st
level and 2
nd
level
models. The first ones are based on laws of physics and chemistry (Landolfo, Di Lorenzo,
Guerrieri 2005). In this case the dissolution of metal and the formation of corrosion products is
evaluated at microscope level in the sense of current (Roberdge 2000). The 2
nd
level models are
useful for engineering application and allow to evaluate the corrosion rate as a function of mass
and/or thickness loss with time, being obtained from the observation and the interpolation of
experimental data.
In the following a review of 2
nd
level corrosion models for structural applications is
presented. In this case the relationship between the corrosion rate and the levels of pollutants is
expressed in combination with different climatic parameters (Kucera 2004). Considered
variables affecting the corrosion rate over time are the time of wetting, the frequency and
duration of drying out periods, relative humidity, temperature and temperature variation, and the
composition of the atmosphere. The corrosion rate is usually expressed as the mass loss per unit
area per unit time, or as the rate of penetration, by means of the thickness loss. It is important to
note that the thickness loss reported in corrosion studies is usually the average of the thickness
losses of the exposed and ground ward surfaces of a specimen (Albrecht & Hall 2003).
Corrosion models usually describe the corrosion depth as a function of time in the form of a
power model:
d(t)= At
B
(2)
where d(t) = corrosion depth [m, g/m
2
], t = exposure time [years], A = corrosion rate in the
first year of exposure, B= corrosion rate long-term decrease.
Because of the formation of corrosion products on the metal surface, the initial corrosion rate
usually decreases on long term period. If B is smaller than 0.5, the corrosion products show
protective, passivating characteristics, otherwise B is greater than 0.5.
Some simplified models have been proposed which take into account the environmental
effects influencing the corrosion rate by means of constant values of coefficients A and B. Such
models express the thickness loss only as a function of time and are usually calibrated on data
obtained from short and long term field test exposure. As a consequence, if they are used for
environments which differ from the one where the model has been calibrated, the predicted
value of the corrosion rate is often inaccurate.
A first attempt to develop general models has been provided by International Standard ISO
9224 (1992), which specifies the long term corrosion rates for standard structural materials in
the five corrosivity classes C1C5. According to the Standard the average corrosion rate of each
material follows a bi-linear low. During the first 10 years the corrosion depth is given by
formula:
d
1
(t) =r
av
t t <10 years (3)
where d
1
(t) = corrosion depth after the first 10 years of exposure (micrometers); r
av
= average
corrosion rate (micrometers per year); t = time at which the exposure ends.

57
After 10 years of exposure, the corrosion rate is assumed to be constant with time and the
thickness loss is given by formula:
d(t) =r
av
10 + r
lin
(t-10) t >= 10 years (4)
where d(t) = corrosion depth for the considered time interval (micrometers); r
lin
= steady state
corrosion rate (micrometers per year); t = time in the linear region of the curve of uniform
corrosion as function of time.
The standard provides the guiding values of both r
av
and r
lin
for carbon steel, weathering steel,
zinc, copper and aluminium. In Figure 2 a representation of corrosivity band for carbon and
weathering steel is shown.

a) b)
Figure 2. An application of ISO 9224. Thickness loss as a function of time for carbon steel (a) and
weathering steel (b) for different corrosiveness classes.
Some adjustment to these corrosion laws have been reported in Albrecht et Hall (2003),
where the authors have proposed a new bi-linear model which accounts for a modified corrosion
rate during the first year of exposure and a steady state during the subsequent years. An
application of the modified bi-linear law for carbon and weathering steel for the upper and
lower bound in the medium corrosivity category C3 is reported in Figure 3.

a) b)
Figure 3. Thickness loss as a function of time for carbon steel (a) and weathering steel (b) for corrosivity
class C3.
Recently, different models have been developed with the aim to generalize the corrosion loss
over time for different environments, reporting the climate and pollutants variables as
independent factors. In the following two of these models are presented.
Several dose-response functions have been developed within the International Cooperative
Programme on Effects on Materials, including Historic and Cultural Monuments in the
framework of the UN ECE convention on long range transboundary air pollution (Kucera

58
2004). These functions have been formulated for different metal materials (Table 4) and are
based on both long term exposures and trend analysis based on repeated 1-year measurements,
taking also into account the unsheltered (eq. 5.1 - 5.4) or sheltered (eq 5.5) exposure.
Table 4. Dose response function, ICP Materials.
Material Mass loss [g/m
2
] Equation n
Weathering
steel (unsheltered)
ML = 34[SO
2
]
0.33
exp{0.020Rh + f(T)}t
0.33
(5.1)
Zinc ML=1.4[SO
2
]
0.22
exp{0.018Rh+f(T)}t
0.85
+0.029R
ain
[H
+
]t (5.2)
Aluminium ML=0.0021[SO
2
]
0.23
Rh exp{f(T)}t
1.2
+0.000023R
ain
[Cl
-
]t (5.3)
Copper ML=0.0027[SO
2
]
0.32
[O
3
]
0.79
Rhexp{f(T)}t
0.78
+0.050R
ain
[H
+
]t
0.89
(5.4)
Weathering
Steel (sheltered)
ML = 8.2[SO
2
]
0.24
exp{0.025Rh + f(T)}t
0.66
(5.5)

Klinesmith et al. (2007) developed a model for the atmospheric corrosion of carbon steel,
zinc, copper and aluminium, taking into account the effects of four environmental variables
(TOW; sulphur dioxide, salinity and temperature). The data of the ISOCORRAG program have
been used to calibrate the coefficients of the proposed models.
The general form of the degradation model is the following:
( )
0
1 1
2
T T J
H F D
B
e
G
Cl
E
SO
C
TOW
t A y
+
|
.
|
\
|
+
|
.
|
\
|
+
|
.
|
\
|
= (6)
where y = corrosion loss (m); t = exposure time (years); TOW = time-of-wetness (h/year); SO
2

= sulfur dioxide concentration (g/m
3
); Cl is chloride deposition rate (mg/m
2
/day); T = air
temperature (C); and A, B, C, D, E, F, G, H, J, T
0
= empirical coefficients (Table 5).
Table 5. Coefficients of Eq. (6) calibrated with ISO CORRAG data (Dean & Reiser 2002) for carbon
steel and zinc materials.
Equation coefficients
Material
A B C D E F G H J T
0

Carbon
steel
13.4 0.98 3800 0.46 25 0.62 50 0.34 0.016 20
Zinc 0.16 0.36 3800 0.24 25 0.82 50 0.44 0.05 20

In order to evaluate the differences among the selected corrosion models, a comparison has
been carried out for carbon and weathering steel. The considered corrosivity class is C3 which is
characterized, according to EN12500 (Table 2), by a temperate climate, medium pollutant
contamination and low chloride influences. The TOW selected levels range from 2500 to 4200,
while sulphur dioxides and chlorides have been chosen within class P1 and S1 respectively, as
defined in ISO 9223. In Figures 4 and 5 the comparison among selected models and corrosion
rates is shown.
A good agreement among standards and selected corrosion models can be observed in the
case of weathering steel for both short and long term exposures.
On the contrary, as far as carbon steel is concerned, it can be noted that Standard ISO 9224
underestimates the corrosion loss for both short and long term exposures, as also observed by
Albrecht & Hall. With this regard, it should be observed also that the corrosion rates for carbon
steel provided by ISO 9224 are similar to the ones given for weathering steel, which instead is
characterized by higher resistance to atmospheric corrosion.

59

a) b)
Figure 4. An application of the selected models to carbon steel for 1 year (a) and 20 years (b), corrosivity
class C3.
As far as Klinesmiths corrosion model, the values predicted for one year exposure are very
close to EN 12500, but on long term it gives higher corrosion rates than the other models,
providing an average thickness loss for a design life of 50 years of about 1.7 mm.

a) b)
Figure 5. An application of the selected models to weathering for 5 years (a) and 20 years (b), corrosivity
class C3.
6 CONCLUDING REMARKS
In this paper a report on corrosion models which are available in both the literature and
international standards has been presented.
The study showed that the largest part of corrosion models are calibrated on the basis of local
field test exposure data. Since the empirical coefficients that characterize each model are
strongly influenced by the exposure location, condition and duration, such models can be used
just for similar environmental conditions.
Only recently general formulations have been proposed with the aim of providing the
thickness loss for different environmental parameters.
The damage models available in literature have been compared with corrosion depth
considered by standards, selecting the different influencing parameters in order to simulate
similar classes of environmental corrosivity.
Comparison performed with standard provisions showed a large scattering of predicted
corrosion depth. In particular, with regard to carbon steel, it has been shown that standards
underestimate the value of thickness loss.
Of course, the scattering of results influences the reliability of models and the value of related
safety factors in a durability design procedure.

60
The probabilistic characterization of the corrosion development over time, relating the
average value of the corrosion depth with the standard deviation, represents the further step of
this study. Once the degradation models have been determined, it could be possible to upgrade
the ordinary mechanic design with durability based performance design by defining the lifetime
safety factors relevant to characteristic service life both for serviceability and ultimate limit
states.
REFERENCES
Albrecht P.& Hall T.T. 2003. Atmospheric corrosion resistance of structural steels. Journal of Materials
in Civil Engineering.
C.C. Technologies Laboratories & NACE International 2001. Corrosion costs and preventive strategies
in the United States. Office of Infrastructure Research and Development Federal Highway
Administration. USA.
CEN 1992. EN ISO 9223. Corrosion of Metals and Alloys. Corrosivity of Atmospheres. Classification.
CEN 1992. EN ISO 9224. Corrosion of metals and alloys. Corrosivity of atmospheres. Guiding values
for the corrosivity categories.
CEN 1992. EN ISO 9225. Corrosion of metals and alloys. Corrosivity of atmospheres. Measurement of
pollution.
CEN 1992. EN ISO 9226. Corrosion of metals and alloys. Corrosivity of Atmospheres. Determination of
corrosion rate of standard specimens for the evaluation of corrosivity.
CEN 1999. EN ISO 8044. Corrosion of metals and alloys. Basic terms and definitions.
CEN 1999. EN ISO 14713. Zinc and aluminum coatings.
CEN 2000. EN 12500. Corrosion likelihood in atmospheric environment.
Dean, S. & Reiser, D.B. 2002. Analysis of long term atmospheric corrosion results from ISO CORRAG
Program. West Conshohocken: ASTM.
Italian Government 2005 Ministerial Decree 14 September 2005.(D.M 2005). Design Rules for buildings.
Klinesmith, D.E., McCuen R. & Albrecht P. 2007. Effect of environmental condition on corrosion rate.
Kucera V. 2004. Mapping effects on materials in Manual mapping critical load. ICP Materials
Coordination Centre, Sweden.
Landolfo R., Cascini L. & Portioli F. 2007. Durability design of metal structures based on lifetime safety
factor method. Proceeding of the International Congress Sustainable Construction, Materials and
Practices - Challenges of the industry for the new Millennium: Portugal SB07. Lisbon, Portugal.
Landolfo, R., Di Lorenzo, G. & Guerrieri M.R. 2005. Modelling of the damage induced by atmospheric
corrosion on 19
th
century iron structures. Proc. of Italian National Conference on Corrosion and
Protection, Senigallia. Italy.
NACE International. 2002. Glossary of Corrosion Related Terms. Houston: NACE International.
National Physical Laboratory (NPL) UK. Guide to the conservation of metals. http://www.npl.co.uk
Roberdge P.R. 2000, Handbook of Corrosion Engineering, McGraw-Hill, New York.

61
U. Kuhlmann, H.-P. Gnther, J. Raichle, & M. Euler
Institute of Structural Design, University of Stuttgart, Stuttgart, Germany
Fatigue of Steel and Composite Bridges
1 INTRODUCTION
The service life of new and existing bridges is invariably governed by their fatigue performance.
Fatigue, as one major degradation process of steel, is therefore a key issue within the life cycle
assessment of steel bridges. Appropriate and efficient fatigue design of various elements and
details in bridge structures is thus important but however often a complex procedure. Current
fatigue design rules e.g. provide a classification system that allocates structural details to
particular S-N curves representing the fatigue resistance of the corresponding detail based on
nominal stresses. These design rules combined with present construction guidelines are in many
cases inadequate for modern types of bridge structures as they do not offer enough information
and possibilities in order to take into account e.g. specific detail arrangements or the use of high
strength steels. This paper thus aims to give a short overview of the fatigue design concept for
steel and composite bridges according to Eurocode and to summarise some recent research
results focusing on several fatigue aspects such as fatigue of high strength steels and post-weld
improvement methods, application of advanced fatigue assessment methods e.g. for hollow
section joints and fatigue of composite bridge girders having stud connections.
2 FATIGUE DESIGN CONCEPT FOR STEEL AND COMPOSITE BRIDGES
The fatigue design of steel and steel-concrete composite bridges is covered in the relevant
Eurocode application parts: for steel bridges in EN 1993-2 (2005) and for composite bridges in
EN 1994-2 (2005). However, these application parts themselves cover, related to their particular
fields, only specific aspects concerning e.g. fatigue loading, whereas for the general fatigue
design concept or the fatigue resistance both parts refer to EN 1993-1-9 (2005) as part of the
general rules for steel structures.
The fatigue design procedure according to Eurocode is based on the rule of linear damage
accumulation in line with Miners law. The verification itself can be done using either a general
procedure based on a damage verification or a simplified procedure based on a stress
verification, see Figure 1. Within the general procedure the fatigue verification should be done
using a representative fatigue load model that results in a certain stress history at the specific
detail, where then - by using a cycle counting method (e.g. reservoir method) - the design stress
range spectrum is determined for the final damage verification, based on a S-N curve for the
relevant constructional detail, see Figure 1 left hand side. Since this method is very time
consuming for a standard fatigue design process, a simplified fatigue verification procedure, see
Figure 1, right hand side, has been set up and included as standard procedure within the code.
For road bridges this procedure uses e.g. a simple single vehicle model consisting of 4 axes
120 kN for determining the relevant fatigue stress range Ao
p
. This stress range is subsequently
transformed into the so called equivalent constant-amplitude stress range Ao
E,2
that would result

62
in the same fatigue damage for 210
6
cycles as the real design spectrum, provided that the
comparison is based on a Miner's summation. Within the design process Ao
E,2
can easily be
calculated using the equivalent factor that has been derived based on the General Procedure.
Knowing the equivalent constant-amplitude stress range Ao
E,2
the fatigue verification can
consequently be performed based on a stress level, showing that Ao
E,2
is smaller than the
corresponding value of Ao
C
of the relevant S-N curve.

Figure 1. Fatigue verification procedures according to Eurocode (characteristic level).
According to EN 1993-1-9 (2005) the fatigue resistance is given by a series of 14 standardized
fatigue strength curves (S-N curves) for direct stresses and two for shear stresses, each applying
to a typical detail category. For direct stresses the S-N curves are defined for detail categories
ranging form 36 to 160, where the number indicates the characteristic fatigue strength Ao
C
at
N = 210
6
cycles. For direct stresses a two-slope curve with m = 3 and 5 is used in combination
with a cut-off-limit at 110
8
cycles, see Figure 2.
The detail catalogue of EN 1993-1-9 (2005) currently comprises about more than 70 different
constructional details.
For composite bridges the fatigue resistance of shear connectors between the reinforced
concrete slab and the steel girder is dealt in EN 1994-2 (2005). However, currently the fatigue
resistance is only given for headed studs with a detail category of At
C
= 90 N/mm
2
in
combination with a rather flat slope of m = 8, see Figure 2.

63

Figure 2. S-N curves according to Eurocode for the detail of a transverse stiffener and headed studs.
The fatigue verification uses the so called nominal stress which is defined as a stress adjacent to
the potential crack location calculated in accordance with elastic theory excluding all stress
concentrations. For constructional details that are not covered within the detail catalogue or for
which it is difficult to define a nominal stress the fatigue verification can also be done using the
hot-spot-stress method, see also section 4.3.
Regarding safety and reliability EN 1993-1-9 (2005) offers the possibility to choose the
partial safety factor for fatigue strength
Mf
between 1.0 to 1.35 depending on the safety concept
and the consequence of failure. Thereby it is foreseen to give the designers and the owners the
chance to decide by there own where to put the effort in: either by using a low safety factor in
combination with a mandatory inspection program or to use a high safety factor without
maintenance and inspection.
3 FATIGUE OF HIGH STRENGTH STEELS AND POST WELD TREATMENT
METHODS
3.1 High strength steels in bridges
Over the past decades high strength steels (HSS) have gained ground in the market of steel
structures. This is because they combine the beneficial properties of a high static yield strength,
a low self-weight, a good weldability and a high ductility. However, in welded structures
subjected to fatigue, as e.g. bridges, the advantage of the higher static strength can only be
exploited in a limited way because slender structures lead to comparatively higher fatigue
stresses and the fatigue strength of welded structures does not increase in the same manner as
the steel tensile strength. Consequently, the possibility that the fatigue criteria become decisive
for the dimensioning of a structure increases with the application of high strength steels.
Within the field of bridge design it has been shown that the application of HSS is likely to be
economically interesting especially for the case of composite road bridges Kuhlmann et al.
(2006), because of the lower proportion of live load to dead load. These investigations however
also showed that in order to achieve a maximum utilization of the higher yield strength of HSS
an improvement of fatigue strength is necessary. In general, it is sufficient to improve the
fatigue strength at critical construction details locally limited in regions of high stress
ranges Ao.
Very effective methods to enhance the fatigue resistance of welded structures are post-weld
treatment methods. However, up to now it has not been possible to apply their positive effects

64
on the fatigue resistance according to missing rules in the present fatigue design standard EN
1993-1-9 (2005).
3.2 Investigations of fatigue resistance of HSS in the as-welded and post-weld treated
condition
In order to get a deeper inside into the beneficial effects of post-weld treatment methods on the
fatigue resistance, a test program was set up by Kuhlmann et al. (2006) especially focusing on
high strength steels. This test program comprised two post-weld treatment methods: first the
well-known TIG-dressing method and secondly the relatively new post-weld treatment method
called Ultrasonic Impact Treatment (UIT), see Figure 3. UIT is a mechanical hammering
technique at a frequency of around 200 Hz superposed by a ultrasonic treatment at a frequency
of around 27 kHz. These combined impacts result in plastic deformations at the treated surface
leading to beneficial compressive stresses in the surface layer and a smoothening of the weld toe
transition.

weld toe before and after TIG-dressing UIT tool and application
Figure 3. Post-weld treatment methods.
The overall test program comprised about 250 small scale specimens and 12 large scale girder
tests. Within the experimental test program various aspects, e.g. the steel grade S355, S460 and
S690, the post-weld treatment condition, the stress ratio as well as the plate thickness were
varied. All investigations took place on the constructional detail of a transverse stiffener, as it
was found out in the aforementioned investigations, that this detail is typically located in critical
sections of steel and composite bridges.
The results of the tests showed that for all steel grades the application of post-weld treatment
methods leads to a considerable improvement of the fatigue resistance, especially in the region
of high-cycle fatigue. For steel grade S460 some test results are shown in Figure 4 in form of S-
N curves as mean life lines obtained by a linear regression analysis. The tests were conducted
with a constant amplitude fatigue loading on a stress ratio of R = 0.1. One can see that in
comparison with as-welded state the beneficial influence of a post-weld treatment method on the
fatigue strength can be represented by a transition and/or a rotation of the fatigue resistance
curve. Regarding the fatigue strength at 210
6
cycles and compared to the as-welded state an
increase of about 50% could be achieved by applying TIG-dressing and increase of about 80%
when applying UIT.
Concerning the UIT method it could further be observed that the fatigue resistance increased
with increasing yield strength. The improvement of the fatigue strength for the steel grade S690
was even so high, that the failure of the specimens did not occur anymore in the region of the
weld but moved in the base material, see Figure 4.
Based on the experimental test program and additional numerical simulations within the
thesis of Drr (2006) it was finally possible to derive simple design recommendations for the
application of the UIT method on the construction detail of the transverse stiffener. According
to these recommendations, the fatigue strength for the as-welded state can be improved by one
detail category for the steel grade S355, by 2 detail categories for the steel grade S460 and by
3 detail categories for the steel grade S690, in combination with a change of the slope of the
fatigue strength curve from m = 3 to m = 5.

65

Figure 4. S-N curves and test specimen with base material failure.
4 FATIGUE OF WELDED CIRCULAR HOLLOW SECTION JOINTS
4.1 General
The application of circular hollow sections (CHS) in highway-bridge design is relatively new.
For about 15 years steel-concrete composite bridges including CHS trusses have developed as
an innovative and aesthetic alternative to conventional all-steel bridges. Outstanding examples
are the bridges of Lully (1997) and Dttwil (2001), both Switzerland.
CHS trusses with diagonals and without verticals have turned out to be most efficient. In this
case the braces run at an angle of 45 to 60 toward the continuous bottom chord forming K-
joints in planar trusses and KK-joints in spatial trusses. The top chord (in case of spatial trusses:
the two top chords) may be integrated within the concrete slab of the bridge, see Figure 5.

Figure 5. Composite bridge with KK-joints and hot spots of a welded joint (FE-plot).
4.2 Truss joints
From the viewpoint of economics and ease of construction it is preferable to weld the braces
directly to the bottom chord. The resulting welded joints lead to high stress concentrations (hot-
spot stresses) because of the coincidence of the metallurgical discontinuity due to the weld and
the stress-rising effect caused by the geometrical intersection of the truss members. Therefore,
welded joints demand a precise adjustment of design, fabrication and quality assurance.

66
In cast-iron joints the welds are located outside the critical chord-to-brace intersection and are
of simpler shape. Aside of the economical aspect one disadvantage of cast iron compared to
welded joints is that the bottom chord is cut for each joint twice. That may cause a drawback on
the reliability of the entire structure.
4.3 Verification
The fatigue verification of cast iron joints is generally based on the S-N method providing that
the detailed joints are covered by an existing detail category. The influence of backing rings,
scale effect etc. has to be taken into account. Eventually, the nominal stresses have to be
modified accounting for offsets etc. In order to verify a welded CHS joint the S-N method does
not suffice any more because no nominal stress can be defined. Due to the non-uniform and
complex stress field in proximity of the weld along the brace-to-chord intersection the more
sophisticated hot-spot stress method has to be performed. For more than 30 years positive
experiences have been gathered in off-shore structures using this method.
The hot-spot stresses can be computed by a Finite-Element analysis. This approach requires a
sufficient expertise of the analyst and is rather time spending. The application of stress
concentration factors (SCF) can be considered as a facility. The SCF are highly dependent on
the so-called non-dimensional geometrical parameters (for example: ratio of brace and chord
diameter, ratio of wall thicknesses etc.). Unfortunately, highway bridges fall out of the
application range of current design specifications such as CIDECT etc.
Highway-bridge specific SCF for K-joints are given in Schumacher (2003). In contrast to K-
joints, SCF for KK-joints are not available in tables for the adequate range of geometrical
parameters typical of highway bridges. In an on-going research project fostered by the German
federal authority Bundesanstalt fuer Strassenwesen highway-bridge specific SCF-tables are set
up. The underlying approach is based on Schumacher (2003). In addition construction rules will
be summarised.
5 FATIGUE OF COMPOSITE BRIDGES WITH HORIZONTALLY LYING SHEAR
STUDS
5.1 General
Regarding composite bridges, the fatigue behaviour of the shear connections devices, which are
in most cases headed studs, are of major importance. While a lot of research has already been
done on headed studs in a normal vertical position, till now only few knowledge is available on
headed studs in a horizontally lying position.
In comparison to normal vertical headed studs, horizontally lying headed studs feature a
small edge distance a
r
of the headed stud to the concrete surface. Depending on the load
direction parallel or orthogonal to the free concrete edge the behaviour and resistance are
different. They allow for new interesting cross sections and applications, see Figure 6.
Numerous experimental and numerical investigations e.g. in Kuhlmann & Krschner (2006)
lead to design rules for static resistance for both load directions and furthermore for fatigue
resistance under longitudinal shear which meanwhile have been implemented in EN 1994-2.
The resistance is influenced by the reinforcement. Therefore beside the load direction the
resistance is specified depending on the effective edge distance a
r
, see Figure 6. With
decreasing of the effective edge distance cracking of the concrete becomes more important and
stud failure is dominated by the concrete cracking.
c + d /2
v s
a' = a'
r,o r
a
r

Figure 6. Loading of horizontally lying headed studs and geometrical parameters.

67
5.2 Investigations on the fatigue strength under transverse loading
Substantial fatigue loads in vertical direction derive from concentrated loads like wheel loads,
see Figure 6. Two test series on push-out specimens with different effective edge distance a
r

were performed to investigate the influence of the effective edge distance a
r,o
.
The specimens were loaded with a constant amplitude load range at similar maximum load
levels and different load ranges.
The existing experiences concerning the fatigue behaviour of headed studs only show
shearing of the studs as stop criterion. Contrarily, in these tests three different failure modes
occurred. Some specimens failed by shearing of the studs, too. But most specimens failed by
concrete outbreak. Two specimens showed a fatigue failure of the reinforcement. In both cases
one stirrup failed in the upper bend.
Contrary to headed studs without an edge influence only a slight concrete crushing at the
weld collar could be observed. Concrete cracking at an early stage is essential for the fatigue
behaviour. The resulting increasing slip causes bending in the stud shank.
A statistical evaluation comes to S-N curves for the 95%-fractile. A comparison with the
fatigue strength of headed studs without edge influence is given in Figure 7. The effective edge
distance a
r,o
obviously has a significant effect on the fatigue strength which is lower as the
fatigue strength of headed studs without an edge influence.

Figure 7. Comparison of the fatigue strength of horizontally lying studs under vertical loading and
headed studs without edge influence and typical failure mode.
5.3 Residual static resistance
The design code EN 1994-2:2005 gives rules to determine the static resistance of headed studs
independently from the fatigue strength. This implies the assumption that the entire static
resistance is valid over the whole life time. The design rules are derived from push-out tests
which use shearing of the studs as failure mode. A crack may appear much earlier. This leads to
the question whether a separate determination of the static resistance and fatigue strength may
lead to unsafe results.
First systematical investigations were made by Oehlers (1990). To this day Hanswille et al.
(2007) carried out the most comprehensive test program concerning this topic. With increasing
number of cycles there is a decreasing residual static resistance observed.
To investigate the residual static resistance for horizontally lying shear studs under vertical
loading a further test series has been performed. After a variable range of load cycles the
different test specimens were tested statically in controlled manner. Although there is a slight
decreasing residual static resistance with increasing number of cycles, all specimens exceeded
the expected static resistance on mean level. Despite the great residual static resistance the
headed studs showed remarkable fatigue cracks.

68
The high residual static resistance of horizontally lying shear studs under vertical loading
after a high cycle preloading may be explained with the low resistance of the concrete for a
mere static loading. As for horizontally lying shear studs under vertical shear concrete failure
limits the static loading decisively, the utilization of the partially cracked stud sections under
cyclic loading is not very high and allows for a higher residual strength compared to studs
without an edge influence.
6 SUMMARY
Fatigue is a major aspect within the design and maintenance of bridges during the whole life
time. The current fatigue design procedure for steel and composite according to the Eurocodes
offers a quite up to date concept including e.g.:
- different verification procedures ranging from simplified to quite sophisticated,
- different safety concepts taking into account the owners strategy concerning
maintenance and inspection and
- a detail catalogue with more than 70 different constructional details that allows in most
cases a proper and durable design over the whole life time.
All these are vital elements with regard to a sustainable engineering.
However, modern trends within the design of steel and composite bridges take place in the
area of lightweight, slender and aesthetically attracting structures, e.g. by using high strength
steels, hollow section joints or new shear connection devices for composite bridges. In this
context an overview of ongoing research work has been presented showing that:
- the application of high strength steels in combination with the allowance of modern post-
weld treatment method is very beneficial in order to reduce the self-weight and thereby
the vast of natural resources,
- the application of circular hollow sections leads to very economical and durable bridge
structures, when refined methods are used that allow for a proper fatigue design of
welded joints, and that
- the existing regulations concerning the fatigue resistance of shear connection devices in
composite bridges have to be modified in order to take into account e.g. the position and
loading direction of the headed stud (horizontally lying studs).
REFERENCES
Hanswille, G. & Porsch, M. & stndag, C. 2007. Resistance of headed studs subjected to fatigue
loading, Part I: Experimental study, Part II: Analytical study. In: Journal of Constructional Steel
Research 63, p. 475 - 484 and p. 485 - 493, DOI 10.1016/j.jcsr.2006.06.035 and
10.1016/j.jcsr.2006.06.036.
Kuhlmann, U. & Krschner, K. 2006. Structural behaviour of horizontally lying shear studs. In: Leon, R.
T.; Lange, J.: Composite construction in steel and concrete V, July 18-23, 2004, Kruger National Park,
South Africa, American Society of Civil Engineers (ASCE), p. 534 - 543. - ISBN 0-07844-0826-2.
Oehlers, D. J. 1990. Deterioration in strength of stud shear connectors in composite beams. In: Journal of
structural engineering 116, No. 12, p. 3417 - 3431.
Kuhlmann U., Bergmann J., Drr A. and Thumser, R. 2006. Effizienter Stahlbau aus hherfesten Sthlen
unter Ermdungsbeanspruchung, Forschungsvorhaben im Auftrag der Forschungsvereinigung
Stahlanwendung e.V., AiF-Forschungsvorhaben-Nr. 13866 B, 2006.
Drr, A. 2006. Zur Ermdungsfestigkeit von Schweikonstruktionen aus hherfesten Bausthlen bei
Anwendung von UIT-Nachbehandlung, Phd-Thesis, University of Stuttgart, Institute of Structural
Design, No. 2006-3 (in German).
Schumacher, A. 2003. Fatigue behaviour of welded circular hollow section joint in bridges. EPFL Thesis
No. 2727, Lausanne, 2003.
Bundesanstalt fr Straenwesen (BASt) 2007. Entwicklung von Entwurfs-, Bemessungs- und
Konstruktionsrichtlinien fr geschweite Hohlprofilknoten mit KK-Knoten im Straenbrckenbau.
Bearbeitet durch Universitt Stuttgart, Institut fr Konstruktion und Entwurf (Prof. Dr.-Ing. U.
Kuhlmann).

69
Z. Lj. Bozinovski
Institute of Earthquake Engineering and Engineering Seismology, University "St. Cyril and Methodius",
Skopje, fyr of Macedonia
J. Lahdensivu
Tampere University of Technology, Tampere, Finland
E. Vesikari
Degradation modes and models for masonry structures
1 INTRODUCTION
Massive masonry buildings are mainly designed and constructed to sustain vertical - gravity
loads. With a high bearing capacity of walls under pressure, insufficient bearing capacity under
tension, low ductility capacity and inadequate connection of structural elements into a whole,
particularly inadequate interconnection of bearing walls in both orthogonal directions and
inadequate connection of bearing walls with the floor structures, their capacity to sustain
external gravity and dynamic effects may be defective leading possibly to a limited service life
in seismically active areas.
Layered masonry walls came into use in Northern Europe in the late 1960s. In those
buildings the external wall assembly consists of bearing inner layer, usually brickwork or
concrete, insulation layer, usually mineral wool, and outer brick lining. The bearing structure
can also be made of wood (so called Swedish external wall).
From 1980s in some small houses the external wall structure consisted of blocks of
autoclave aerated concrete (AAC) or breeze-blocks which have an insulation layer between two
breeze-blocks. This kind of wall structure is still quite popular in rendered small houses in
Finland.
The outer surface of a masonry external wall can be treated in several different ways. It can
be fair-faced or rendered in many different techniques, of which the three coat rendering is the
most traditional. The surface treatment of a masonry structure affects a lot the water absorption
and drying processes of masonry.
In cold weather countries the fact that burnt bricks, concrete blocks, mortars and renderings
are porous, water absorbing and brittle, causes masonry faades subject to frost attack. The
degradation of masonry structures is a very complicated process depending on many features of
the actual case, such as the environmental conditions at the site, compass point, possible shelter
given by other structures, physical performance of the cross section of the structure, connecting
materials, etc. All details of degradation are not yet fully known. However, the temperature and
water content in materials under outdoor exposure can be monitored or simulated by numeric
methods. Connecting the theory of critical degree of water saturation to the analysis of
temperature and moisture content of materials the degradation of masonry structures can in
principle be modeled and the service life can be roughly evaluated.

70
2 STRUCTURAL DETERIORATION BY SEISMIC EFFECTS
While designing masonry structures, it is necessary to satisfy the requirements related to
providing of an appropriate, defined by codes, strength, stiffness, deformability and capability
for energy dissipation of structural elements and the system as a whole.
In the course of time, the designed hysteretic force-displacement relationship is the main
parameter to be used in defining the structural response to external gravity and dynamic effects.
The strength, stiffness and deformability capacity of the structural elements and the system as
a whole changes in the course of time, or more precisely, decreases due to a number of reasons.
The reasons include ageing of the material, aggressiveness of the environment, extreme climatic
conditions, interventions done on the structural elements as well as abrupt and unexpected
dynamic effects as are the effects of explosions and earthquakes.
Generally, the strength and the deformability capacity of the structure decreases during its
serviceability life resulting in a structural response with increased required displacement under
an earthquake effect.
The reduced deformability capacity and the increased required displacement can reach a
certain defined moment, most frequently referred to as the ultimate allowable deformation, after
which the service life of the structure can be considered to be at its end and a structural
intervention, i.e., strengthening of the structure is necessary.
The strengthening compensates for the reduced strength, stiffness and deformability capacity
and the ability of the elements to dissipate seismic energy through nonlinear behaviour.
2.1 Design conditions
The design conditions contain several parameters associated with the bearing structural system,
the purpose, the required seismic protection, the location of the structure and other parameters in
accordance with the valid technical regulations. The purpose of the structure is important from
the aspect of definition of the required level of resistance to seismic effects, in accordance with
the mentioned regulations. The level of required seismic protection is given in the seismological
map. It is defined by special seismological-geological investigations of the considered site.
The strength, stiffness and deformability characteristics of the materials as well as the ability
of elements to dissipate seismic energy through linear and nonlinear behavior have a different
influence on the behavior of the building structure exposed to vertical - static and horizontal -
dynamic loads. To provide conditions for analysis of safety of existing structures as well as
proper and most adequate design of repair and strengthening of the principal structural system,
it is necessary to define the strength and deformability characteristics of the bearing walls under
vertical and horizontal loads as well as the behavior of the structure under an earthquake.
Definition of the characteristics of materials, particularly strength and deformability capacities
is done analytically or experimentally. The most appropriate method is the experimental that can
be carried out in laboratory conditions or "in situ" on a certain bearing wall (Fig. 1). The
minimum necessary parameters that should be defined are: ultimate compressive stress of
masonry, ultimate tensile stress of masonry, elasticity modulus of masonry and sliding modulus
of masonry.
2.2 Analysis of the structure
During the serviceability period, the structures are exposed to the effect of vertical and
horizontal, static and dynamic, permanent and temporary, normal and abnormal loads. The
vertical loads, dead or life, are mainly static loads depending on the purpose of the structure, the
finishing of the floor and facade structures, the insulation of the internal and external walls and
the roof structure. The vertical loads should be more precisely defined, since the intensity of
seismic forces and the response of the structure are directly related to the amount of vertical
loads, i.e., masses of the structure.
The mathematical model of the structure is defined as a system with masses concentrated at
the level of floor structures, connected by springs and dampers.

71

Figure 1. P-o diagram, Palace Sponza, obtained during 'in situ' test.
For the defined vertical and horizontal loads, linear static and dynamic analysis is performed for
the purpose of obtaining the periods, the mode shapes, the storey stiffness and the relative
displacements. The static quantities, bending moments, shear and axial forces are checked for
characteristic frames in both orthogonal directions. Based on the storey stiffnesses, defined on
the basis of geometrical characteristics of bearing elements, the defined relative storey
displacements, the defined seismic effects on the considered location with intensity and
frequency content for different return periods and the corresponding design criteria, first of all
the ductility capacities of the bearing elements, through the dynamic response of the building
structure, optimization of equivalent seismic forces along the height of the building is performed
and hence optimization of strength, stiffness and deformability of the structure. Static analysis
of the building structure under vertical loads and optimal equivalent seismic forces is performed
in both the orthogonal directions. Defined are the static quantities, bending moments, shear and
axial forces, for each element and each direction.
The above holds for design of new structures. For analysis of repaired and strengthened
masonry structures, static analysis is performed for vertical loads and equivalent seismic forces
according to regulations. The main purpose of the static analysis, when analysis of elements up
to ultimate states of strength and deformability is performed, is to distribute the seismic forces
per bearing elements and find out the axial forces in the vertical elements due to the gravity
loads of the floor and roof structures, along with the dead weight of the massive walls.
For the structural elements with geometric characteristics, characteristics of materials and
position in the structure, analysis of the elements is done and hence analysis of the structure is
performed up to ultimate states of strength and deformability. Involved in the analysis are
several types of elements characteristic for masonry structures. For several types of walls, as are
stone walls, brick walls, stone or brick walls with reinforced-concrete jackets, framed brick
masonry with reinforced concrete vertical and horizontal belt courses and stone walls with a
concrete coating, there is a simple, but sufficiently exact way of determining the strength and
stiffness capacities in the linear range of behavior. The deformations in the same range are
defined by the linear strength - stiffness relationship. This way of defining the ultimate states is
not sufficiently exact for the reinforced-concrete elements as are columns and walls due to the
impossibility of controlling the failure mechanism, particularly from the aspect of defining the
deformation at which it occurs. Generally, for all possible elements occurring in the masonry
structures, analysis of strength capacity can be done in the simple way of defining strength and
stiffness capacity, whereas care should be taken as to deformability, knowing its importance for
the behavior of the structure under an earthquake. Considering the complicated behavior of
reinforced-concrete elements, a more precise way of determining their ultimate strength and
deformability value is given, with control of the mechanism of behavior from the beginning of
loading up to failure. This is important because we are often forced to define simultaneously the

72
ultimate states of masonry and reinforced-concrete elements so that, by their superposition, one
arrives at storey strength, stiffness and deformability capacities in both the orthogonal directions
of the structure (Fig. 2).
Representative bilinear storey diagram
Cumulative storey
diagram - RC elements
Cumulative storey diagram -
Masonry wall elements
Individual RC elements
Individual masonry wall elements
K=4160 [kN]
LP=0.11
o
y=1.13 [cm]
o
u=4.20 [cm]
Qy=4700 [kN]
Qu=6100 [kN]
D=3.72

Figure 2. Force - displacement diagram for masonry and R/C elements.
The most important for the behavior of the structures is their response under static and dynamic
loads. Masonry structural systems, as well as other systems, exposed to dynamic - seismic
effects of moderate and high intensity, suffer nonlinear deformations of the constituent
elements, by which the initial strength and stiffness is decreased considerably, depending on the
size of deformations and the number of times they are iterated.
The strength, stiffness and deformability capacity of the structural elements and the system as
a whole changes in the course of time, or more precisely, decreases due to a number of reasons
(Fig. 3 - 4).

Figure 3. Relationship strength - displacement.

73

Figure 4. Relationship strength - displacement.
2.3 Evaluation of the stability of the structure and the need for repair and strengthening
Based on the analysis of the existing state of the building structure and damages to the structure,
the elements of occurred damages and the reason of occurred damages are defined. Such
considerations are important for selection of possible and necessary measures for repair and
strengthening of the structural system.
The strength and deformability capacities of the bearing elements and the system as a whole
are compared to those required (according to the regulations) and those required for the
analyzed structural response under expected earthquakes on the considered site, with intensity
and frequency content. If the strength and the deformability capacity is less than the required, it
is concluded that the building structure does not have sufficient strength and deformability and
therefore it needs repair and strengthening.

Figure 5. Bozinovskis hysteretic model.

74
Each concrete structure is a case for itself and there are several ways of repairing and
strengthening it. The solution for repair and/or strengthening depends on: seismicity of the site,
local soil conditions, type and age of the structure, level and type of damages, time available for
repair and/or strengthening, equipment and man power, restoration and architectural conditions
and requirements, economic criteria and necessary seismic safety. Selected are several variant
solutions for repair and strengthening.
Analysis of each solution is done and an insight into the advantages and disadvantages is
obtained from several aspects. Out of these, selected is the most adequate solution from the
economic aspect and the stability aspect according to the required seismic protection. Out of
several analyzed possible solutions for repair and strengthening of the main structural system,
selected is the most favorable from the aspect of: stability, i.e., fulfillment of the design criteria
according to regulations, possibility for realization of the solution, available materials, economic
justification, fulfillment of social requirements and satisfying of aesthetic requirements (Fig. 5).
2.4 Conclusion on the stability of the structure and the vulnerability level
Based on the strength and deformability capacities of the bearing elements and the structure as a
whole and on the basis of required strength and deformability for expected seismic effects with
intensity and frequency content, conclusions are drawn regarding the stability of the structure
and its service life. It is of exceptional importance to bring the strength, stiffness and
deformability of the structure within the frames of the requirements according to the valid
technical regulations and latest knowledge on the behavior of masonry structures exposed to
gravity and seismic effects.
3 FROST ATTACK
3.1 Frost attack of masonry structures in cold weather countries
Frost attack due to a high moisture load is the most common reason for deterioration of masonry
structures in Nordic outdoor climate. Mortars in masonry and rendering as well as bricks are
porous materials whose pore system may, depending on the conditions, hold varying amounts of
water. As the water in the pore system freezes, it expands about 9% by volume which creates
hydraulic pressure in the system. If the level of water saturation of the system is high, the
overpressure cannot escape into air-filled pores and thus damages the internal structure of the
concrete resulting in its degradation.
It has been presented more than 15 different theories or explanation for frost attack of porous
materials during time (Kuosa & Vesikari 2000). It can be noticed that frost weathering is a
complex process, and frost damage can occur in many different ways (Fagerlund 1997).
Probably the most widely known frost damage theory is the hydraulic pressure theory by
Powers published in 1949. Accordingly, damage occurs as freezing water expands creating
hydraulic pressure within the pore structure of a porous material. The pressure is created when
part of the water in a capillary pore freezes and expands forcing thereby the unfrozen water out
of the pore. The migration of water causes localised internal tensions in the material whereby its
strength may fail resulting in cracking (Powers 1949).
Theory of volume chances in microscopic ice crystal was developed to complete hydraulic
pressure theory (Powers & Helmuth 1953). During freezing process, in capillary pores become
small ice crystals which tend to grow. Growth of ice crystals follow from lower chemical
potential in ice crystals than supercooled pore water in smaller gel pores. If there is not enough
empty space for growing ice crystals, the growing ice crystals will cause pressure to pore
structure an might lead to cracking of pore structure. It is peculiar to growing microscopic ice
crystals, that growing will going on despite the lowering of temperature will stop during
freezing phase.
Theory of osmotic pressure complete two previous theories taking into account also
migration of dissolved chemicals, mainly the alkalies Na
2
O and K
2
O, in the pore water. Those
dissolved chemicals lower the freezing temperature of pore water and increase the concentration
of salts in the water surrounding ice. The concentration of dissolved chemicals tries to get

75
equilibrium between different pore water solutions causing osmotic pressure to pore structure
(Pigeon & Pleau 1995).
Litvan noticed in the beginning of 1970s that water in pore structure of porous material does
not freeze immediately when temperature decreases below 0 C. Freezing happens first in
bigger capillary and gravitation pores. In smaller gel pores water begins to freeze when
temperature is approximately -15 - -20 C. The unfrozen pore water is thus supercooled, which
tends to cause drying of paste because the saturation pressure over supercooled water is higher
than that over ice. According to Litvan, mechanical damage happens when moisture transfer
cannot occur in an orderly manner, i.e. when the rate of freezing is too high, or the distance that
water must travel to reach an external surface and freeze is too long (Pigeon &Pleau 1995).
Litvan has made lot of work concerning action of frost during 1970s.
Increasing of temperature on faade also increases the temperature of ice in pore structure.
First when temperature decrease, ice in pore structure shrinks, and supercooled pore water can
flow capillary to new empty space of pores and freezes rapidly. When temperature increases,
ice will need more space than before. This causes remarkable hydraulic pressure to pore
structure (Penttala 1998). The temperature factor of ice is 50 x 10
-6
K
-1
, it is 5 to 10 times more
than temperature factor of burned bricks or mortars.
Frost weathering is manifested as reduction in strength of the mortars and bricks, loss of
adhesion, or crazing or chipping off of the surface. In Nordic outdoor climatic conditions frost
weathering is the most common cause of damage.
The degree of frost weathering may vary in different parts of the facade - depending on, for
instance, the load and variation in material properties - as well as thorough the thickness of the
rendering. Weathering due to high local moisture load may affect only a very limited area. On
the other hand, an improper surface treatment may result in deterioration across most of the wall
surface.
3.2 Theory of critical degree of water saturation
The theory of critical degree of water saturation was developed in the early 1970s (Fagerlund
1977). The basic idea of this theory is that there is a critical degree of water saturation above
which the porous and brittle material is damaged while freezing. If the actual water saturation is
below the critical one no damage takes place during freezing.
The theory of critical degree of water saturation is easily adaptable to a building physical
simulation programme. By an application that simulates temperature and moisture content in a
structure the number of critical freezing events, i.e. the events when the critical degree of water
saturation is exceeded simultaneously with freezing, can be counted. The progress of the
damage as a function of critical freezing events can be evaluated by the theory of Fagerlund.
When the degree of damage exceeds the prescribed limit state the service life is considered to be
ended.
If the degree of saturation is smaller than the critical saturation practically no reduction in the
dynamic modulus of elasticity of a material sample takes place as a result of freezing. If the
degree of saturation is greater than the critical one the diminishing of dynamic modulus is
proportional to the difference between the actual degree of saturation and the critical degree of
saturation. According to Fagerlund the degree of damage may be defined in the following way
(Fagerlund 1977):
) ( 1
0
crit N
N
S S K
E
E
D = = (1)
where D is degree of damage (0 < D < 1)
N number of critical freezing events,
S degree of saturation at the moment of freezing,
S
crit
critical degree of saturation,
K
N
Coefficient of degradation.
In the derivation of Eq. 1 an assumed has been that the degree of water saturation stays
constant at successive critical freezing events. In natural weather conditions, however, the
moisture content of concrete varies all the time. So the degree of water saturation is different in

76
successive freezing events and Eq. 1 cannot be applied as such. That is why a modified formula
was created to be applied in normal outdoor conditions:
D
K
w
w w
N
N
sat
N crit
N
N
=
=
A
( )
1
(2)
where D
N
is degree of damage after N critical freezing events,
w
N
moisture content of concrete at the moment of N
th
critical freezing, kg/m
3
,
w
crit
critical moisture content (corresponding to critical degree of saturation),
kg/m
3
and
w
sat
moisture content of completely saturated concrete (corresponding to total
volume of pores), kg/m
3
.
As the number of critical freezing events is related to the age of the structure the service life
of the structure can be determined from Eq. 2 when the maximum allowable degree of damage
is defined.
) D D ( t t
max N L
= = (3)
where t
L
is service life, and
D
max
maximum allowable degree of damage.
3.3 Tests for bricks and mortars
For building physical simulation it is essential to do tests of the simulated materials. The tests
are designed to yield input data for the simulation of moisture transfer and frost attack. Both the
brick and the mortar may be attacked by frost. So, both materials should be tested. In the
following example two brick types (A and B) and two cement based mortars (O and P) were
tested. The properties of these bricks and mortars according to quality control tests are presented
in Tables 1 and 2.
Table 1. Results of quality control tests for bricks A and B.
Brick Water
absoption
%
Compressive
strength
MN/m
2

Density

kg/m
3

A 20,7 45.7 1540
B 9.3 37.7 1250

Table 2. Results of quality control tests for mortars O and P.
Mortar Flexural strength
N/mm
2

Compressive strength
N/mm
2

7d 28d 7d 28d
O 2,3 3,1 8,3 11,2
P 2,0 3,1 7,7 11,2

The test programme for both bricks and mortars included:
- Capillary water uptake test (Fagerlund 1977)
- Drying test
- Critical degree of saturation test (Fagerlund 1977).
The capillary water uptake test is performed using flat specimens with about 25 mm thickness.
The specimens are placed on a grating so that they can absorb water through their lower surface.
The increase of moisture content is followed with time and the results are normally presented as
a function of square root of time. Then the nick-point of the absorption curve can be
determined. The test simulates conditions when the structure is exposed to rain or free water in
general.

77
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 2 4 6 8 10 12 14 16 18 20
Square root of time, \h
D
e
g
r
e
e

o
f

w
a
t
e
r

s
a
t
u
r
a
t
i
o
n
,

S
Mortar O
Mortar P
Brick B
Brick A

Figure 6. Results of the water uptake tests.
The drying test is performed in a room with the relative humidity of 70% and the temperature
20 C. The specimens are about 20 mm thick and the drying can take place from both opposite
surfaces. In the beginning of the test the specimens are saturated in 100% RH. The drying test
simulates drying of a material in atmospheric conditions.
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0
20.0
0 20 40 60 80 100
Time, d
M
o
i
s
t
u
r
e

c
o
n
t
e
n
t
,

w
-
%
Brick A
Brick B
Mortar O
Mortar P

Figure 7. Results of the drying tests.
In the critical degree of saturation test the number of test specimens is between 10 - 20. The
moisture content of specimens is adjusted to different levels of water saturation, between 0.6 -
1.0. Then the specimens are wrapped in plastics and frozen to -20 C. The damage in specimens
is checked by ultra sonic measurement after 6 - 9 freeze-thaw-cycles.

78
Brick A
(Af ter 6 and 9 f reeze-thaw-cycles)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Degree of water satusation, S
D
a
m
a
g
e

(
E
n
/
E
o
)
6 cy
9 cy

Brick B
(Af ter 6 and 9 f reeze-thaw-cycles)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Degree of water saturation, S
D
a
m
a
g
e

(
E
n
/
E
o
)
6 cy
9 cy

Mortar O
(Af ter 6 freeze-thaw cycles)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
D
a
m
a
g
e

(
E
n
/
E
0
)

Mortar P
(Af ter 6 f reeze-thaw cycles)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
D
a
m
a
g
e

(
E
n
/
E
0
)

Figure 8. Results of critical degree of water saturation tests.
3.4 Simulation
The simulation in the next example was performed using the data of Otaniemi weather station
(near Helsinki) and the material properties presented above. However, the simulation based on a
typical facade structure under typical driving rain conditions showed no exceeding of the critical
moisture content. Only by increasing the driving rain load multiple folded as compared to the
normal load critical freezing events (freezing events with moisture content above the critical
content) could be observed. Figure 9 shows the calculated freezing events during one year with
a 6-fold driving rain load and for the assumed critical moisture contents (Brick B).

79

Figure 9. Freeze-thaw cycles for a masonry facade wall (the thickness of the outer core is 130 mm). The
curve for all cycles is depicted with a dotted line. The curves for critical freezing events with different
assumed critical moisture contents are drawn with continuous lines.
3.5 Conclusions
The moisture transfer properties and critical degree of water saturation of bricks and mortars are
crucial for the degradation and service life of masonry structures in cold weather countries. The
theory of critical degree of saturation as combined with numeric simulation of moisture content
can be effective in the evaluation of degradation rate and the service life for masonry structures
as for other porous and brittle facade materials too. However, the models in numeric simulation
should still be developed and more data on properties of materials should be gathered to get the
method work reliably.
REFERENCES
Bozinovski, Lj. Z. 1997. Procedure of analysis of masonry building structures to be constructed in
seismic prone areas, Fourth National Conference on Earthquake Engineering, Ankara, Turkey, 17-19
September, 1997.
Bozinovski, Lj. Z., Milutinovic, Z. & Trendafiloski, G. 2002. Disaster Prevention, Mitigation, Diagnosis,
Urgent Measures And Education Of Users For School Buildings, The Twelfth European Conference
on Earthquake Engineering, UK, London, September 9-13, 2002.
Bozinovski, Lj. Z. 2002. Repair and Strengthening of School Buildings Based on Experimentally Defined
Characteristics of Bearing Walls, The Twelfth European Conference on Earthquake Engineering, UK,
London, September 9-13, 2002.
Fagerlund, G. 1977. The critical degree of saturation method of assessing the freze/thaw resistance of
concrete. Tentative RILEM recommendation. Prepared on behalf of RILEM Committee 4 CDC.
Materiaux et Constructions 1977 no 58. pp. 217 - 229.

80
Fagerlund, G. 1997. Internal frost attack - state of the art. In: RILEM Proceedings 34, Frost resistant of
concrete. Setzer, M. J. & Auberg, R. (eds). pp. 321 - 338.
Kuosa, H. & Vesikari, E. 2000. Betonin pakkasenkestvyyden varmistaminen. Osa 1. Perusteet ja
kyttikmitoitus ( Ensuring of concrete frost resistance Part 1: Basic data and service life design).
VTT Technical Research Centre of Finland. Research notes 2056. 141 p. (In Finnish)
Penttala V. 1998. Freezing-induced strains and pressures in wet porous materials and especially in
concrete mortars. Advanced Cement Based Materials 7/1998. pp. 8 - 19.
Pigeon M. & Peau R. 1995. Durability of concrete in cold climates. Suffolk. E & FN Spon. 244 p.
Powers T. C. 1949. The air requirement of frost-resistant concrete. Chicago: Portland Cement
Association, Research and Development laboratories, Development Department. Bulletin 33.
Powers T. C. & Helmuth R. A. 1953. Theory of volume changes in hardened Portland cement pastes
during freezing. In: Proceedings of the Highway Research Board 32. pp. 285 - 295.
Vesikari, E., Rautiainen L., Hkk-Rnnholm, E., Silvennoinen K., Salonvaara M., & Viitanen, H. 2001.
Julkisivujen ja katteiden kyttin ennakointi (Service life prediction of facades and roof coverings).
VTT Technical Research Centre of Finland. VTT Publications 850. 158 p. (In Finnish)
Vesikari, E. & Kuosa, H. 2000. Tiilien ja muurauslaastien kosteudensiirto-ominaisuudet ja
pakkasvaurioituminen (Properties of moisture transfer and frost attack for bricks and mortars). VTT
Technical Research Centre of Finland. Report RTE37-IR-11/2000. 24 p. (In Finnish)

81
Durability and service life of wood structures and components -
State of the art
1 INTRODUCTION
The long term durability of building structures depend typically on several factors, but the first
stage for evaluation of durability and service life consist on evaluation of the exposure
conditions (ISO 15686-1): e.g. climate, type of local environment, building type and orientation,
design and details of the structure. For wood material, the microbes play often a key role for the
durability of material, especially in high humidity conditions. The microbial activity is often
highest in the tropical and subtropical climate and lowest in the boreal and arctic climate. There
are several climatic approach for biological activity. Koeppen's climate classification was
originally developed for the botanical and agricultural use, but it will give an overview on the
world macro-climate mapping for environmental biological activity. There are several main
climate areas based on temperature and precipitation. A new version of the climate classification
was presented by Kottek et al (2006).
For evaluation of the effect climate on decay development, the Scheffer index was developed
(Scheffer 1971), and it has long been used for mapping the decay hazard areas in the USA. In
Europe, EuroIndex for decay development has been presented (Grinda and Carey 2004,
VanAcker 2003). Brischke and Rapp (2007) found a poor correlation between decay rate and
cumulative Scheffer index values over the 4 years double decking test at 22 sites in Europe.
Dawson et al (2005) used two Climate Index (CI) models to compare the climate of
Braunschweig, Germany and Rotorua, New Zealand: a) CI
EU
and b) CI
J
, which are based on a)
global radiation, days of rainfall and total precipitation, or b) mean of monthly highest
temperature, total sunshine and number of rainy days. In Australia the decay development in
above ground in different climatic conditions was modelled using lap-joint field test results and
weather conditions of the sites, and also a software for calculating the decay risk was developed
(Wang et al. 2008, Wang and Leicester 2008).
The microclimatic means the climate conditions close the materials and structure, and it is a
result of several simultaneous factors: macroclimate (rainfall, temperature, humidity, air
pressure conditions etc.), and meso-climate (location of the building, structural details and the
materials used). The micro-climate conditions are the basic level for building physical and
microbial activity evaluations. There are several different programmes for evaluate the moisture
and temperature behaviour of structure (Ojanen et.al 1994, Knzel 1995). Mathematical models
on mould development have been introduced to these programmes to evaluate the eventual risk
of mould growth (Ojanen and Salonvaara 2000, Viitanen and Salonvaara 2001, Sedbauer 2001,
Moon 2005, Viitanen et al. 2009).
For durability aspects, there are so many factors that mathematical models are needed to
handle the complicated relations (Leicester et al 2003, Wang et al 2008). For micro-climate
level close the studied structure the time of wetness is a useful factor when evaluating e.g. risks
for corrosion of steel structures or mould growth on materials, but it alone does not give
adequate information about the durability risks for organic, wooden materials. Long period,
high moisture levels may start biological growth on timber surfaces, first mould or stain fungi

82
and finally decay (Viitanen 1996). The time of wetness in the exterior climate, however, does
not necessary correspond with the time of wetness in different parts of the building envelope.
The ISO factor method includes several general factors in order to evaluate the service life of
building components for different performance requirement levels (ISO 15686-1, 2006). Service
life means the period of time after installation during which a building or its parts meets or
exceeds the performance requirements, which means the minimum acceptable level of a critical
property, and can be defined as limit states. The life time expectations and analyses of different
building products will need more data on the durability of products, service life and resistance
against mould and decay, not only data on wood material itself. The complicated interaction of
different factors may be analyzed using different mathematical models.
2 CALCULATIONS OF THE EXPOSURE CONDITIONS FOR SERVICE LIFE
The type of data which is needed depends on the type of exposure and degradation mechanism
considered. Moisture and temperature is generally very important factors for biological and
chemical processes, the acting factor for the durability of materials is the humidity/moisture and
the temperature close the materials. In the first stage, climate data are needed at the boundary of
the wood element for evaluate the microclimate conditions. The conditions of microclimate
depend on varied and many factors. The starting point here is meteorological data defining the
regional climate in the area where the building is situated. Examples of data are temperature,
relative humidity, solar radiation, rain and wind intensity and duration. The next step is to
define the local climate, i.e. the climate conditions close to the building but still undisturbed
by the properties of the wood material and the shape of the structure. Local climate depends on
e.g. building components shadowing solar radiation and rain such as a roof overhang.
The micro climate, which could be evaluated from regional and local climates, can be
expressed as
- The equivalent air temperature close to the structure and the temperature distribution
- Humidity and wetness at the surfaces and moisture conditions of the materials
- Solar radiation on the surface.
The climatic parameters need to be estimated in terms of variability, extreme values and time
variation. Data are also needed for critical states leading to decay, mould growth or other
undesirable effects (Viitanen 1996). This can be a critical moisture threshold often dependent on
the duration of the moisture exposure, temperature, type of wood material considered etc. This
may be seen as a material property, whose variability also has to be taken into account in a risk
based design procedure (Isaksson 2008).
The time of wetness can be calculated by summing up the length of time over one year when
the relative humidity RH is above 80 or 95% at the same time when temperature is above 0 C.
The atmosphere can be characterized in relation to its fatality by calculating the time of wetness
using the temperature and the relative humidity of the ambient air. The time of wetness in the
building envelope parts, however, does not necessarily correspond with the time of wetness in
the exterior climate. The time of wetness may be high in the building structure even in cold and
dry climates if the moisture performance of the structure is poor.
A basis for a quantitative probability based design is the definition of a so called limit state or
performance degree (Figure 1). This is a more or less precise definition of the limit between
acceptable performance and non-acceptable performance. An example is onset of mould growth
in materials in the building envelope, which can be regarded as non-acceptable since it may
create aesthetic and health problems in a building. This is comparable to a serviceability limit
state for structures. Another example is attack from decay fungi, which will reduce the capacity
of a load bearing structure. During the life time of a building, different level of maintenance and
repair will be needed for acceptable performance. For wood material in exterior use, the
maintenance is normally involved for the service life evaluation depending on the exposure
conditions and structure.

83
Building performance and life cycle
Operation over time
Q
u
a
l
i
t
y

/

F
u
n
c
t
i
o
n
PD 0
PD 2
PD 3
PD 4
PD 1
Performance without preventive actions
Replacement
Maintenance
Repair Refurbishment
Aging / Failure

Figure 1. Building performance life cycle as a function of quality, performance degree (PD), failure,
maintenance, refurbishment, repair and replacement (based on ISO 15686-7, 2006).
The Scheffer (1971) index is developed to evaluate the decay risk in different parts of the USA
based on temperature and distribution of rainfall. The Scheffer index may give an overview on
the macroclimate conditions for decay development but it does not take account the local and
macroclimate conditions. Morris et al (2008) presented a new map based on the index.
In Europe, Francis and Norton (2006) found the Scheffer Index to correlate better with decay
rates in L-joint tests when data from one hot day site was excluded. In Australia the decay
development in above ground in different climatic conditions is modelled and also a software
for calculating the decay risk is developed (Mackenzie et al 2007, Wang et al. 2008). The
climatic data used for the modelling (Wang and Leicester 2008) were: annual rainfall
(mm/year), number of dry months (months/year), number of rain days (days/year), time of
wetness (hours/years), dry-bulb temperature (C), wet-bulb temperature (C), wind speed
(km/hr) and wind direction (degrees from the north). Also the data on decay in ground contact
and termite attack connected to climatic data is used for modeling (Leicester et al. 2003a, b). In
the Australian studies (Wang et al 2008), an exposure mapping of decay development in
different part of Australia was based on the results of outdoor field tests conducted in the
different part of Australia using the climatic data mentioned above.
Dawson et al (2005) used two models to compare the climate of Braunschweig, Germany and
Rotorua, New Zealand: a) CI
EU
and b) CI
J
, which are based on a) global radiation, days of
rainfall and total precipitation, or b) mean of monthly highest temperature, total sunshine and
number of rainy days. It was found, that rain fall, sunshine, mean daily global irradiance and
mean daily temperature were higher in Rotorua than these in Braunshweig, but the mean daily
relative humidity was higher in Germany. The New Zealand site had a climatic index 50% more
challenging for performance of wooden products.
In Europe, Brischke and Rapp (2005, 2007) and Brischke (2007) have evaluated the effect of
different test conditions on the decay development in different wood species. In sapwood and
heartwood, the decay development was different depending on the exposure conditions, and the
average weather data was not sufficient for estimation decay development. Brischke and Rapp
proposed a roadmap to specify the performance of wood durability, and a COST decay index
has been started to develop. An example of the European work is the on-going Woodexter
project, where a new format of the basic decay model according to Viitanen (1997b) was
developed (Viitanen et al. 2009).

84
3 EVALUATION AND MODELLING THE EXPOSURE CONDITIONS FOR MOULD
AND DECAY DEVELOPMENT
3.1 Critical part of the building as an exposure model
The technical function of the cladding or facade is to protect building against weather. Cladding
can be fully or partly protected/exposed depending on the design (Figure 2). According to the
standard EN 335-1, the use conditions of wood and wooden products are classified for the
evaluation of the activity of different biological agents to wooden structure. The classification
is, however, very general, and in reality, many details are often important for the durability and
service life of the whole structure or building. The evaluation of the exposure conditions can be
performed at general macroclimate level like using European Climatic simulation model
(Uppala et al 2005) and at a building component level using building physics calculation
programme like WuFi (WuFi).
The aim of the evaluation work of service life of wood products is to simulate and calculate
the complicated interactions of different factors of exposure of gladding and decking. Climate
data simulation will give an overall estimation of the macroclimate part of the exposure of ideal
or model structure. The more complicated interaction of the meso-climate like the effect of
environment and structural components (driving rain, details) will be evaluated using a WuFi
programme. The response of the exposure to decay development will be evaluated using a new
model developed on the basic work (Viitanen 1996).

Attic
Covered
Balcony
Cladding
Terrace
Uncovered
Cladding
Balcony
UC 2
UC 3.1 - 3.2
UC 2 - 3.1
UC 3.1 - 3.2
Terrace
UC 3.2
UC 3.1 -3.2
UC 3.2 - 4
Terrace / Jetty UC 3.2 - 4
Wood in ground
UC 4
Environmental factors:
Clima, Geography, Terrain,
other buildings, vegetation etc.
Structural factors:
Design, Work execution
Using factors:
Exposure, Maintenance
Fence UC 3.2 - 4

Figure 2. A general exposure situation of wooden products used in different structure and exposure types.
The suggestion of use class evaluation for develop the use condition classification is also shown
(Viitanen et al 2009).
3.2 Modeling development of mould or decay
3.2.1 Mould
Ambient microclimatic conditions, especially moisture conditions, are the most important
factors for durability of wood and the classification of use conditions is based on the evaluation
of the water exposure during the use. One of the first biological signs of ageing is mould growth
that does not affect durability as such but can cause discoloration and health problems.
Discoloration is often connected also to paint and surface treatment properties. The first isoplets
(graphs on critical humidity and temperature) or models for mould growth were based on the
laboratory studies on agar culture using different mould fungi (Ayerst 1969, Block 1953, Smith

85
and Hill 1982, Grant et al. 1989). On the base of results, different functions or s.c. isoplets and
critical humidity and temperature levels for spore germination and mould growth were
developed (Figure 3).

Figure 3. Relative humidity and temperature for the start of mould growth (Sedbauer 2001).
The time-of-wetness or varied humidity and temperature conditions is important for the mould
development (Adan 1994, Viitanen 1996). The TOW is defined by the ratio of the cyclic wet
period (RH> = 80%) and the cyclic dry periods. For mould development, the ambient critical
humidity level of microclimate is between RH 80 and 95% (Fig. 4a) and time of high humidity
followed by dry periods is important (Fig. 4b) and varied cyclic conditions can be taken under
consideration and basic information for modeling (Hukka and Viitanen 1999).
70
75
80
85
90
95
100
0 5 10 15 20 25 30
Time (weeks)
R
H

(
%
)
1 C
5 C
10 C
20 C

0
1
2
3
4
5
6
0 28 56 84 112 140 168 196
Time (days)
M
o
u
l
d

i
n
d
e
x

A Constant RH
12/12 h 97/75 RH
6/12 h 97/75 RH
3 / 21 h 97/75 RH
Figure 4 a, b. RH and temperature isoplets as a function of time for start of mould growth (left) and effect
of wet time (high humidity) on the development of mould growth (right) in pine sapwood (Viitanen 1996,
Viitanen et al 2003). Modeling is based on large laboratory work (Viitanen 1997a).
The models on biodeterioration and mould growth can be used as a tool for building physic
performance and service life evaluation. Sedlbauer (2001) has evaluated the spore moisture
content and germination time based on calculated time courses of temperature and relative
humidity in various positions of the exterior plaster of an external wall. This model has been
incorporated in a hygrothermal calculation tool Wufi (Wufi). Ojanen and Salonvaara (2000)
have used the VTT mould growth model implemented in another building physic simulation
model TCCC2D for evaluate the risk of mould growth in different humidity exposure conditions
in building envelope. Isaksson (2008) has presented the state of art situation on methods to
predict wood durability.
3.2.2 Decay
Decay is the more severe result of high moisture exposure of wooden structures when the
materials are wet for long periods. Mould growth and decay development are separated
processes and also the models will be different. For start of the growth of decay fungi and decay

86
development, the ambient critical humidity level of microclimate should be above RH 95 -
100% and moisture content of pine sapwood above 25 - 30% (figure 5).
70
75
80
85
90
95
100
0 4 8 12 16 20 24
Time (months)
R
H

(
%
)
0 C
5 C
10 C
20 C

Figure 5. RH and temperature isoplets as a function of time for early stage of decay development in pine
sapwood (Viitanen 1996, 1997b).
The humidity and moisture limits were based on large laboratory work on pine and spruce
sapwood (Viitanen 1996). According to experience, the decay will develop when moisture
content of wood excess the fibre saturation point (RH above 99.9% or wood moisture content
30%. Morris et al. (2006) have modelled decay development in wooden sheathing and found the
critical ambient humidity condition for decay development is around RH 98 - 99%, depending
on the temperature and exposure time.
On the basis of the laboratory work on the decay development of brown rot in spruce and
pine sapwood in different constant relative humidity and temperature conditions a decay model
was presented by Viitanen (1996). The decay growth model, expressed as mass loss, thus only
applies in temperature ranges of T = 0 - 30 C and in relative humidity's 95% and above. It is
noted that mass loss does not occur immediately when the wood is exposed to these
environments (Figure 4). Thus, there is a time lag or as named here, and activation period in the
beginning. Based on the experimental findings presented above, a model for variable conditions
is proposed. This model is a time stepping scheme. The development of decay is modeled with
two process:
a) Activation process:
This is termed as parameter, which is initially 0 and gradually grows depending on the
air conditions to a limit value of 1. This process is able to recover favorable conditions
(dry air) at a given rate. (although no experimental evidence of recovery is available).
b) Mass loss process:
This occurs when the activation process has fully developed ( = 1) otherwise it does not
develop. This process is naturally irrecoverable.
These processes only occur when the temperature is 0 - 30 C and the relative humidity is 95%
or above. Outside these condition bounds, the activation process may recover, but the mass loss
process is simply stopped. The activation process is as given in the following equations. The
recovery rate (when the air conditions are out of the bounds given above) is assumed to be
17520 [h] (2 years). This time means that may recover from a value of 1 back to 0 in this time,
when the conditions are outside the bounds of decay growth.

87
( )
( )
( )
] [ 24 30
45 . 0 14 . 0 0 . 42
024 . 0 035 . 0 3 . 2
) , (
17520
) , (
0
,
0
) (
1 .. 0
hours
RH T
RH T RH T
T RH
crit
t
decay of conditions e unfavorabl in
t
decay of conditions favorable in or
T RH
crit
t
t
t
where
t
d t
process Activation

+ +
+
=
A
= A
A
= A
A = =
=
(
}
o
o
o o o
o
(2)
The mass loss process proceeds the activation process, when has reached 1.
] / [%
4
10 25 . 6
4
10 96 . 1
2
10 96 . 5
) , (
) , ( ) , (
) (
1
1 1
hour RH T
dt
T RH ML
t
t
dt
T RH ML
t
dt
dt
T RH ML
t ML
when process loss Mass
at t at t
'
A =
}
'
= '
>
=
|
.
|
\
|
= o o
o
(3)
3.3 Modeling the exposure conditions for decay resistance
For evaluation the climatic exposure conditions, the empirical wood decay model presented in
the chapter 3.2.2 can be used for the ERA-40 data for air temperature, humidity and
precipitation at 6 hour intervals. ERA-40 is a massive data archive produced by the European
Centre of Medium-Range Weather Forecasts (ECMWF). The reanalysis involves a
comprehensive use of a wide range of observational systems including, of course, the basic
synoptic surface weather measurements. The ERA-40 domain covers all of Europe and has a
grid spacing of approximately 270 km. The nature of the data and the reanalysis methods of
ERA-40 are described in detail in Uppala et al. (2005).
The resulting modelled mass loss in 1961 - 1970 at the calculation points of the ERA-40 grid
were analyzed by a chart production software producing a maps of wood decay in Europe
(Figure 6). In these calculations, the factor of the empirical wood decay model was reduced
during non-decay periods by the rate that corresponds to the recovery time of two years.

Figure 6. Modelled mass loss (in %) of small pieces of pine wood that exposed to rain in 10 years in
Europe (according to Toratti et al 2009).

88
In the simulation, the calculation was based on the relative humidity and temperature in air so
that the humidity of air was set to 100% during precipitation, at non-freezing temperatures. The
present maps on Europe of decay development are based on evaluation of decay activity studied
in laboratory. However, they will give theoretical evaluation on the effect of climatic conditions
on the decay development in different geographical area.
3.4 Using a model in variable temperature and relative humidity
The model presented in chapter 3.2.2 is used here to see how the decay growth would develop
in varied humidity and temperature conditions. In the first case, the measured weather
conditions in Helsinki area of temperature and relative humidity is given (origin of data to be
given). This climate is shown in the figure 7 for a one year period. According to the model, this
climate seems to induce a low mass loss of 1.1% in 4 years (Figures 8 and 9).
During the first year, no decay development will occur in untreated pine sapwood. After 3
and 4 years exposure, some early findings on decay process only in the surface of unprotected
pine sapwood can be expected. Under normal use conditions, the cladding is protected using
paints or other coatings when water will not affect direct to wood surface and decay
development will be significant retarded or even neglected in the surface.
-40
-20
0
20
40
60
80
100
0 2000 4000 6000 8000 10000
Ti me [h]
T
e
m
p

C

o
r

R
H
%
RH
temp

Figure 7. Measured climate data (Helsinki) used in the decay model for one year.
0,0000
0,0500
0,1000
0,1500
0,2000
0,2500
0,3000
0,3500
0,4000
0,4500
0,5000
0 2000 4000 6000 8000 10000
Time [hours]
a
l
f
a
-
p
a
r
a
m
e
t
e
r
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1
m
a
s
s

l
o
s
s

[
%
]
alfa
Mass loss [%]

Figure 8. No activation of growth or decay development during the first and second years.

89
0,8200
0,8400
0,8600
0,8800
0,9000
0,9200
0,9400
0,9600
0,9800
1,0000
1,0200
0 2000 4000 6000 8000 10000
Time [hours]
a
l
f
a
-
p
a
r
a
m
e
t
e
r
0
0,2
0,4
0,6
0,8
1
1,2
m
a
s
s

l
o
s
s

[
%
]
alfa
Mass loss [ %]

Figure 9. An activation of decay process after 4 years exposure may be expected (Viitanen et al 2009).
When applying the models for evaluation of service life in certain microclimatic conditions, the
great natural variability of materials, coatings, structures, different treatments and organisms
should also be taken into consideration (VanAcker et al 1999). Different type of microbial
growth will be found on stone based material and on insulation material than that on wood
material. The ageing of material and accumulation of dust and other material on the surface of
building material will change the response of the material to moisture and biological processes.
(Theander et al 1993, Sedbauer 2001, Viitanen et al 2003.
4 SERVICE LIFE EVALUATION OF WOOD PRODUCTS AND COMPONENTS
The long term durability of building structures depend typically on several factors, but the
important stage for evaluation of durability and service life consist on evaluation of the
exposure conditions. The ISO 15686 identifies a wide range of parameters important to Service
Life Prediction:
A= Quality of components e.g. wood natural durability, treatment and coatings
B = Design level e.g. protection by design
C= Work execution e.g. joints and details
D= Indoor environment, e.g. temperature, RH, condensation
E= Outdoor Environment e.g. climate, driving rain, shadow
F= In use conditions e.g. wear, mechanical impacts
G= Maintenance level e.g. repair, revisions, repainting.
For wood materials and components, the factors B, C, E and G are obvious the most important
(Vesikari et al 2001, Viitanen 2005). However, values for these parameters could be determined
regionally, taking into account all the local effects of the external factors considered to be
important. The quality of components means different wood species, including sap and
heartwood parts (Table 1). Natural durability of wood, however is mainly based on heartwood
only (EN 350-2, 1994). Wood based products, however, include different impregnated and
modified products which give more vide use conditions. According to Brischke et al (2006), the
overall consideration of all possible influences of decay and service life of wood products
should be taken care for reliable evaluation and database of wood durability. For normal use
condition in Nordic Countries, the coated spruce boards and paint base film coated birch
plywood have been used for long time in exterior conditions without any significant durability
problems when best practices are followed (Viitanen et al 2008b). In the table 1 a list of factors
for service life of wooden claddings are presented. The list is long and shorter list for practical
solution and use may be needed. For decking, the effect of wood material is more important
than that for cladding. For decking material, most often impregnated or modified wood is used.

90
In Task Force Group of COST E 37 (2007) much of the discussion ranged around the subject
of service life evaluation and durability testing. It was accepted that a great deal of work needs
to be done to determine values for the relevant parameters. However, it was recognized that if
certain parameters are understood and controlled they should be taken into account when
estimating service life or specifying the level of durability required. What would be needed to
align with ISO 15686 is a system that allows establishment of a durability rating from the
existing test methods. These would then be used to set a value for Factor A (quality of
components) to cover e.g. treated commodities in each Use Class. The durability classes would
then be related to accepted reference products used within the biological tests. It is increasingly
recognized that there is a need to determine a range of effectiveness and define durability of
components within a range, as is done with natural durability of wood. For relationship of
damage and strength properties of wood, decay have an important effect. The effect of mass
loss, however, is depending on the timing, decay type and the sample size focused (Van de
Kuilen 2007).
Use classes presented in standard EN 335-1 (see Figure 2) will give a general overview on
exposure conditions of wood and wooden products in different intended use conditions
classified according to the expected biodeterioration risks (discoloration and decay fungi,
insects, termites and marine organisms) and ambient humidity conditions close the evaluated
material and structure (microclimate conditions). The use classes can be used as a guide for
designing and manufacture of wooden products. Service life evaluation, however, is more fitted
to the actual situation of material and structure. The evaluation may be more complicated (as
shown in the Table 1) or more simple based on the classified simple construction and material
types.
Table 1. Factors for service life of wooden facades.
Code Factor Parameters/factors for estimated service life
A 1 Wood material Wood species, decay and weather resistance, water permeability,
board quality, dimension, modification, preservation
A 2 Coating Coating type and properties (thickness, opacity, color), needs for
repainting (maintenance)
B Structure, design,
especially details
Structure of the houses: eaves, height of the wall and foundations.
Structure of the faade; board type, bonds and joints, ventilation,
protection of joints and end grains, fixing
C Work execution Achievements and treatments details, fixing, wood moisture
content, storage condition
D Exposure conditions
(weather, environment)
Point of compass, type of environment (protective - exposed)
macroclimate- > microclimate, exposure to driving rains
E Use conditions Indoor environment, moisture stress, mechanical inkuries
F Maintenance Care of accidental damage, serviceable, repainting (opaque - stains)
time of repaint, cleaning of the old surface

The service life evaluation model of EnnusPuu (Viitanen 2005) developed at VTT is based on
the factors shown in the table 1. A simplified figure on the systems is shown in the figure 10.
5 CONCLUSIONS
For service life of wood products, modeling of biological deterioration will give a tool to
evaluate the effect of environmental exposure to performance of wood in exterior conditions.
These models can also be as basic knowledge for developing the service life evaluation methods
for wood products and structure. Factor method presented in the ISO 15686 can be used as a
basic methods for evaluate the service life of wooden cladding.
6 ACKNOWLEDGEMENTS
The authors gratefully acknowledge the financial support of WoodWisdom-Net and the wood
industry partnership Building with Wood for funding. The WoodExter project partners are

91
thanked for their collaboration in this project. The ERA-40 data have been provided by
ECMWF

Is this a decking (horizontal structure) ? Tick here if yes
Otherwise this is a cladding (vertical structure)
Is the solution acceptable Acceptable
Service life factor [OK if 1] 0,9
Cons RSLC 1,1
A1 Materiaali 0,84
A2 0,64
B Suunnittelu 0,90
C Tyst / Toteutus 1,17
D 0,96
F 1,50
VTT
date
16.9.2009
Consequence class
Hannu Viitanen, Lasse Makkonen, Ruut Peuhkuri, Tomi Toratti
Service life factor = Consequence class factor x A1 x A2 x B x C x D x F
Consequence class factor
Wood material
Design /structural details
k t t
Construction works
Climate conditions
Repair and service procedures
Service life estimator for claddings and deckings
Surface coating
2

Figure 10. An example on service life evaluation calculation based on EnnusPuu VTT programme (a
first version presented by Viitanen 2005).
REFERENCES
Adan, O.C.G. 1994. On the fungal defacement of interior finishes. Eindhoven University of Technology.
Thesis. Eindhoven. pp 83 - 185.
Ayerst, G. 1969. The effects of moisture and temperature on growth and spore germination in some
fungi. J. Stored Prod. Res. 5:127 - 141.
Block, S.S. 1953. Humidity requirements for mould growth. Applied Microbiology 1(6), 287 - 293.
Brischke, C. 2007. Investigation of decay influencing factors for service life prediction of exposed
wooden components. Dissertation. In German. University of Hamburg. 321 p.
Brischke, C. and Rapp, A.O. 2005. Relation between lab tests, field tests, and in service performance and
their contribution to SLP. Proceedings of COST E37 Workshop, Olso, Norway.
Brischke, C; Bayerbach, R. and Rapp A.O. 2006.Decay-influencing factors: A basis for service life
prediction of wood and wood-based products. Wood Material Science and Engineering 1: 91 - 107.
Brischke, C. and Rapp, A.O. 2007. Dose-response relationships for service life prediction of wood. Wood
Science and Technology (In Brischke C. 2007).
COST Action E37. Report. Task Force Performance Classification. 2007.
Dawson B.S.W; Gottgens A. and Hora G. 2005. Natural weathering performance of exterior wood
coatings on Pinus sylvestris and Pinus radiata in Germany and New Zealand. JCT CoatingsTech 2005,
2(7): 539 - 546
EN 113. 1997. Wood preservatives - The method for determining the protective effectiveness against
wood destroying Basidiomycetes. Determination of toxic values. CEN. European committee for
standardization.
EN 335-1. 2006. Durability of wood and wood based products - Definition of use classes - Part 1:
General. CEN. European committee for standardization.

92
EN 350-2. 1994. Durability of wood and wood based products. Part 2: Guide to natural durability and
treatability of selected wood species ot importance of Europe. CEN. European committee for
standardization.
Francis, L.P. and Norton, J. 2006. Predicting of decay resistance of timber above ground. I: climate
effects. IRGWP/06-20330. Stockholm, Sweden. International Research Group on Wood Protection.
Grant, C., Hunter, C.A., Flannigan, B. and Bravery, A.F. 1989. The moisture requirements of moulds
isolated from domestic dwellings. Internat. Biodet. 25, 259 - 284.
Grinda, M. and Carey, J. 2004. The COST Euro index for fungal decay: Five years results. COST Action
E22 Final Workshop. March 2004.
Hukka A. & Viitanen H. 1999. A mathematical model of mould growth on wooden material. Wood
Science and Technology 33(6), 475 - 485.
Isaksson, T. 2008. Methods for predicting durability of wood. State of the art. Draft report. TVBK -
30XX. Division of Structural Engineering. Lunds Institute of Technology, Lund University.
ISO 15686-1. 2006. Building and construction assets - Service life planning - Part 1: General principles.
ISO 15686-7. 2006. Building and construction assets - Service life planning - Part 7: Performance
evaluation for feedback of service life data from practice.
Kottek, M.J., Grieser, C., Beck, B.R. and Robel, F. 2006. World map of the Kppen-Geiger climate
classification updated. Meteorol. Z 15, 259 - 263.
Knzel, H.M. 1995. Simultaneous Heat and Moistur Transport in Building Components - One- and Two-
dimensional calculations using simple parameters. IRP Verlag, Stuttgart.
Leicester, R.H., Wang, C-H. and Cookson, L.J. 2003. A risk model for termite attack in Australia.
Document No IRGWP 03-10468. Stockholm, International Research Group on Wood Protection. 15
p.
Leicester, R.H., Wang, C-H. Ngyen M.N., Thornton, J.D., Johnson G., Gardner, D., Foliente G.C., and
MacKenzie C. 2003. An engineering model for the decay in timber in ground contact. Document No
IRGWP 03-20260. 21 p. Stockholm, Sweden. International Research Group on Wood Protection.
Mackenzie, C.E., Wang, C-H., Leicester, R.H., Foliente, G.C. and Ngyen, M.N. 2007. Timber service life
design guide. Report Project No PN07.1052. Forest and Wood Products Australia Limited. 109 p.
Moon, H.J. 2005. Assessing mould risk in buildings under uncertainty. Doctoral thesis Georgia institute
of Technology, USA.
Morris, P., Symons P, and Clark, J. (2006) Resistance of wood sheating to decay. Wood protection 2006.
March 21-23, 2006. New Orleans, Lousiana, USA.
Morris, P.I, McFarling, S and Wang J. 2008. A new decay hazard map of North America using the
Scheffer index. IRGWP 08-10672. Stockholm, Sweden. International Research Group on Wood
Protection.
Ojanen T., Kohonen R. & Kumaran M. 1994. Modeling HAM transport through building materials and
components. Moisture Control in Buildings. Ed. H.R. Trechsel. ASTM. Philadelphia , pp. 18 - 34.
Ojanen, T and Salonvaara, M. 2000. Numerical simulation of mould growth in timber frame walls.
Healthy Buildings. Espoo, 6 - 10 Aug. 2000. Seppnen, O. & Steri, J. (ed). Vol. 1. FiSIAQ.
Scheffer, TC. 1971. A climate index for estimating potential for decay in wood structure above ground.
Forest Products Journal 21, 25 - 31.
Sedlbauer K. 2001. Prediction of mould fungus formation on the surface of/and inside building
components. University of Stuttgart, Fraunhofer Institute for building Physics, Doctoral thesis.
Stuttgart. Germany.
Smith, S.L. and Hill, S.T. 1982. Influence of temperature and water activity on germination and growth
of Aspergillus restrictus and Aspergillus versicolor. Transactions of the British Mycological Society
Vol 79. H 3 p. 558 - 560.
Theander, O., Bjurman, J and Boutelje, J. 1993. Increase in the content of low-molecular carbofydrates at
lumber surfaces during drying and correlation with nitrogen content, yellowing and mould growth.
Wood Science and Technology 27, 381 - 389.
Toratti, T; Viitanen, H; Peuhkuri, R; Makkonen, L; Ojanen, T; Jms, S. 2009. Modelling of durability of
wooden structures. 4th International Building Physics conference IBPC 2009, Istanbul, Turkey, 15-18
June 2009. Istanbul Technical University. Istanbul, Turkey, 127 - 134.
Uppala, S.M., Kllberg, P.W., Simmons, A.J., Andrae, U., da Costa Bechtold, V., Fiorino, M., Gibson,
J.K., Haseler, J., Hernandez, A., Kelly, G.A., Li, X., Onogi, K., Saarinen, S., Sokka, N., Allan, R.P.,
Andersson, E., Arpe, K., Balmaseda, M.A., Beljaars, A.C.M., van de Berg, L., Bidlot, J., Bormann, N.,
Caires, S., Chevallier, F., Dethof, A., Dragosavac, M., Fisher, M., Fuentes, M., Hagemann, S., Hlm,
E., Hoskins, B.J., Isaksen, L., Janssen, P.A.E.M., Jenne, R., McNally, A.P., Mahfouf, J.-F., Morcrette,
J.-J., Rayner, N.A., Saunders, R.W., Simon, P., Sterl, A., Trenberth, K.E., Untch, A., Vasiljevic, D.,
Viterbo, P., and Woollen, J. 2005. The ERA-40 re-analysis. Quarterly Journal of the Royal
Meteoroogical Society, 131, 2961 - 3012.

93
Van Acker, J. Militz, H. and Stevens, M. 1999. The significance of accelerated laboratory testing
methods determining the natural durability of wood. Holzforschung 53 (5), 449 - 458.
Van Acker, J. 2003. Service life prediction and creation of a Euro decay index. Exterior applications out
of ground contact. COST Action E 37. Workshop September 2003.
Van de Kuilen, J.W.G. 2007. Service life modelling of timber structures. Material and structures 40: 151
- 161.
Vesikari, E., Rautiainen, L., Hkk-Rnnholm, E., Silvennoinen, K.; Salonvaara, M. Viitanen, H.
Julkisivujen ja katteiden kyttin ennakointi. Service life prediction of facades and roofs. Espoo,
VTT Rakennus- ja yhdyskuntatekniikka, 2001. 158 s. VTT Julkaisuja - Publications; 850 (In Finnish,
English abstract).
Viitanen H. 1996. Factors affecting the development of mould and brown rot decay in wooden material
and wooden structures. Dissertation, dep. of forest Products, The Swedish University of Agricultural
sciences, Uppsala, Sweden.
Viitanen, H. 1997a. Modelling the time factor in the development of mould fungi in wood - the effect of
critical humidity and temperature conditions. Holzforschung 51 (1), 6 - 14.
Viitanen, H. 1997b. Modelling the time factor in the development of brown rot decay in Pine ans Spruce
sapwood- The effect of critical humidity and temperature conditions. Holzforschung 51 (1997) 99 -
106.
Viitanen, H. 2005. Tests of resistance and durability of wooden products; an example of evaluation of
service life for wooden faade. COST Action E37. Sustainability through new technologies for
enhanced wood durability. Workshop. Service life prediction: The contribution to Sustainability. Oslo
19th - 21th June 2005. WG2 Performance Seminar - "Service Life Prediction Tools".
Viitanen, H. and Salonvaara, M. 2001. Failure criteria. In Trechsel, E (Ed.). Moisture analysis and
condensation control in building envelopes. 2001. American Society for Testing and Materials ASTM
MNL40, pp. 66 - 80.
Viitanen, H., Ritschkoff, A-C, Ojanen, T., Salonvaara, M. 2003. Moisture conditions and biodeterioration
risk of building materials and structure. Proceedings of the 2nd International Symbosium ILCDES
2003. Integrated Lifetime Engineering of Buildings and Civil Infrastructures, Kuopio, 1-3 Dec. 2003
RIL, VTT, RILEM, IABSE, ECCE, ASCE. Espoo, 151 - 156.
Viitanen, H., Nurmi, A., Mets-Kortelainen, S., Jms, S and Paajanen, L. 2008a. Effect of coatings on
the performance and durability of birch plywood - Results after outdoor weathering and accelerated
decay resistance assessment. IPPS conference 24.-26. September, Helsinki 2008.
Viitanen H., Vinha J., Salminen K., Ojanen T., Peuhkuri R., Paajanen L. & Lhdesmki K. (2008b).
Moisture and biodeterioration risk of building materials and structures. BEST 1 Conference,
Minneapolis, USA, June 10-12, 2008.
Viitanen H., Toratti, T., Peuhkuri, R., Ojanen, T., and Makkonen, L. 2009. Evaluation of exposure
conditions for wooden facades and decking. Document No IRG/WP 09-20408. Stockholm, Sweden.
International Research Group on Wood Protection 18 p.
Wang, C.H., Leicester, R.H. and Ngyen, M.N. 2008. Timber Service life design. Manual No 4. Decay
above ground. CSIRO. 109 p.
Wang, C-H. and Leicester, R.H. 2008. Timber Service life design. Manual No 1. Processed climate data
for timber service life prediction modelling. CSIRO. 144 p.
Wufi (Wrme und Feuchte instationr - Transient Heat and Moisture) 4.1 Pro software, The Fraunhofer
Institute for Building Physics IBP.

94
Sz. Wolinski
Rzeszow University of Technology, Rzeszow, Poland
Service life design of reinforced concrete structures - Alternative
probabilistic approaches
1 INTRODUCTION
There is commonly known statement made by De Sitter in 1984: If no maintenance is carried
out the later repair costs will be five times the saved maintenance costs. If no repair is carried
out, the cost of renovation will be five times the money saved by not repair. Today, design of
concrete structures for durability or more generally for service life has wined wide recognition
and is as important as design for safety and serviceability. More general concept of Service Life
Planning (SLP) of construction works is the subject of ISO 15686:2000 standard series
Building and construction assets - Service life planning and ISO 13823:2008 General
principles on the design of structures for durability. The fib Model Code for Service life
Design (MC - SLD) addresses service life design for plain concrete, reinforced concrete and
pre-stressed concrete structures with a special focus on design provisions for managing the
adverse effects of degradation (fib Bulletin No.34, 2006).
This paper presents some general remarks on SLD of reinforced concrete structures design
for service life with emphasis on probabilistic approach. Phases of the deterioration course due
to carbonation, chlorides and other aggressive agents attack, freeze-thaw deterioration of
concrete, initiation and evolution of the reinforcement corrosion, cracking and spalling of the
concrete cover are discussed in brief. General overview of the modeling of different corrosion
deterioration mechanisms affecting the reinforced concrete elements durability and time to
critical deterioration are presented. To deal with uncertainties involved in modeling
deterioration mechanisms of reinforced concrete and performance decrement, probabilistic
methods are the most suitable. By quantifying durability performance, the reliability measures:
partial safety factors, reliability index or probability of survival are set to make sure of the
safety level. Design service life is defined by relevant limit states, reference period (number of
years), level of probability of not passing limit states during this period and anticipated
maintenance.
Three levels of probabilistic methods are taken into consideration: semi-probabilistic (Partial
Factors Method), simplified probabilistic (First Order Reliability Method FORM and Second
Order Reliability Method SORM) and full probabilistic. In order to guarantee sufficient lifetime
durability for concrete structures, time-variant and fully probabilistic methods should be used.
In some cases time dependent failure modes may be transferred into corresponding time
independent mode for instance: when individual time-varying variables can be modeled by the
distribution of the maximum or minimum value within a given period, in case of the continuous
ergodic processes the outcrossing approach can be applied, e.g. in the form of Rices formula,
application of intermittent processes, finite elements method, spectral analysis for nonlinear
systems, and other special methods. The Monte Carlo simulation technique is recommended as
powerful and friendly tool which makes possible practical applications of the full probabilistic

Application of Degradation Models to Service Life Design
95
approach to SLD. The levels of reliability related to structural resistance, serviceability and
durability can be achieved by combinations of measures relating to: design calculations, quality
management, accuracy of the mechanical models used, detailing, adequate inspection, etc. The
presented approach is illustrated through an example application.
2 CAUSES AND MECHANISMS OF REINFORCED CONCRETE DETERIORATION
Deterioration of reinforced concrete elements and structures is a natural consequence of the
ageing process. Structures are prone to deterioration due to: reinforcement corrosion, chemical
influences, ettringite formation, alkali silica reaction and freeze-thaw effects.
Typical phases of reinforced concrete deterioration for different mechanisms of deterioration
are as follows (Czarnecki & Emmons, 2002):
a) Deterioration due to reinforcement corrosion:
- carbonation or chloride penetration,
- initiation of the reinforcement corrosion,
- evolution of the corrosion (uniform or localized),
- initial cracking of the concrete cover,
- evolution of cracks in the concrete,
- initiation and evolution of concrete spalling,
- reduction of the reinforcement area and bond between reinforcing bars and concrete,
- reduction of the stiffness and resistance of an element (limit states, collapse)
b) Deterioration due to chemical influences:
- acids attack; reaction of acid with cement paste and destruction of its microstructure,
- attack by sulphates; reaction of sulphates with calcium and aluminum ions causes
expansion of its products and formation of cracks,
- ettringite formation; expansion during crystallization and separation of concrete
components,
- alkali reaction with silica aggregates; expansion of reaction products and surface
map cracking
c) Freeze-thaw deterioration due to repeated cycles of freezing and thawing of the free
water in combination with deicing salts or not:
- cracking, scaling and spalling leading to loss of structural integrity and strength of
concrete.
Recently there is generally recognition that damage due to reinforcement corrosion caused by
carbonation and penetration of chlorides are the most important causes of reinforced concrete
structures and a substantial number of research projects have been focused on description of
their mechanisms (DuraCrete, 2000, Rostam, 2001).
3 REMARKS ON MODELLING OF THE DETERIORATION
In most cases mechanisms that causes deterioration of reinforced concrete are based on
penetration of chemicals in concrete. In this paper only corrosion of the reinforcement due to
carbonation and chlorides penetration through the concrete cover, cracking and spalling of the
concrete as well as reduction of the diameter of rebar are considered. The development of
corrosion over time may be described as a multi-stage process shown in Fig. 1. (Bertolini et. al.,
2004).
Corrosion initiation time t
cor
is defined as the period during which carbon dioxide or chloride
ions penetrate the concrete cover and reach the reinforcing steel bars. Using the Ficks law of
diffusion the rate of aggressive agents penetration into concrete as a function of time and depth
can be calculated with:
| | t D x x erf C t x C
c
/ 2 / ( 1 ) , (
0
= (1)

96
where ) , ( t x C is the aggressive agent concentration at the distance x from the surface after time
of exposure t, C
0
is the aggressive agent concentration at the surface, D
c
is the diffusion
coefficient, erf(.) is the error function.
Many different modifications of the formula (1) were proposed in the literature, for example
the Duracrete model ((DuraCrete, 2000):
(
= ]) ) / ( [ 2 / ( 1 ) , (
0 0
t t t KD x erf C t x C
Cl
n
c
(2)
where K is an empirical coefficient (taking account of various influences), t
0
is the reference
time and
Cl
n is the material dependant reduction exponent.
stage of corrosion t
cor
corrosion initiation time
t
crc
crack initiation time
propagation t
spl
spalling of concrete cover time
t
col
collapse time
initiation

time
t
cor
t
crc
t
spl
t
col
Figure 1. Development of reinforcement corrosion in concrete.
Using the concept of concrete cover degradation as the result of corrosion products volume
increase, the time to crack initiation t
ccr
can be calculated from the following equation (Liu &
Weyers, 1998):
) ( / ) ( / ) ( t W t k dt t dW
rust rust rust
= (3)
where W
rust
is the amount of rust products in time t, ) (t k
rust
is the factor proportional to the
annual mean corrosion rate.
Evolution of cracks and reduction of the rebar diameter can assessed using the relationship
between the rate of rust production and time, e.g. equation (3), or by means of semi-empirical or
empirical formulas (Thof-Christensen, 2000).
The existing analytic models of reinforced concrete deterioration contain a number of
parameters. For those parameters a number of indicative values have been given. However,
these values has not yet been sufficiently verified.
4 STATISTICAL QUANTIFICATION OF THE VARIABLES
Many sources of uncertainty are inherent in models of reinforced concrete deterioration. Most
variables and parameters in these models are random variables. Moreover, in many cases they
are time dependant and show spatial variation.
According to survey of published results, distribution types and parameters of more important
random variables determining life service of concrete structures can be summarized as follows
(Akiyama et.al., 2006, Wolinski, 2008a):
- the concentration of aggressive agents at the concrete surface; normal (N) or log-normal
(LN) distribution,
- the diffusion coefficient; N or LN distribution, coefficient of variation v = 0.17 0.91,
- the critical concentration of aggressive agents at rebar surface; N distribution, v = 0.17
0.375,
- the thickness of concrete cover; N distribution, v = 0.20 0.50,
- the corrosion initiation time t
co
; Weibulls distribution (W), ~ v 0.40, coefficient of
skewness ~
1
+1.8,
- the crack initiation time t
crc
; W distribution, ~ v 0.60, ~
1
+ 0.20,
- the spalling of concrete cover time t
spl
; W distribution, ~ v 0.45, ~
1
+ 0.50.

97
In most cases mechanisms that causes deterioration are based on penetration of chemicals in
concrete. Unfortunately, the variation in concentration of chemicals (e.g.
Cl ions and
2
CO ),
the variation in microstructure of concrete and the variation in environmental conditions
(temperature, relative humidity, wind, solar radiation, etc.), as well as changes in climate during
the lifetime of structures, play an important role in the analysis of reinforced concrete durability.
In Fig. 2 scatter in chloride penetration between cores taken from 1m
2
of a quay wall is shown,
according to CUR Committee B82.
Chloride as % of cement weight
3,0
2,5
2,0
1,5
1,0
0,5
0,0
0 10 20 30 40 50 60
depth [mm]

Figure 2. Scatter in chloride penetration within an area of 1m
2
(according to CUR Comm. B82).
Many parameters in degradation models show spatial and time variability that is usually linked
to dependencies on temperature, water binder ratio, humidity and workmanship. Spatial
variability of physical properties includes systematic and random variation. The influence of the
spatial variation of physical quantities can lead to significant decrease in structural reliability
and in service life of concrete structures (Li & Vrouwenvelder, 2007). The background of
spatial variation in chemical is not yet well known and should be further investigated. The full
description of the spatial random field of a considered quantity requires modeling with respect
to its correlation structure in space. The homogeneous Gaussian field may be a helpful tool in
the statistical description of the fluctuation pattern of spatial random properties. To describe
such a type of field we need a value for the mean, a value for the standard deviation and the
correlation function. Unfortunately, there are no reliable test results to estimate these parameters
in practice. Time variability of a random quantity considered in the service life design of
concrete structures has been studied in two aspects; as variability connected with concrete age
(Meijers, 2003, Rooij & Polder & Oosten, 2007), and as variability connected with changes in
climate (Nijland et. al., 2009). By way of example, Fig. 3 shows the variation of the chloride
ions diffusion coefficient as a function of concrete age, from extensive testing in 15 laboratories
(according to Rooij & Polder & Oosten, 2007).
Changing in climate parameters, such as a temperature, amount and intensity of precipitation,
evaporation, wind speed, solar radiation, etc., will affect a random properties and durability of
concrete and reinforcing steel, and should be considered in an extensive analysis of service life
of reinforced concrete structures. It is remarkable that in last years the concentration of CO
2
,
which is an important factor causing corrosion of reinforcement due to carbonation of the
concrete cover, is showing tendency to rapid increase (Fig. 4).

98

Figure 3. Variation of the diffusion coefficient of the
Cl ions as a function of concrete age. Solid line is

the the best fit (according to Rooij & Polder & Oosten, 2007).

Figure 4. Concentration of
2
CO in parts per million (ppm) as a function of time (fib Bulletin, 2006).
5 DESIGN FOR SERVICE LIFE
In the fib Model Code for Service Life Design (SLD) contemporary ideas and procedures
referred to service life design of structural concrete are presented with a special focus on design
provisions for managing the adverse effects of degradation (fib Bulletin No.34, 2006). Four
different options for SLD are recommended:
a) Avoidance of deterioration; it means that sensitivity of the structure should be reduced
by:
- application of materials with low sensitivity to deterioration,
- application of coating or membranes,
- prevention against damaging reactions
b) Deemed to satisfy rules (recipe based approach); the design criteria that should be
fulfilled guarantee a specified service life; for instance: minimum rebar cover, proper
composition of concrete, limited cracks width, etc.
c) Semi-probabilistic approach (partial factor design):

99
- reliability level of a structural element is checked by means of partial safety factors;
d k m
f f / =
and
k d f
S S / =
,
d k R
R R / = , where: f
k
,
d d d k k k
S R f S R f , , , , , are
characteristic and design values of material properties, resistance and actions,
- within the partial factors design the effect of deterioration of structural components is
usually considered by multiplying the partial safety factor
R
by modification factors
corresponding to different types of deterioration processes and the degree of
knowledge about the future degradation:
M P D R R
=
0
(4)
where
0
R
is the initial value of partial safety factor corresponding to a certain design

service life and a specific safety class, are
, ,
D P M

the modification factors
corresponding to a given deterioration process, protection and maintenance strategy.
d) Simplified and full probabilistic approach; these methods are discussed in the next
chapter.
For the practical application of semi-probabilistic and probabilistic methods, service life
design should be defined by:
- relevant limit states and limit states criteria,
- a reference period (number of service life years),
- a definition of the end of service life,
- a target value of reliability (or safety) measure, - a strategy for maintenance and repairs,
quality system control.
6 PROBABILISTIC METHODS OF DESIGN FOR SERVICE LIFE
There are at least three most important reasons for application of probabilistic methods of
design for service life:
- the variation in properties of the concrete and its microstructure, the variation in the
environmental conditions (relative humidity, temperature, wind, etc.) and variation in the
concentration of chemicals penetrating through concrete (e.g.
Cl ions and
2
CO ),
- the resistance ) (t R , loads and environmental actions as well as their effects ) (t E are
random functions of time (stochastic processes),
- the failure probability of an individual section or element usually are usually not
representative for the failure probability of the entire structural system.
Service life of a structure is usually defined (Rostam, 2001) as the time till the deteriorating
structure can carry loads with specific degree of reliability (see Fig. 5).

Figure 5. Illustration of probabilistic determination of service life.
Probabilistic methods can be used for structural design when the set of random events and
variables can be identified. Two levels of probabilistic methods are usually distinguished:
simplified probabilistic methods (First Order Reliability Method FORM and Second Order
Reliability Method SORM) and full probabilistic method (Nowak & Collins, 2000, ISO 2394,
1988).

100
a) Simplified probabilistic methods
Commonly known and recommended in contemporary standards (ISO 2394, 1988, EN 1990,
2002) simplified probabilistic method is FORM which is usually identified with the reliability
index method ( | - index theory). Reliability index is defined as:
Z Z
m o | / = where:
E R Z = is a performance function, R is the resistance, E is the action effect,
Z
m and
Z
o are
the mean value and the standard deviation of a random variable Z, Using the reliability index
measure, the influence of uncertainty can be taken into consideration by means of the resistance
correction (Thof-Christensen, 2000, Faber & Melchers, 2001):
2 2 2
2 1 2 1
) ... ( / ) ... (
E R n E R n
m m o o | + = (5)
where:
E R E R
m m o o , , , are the mean values and standard deviations of the resistance and
action effects,
n
,..., ,
2 1
are correction factors corresponding to a given deterioration
processes, quality of execution, protection and maintenance strategy, etc.
The |-index design method is based on two assumptions: the probability of failure should be
equal for each structural element and partial safety indices should be proportional to sensitivity
factors of basic variables. Three optional approaches: distribution-free First-Order Second-
Moment method (FOSM), First-Order Reliability Method (FORM), Second-Order Reliability
Method (SORM) and Approximate Full Distribution Approach (AFDA) can be applied in
practical design using the |-index method. However, this method can give nonsensical results in
some quite real cases. Other formally correct simplified probabilistic methods have also been
presented and discussed in the literature. For instance, the k-index method which is based on the
following assumptions: the probability of failure should be minimal for each structural element
and partial safety factors for any particular load should be equal to the load effect factor of
simultaneous load combination and a partial safety factor of any component material should be
equal to the resistance factor of the whole composite element (Murzewski, 1998). Nevertheless,
the conventional |-index method is well developed, commonly known in the world and
recommended in the ISO and European standards.
b) Full probabilistic methods
In the case of service life design of reinforced concrete reliability aspects are time dependent
as a result of time dependant degrading mechanisms and fluctuating actions. The point in time
probability of failure ) (t p
f
at time t in those cases can be formulated as:
} 0 ) ), ( ( { ) ( s = t t X g P t p
f
or
}
s
=
0 ) , (
) ( ) ( ) (
t X g
X f
t dx t x f t p (6)
where )) ( ( t X g is time dependent limit state function and X(t) is stochastic process.
In time dependent service life design problems, interest usually lies in estimating the
f
p -
value over a time interval from 0 to an assumed t
sl
. For at last one case of failure in the time [0,
t
sl
]:
0 ) ), ( ( { ) ( s = t t X g P t p
sl f
(7)
The ) (
sl f
t p - value could be obtain by integrating ) (t p
f
over the time interval [0,
sl
t ], taking
into account inter-correlation characteristics in time of the process X(t). Time dependency
results from time dependent actions, including deterioration of reinforced concrete. Typical
models of time dependency are as follows: continuous and differentiable process, random
sequence, point pulse process with random intervals, rectangular wave process with random
intervals, etc.
In some cases time dependent failure modes may be transformed into a corresponding time
independent mode. There are known several methods of such a transformation that depend
usually on the models of time dependency of major random variables (Karadeniz &
Vrouwenvelder, 2003).
- When the mechanisms of deterioration are known in advance and can be expressed in the
form of degradation function ) (t g describing evolution of the structural deterioration in
course of time, the time-variant resistance R(t) may be defined by:

101
) ( ) (
0
t g R t R = (8)
where
0
R is the initial resistance, and the full probabilistic approach to durability assessment
may be applied. An illustrative example presented in chapter 8 shows a practical application of
this method. Preventive and mitigating measures can also be introduced by differentiation the
time of degradation initiation, and the maintenance or repair may be taken into account by
modification of the initial resistance.
- In case where an individual time-varying variable is the random process the failure
probability ) (t p
f
, can be modeled by the distribution of the maximum or minimum
value within an assumed period of service life
sl
t :
)} ( { max t p X
f sl
= or )} ( { min t p X
f sl
= (9)
For continuous processes, the largest extreme may be approximated by one of the asymptotic
extreme value distributions. In situations where two or more variables are simultaneously
random processes of time, the Turkstras rule suggests that approximately the maximum value
of the sum of some independent processes attains its maximum value (Nowak & Collins, 2000):
+ + +
+ + +
+ + +
=
)} ( { max ... ) ( ) (
... .......... .......... .......... .......... ..........
) ( ... )} ( { max ) (
) ( ... ) ( )} ( { max
max
2 1
2 1
2 1
t p t p t p
t p t p t p
t p t p t p
X
nf sl f f
nf f sl f
nf f f sl
(10)
- In case where an individual time-varying variable is the ergodic process X the out-
crossing approach can be applied to calculate the up-crossing frequency in service life
time ) / (
sl
t N v . For a continuous Gaussian process the Rices formula can be applied:
)
2
exp( ) 0 (
2
1
) / (
2
' '
k
t N v
sl
=
t
(11)
where
X
is the correlation function and
2
/ ) (
X
X N k = .
In cases where a time-varying variable are the nonergodic process or/and in a so called
hierarchical case where the process contains constant in time or/and slowly and rapidly varying
parts, the out-crossing approach can be also applied but the calculations are quite cumbersome
and the upper-bond approximation may be often satisfying.
- Other methods are also applied for analysis of time dependent reliability problems,
among other things: nested reliability method, intermittent processes, spectral analysis,
finite element methods (Karadeniz & Vrouwenvelder, 2003).
c) Techniques of probabilistic calculations
Several techniques may be used to calculate the failure probability related to structural
reliability problems. Analytical integration can only be eligible in some simple cases. Numerical
solutions are based on application of a transition probability density on closed form valid for
small time increments and are feasible only for a limited number of random state variables.
However, application of the simulation-based Monte Carlo methods make possible to obtain
approximate solutions in case time-variant problems for highly nonlinear performance
functions, unknown in advance area of integration and large number of random variables taken
into account (Nowak & Collins, 2000, Wolinski, 2008a, Wolinski, 2008b, Karadeniz &
Vrouwenvelder, 2003).

102
d) Element and structure reliability
In the practical design of structures we focused almost exclusively on the calculation of the
failure probability of individual cross sections or elements of a whole structure. Since most
structures consist of a finite number of interconnected sections and elements, the failure
probability of an individual section or element usually may not be representative of the failure
probability of the entire structural system. From mathematical point of view, a structure forms a
system of structural elements. Three idealized types of systems can be distinguished: series,
parallel and mixed or combined. Each type of the structural system may consists of brittle
or/and ductile elements which may be correlated or/and uncorrelated.
Most structures should be considered as combined systems consist of elements or groups of
elements connected in series and in parallel. The mixed system fails either by a failure of
anyone of elements connected in series or by a failure of all the elements connected in parallel.
The entire system can be described by a finite (but usually very large) number of the paths of
failure. The system is efficient if at last one of the paths is efficient. It should be noticed that the
effect of correlation on the probability of failure of a structural system is important in many
cases (Nowak & Collins, 2000, Wolinski, 2008b).
e) Quality of execution
Every design methods should take into account a number of factors which influence the
service life. Some more important factor listed in sequence with their occurrence are: the
conceptual design, the environmental conditions, the choice of structural materials, detailing,
execution, maintenance, the recorded data quality, etc. Results of the initial inspection and data
recorded during the life of a structure are essential. It is advised to test the condition of a new
structure and to record the corresponding data in something of the birth certificate sort
(Walrawen, 2008). Tests, inspections and study results relating to realization of random
variables may be used for updating the available data by assuming the distribution parameters of
the distributions used in the probabilistic modelling. New results should be utilized to update the
probability distribution functions of these distribution parameters. The prior distribution
functions may be updated by Bayesian reasoning.
7 ACCEPTABLE PROBABILITY OF EXCEEDING LIMIT STATES
In both the simplified probabilistic and full probabilistic methods the measure of reliability is
identified with the survival probability P
s
= 1 P
fd
, where P
fd
is the target probability of failure
which depends on the category of limit state (ultimate and serviceability limit states: ULS and
SLS), various design situations, failure modes and reference periods.
In the Eurocode (EN 1990, 2002) recommended values of target probability depend on the
reliability class of a structure (RC1, RC2, RC3), the length of reference period (1 year and 50
years) and are related to the consequence classes (CC1, CC2, CC3). In the ISO standard (ISO
2394, 1998) target probabilities are referred to the relative costs of safety measures and
consequences of failure. The numerical values of
fd
P are usually described as formal or
notional numbers, intended as a tool for developing consistent design rules and not as a
description of the frequency of structural failures and established on the basis of the past
experience.
A target value of reliability measures in for service life design should be assumed with the
reference to relevant limit states. In the MC-SLD several examples of limit states connected to
the corrosion of reinforcement are given: depassivation of reinforcement, formation of the
corrosion crack, spalling of the concrete cover, reduction of the reinforcement area, reduction of
the bond between reinforcing bars and concrete, collapse(fib Bulletin No.34, 2006).
Recommended target values of the reliability index | and corresponding target values of
failure probability P
fd
according to the MC-SLD are presented in Table 1.

103
Table 1. Target values of the reliability index | (and P
fd
- values) for the durability limit states (fib
Bulletin No.34, 2006).
Environment Action Reliability SLS ULS
class class Depassivation | (P
fd
) Collapse | (P
fd
)

XC Carbonation RC1 1.3 ( ~
1
10
) 3.7 ( ~
4
10
)
XD Deicing RC2 1.3 ( ~
1
10
) 4.2 ( ~
5
10
)
XS Sea Water RC3 1.3 ( ~
1
10
) 4.4 ( ~
6
10
)

8 ILLUSTRATIVE EXAMPLE
Consider a simply supported reinforced concrete slab subjected to the combination of dead and
live loads and to deterioration such as: wear due to erosion of the upper surface of concrete,
corrosion of reinforcing steel, cracking and spalling of concrete cover. Using the probabilistic
approach, calculate the area of reinforcement steel
s
A in the critical cross-section of the slab.
- Input data:
- dead load g: N (5, 0.5) kN/m,
- long lasting live load p: G (5, 1.75, 0.8) kN/m,
- compressive strength of concrete f
c
: LN (28, 4.86) N/mm,
-
-
- yield stress of reinforcing steel f
y
: N (462, 35.7) N/mm
2
,effective depth of the cross-
section d
0
: N (130, 6.5) mm,
- concrete cover c: N (30, 3.0) mm,
- coefficient of carbonation : LN (4, 0.82),
- empirical coefficien of concrete wear : N (3.8, 0.33),
- effective span: L
eff
= 3.0 m.
Types of distributions: N - normal, G - gamma, LN - log-normal.
Characteristics of random variables: the mean value, the standard deviation and the
coefficient of skewness, and they are represented by truncated bounded histograms.Physical
modeling; three models for the three mechanisms of deterioration were defined:
- erosion wear of the concrete surface t d t d k =
0
) ( ,
- corrosion initiation time
2
) / ( o d t
cor
= ,
- crack initiation time
rust r crc
k W t 2 /
2
= , symbols according to formula (3),
- reduction of the rebar diameter t i c d t d
cor cor
=
0
) ( , where
cor
c and
cor
i are
empirical coefficients,
- load effect combination
2
) ( 125 . 0
eff
L p g M E + = = ,
- resistance at the critical section bd f A M R
y s
= = , where b = 1 m, is the relative
value of the effective depth of compression zone.
- Probabilistic calculation were performed using the Monte Carlo simulation method.
Results in the form of relationships between the area of tension reinforcement area
required with the probability of collapse
4
10
~
fd
P ( 7 . 3 ~ | , for environment class XC,
carbonation and ULS of collapse) and time are presented in Fig. 3. The mean values of
characteristic moments of the slab service time are: the corrosion initiation time is
9 . 26 =
cor
t years, the crack initiation time is 36 =
cor
t years and the spalling of concrete
cover time is 5 . 53 =
spl
t years.

104
0 10 20 30 40 50 [YEARS]
TIME
900
800
700
600
500
553
610
632
654
664
763
890
(663)
(683)
(704)
with deterioration
of concrete
with corrosion of steel
and deterioration
of concrete
A (t,P =7.2 x10 )
s f
As [mm ]
[YEARS]
t =26.9
2
-5
cr
Figure 6. Diagram of the tensile reinforcement area required with the probability of collapse 10
-4
and time
of service life for the reinforced slab in bending.
The probability of failure as a function of service time for the slab with the initial reinforcement
area 552.8 mm
2
is shown in Fig. 7. Service time of the considered slab is 32 years and P
f
[t =
32] = 0.005 < P
fd
= 0.000072 ( 80 . 3 645 . 1
2
= < =
RC
| | ).
P
[YEARS]
TIME
0 10 20 30 40 50
t =26.9
carb
1.0E-01
1.0E-05
1.0E-04
1.0E-03
1.0E-02
5.7E-02
2.7E-02
5.0E-03
2.0E-03
1.1E-03
5.0E-04
7.2E-05
f

Figure 7. Diagram of the failure probability versus service time for the reinforced concrete slab subjected
to the standard loads and to the deterioration processes.
9 CONCLUSIONS
Recently there is generally recognition that the design of concrete structures for service life has
gained the same level of importance as design for reliability and serviceability. Economic,
social and environmental aspects should be involve in the design for service life procedures.
Moreover, all these aspects should be addressed explicitly or in other words, they should be
expressed clearly and fully quantitatively.
The effects of scatter in deterioration of reinforced concrete should be treated in a similar way
as mechanical failure mechanisms and should be discussed, studied and regarded in
deterioration models.
The quality of execution has a large influence of on service life of concrete structures. The
formulation of demands for initial and further controls as well as data recording and updating
should get high priority.

105
In order to guarantee sufficient lifetime durability for concrete structures, time-variant and
fully probabilistic methods are the most suitable. Monte Carlo simulation procedures that allow
for direct, transparent and efficient applications of probabilistic approach are recommended for
the numerical calculations. However, for a large part of standard and simple structures the
deemed to satisfy rules or semi-probabilistic method (partial factor design) can be sufficient.
REFERENCES
Akiyama, M., Suzuki, M., Matsuzaki, H., Dang. T.H., Li, Y.F. 2006. Application of reliability theory to
the durability-design of concrete structures subjected to corrosion. Proceedings of the 2
nd
fib
International Congress. June 5-8, 2006. Naples, Italy (full version on attached CD-room).
Bertolini, L., et. al. 2004. Corrosion of steel in concrete: prevention, diagnosis, repair. Wiley-VCH
Verlag GmbH & Co. KgaA, Weinheim, ISBN 3-527-30800-8, 392 pages.
Czarnecki, L., Emmons, P.H. 2002. Structural concrete repair and maintenance. Polski Cement,
Krakow, 434, (in Polish and in English).
CUR Committee B82: Durability of concrete structures in marine environment. CUR, Gouda, The
Netherlands, 2002.
De Sitter, W.R. 1984. Costs for service life optimization: The law of fives. Workshop Report on
Durability of Concrete Structures; Copenhagen, 18-20 May 1984. Ed. Steen Rostman, Copenhagen:
131 - 134.
DuraCrete - Brite EuRam III 2000. DuraCrete, Final Technical Report: Probabilistic performance based
durability design of concrete structures. The European Union; Project BE95-1347, Gouda, May2000.
EN 1990:2002 E. Eurocode - Basis of structural design. Brussels: CEN, 2002.
Faber M., Melchers, R. 2001. Aspects of safety in design and assessment of deteriorating structures.
Conf. Report, Int. Conf. on Safety Reliability. Malta, 21-23 March 2001: 161 - 166.
fib Bulletin No. 34. 2006. Model Code for Service Life Design. EPFL Lausanne, 2006.
ISO 15686: 2000. Buildings and constructed assets - Service life planning. Part 1: General principles.
Part 2: Service life prediction procedures.
ISO 13823: 2008 (E). General principles on the design of structures for durability.
ISO 2394: 1988(E). General principles on reliability for structures.
Karadeniz, H., Vrouwenvelder, T. 2003. Overview reliability methods. SAFERELNET Task 5.1, TU
Delft, The Netherlands, October 2003.
Li Y., Vrouwenvelder T. 2007. Service life prediction and repair of concrete structures with spatial
variability. Heron Journal, Vol. 52, No. 4, 2007, 251 - 267.
Meijers S. 2003. Computational modelling of chloride ingress in concrete. PhD thesis, TU Delft.
Murzewski J. 1998. A more fundamental reassessment of probabilistic design methods. Proceedings of
the Eigh IFIP WG7.5 working conference on reliability and optimisation of structural systems, 1998.
The University of Michigen, Ann Arbor, USA, May 11-13, 1998, 39 - 50.
Nijland T., Adan O., van Hees R., van Etten B. 2009. Evaluation of the effects of expected climate
changes on the durability of building materials with suggestion for adaptation. Heron 54, No. 1, 2009,
37 - 48.
Nowak, A., Collins, K. 2000. Reliability of Structures. Int. Edition: McGraw-Hill: 338.
Rooij de M., Polder R., Oosten van H. 2007. Validation of service life performance of in situ concrete by
TEM and RCM measurement. Heron Journal, Vol. 52, No. 4, 2007, 225 - 238.
Rostam, S. 2001. Performance based design of concrete structures: the challenge of integrating strength,
Durability and sustainability. Proceedings of fib Symposium Concrete and Environment. Berlin 3-5
October.
Thoft - Christensen, P. 2000. Modeling of the deterioration of reinforced structures. Proc. 9
th
IFIP WG
7.5 Working Conference on Reliability and Optimization of Structural Systems. The University of
Michigan, Ann Arbor: 15 - 26.
Walrawen, J. 2008: Design for service life: a challenge for the future. AMCM2008, 6
th
International
Conference Analytical Models and New Concepts in Concrete and Masonry Structures. Lodz,
Poland, 9-11 June, 2008: 187 - 198.
Wolinski, Sz. 2008a. Probabilistic methods for durability assessment of reinforced concrete structures.
Accepted for publication: chapter in the monograph of Polish Academy of Science (in Polish).
Wolinski Sz. 2008b: Risk-based design and assessment of concrete structures. AMCM2008, 6
th

International Conference Analytical Models and New Concepts in Concrete and Masonry Structures.
Lodz, Poland, 9-11 June, 2008: 345 - 346.

106
L. Cascini, F. Portioli, R. Landolfo
Service Life Design of Metal Structures Based on Lifetime Safety
Factor Method
1 INTRODUCTION
Sustainable Development (SD) was defined for the first time in the last 80s in the Brundtland
Report (WCED 1987) as the development that meets the needs of the present without
compromising the ability of future generations to meet their own needs. In the beginnings the
strategic objectives just focused on environmental protection, while nowadays the field of
interest has greatly enlarged, including also economic and social concerns.
Among different human activities, which interact with the three SD dimensions, the
construction sector plays an important role in delivering sustainable development requirements.
Review of literature shows that the definition of a sustainable construction implies several
statements, such as the efficient use of raw materials, the minimum use of energy and emissions
during service life, the life duration and robustness, and a prolonged service life as target (CIB
1999).
Metal constructions can easily satisfy the above mentioned requirements thanks to different
features. They are recyclable and have high structural efficiency, design and manufacturing
flexibility and speed of building. Such advantages provide lower raw material consumption,
facilitate changing requirements avoiding obsolescence, and reduce the construction activities
impact on the local environment by reducing emissions and noises.
Despite all the previous advantages, metal structures present some drawbacks, which are
mainly related to durability, whose evaluation is necessary for a suitable maintenance and
rehabilitation planning during life span.
Among several factors that affect the life duration of metal structures, the atmospheric
corrosion is recognized to be one of the major risk which decreases the performance of
constructions, resulting in huge economic and societal losses.
In terms of structural effects, the atmospheric corrosion causes the thickness loss of the cross
section that leads to a smaller resistant area. The loading capacity of the element itself is
reduced and the safety margin rapidly decreases.
Besides the effects on coatings, which could modify the exterior appearance of constructions
reducing their aesthetic and economic value, corrosion represents an additional critical load
which is able to lead the structures to collapse, specially when it is associated with high stress
rates and cyclic loads, such as in the cases of stress and fatigue corrosion (Figure 1).

107

Figure 1. Collapse of the Silver Bridge on the Ohio River due to stress and fatigue corrosion (USA,
1967).
With regard to life duration, even if international standards and codes provide general
recommendations for preventing early ageing of constructions, the durability design of metal
structures has still to be fully defined and codified. In particular, a procedure for the evaluation
of original and residual service life should be developed for each material under assigned
environmental loads on the basis of available degradation models.
This paper represents a first attempt to propose and develop a durability design procedure for
both new and existing metal structures with the aim of evaluating the residual service life of
constructions with respect to corrosion degradation.
2 DURABILITY CRITERIA IN STANDARDS AND CODES
The main European standards and codes on constructions give only few qualitative and common
provisions for durability of metal structures, that are mainly concerned with coating corrosion or
with general recommendations on structural material redundancy. No specific design procedure
is provided for the evaluation of design life time of constructions under the identified
environmental loads.
A specific definition of durability with respect to the corrosion of metal structures is provided
in ISO 8044 (1999), stating that durability is the capability of corrosion system, that is the metal
itself or the coating, to fulfill the serviceability requirements for a specific period of time, when
adequate maintenance actions are performed.
In EN 1993-1-1 (2004) only few common principles are stated for durability of metal
structures and in particular for preventing steel buildings from possible causes of corrosion
damage. The code refers to EN 1990 (2001) for durability in general and gives some
recommendations such as the opportunity of providing corrosion protection measures by means
of surface protection systems, improving the use of weathering or stainless steel and resorting to
structural redundancy. However, in such a case no references are made to models able to
estimate the corrosion depth as a function of time and of the different factors influencing the
degradation rate.
Even if durability is analyzed in general terms, it should be noted also that in some codes for
metal structures (DM 2005), an important innovation is being introduced concerning corrosion,
which is expressly included among the different loads acting on constructions. In these
standards corrosion is classified as a type of entropic load, which comprises deteriorating
actions, caused by natural degradation mechanism of materials, and environmental loads, which
affects the structural integrity.
Some specific references to durability are reported in the international standards with respect
to life duration of the different coating protective systems.
EN ISO 12944-1 (1998) sets three durability classes with regard to protective paint system. In
particular, durability means how long the protective coating is effective before a maintenance
provisions have to be performed. The standard defines three durability classes: Low (from 2
years up to 5), Medium (from 5 years up to 15) and High (more than 15 years).

108
In EN ISO 14713 (1999), the life duration of zinc and aluminium coatings is related both to
thickness loss and corrosiveness of environment. In particular, specific recommendations are
given for each corrosivity class with respect to different coating typologies. However, it should
be noted that in this case the design life of coating thickness is evaluated on the basis of linear
extrapolation of corrosion rates per year.
Many other references (EN ISO 9226 1992) can be found in international standards, but a
design procedure has still to be codified for predicting and preventing the potential damage that
a specific environment could lead to both coatings and structural materials, during the entire
service life. Taking into account the lack of codified durability design procedures, the
application of the life time safety factor method to metal structures is discussed in the following.
3 DURABILITY ANALYSIS BASED ON LIFETIME SAFETY FACTOR METHOD
The lifetime safety factor method in durability design was first time presented in the RILEM
Report on concrete structures (Sarja & Vesikari 1996) and then developed within the framework
of the EU Project LIFECON (Sarja 2004).
The method is used for the calculation of design life of constructions and it is based on
probabilistic degradation models, which consider the decrease of structural resistance caused by
different classes of environmental loads. It is so called because of the safety factor applied to the
average or the characteristic value of design life. The latter is calculated by degradation models,
taking into account both the statistical values of resistance and loads and it is calibrated on the
maximum allowable failure probability related to the considered limit states.
Such design approach belongs to semi-probabilistic methods. In this case, both the resistance
R and the action effects E are considered as independent random variables. The failure
probability P
f
is conventionally defined by the reliability index |, which is related to P
f
by the
equation P
f
= (|), where is the cumulative distribution function of the standardized normal
distribution. According to semi-probabilistic methods, the design values of the basic variables
X
d
and F
d
influencing both R and E are introduced with their characteristic values divided or
multiplied by partial factors. Such partial safety factors are calibrated in order to satisfy the
equation P
f
P
*

f
when E
d
R
d
, where P
*

f
is the maximum allowable failure probability
relevant to the considered limit state.
The durability design based on lifetime safety factor method is analogous with the static limit
state design. In particular, it is related to control the failure probability by considering the
effects on R of the environmental loads acting during the entire life time cycle, while static limit
state design is devoted to control the structural reliability of constructions under external
mechanical loading. In durability limit state design the resistance R is considered as a time
dependant variable, contrary to static limit state design, where the effects of time are usually
neglected for R (Figure 2).
In particular, a deterioration function D can be formulated on the basis of the time dependant
resistance R according to formula:
D(t)= R(0)-R(t) (1)
where: D(t) is the deterioration at time t; R(0), R(t) are respectively the resistance at t = 0 and at
the generic time t of the life cycle.
Because of the different sources of uncertainties that are involved in the definition of the
variation of capacity with time, the values in the previous equation are usually taken as mean or
characteristic values. The load S is usually adopted to be constant with time, and its design
value is also taken as a mean or characteristic value, multiplied by the relevant safety factors.
The failure event corresponds to time t
max
when the capacity R is equal to load S. The difference
R(t)-S represents the reduction with time of the safety margin:
R(t
max
)=S (2)
On the basis of previous considerations, a durability design procedure organized into different
steps can be formulated, as specified in the following.

109

Figure 2. Representation of the life time safety factor design approach by R, S, t and D, t variables.
First, the target service life t
g
has to be defined for the considered constructions. The target
service life must be specified as a basis for assessing statistically variable actions and to
evaluate the reliability with respect to durability. Target service life is assumed to be the time
period for which a structure or part of it is to be used for its intended purpose with anticipated
maintenance but without major repair being necessary (EN 1990 2001). The numerical
reference values are selected according to common standards, regulations and codes. The
standard EN 1990, with respect to target service life, defines different indicative values on the
basis of the construction characteristics, typology and use (e.g. 50 years for building structures
and other common structures, 100 years for monumental building structures, bridges, and
others). The reference value of the target service life has to be selected according to mechanical
design procedure.
Once the reference period has been stated, it is possible to identify the environmental loads S
that will likely act onto structure. Each environmental load is analysed and quantified, where
relevant, in a statistic way. The analysis of the environmental condition has to be performed in
order to define the project background. With regard to atmospheric corrosion of metal structures
the identification of both the climatic conditions (such as temperature, rain, condensation of
moisture, freezing, solar radiation and air pollution), and the geological conditions (such as the
location of ground water, possible contact with sea water, contamination of the soil by
aggressive agents like sulphates and chlorides) has to be provided.
On the basis of the identification of environmental loads, the degradation factors and
mechanisms should be evaluated.
Once deterioration mechanisms that could act onto structures during the life cycle have been
identified, corresponding damage curves should be considered as a function of time in the form:
D
m
(t) = t
n
(3)
where: D
m
(t) is the mean value of degradation; is a constant coefficient; t time; n degradation
mode coefficient.
Substituting t
max
in eq. (3), we have:
D
m
(t
max
) = D
max
(4)
where D
max
= R(0)-R(t
max
) = R(0) - S

represents the maximum allowable value of degradation
(e.g. the maximum allowable mass loss and/or corrosion depth in the cross section of a beam).
Durability requirements are fulfilled if the failure event (4) occurs after the design service life
had expired, with a proper safety margin. That could be expressed according to formula:
t
d
= t
max
/
t0
> = t
g
(5)
where: t
d
is the design service life; t
max
is the calculated mean value of the service life
corresponding to D(t
max
) = D
max
;
t0
is the central lifetime safety factor; t
g
is the target service
life.

110
Assuming that degradation is normally distributed and the standard deviation of D is
proportional to the mean degradation, the coefficient of variation V
D
being constant, it can be
shown that the central lifetime safety factor of the design life depends only on safety reliability
index, the coefficient of variation of D and the exponent n, according to the formula:
t0
= (V
D
+1)
1/n
(6)
On the basis of such assumptions, the safety reliability index depends on both the maximum
allowable failure probability for the selected limit state and the degradation function. In
particular, it is a function of the mean values and standard deviations of R and S, as follows:
( ) ) /(
2 2
S R S R
o o | + = (7)
In the following sections, all the phases of the durability design procedure for metal structures
are discussed in detail with respect to degradation due to atmospheric corrosion.
4 ENVIRONMENTAL LOADS: THE ATMOSPHERIC CORROSION
Corrosion is defined as a deterioration of metal materials that results from a reaction with its
environment (NACE 2002), causing the degradation of both.
The deterioration is caused by a chemical and/or electrochemical attack of the metal surface.
Corrosion processes are influenced by several factors, that could be divided in two broad
classes: the endogenous and the exogenous factors. The first ones are related to the metal itself,
e.g. the composition of the metal, the chemical and physical homogeneity of the surface. The
exogenous factors are associated to the atmospheric composition, such as temperature, relative
humidity and concentration of pollutants (Landolfo & Di Lorenzo & Guerrieri 2005).
Depending on the physical and chemical features of the environment, the corrosion
phenomena could develop in different forms, and may be uniform or localised.
The atmospheric corrosion is mainly an electrochemical process that occurs when a thin
electrolytic layer is formed on the metal surface; the rate at which the corrosion attack proceeds
is strictly related to the corrosiveness of the environment. Relative humidity rate, the pollutants
(sulphur dioxide, sodium chloride, ammonium sulphate, etc..) the airborne particles, dust, the
weather condition, the wind and the temperature average are the main features that characterize
the outdoor atmosphere.
5 CLASSIFICATION OF ENVIRONMENTAL LOADS
Within a durability design method, the classification of environmental loads is necessary for
evaluating all the factors which affect the degradation rate during life time of constructions.
Several classification systems are provided by standards to assess the corrosiveness of
environment. The codes usually provide the corrosivity of the environment on the basis of the
mass loss after one-year exposure. It should be noted that these values represent the result of the
corrosion attack after one-year and they cannot be extrapolated to estimate the extent of damage
in a time t of the service life.
EN ISO 9223 (1992) defines five corrosivity classes C1-C5 on the basis of three key factors:
the TOW (time of wetness), the deposition rate of chlorides and sulphures dioxide. EN 12500
(2000) defines the corrosiveness of the environment, according to ISO 9223, by assessing the
mass loss of standard samples, after 1 year exposure and it establish five corrosivity categories
from C1, very low corrosivity environment, up to C5, very severe corrosiveness atmosphere.
EN 12500 recommends to evaluate the corrosiveness of the environment quantitatively on the
basis of the sample exposure but whether impossible, it could be used to assess qualitatively the
atmospheric category. The standard points out that the qualitative classification of the
environment could mislead designers in selecting protective measures, field test exposure are
strongly recommended. According to EN 12550 the designer should run field test on standard
carbon steel, zinc, copper and aluminium samples measuring, after 1 year exposure, the mass

111
loss of the samples. The corrosiveness class of the local environment is thus established
comparing the measured values with the guiding values reported in the following Table:
Table 1. Mass loss (g/m
2
) for one year field test exposure. EN 12550.
MASS LOSS g/m
2

Carbon Steel Zinc Copper Aluminium
Corrosiveness category
10 0.7 0.9 Negligible Very low C1
10 200 0.7 - 5 0.9 - 5 0.6 Low C2
200 400 5 - 15 5 - 12 0.6 - 2 Medium C3
400 650 15 - 30 12 - 25 2 - 5 High C4
650 - 1500 30 - 60 25 - 50 5 - 10 Very high C5

As an alternative to the previous procedure, the corrosiveness of the local environment could be
assessed qualitatively on the basis on climatic and/or environmental parameters and/or
construction location (EN 12500 2000). The qualitative classification of the atmospheres ranges
from Rural to Marine industrial types and it is based on a general description of corrosive agents
and degradation rates.
6 DEGRADATION MODELS
Several degradation models can be found in literature which have been defined for metallic
materials. These models provide the corrosion depth or the mass loss with time for different
corrosive environments. In general, they are formulated according to a probabilistic approach
because of the uncertainties which affect degradation rates.
In order to use such models in the durability design of metal structures, it is necessary to
relate the classification of the environmental loads to the corresponding corrosion rate. The
thickness loss during structural life time can be predicted by means of the considered
degradation law and on the basis of assigned environmental conditions.
At this stage of the study, it should be noted that no clear relationship between corrosion
models and atmosphere classes are provided by the codes.
In the following, an application of degradation models provided in literature to corrosiveness
classes defined in standards is reported. In particular, the corrosion models developed by
Klinesmith (2007) are used. Such models were formulated for different materials taking into
account the effects of four environmental variables, which are time-of-wetness, sulfur dioxide,
salinity, and temperature.
The form of the degradation model is the following:
( )
0
1 1
2
T T J
H F D
B
e
G
Cl
E
SO
C
TOW
t A y
+
|
.
|
\
|
+
|
.
|
\
|
+
|
.
|
\
|
= (8)
where y = corrosion loss (m); t = exposure time (years); TOW = time-of-wetness (h/year); SO
2

= sulfur dioxide concentration (g/m
3
); Cl is chloride deposition rate (mg/m
2
/day); T = air
temperature (C); and A, B, C, D, E, F, G, H, J, and T
0
= empirical coefficients.

112
Table 2. Coefficients of Eq. (8) calibrated with ISO CORRAG data (Dean & Reiser 2002) for carbon
steel and zinc materials.
Equation coefficients
Material
A B C D E F G H J T
0

Carbon
steel
13.4 0.98 3800 0.46 25 0.62 50 0.34 0.016 20
Zinc 0.16 0.36 3800 0.24 25 0.82 50 0.44 0.05 20

The degradation curves for zinc and carbon steel corresponding to the selected model (Table 2)
are shown in Figure 3. The environmental variables have been considered with values
corresponding to the different corrosiveness classes as defined in EN 12500 (2000).

a) b)
Figure 3. Thickness loss as a function of time for zinc (a) and carbon steel (b) for different corrosiveness
classes, according to selected degradation models.
7 DURABILITY LIMIT STATES
In order to complete the development of a durability design procedure based on the partial
safety factor method, both serviceability and ultimate limit states related to basic requirements
need to be defined, such as functionality in use and structural safety. In particular, the definition
of the maximum allowable degradation value for each material and limit state and of the
relevant reliability index is necessary for the calculation of design life t
d
and of relevant safety
factor
t0
.
Durability limit states could be the same of the ones used for mechanical design, but some
specific limit states should be defined for durability (Sarja 2004).
Serviceability and ultimate limit states in durability design can be referred to the thickness
loss of coating and structural material respectively. In particular, serviceability limit states are
related to changes in functionality or aesthetics and they are based on maintainability, economy
and environmental impacts. As a consequence, such limit states can be ascribed to the
attainment of partial or total loss of coating thickness. On the contrary, ultimate limit states are
related to the thickness loss of structural material which compromises the mechanical safety of
constructions.
With regard to serviceability limit states (SLS) relevant to zinc coating corrosion, it is
suggested to use the definition provided by British Steel Construction Institute research (Popo-
Ola & Biddle & Lawson 2000) on durability of galvanized cold formed steel section used in
housing. In this case, different limit states have been defined depending on the possibility of a
building component to be inspected. In particular, if the use conditions do not allow regular
inspections, such as in case of wall frames and wall ties, the limit state is reached when 50% of
the weight of zinc has been lost, otherwise for roof trusses and internal floors the limit state is
defined as the 80% of the weight of zinc loss.

113
As far as ultimate limit states (ULS) are concerned, it is proposed to carry out a durability
design which is separated from the mechanical one. In particular, the safety of constructions at
the end of their life-time should be evaluated by considering the possible thickness loss of
structural material which starts after the complete corrosion of coating. Such thickness should
be added to the one obtained from mechanical calculation in design phase of new constructions,
increasing the structural redundancy, or should be used for evaluating the actual safety factor
and the residual service life in existing ones.
In order to evaluate the design service life t
d
for different limit states, the central safety
factors have to be calculated on the basis of both Eq. (6) and corresponding reliability indexes.
With respect to reliability index , different values are given by standards and codes for
serviceability and ultimate limit states. The reliability indexes to be used in the durability design
should be defined close to the mechanical ones but, in order to balance the costs with the
benefits, lower reliability level should be tolerated (Sarja 2004). In particular, the suggested
values for ultimate and serviceability limit states are equal to 4.3 and 1.5 respectively, in case of
reliability class RC2 as defined in EN1990.
By applying the proposed procedure in case of carbon steel and assuming V
d
= 0.3, the central
safety factors calculated by Eq. (6) range from 1.5 to 2.5 for SLS and ULS, respectively. As far
as zinc corrosion is concerned, the relevant central safety factor for SLS is equal to 2.8.
8 CONCLUSIONS
This paper focuses on a proposal for durability design of metal structures against corrosion. The
different phases of the design procedure have been investigated considering the environmental
corrosivity classes defined in current standards and by using degradation models which are
available in literature for the calculation of the corresponding thickness loss.
Different limit states corresponding life time safety factors have been proposed, also on the
basis of statistical values of considered degradation models.
REFERENCES
CIB 1999. Agenda21 on Sustainable Construction. CIB Report Publication 237. Rotterdam: CIB.
CEN 1992. EN ISO 9223.Corrosion of Metals and Alloys. Corrosivity of Atmospheres. Classification.
CEN 1992. EN ISO 9226 Corrosion of metals and alloys. Corrosivity of Atmospheres. Determination of
corrosion rate of standard specimens for the evaluation of corrosivity.
CEN 1998. EN ISO 12944-1. Corrosion protection of steel structures by protective paint system.
CEN 1999. EN ISO 8044. Corrosion of metals and alloys. Basic terms and definitions.
CEN 1999. EN ISO 14713. Zinc and aluminum coatings.
CEN 2000. EN 12500. Corrosion likelihood in atmospheric environment.
CEN 2001. EN 1990. Eurocode. Basis of structural design.
Dean, S. & Reiser, D.B. 2002. Analysis of long term atmospheric corrosion results from ISO CORRAG
Program. West Conshohocken: ASTM.
Italian Government 2005 Ministerial Decree 14 September 2005.(D.M 2005). Design Rules for buildings.
Klinesmith, D.E. & McCuen R. & Albrecht P. 2007. Effect of environmental condition on corrosion rate.
Landolfo, R.& Di Lorenzo, G.& Guerrieri M.R. 2005. Modelling of the damage induced by atmospheric
corrosion on 19
th
century iron structures. Proc. of Italian National Conference on Corrosion and
Protection, Senigallia.
NACE International. 2002. Glossary of Corrosion Related Terms. Houston: NACE International.
Popo-Ola, S.O.& Biddle, A.R. & Lawson R.M.2000. Durability of light steel framing in residential
buildings. SCI Publication P262.Ascot: The Steel Construction Institute.
Sarja, A .2004. Lifecon deliverable D2.1. Reliability based methodology for lifetime management of
structures. Finland.
Sarja, A. & Vesikari, E. 1996. Durability design of concrete structures. RILEM Report of TC 130-
CSL.RILEM Report Series 14.
World Commission on Environment and Development(WCED) 1987. Our Common Future. Report
A/42/427. United Nation.

Chapter 3. Survey and condition assessment of structures

Chapter 3. Survey and condition assessment of structures:
Introduction
117
G. Hauf
University of Stuttgart, Institute of Structural Design
O. Hechler
ArcelorMittal, Commercial Sections, Luxembourg
Introduction
Monitoring is more and more accepted since it has become an efficient instrument for structural
assessment. Today monitoring is found for various purposes in maintenance but also
rehabilitation of buildings, bridges and structural facilities.
The following document provides information about the state of art, the specification,
implementation and operation of monitoring systems.
Beginning with a short introduction the different applicable insturmental techniques are
presented. The techniques for monitoring have been developed to reduce inspection and
maintenance costs and as tools to extend significantly the service life of structures. These
monitoring techniques and tools are of special interest today due to the increasing number of
structures, infrastructure operators have to manage and maintain, as well as the increasing
number of structures, that are approaching their design lifespan and require a decision about
their future use (e.g. the rehabilitation or renovation for bridges under fatigue behaviour).
The following contribution gives an overview on the most common monitoring concepts and
techniques which are used today for civil engineering structures.
Monitoring systems are mainly used for the damage control of materials and structures. In
this context the item of the structural health monitoring (SHM) has to be mentioned. Structural
Health Monitoring (SHM) is defined as the use of in-situ, non-destructive sensing and analysis
of structural characteristics in order to identify damage presence, location, severity and
prognosis. It is a rather young practice in civil engineering, but the recent technological and
sensing developments have given an impulse to its dissemination. SHM takes advantage of the
new technologies in sensing, instrumentation, communication and modeling in order to integrate
them into an intelligent system which provides several opportunities in the field of construction
management and maintenance. They basically consist of reduction of inspection and
maintenance costs, seismic protection, real or near real-time observation of the structural
response and of evolution of damage, possibility to develop post-earthquake scenarios and
support rescue operations. Thus, effectiveness of hardware systems and data processing
procedures plays a relevant role for a successful implementation of such systems. Several
applications of Structural Health Monitoring in the civil engineering field are reported in the
literature. Bridges and buildings are the structural typologies usually monitored, even if an
increasing number of applications focus also on geotechnical aspects. In the present paper, the
main aspects of integrated SHM system design, implementation and management will be
reviewed starting from description of some experiences carried out in this field by the research
group of the Structural and Geotechnical Dynamics Lab at University of Molise. Some issues
related to automated data processing for damage detection purposes will be also illustrated and
an algorithm for reliable and fully automated modal parameter identification for health
assessment purposes will be described through its application to a number of case studies
characterized by different levels of complexity.
As the implementation of monitoring systems is often planned already during the design of
construction and due to the necessity for long-lasting monitoring concepts the development of


118
monitoring systems offers new and easy applicable sensor techniques like wireless sensor
networks.
The main advantages of wireless sensor networks are fast deployment, little interference and
self-organization. However, since wireless sensor nodes are battery powered, in long term
monitoring applications the power management influences significantly the operation of a
wireless sensor network. In data intensive applications, e.g. vibration based monitoring, low
power hardware, duty cycle operation and efficient communication policies are not sufficient
for achieving a sustainable system lifetime. Since data communication is the most energy
consuming task, sustainable system lifetimes can be achieved by a significant data reduction in
the nodes. This data reduction is a challenging task, since it has to be performed with very
limited computational and memory resources and in competition with tasks providing the basic
network functionality. The objective of the paper is to provide a brief overview of the wireless
sensor network technology and to present our experience over the past three years with data
intensive structural monitoring using wireless sensor networks. The deployment on a bridge
(natural frequency tracking of stay cables) is illustrated and specific aspects of sensing, data
quality, stability, availability, and system lifetime are analyzed.
In the next contribution MEDEA (Manual for Earthquake Damage Evaluation and safety
Assessment) will be presented. The evaluation of damage due to a seismic event on both
masonry and reinforced concrete buildings is a central problem at different levels within the
scientific community. The paper presents MEDEA, a self-training multimedia tool for seismic
damage evaluation. The different sections of this didactic tool represent a guided, step-by-step
training path, which allows the users to develop knowledge either to understand the typological
characteristics of the buildings or to evaluate the damage; both of which are very important
aspects in the final judgment of safety. The CD-Rom has been produced with particular care in
designing a very simple and easy interface, combining the applications contained to the best
effect. MEDEA is divided into main sections, following a principle of logical continuity.
These published contributions show the wide spread of application of sensor techniques and
the damage monitoring of materials and structures. In the next step the results of the monitoring
have to be evaluated and concluded to a condition assessment of the structure. As more than
half of the budget for the infrastructural development in Europe is nowadays invested in
maintenance and modernization of the existing infrastructure, thus, less than half of the budget
is left for the extension and renewing of infrastructure. Due to this fact, design engineer needs to
know more than the rules and recommendations for the design of new structures - knowledge
and experience in the assessment of existing structures are required, in particular in fatigue
loaded structures with a limited design service life. As a consequence a working group has been
created in 1993 within the framework of the ECCS Technical Committee 6 Fatigue in order to
produce recommendations for the general practitioner and to prepare a draft of a future
extension of the Eurocodes covering the assessment of existing steel structures.
Of course the monitoring systems can also be used for the condition assessment and the
deterioration of concrete components. The deterioration mechanisms for concrete is classified
into physical, chemical and biological mechanism and is discussed in this contribution mainly
under the prism of a short presentation of their driving potentials, their characteristics and the
factors influencing their initiation and development. Also, a short reference to their impact on
concrete components is made. In the second part of the paper, the analysis focuses on the
inspection of deteriorated concrete, on the specification and the assessment of the concrete
components deterioration type and degree respectively. Not only methods that can be applied
on site for the investigation and the assessment of the damage, but also laboratorial tests and
methods are referred to and discussed. The range of the methods presented varies from simple
tests to investigations employing complex and technologically advanced equipment.
Beneath the condition assessment the service-life prediction appears as the second aim for the
issues of monitoring. For a sustainable lifetime management of buildings and of civil
infrastructures, methods for predicting how the condition is likely to evolve with time are
needed. These methods enable the owner or the facility manager to plan adequate maintenance
and repair actions and are particularly useful when large assets of structures have to be
managed.
Hereby the contribution presents a method for condition assessment and for service-life
prediction, which can be applied to aged as well as to new reinforced concrete structures, whose

Introduction
119
degradation is mainly promoted by carbonation of the cover concrete, as it is the case for
concrete buildings. Its theoretical background relies upon concepts of reliability theory, in
particular upon the computation of the probability of reaching a predefined limit state. For
carbonation induced degradation the limit state corresponds to the onset of reinforcement steel-
bars depassivation. It is assumed that the depth of the carbonation front is proportional to the
square root of time.
In the case of aged structures, the method requires as input variables the distribution
functions of the carbonation depth and the one of the thickness of the concrete cover, whereas
for new structures carbonation depth is replaced by the air permeability of the concrete cover.
Cover thickness and air permeability data are collected on site non-destructively by means of a
cover meter and of a Torrent permeability tester. The carbonation depth is measured by the
phenolphthalein method on core samples taken out from the structure.
The paper also presents case studies in which the condition assessment and service life
prediction method has been applied. Most of the investigated structures belong to the same
building asset, which is owned by the government of Canton Ticino and which comprises more
than 800 buildings. It is illustrated how the facility managers, which are in charge of this asset,
may use the knowledge about the temporal evolution of the structures conditions in order to
define intervention priorities and preventive maintenance strategies.
The last paper dealing with condition assessment and service-life prediction illustrates the
main objectives in structural damage and vulnerability assessment as a part of LCA (Life Cycle
Assessment) and LCC (Life Cycle Costs) analyses. A proper analysis of vulnerability factors
due to degrade conditions and to construction features of buildings helps to guide the design
choices for refurbishment interventions, introducing assessments on the appropriateness of the
retrofit action when the damage or vulnerability level is too high. From this point of view, the
introduction of new tools and methodologies is needed, in order to support the shared evaluation
models with specific assessments related to the various building components. The paper shows
the approach used in MEDEA (Manual for Earthquake Damage Evaluation and safety
Assessment) developed by PLINIVS Study Centre, University of Naples Federico II, in
collaboration with Italian Civil Protection Department.


120
G. Hauf
O. Hechler
Monitoring an introduction
1 INTRODUCTION
1.1 Definition of Monitoring
The overall objective of monitoring is to improve the knowledge of the real condition and/or
behavior of constructions. Structural monitoring is in this connection a term that increasingly
occurs for structural engineering. Though it refers to a great variety of measurement systems
implemented on full-scale civil structures with the purpose to inform about the actual condition
of their structures. The information obtained by monitoring can be used for very different
objectives:
condition assessment of structures (e.g. monitoring as support to visual inspection in
unapproachable areas),
monitoring of structural performance (e.g. bearing capacity),
remaining life-time prediction (input values for fatigue analysis),
model identification and correlation,
long term structural health monitoring (SHM: real time information) (e.g. monitoring as
Red-alert-system to send alarm when a pre-defined threshold value has been
exceeded).

Monitoring of structures is an interdisciplinary activity that requires expert knowledge from
different disciplines: structural engineering and electronical engineering, modeling and
simulation techniques, data processing, communication and archiving techniques, software
design and implementation techniques. A common detailed analysis and planning, and
communication between the experts is therefore a mandatory necessity for assuring a common
understanding of the goals and a successful monitoring activity (Rauert et al. 2007), (Feltrin et
al. 2007 (a)).

1.2 Basic concepts of technical monitoring
Monitoring activities can be differentiated into:
- Representative monitoring / Individual monitoring,
- Long-term monitoring / Short term monitoring,
- The monitoring concepts Action monitoring, Performance monitoring, Stress
monitoring and Health monitoring.

A representative monitoring campaign is focused on the monitoring of one single object, which
is considered to be representative for similar details or structures. It is assumed, that the

Survey and condition assessment of structures by monitoring
121
behaviour of the monitored object is identical to the behaviour of this kind of detail/structure
and the results can be transferred. As buildings are normally one-of-a-kind structural facilities
the possibilities to apply representative monitoring are limited.
The contrast to the representative monitoring is individual monitoring, limited to draw
conclusions for the monitored object only, e.g. due to unique boundary conditions or stresses in
one single structural element.
The second differentiation is based on the time range of a monitoring campaign. Long-term
monitoring is carried out permanently or in intervals over a defined time period (even from
construction to end of service life). Short-term monitoring is performed to observe a response
during a special event, e.g. the monitoring of bridge deflections during crossing of an abnormal
heavy load to avoid collapse or during the change of the bearing.
Further, concepts of technical monitoring have been defined based on the objective of the
monitoring campaign. Depending on the analysis task however, one or even more of these
concepts are applied. The monitoring concepts are defined as:
- Action monitoring
Action monitoring allows assessing the magnitude as well as the spatial and temporal
distribution of specific forces acting on a structure or a structural component.
- Performance monitoring
Performance monitoring allows assessing whether a structure or a structural component
meets the performance requirements under specific or any actions.
- Stress monitoring
Stress monitoring allows assessing the state of stress and strain in a structure or a
structural component.
- Health monitoring
Health monitoring allows assessing the health condition of a structure or a structural
component by means of health indicators. The scope of structural heath monitoring is to
provide real-time information to assess the safety and serviceability of a structure or
structural components. This implies that the behaviour of a structure under operating
conditions is sufficiently well understood and that there exist a number of health
indicators which can be measured with sensors and procedures which describe how to
use these health indicators to assess the state of a structure.
Furthermore, the failure scenarios with regard to safety and serviceability, that have to
be avoided during the life-span of a structure, have to be identified and characterized by
quantifiable health indicators that can be estimated through measurements of physical
quantities.
2 TECHNICAL ASPECTS OF MONITORING
2.1 Periodic inspections
Also periodic inspections (e.g. general inspections for bridges according to DIN1076 in a
frequency of every six years) can be interpreted as monitoring for the examination of the state
of condition of constructions.
In dependency of the monitoring concept applied there are different options for this examination
and precision of the inspection in reference to the objective of the campaign and aimed results:
Visual inspection: periodical monitoring of visible degradations: obvious damages on the
structure like deformations or corrosion protection/damage are identified, roughly cracks may
be identified; visual inspection is limited to spot existing or starting damages of the structure;
hidden and internal structural deficiencies cannot be monitored.


122

Figure 1. Visual inspection (bursted concrete superstructure of a bridge)
Mechanical inspection: mechanical inspection include destructive and/or non-destructive
testing; experimental result deliver of the real material properties (e.g. by core hole drilling of
concrete, or steel tensile specimen) and present condition of the structure (e.g. ultrasonic tests of
the welding or identification of reinforcement corrosion); the results from mechanical inspection
are e.g. needed for the recalculation of the existing bearing capacity of structural elements.

Figure 2. Extraction of a concrete core (core drilling)
Inspection with sensors: monitoring with the use of sensor technique provide very exact and
precisely analysis and measurements; the test campaign is in general extensive; as continues
monitoring is as well based on sensor monitoring, the function of various sensor techniques will
be focused in chapter 3 in more detail.

Figure 3. Typical sensors monitoring specimen (strain gauge, extensometer, seismic transducer)
2.2 Continues monitoring with sensors
Continues monitoring, especially sensor monitoring, means a measurement based on the
application of sensor techniques which provide empirical information.

123
The empirical information for buildings, bridges, highways, tunnels, dams, pipelines etc. can
be:
- physical quantities: position, deformation, strain, force, pressure, acceleration, vibrations
displacement, velocity, acceleration, strain, force or pressure,
- temperature, humidity
- chemical quantities: pH-value, chloride-concentration
- environmental parameters: wind speed, wind direction, snow, water levels, water flow
rate, pollutant concentration (Feltrin et al. 2004 (b))

A monitoring system consists of different components: Starting with the sensor which senses a
specific quantity of a physical process and sends an electric output depending on the time and
reflecting output. After conditioning and transmission a physical signal is amplified and filtered
and converted from an analog into a digital output which is necessary to be continued for use,
see the following figure.
Now the digital output can be used for analysis, recalculations or comparison with theoretical
calculated results and be displayed and stored.
Figure 4. Technical conversion of monitoring signal into digital data (Feltrin, G. et al. 2007 (b)).
2.3 The advantages of technical innovations
The quality of the technical monitoring depends highly on the quality of the used devices and
evaluation methodologies (Rauert, T. et al. 2007). The development in these fields is enormous
and the costs of high-end technologies decline more and more.
Even if monitoring is cheaper than in the past, the monitoring concept needs to be coherent
and the aim of the monitoring has to be clear for the post-monitoring analysis (see e.g. Hechler
et al. 2006). Nowadays sensor techniques allow producing such a huge amount of data, so the
engineer has to handle them responsible and hast to reduce the useful data. Especially for long-
term monitoring the data volume rises strongly that today handling becomes more and more
difficult. New evaluation methodologies like output only methods for system identification
provide however considerable advantages within the post processing.
Though, of major importance is to plan monitoring exactly with focus on implementation and
resolving adequate measurement points as well as feasibility of the results and data volume
produced.
2.4 Applications
Application examples of monitoring of civil structures are innumerable and of different
objective, see e.g. table 1 (Rauert, T. et al. 2007). The range is from damage detection,
observing sensitive system parameters and checking design assumptions to acquisition of
structure relevant statistical data like fatigue data, climate data or load characteristics.
Latest developments show that structural health monitoring (SHM) has become very popular.
Often it is used as an instrument to estimate the structural effects of earthquakes and to avoid


124
sudden collapses of e.g. sensitive structures in earthquake zones. Nevertheless these new
monitoring methods only can support regular visual inspections or other local testing methods.
Table 1. Applications of monitoring civil structures (Rauert et al. 2007, reduced table)
Object Monitored parameter/process Conclusions Aim
Public building Smoke emission
Fire in the
building
Protecting human life and the
structure
Traffic Bridge in an earthquake
zone
Accelerations at several points to
determine dynamic system properties of
the bridge structure after earthquakes
Structure
Damages
Protecting human life and the
structure
Steel structure with fatigue
cracks
Ultrasonic response signal near the cracks
due to ultrasound
Fatigue crack
propagation
Determination of service life,
Ensuring the durability of the
structure
Concrete structure with large
cracks
Digital image data of the surface
Crack
propagation
Determination of service life,
structure
Chloride exposed concrete
structure
Chloride concentration in surface areas Corrosion risk
structure
Aerial mast Strains at several points of the structure Wind loads
Verifying the load assumptions,
improving the design process
2.5 Limitations of monitoring
Sensor monitoring and visual inspection are sometimes problematic to be applied, for instance if
construction elements cannot be accessed, if parts of bridges and buildings are under heavy
traffic and/or if parts of the structure are covered and can therefore hardly be inspected visually
(e.g. grouted tendons in bridges). After finishing the construction work an application of the
sensor techniques is often not feasible anymore.
The goal of continuous monitoring hereby is not to supersede periodical inspections but to
improve the efficiency of the inspection process. This can be achieved by providing relevant
information continuously. Thus the number of components subjected to a detailed inspection
can be reduced. If a reduction of the time between inspections cycles is required technical
monitoring can avoid additional special inspections.
Consequently it is recommended, that monitoring concepts should ideally be planned and
applied with the ongoing design and construction of the structure to increase efficiency of the
campaign, and reduce costs and efforts for the installation (although today the aspect of later
necessary monitoring is in many cases neglected). It has to be noted, that post-application of
inspections and monitoring methods are in general more extensive and expensive.
An argument for monitoring is that inspections and monitoring may reduce maintenance costs
and guides to the optimization of date and extent of repair work. Although the optimal
application of monitoring is an individual matter considering the structure itself and the
boundary condition. It is not possible to find or give a general conclusion about monitoring
systematical concepts or solutions. Based on experience however guidance on the use of
different sensor techniques with their individual and possible application can be given, see
following chapter.
3 MONITORING TECHNIQUES
3.1 Overview of monitoring techniques
In the following chapter the main used monitoring techniques are presented and shortly
assessed. Hereby the focus is put on the sensor techniques; visual inspections or mechanical
inspections are not covered.
Table 2 gives an overview of sensors used for buildings and bridges. More detailed
information about sensor techniques can be found in (Feltrin, G. et al. 2007 (b)). Information
about wireless sensor techniques can be found in (Feltrin, G. et al. 2010).
3.2 Strain gauges
Strain gauges are used more or less for all constructional materials (even concrete strain gauges
are available) and consists of metallic wire. Because the resistance of the wire depends on the

125
length and cross section of the metallic wire, a length variation due to change of the strain
changes the electrical resistance. This change is measured and recalculated into a strain of the
material. Strain gauges are used for the localization of strain and stress concentrations.
Table 2. Overview of potential sensors used for monitoring of structure
Sensor Physical quantity see chapter
Strain gauges strain 3.2
Inductive position sensor;
Extensometer
displacement 3.3
Accelerometer acceleration 3.4
Temperature sensors Temperature (degree) 3.5
Mems sensor Mostly all physical quantitiy 3.6
Fibre-optic sensor 3.7
Methods for cracking detection deformation 3.8

The length of the applied strain depends on the demand as well as the material it is used for.
For concrete, long strain gauges should be preferred due to the heterogeneous material behavior.
Several configuration types of the strain gauges exist: uniaxial, biaxial, triaxial rosette as well
as shear pattern (Feltrin et al. 2004 (b)).

Figure 5. strain gauges (uniaxial strain gauges on constructional steel; triaxial rosette)
3.3 Inductive position sensor
Inductive linear position sensors are used as a deformation or displacement sensor. It consists
out of coils with a magnetic core (inductive sensor). The sensor output is given by voltage and
can be recalculated into a displacement or deformation. Therefore the sensor is based and the
displacement is measured relatively to this base, see the following figure.

Figure 6. Inductive position sensor
Extensometers are a modification to inductive position sensor. The difference is the fixed
measured basis length of the extensometer itself. The measured strains are based on a certain
length. By using several extensometers it is possible to measure the strain distribution. The
extensometer shown in figure 7 is qualified for the strain distribution on concrete surfaces to
analyze the effective width.


126

Figure 7. Extensometer
3.4 Accelerometers
The accelerometer is mainly used sensor for mechanical vibration measurements. An
accelerometer consists out of an internal mass, a viscous damping element and a linear spring.
The mass produces due to vibrations of the global structure an internal vibration that is
measured and transformed to a sensor output signal given as acceleration.

Figure 8. Accelerometer (sensor & mechanical model (Feltrin et al. 2004))
A modified accelerometer is given by piezoelectric accelerometer. Hereby the piezoelectric
effect is used (quartz crystal produces under deformation an electrical output). The internal mass
is connected to the quartz crystal and an acceleration of the structure can be measured.
Further modifications can be:
- Capacitive accelerometer
- Force-balanced accelerometer
More information can be found in (Feltrin et al. 2004 (b)).
3.5 Temperature sensor
As temperature affects mostly every physical effect it plays an important role for the sensor
monitoring. Due to temperature, material elongates or shortens or a change of the material
volume occurs; these effects of course have to be considered e.g. in the monitoring of stresses
resulting out of strains.
Therefore temperature sensors are often applied on the structure. With the temperature
information it is possible to exclude the influence of the temperature from the sensor monitoring
results and/or to clear these effects calculative.
The most commonly used sensors for temperature are:
- Resistance thermometer (change of temperature exhibits a change in the electrical
resistance)
- Thermocouples (two dissimilar materials in contact; voltage due to different temperature
at two different thermocouples junctions), see figure 9.

127

Figure 9. Temperature sensor (Feltrin et al. 2004 (b))
3.6 MEM sensor
Microelectromechanical sensors (MEMS) are alternatively to be applied for the monitoring of
structures. MEM sensors belong to so called active sensors which means, that they do not need
an additional electric power supply. MEMS consist out of two parts, the mechanical component
supplies the measurement itself and the electrical component provides the signal conditioning.
MEMS can be used for very wide application range, e.g.:
- Measurement of humidity and temperature
- Measurement of strains, stresses
- Measurement of acceleration

Figure 10. MEMS Accelerator sensor (Feltrin et al. 2004 (b))
Nowadays MEM sensors are mainly used for mechanical engineering applications; the ratio for
civil engineering is quite small.
3.7 Fibre-optic sensors
Monitoring systems enable to replace scheduled maintenance practice and finally to reduce the
costs for the rehabilitation and renovation of building structures. One solution could be a
continuous monitoring system with real-time information. Fibre-optic sensor techniques offer
these essential functions at best.
The link between the sensor and the detector consists out of fibre-glass so that external
electromagnetically interference effects become extinct; whereas conventional sensor
techniques with electric circuit connection can be affected by electromagnetic interference or
atmospheric disturbances.
Latest developments show that conventional sensor techniques are more and more replaced by
fibre-optic sensors. Fibre-optic techniques are mainly used best for loading behavior, corrosion
and degradation for fatigue analysis as real time monitoring systems. They can also be used for
monitoring of strains, vibrations, acoustic emission, pressure, temperature, humidity, wind and
fracturing in civil structures (Feltrin et al. 2004 (b)).


128
3.8 Methods for crack detecting
The deterioration of the concrete due to cracking and carbonation or the corrosion of the
reinforcement or tendons cause high costs for the maintenance of structures each year. Fractures
in constructional steel structures are rare.
To investigate in crack development the width of the crack has to be monitored. For the design
of the structural building the concrete crack width is designed and calculated by a certain ratio
of applied reinforcement. After construction, these assumptions have to be checked; if
applicable first rehabilitations have to be done. Due to rising traffic changing environmental
effects structures are constantly exposed to that impact. Hereby it is necessary that the
development of the cracking and starting degradation is recorded continuously.
Therefore new monitoring methods, destructive (DT) as well as non-destructive (NDT) testing
methods were developed for a steady-monitoring application and a verified degradation state of
the structural building.
These testing methods enable the crack detection or location, the crack width and some also
the measurement of the real depth of cracks. By using destructive testing methods for instance
concrete cores or steel samples are drilled out of the structure which of course means an
extended effort. For the non-destructive testing several methods are applicable:
- visual and optical testing,
- electromagnetic testing,
- acoustic emission testing,
- vibration methods,
- strain gauges, extensometers or inductive position sensors (see chapters above).

In (Feltrin et al. 2004 (b)) more and detailed information about the methods for crack
detecting can be found.
4 LATEST DEVELOPMENTS WIRELESS SENSOR MONITORING
In the last years new monitoring sensors appeared on the market for the inspection of structural
buildings: wireless sensor networks (WSN). The advantage of these new methods are fast
deployment, little interference, self-organization and reduction of installation work due to
abdication of the sensor wires. Wireless sensor techniques simplify the use of monitoring
sensors and will help to raise the application for structural building monitoring.
For further information about the techniques, methods and concepts it is referenced to (Feltrin
et al. 2010).
5 CONCLUSIONS
Nowadays monitoring sensor techniques assist to analyze the building condition, to identify
existing or even also starting damages on the structure and to start rehabilitation work at an
early stage.
This report should give a short overview of existing monitoring objectives and concepts, their
potential and the technical application of sensors mainly used for building structures. The
development is still in progress; new techniques are regularly appearing, e.g. fibre-optic sensors
or wireless sensor techniques. Sensor techniques are an adequate device for monitoring,
although the user has to assemble it carefully and prudentially. Monitoring concepts have to be
designed to inspect the critical locations on the structure. Also the huge amount of data
produced by the sensors has to be checked and filtered and the sensor information should
always be counterchecked by additional visual inspections.
To summarize monitoring becomes more and more important for analyzing constructions
helps to reduce costs for inspections and maintenance works to a minimum and optimizes
necessary repair campaigns.

129
REFERENCES
Feltrin, G. et al. 2007 (a). Monitoring Guidelines for Railway Bridges - Background document D5.2.
Sustainable Bridges. Current European research project FP 6.
Feltrin, G. et al. 2007 (b). Monitoring Instrumentations and Techniques - Background document D5.1.
Sustainable Bridges. Current European research project FP 6.
Feltrin, G., Bischoff, R, Meyer, J., Saukh, O.: Wireless sensor networks: A new tool for structural
monitoring, Final Report. COST C25: Sustainability in Construction Integrated Approach to life-
time Structural Engineering. 2010.
Hechler O., Feldmann M., B. Khn. 2006. Fatigue monitoring of steel railway bridges. IABMAS06 -
Third International Conference on Bridge Maintenance, Safety and Management (Porto)
Hechler O., Khn, B., Lsche, A., Sedlacek, G. 2005. Guidelines for monitoring of steel railway bridges.
WP5-19-T-050331-D-D5.2.1. Sustainable Bridges. Current research project FP 6 - PLT 001653.
T. Rauert, B. Hoffmeister, C. Heinemeyer J. Radic, J. Bleiziffer, O. Hechler: Systematic Monitoring of
Civil Structures, 1
st
Workshop. COST C25: Sustainability in Construction Integrated Approach to
life-time Structural Engineering. Lisbon, 2007.


130
G. Fabbrocino & C. Rainieri
Structural & Geotechnical Dynamics Lab StreGa, University of Molise, Department SAVA, Campobasso,
Italy

Monitoring of structural behaviour of buildings
1 INTRODUCTION
The opportunities that Structural Health Monitoring (SHM) offers in the field of management
and maintenance of civil structures make it very attractive and justify its increasing diffusion.
Reduction of inspection costs, optimization of maintenance actions, seismic protection, real or
near real-time observation of the structural response and of evolution of damage, possibility to
develop post-earthquake scenarios and support rescue operations are common targets of SHM
applied to civil structures.
The structural health assessment is carried out at different levels, from mere identification of
presence of damage to prognosis, across the evaluation of its location and extension (Silkorsky
1999). In-situ, non-destructive sensing and analysis methods are employed to this aim. The
assessment of the health state of a structure by tests and measurements is a common practice
since a long time (Mufti 2001), but SHM takes advantage of the new technologies in sensing,
instrumentation, communication and modeling in order to integrate them into an intelligent
system. Thus, it is a multidisciplinary field, where a number of different skills (seismology,
electronic and civil engineering, computer science) and institutions can work together in order
to increase performance and reliability of such systems, whose promising perspectives are well
established.
SHM systems are useful for short term evaluation of the impact of particular load conditions,
such as ground motions, and for long term assessment of the structural performance and health
state in operational conditions. They provide information about eventual damage conditions, so
that a structural safety evaluation basically in real time and setup of appropriate maintenance,
rehabilitation or strengthening strategies can be carried out. Moreover, reduction of down time
and improvement in reliability enhance the productivity of the structure (Aktan et al. 1999,
Various Authors 1999) and the monitoring results give a deeper insight in the structural
behavior which is useful to enhance design of future structures.
All these objectives require acquisition of data and information from different types of
sensors and their availability and reliable transmission even in critical conditions. However,
measured data often provide low-level information without further analyses. In particular,
damage identification from records of the structural response require large data processing and
analysis efforts and they can take advantage of information related to environmental parameters
such as temperature and humidity. The need to merge heterogeneous information drives the
development of effective data mining and data fusion algorithms. Thus, efficient data processing
procedures play a relevant role for extraction of information from measurements and for
reduction of the amount of data to be transmitted.
In the present paper attention is focused both on hardware and analysis aspects of
implementation of automated SHM systems for civil structures. After a description of the main

Damage monitoring and control of materials and structures
131
issues concerning their design and installation, a number of innovative applications in the field
are reported. Attention is then focused on the development of automated output-only modal
analysis procedures, which represent a key issue for application of modal-based damage
detection procedures (Doebling et al. 1996) to SHM. Finally, some experiences in the field
carried out by the research group of the Structural and Geotechnical Dynamics Lab StreGa at
University of Molise are described.
2 SHM OF CIVIL STRUCTURES
SHM of civil structures concerns detection of sudden or progressive damages at an early stage,
but the structural performance under operational conditions and/or in case of hazardous natural
events, such as earthquakes, is also of interest (Mufti 2001). Thus, a variety of sensors has to be
installed to measure the physical parameters describing the interaction between structure and
environment as well as the structural dynamic characteristics and the response to loads.
Monitoring the structural response in service conditions can lead to the identification damage
phenomena induced, for instance, by fatigue or corrosion.
Evaluation of the structural performance in operation can be useful to investigate vibration
problems and support design of structural interventions. In seismically-prone areas, apart from
checks of the health state in operation, SHM makes possible a field assessment of the
earthquake resistance of a building and a real-time or near real-time evaluation of the level of
damage suffered by the structure. Several applications are therefore related to SHM of civil
structures: design improvement (through model updating techniques), damage evaluation,
maintenance and retrofitting of existing structures, vibration control in operation and structural
control during earthquakes (Aktan et al. 2005, Lynch & Law 2002).
A reduction in inspection and maintenance costs can be also achieved. A review of
worldwide SHM systems, applied to different kinds of structures ranging from buildings (Mita
et al. 2006) to bridges (Seim & Giacomini 2000, Liu et al. 2007) and wind turbine blades
(Srensen et al. 2002), is reported in (Fabbrocino et al. 2008, Rainieri et al. 2008). Each SHM
system is characterized by a given choice about the sensors and the hardware architecture, and
by the adopted data acquisition, processing and transmission solutions. The number of sensors is
widely variable, depending on the characteristics of the structure, the objective of monitoring
and the available budget. They are usually characterised by different types and performance and
require an efficient data acquisition and processing system. The latter acquires sensor output,
processes data and eventually provides an alarm. Thus, data processing, reduction and storage,
sampling frequency and simultaneous sampling are the fundamental issues, especially when a
large number of sensors are installed. A real-time response of the system can be achieved if data
are collected, stored, assessed for validity and processed in a very short time. Sampling
frequency has to be carefully chosen in order to acquire and retain an optimized amount of data.
About the simultaneous sampling, when a single data logger is used data synchronization is
governed by its switch rate (McConnell & Reiley 1987). If a large number of sensors are
installed, even of different type, modular architectures are usually adopted and strategies to
ensure a simultaneous sampling have to be adopted.
The choice of a sensor depends on the structure and the monitoring requirements. No
particular sensor is suitable for every SHM application. A network of different types of sensors
may be necessary for a specific application. An appropriate choice of sensors (strain gauges,
accelerometers, FBG sensors, temperature sensors, anemometers, load cells and so on) and
consideration of the main issues related to installation and data processing (maintenance of
sensors, data volume and processing time) represents a key issue for a successful SHM
implementation. Protection of sensors, wires and connections is fundamental to ensure
durability of the system and data quality. For some types of sensors electromagnetic radiation
(EMR) effects have to be taken into account. A monitoring system usually consists of remote
sensors wired directly to a centralized data acquisition system which aggregates, stores and
analyzes the data. However, a large number of sensors can lead to heavy data streams and a
large computational effort for the server. This circumstance can affect the real-time response of
the system. Another drawback is, conversely, related to the high costs of installation and
maintenance associated with system wires (Lynch 2002), while power harvesting can represent


132
an issue for outdoor applications. Thus, modular architectures are spreading and, whenever
possible, wired systems are replaced by new low-cost wireless sensing units.
The main aspects of the SHM systems implementation for civil structures are related not only
to sensing, power, communication and storage technologies but also to data processing. Raw
data can provide limited information about the health state of a structure. Thus, data mining,
signal processing and health evaluation (damage detection and prognosis) algorithms have to be
considered as well. Strategies to manage data and combine information coming from a variety
of sensors related to different physical variables still represent a challenge in SHM.
Nevertheless, algorithms are able to identify the existence, location and extension of damage on
the basis of structural response measurements are already available. They can be classified into
two main categories, the so-called modal-based algorithms, focused on changes in the
dynamic characteristics of the structure before and after damage (Doebling et al. 1996), and the
approaches based on direct post-processing of the measurement data to detect anomalies (Gul &
Catbas 2009, Noh et al. 2009). In both the cases, the trend is to use methods that are able to
automate the detection process by taking advantage of the recent advances in information
technology (Aktan et al. 2005). Automation of modal identification, however, is not easy and it
represented for long the main obstacle to integration of modal-based damage detection
algorithm into fully automated SHM systems. Recently developed algorithms for automated
output-only modal identification have made modal-based damage detection possible but some
open issues are still there. Additional information about automated modal identification
algorithms and their limitations are given in the following.
Generally the presence of structural damage leads to a loss of stiffness, which results in the
change of both dynamic and static responses. Thus, apart from the traditional vibration-based
methods the use of static data for damage detection is also becoming more widespread with the
development of new model-free strategies for damage detection (Lanata 2008). Finally, reliable
procedures for data reduction are useful to cut the requirements for data transmission, in
particular after an earthquake, when a limited communication bandwidth can be available.
Wavelet-based approaches seem to be particularly promising in this field (Li et al. 2007,
Mizuno & Fujino 2007).
3 INNOVATION IN SHM
Innovative aspects in the field of SHM of civil structures concern its application to geotechnical
structures, its integration with seismic early warning systems, the design of new smart sensors
and the development of a wireless sensor network characterized by distributed data processing
rules for SHM of large structures. SHM systems have been applied, in the civil engineering
field, to a variety of structures. In geotechnical engineering, however, while static control of
displacements and pressures is a quite common construction practice, dynamic monitoring is
relatively limited. Geotechnical applications of fiber optic sensors are reported in (Habel et al.
2007), where such sensors have been used extensively in Geosynthetics and above all in micro
piles for corrosion and damage detection purposes. Some applications of embedded sensors in
piles concern monitoring of steel reinforcement and soil stresses for static purposes (Song &
Zhou 2007). Wireless sensors have been installed during casting of prestressed concrete piles in
order to monitor stresses and accelerations during driving and prevent microcracking, thus
extending life of such elements in a marine environment (Szyniszewski et al. 2008). A limited
number of full-scale dynamic measurement systems are applied to geotechnical structures, while
several data are available in terms of post-earthquake permanent deformations. Even if
vibration-based monitoring applications to geotechnical structure are still not so spread, due to
issues in sensor installation or embedment and the high level or required performance, a number
of case-histories are appearing in the literature, focused on seismic monitoring of earth dam
(Sica 2003), soil-foundation-structure interaction (Steidl & Nigbor 2004, Pitilakis et al. 2005),
dynamic and seismic behavior of flexible retaining walls (Fabbrocino et al. 2008).
These applications point out that SHM can represent a useful tool to enhance knowledge
about both structural and geotechnical aspects of dynamic behaviour of civil structures, but
seismological data can be also integrated into the process by combining SHM and seismic early
warning (Rainieri et al. 2010a). In such a case, unlike traditional seismic monitoring, an event

133
driven monitoring system is not useful. Continuous condition assessment and performance-
based maintenance of civil infrastructures are recommended to assess the short-term impact due
to earthquakes and the long-term deterioration process due to physical aging and routine
operation. Furthermore, anomalies can be detected by processing the incoming data. In the case
of earthquake risk analysis monitoring systems can be used to create a database from
measurements taken during the life of the structure. In the pre-seismic event phase, these data
can be analyzed to evaluate the ability of the building to withstand seismic events on the basis
of tremors, such as those due to traffic or wind excitation, by updating the numerical model.
At the same time, calibration of the available structural models can improve the ability of
structural calculations to make reliable estimations of seismic performance. Finally, an in-depth
knowledge about the seismic characteristics of the site (such as zone of the epicentre, seismicity,
etc.) provides additional useful information. As the database builds over time, this analysis
improves, and the detailed characterization of seismicity and structural behaviour allows more
reliable predictions of the structural response and of the event that will occur, so decreasing the
problem of a false alarm. A combined monitoring and early warning system can also be used for
disaster and emergency management, traffic control, damage evaluation, and post-earthquake
scenario definition. The use of monitoring systems on underground pipelines may be considered
as an example of post-earthquake emergency management: damaged gas utilities can cause
secondary disasters and, as a consequence, cause serious direct and indirect losses. In such a
case, information about abnormal pressure changes in gas pipelines can lead to an emergency
shut-off. Similar controls can affect traffic, if information about structural integrity of
infrastructures are available.
In fact, they can help decision makers arrange routes to the disaster area for rescue of
personnel and goods using still operable bridges. These different applications, related to
operational and extreme event conditions, have different requirements in terms of reliability and
speed of communication. The recent advances in Information and Communication Technology
allow real-time monitoring of structures. However, the consequences of extreme events on data
transmission systems have to be taken into account when early warning applications and
disaster management are considered. This aspect is less critical for traditional structural
monitoring, where backup procedures avoid data losses through local data storage. In order to
deal with the limited bandwidth for data transmission, two strategies are usually considered:
employment of redundant vectors for data transmission and reduction of the amount of data to
be transmitted (Li et al. 2007). Redundant measurement systems are the result of a compromise
between high spatial resolution of sensors, which implies a high volume of data to be processed,
and the need of continuous monitoring, to depict the details of time-variant phenomena.
The amount of installed sensors is currently too low with respect to the number of degree of
freedom of the monitored structures. However, sensor miniaturization and adoption of wireless
architectures will lead in the near future to a bio-inspired distribution of sensors, in analogy with
sensing elements in the human body. Nevertheless, examples of embedded sensors (Fabbrocino
et al. 2008) and of smart materials are already available. Use of cement-based composite for
SHM purposes is rapidly spreading (Ou & Li 2009, Azhari & Banthia 2009). New corrosion
monitoring systems are also appearing: in particular, micro-electro-mechanical self-powered
corrosion sensors (Ou & Li 2009), able to harvest the minor-current during the corrosion
process, represent promising tools for SHM of civil structures. Sensor integration in structural
elements can be easily achieved even by putting fiber optic sensors in fiber reinforced polymer
products (Ou & Li 2009). Whenever sensor miniaturization and embedment is not possible,
adoption of wireless solutions allows cost saving for large installations, where wired solutions
become too expensive. Moreover, since wireless sensors often have on-board analog-to-digital
converters and low-power microcontrollers, data processing tasks can be distributed and carried
out at each sensing node (Lynch et al. 2008). Local data processing allows energy savings,
reduction of data stream in transmission, and easy development of a scalable architecture. Thus,
new distributed computing procedures are being developed for system identification and health
monitoring of civil structures through wireless sensor networks (Lynch et al. 2008).


134
4 AUTOMATED MODAL IDENTIFICATION FOR SHM
An effective Structural Health Monitoring cannot neglect the implementation of proper data
processing procedures in view of health assessment and eventual prognosis. Since hardware
systems provide only raw information, specific data processing procedures must therefore be
evaluated and implemented so that a structure equipped with sensors and data logger becomes a
smart one. As a basis for evaluation of the health state of the structure, any change in structural
properties must be reconstructed on the basis of measured data. Modal-based damage detection
algorithms require monitoring of the dynamic properties (natural frequencies, mode shapes and
damping ratios) of the structure over time. Thus, specific algorithms for automated
identification and tracking of modal parameters play a fundamental role. Recently, some
strategies have been set up in order to automate identification and tracking of modal parameters
(Brincker et al. 2007, Magalhes et al. 2009, Poncelet et al. 2008, Devriendt et al. 2008, Rainieri
& Fabbrocino 2010) and allowing a full integration of modal identification within SHM
systems.
However, the available automated modal identification methods show some relevant
drawbacks: 1) the majority of the methods are based on threshold based peak detection; as a
consequence, a first calibration phase is needed for its proper definition; moreover, just some of
the identified peaks correspond to actual modes; performance of peak detection algorithms can
get worse in presence of measurement noise; 2) identification of actual modes is based on a
number of parameters which require time consuming calibration process for each monitored
structure; moreover, a static identification of thresholds and parameters may be inadequate to
follow the natural changes in modal properties of structures due to damage or environmental
effects.
Thus, an alternative strategy could be the definition of some objective criteria for
identification of mode bandwidth before modal parameter extraction. A new FDD-based
automated modal identification algorithm has been developed to this aim (Rainieri &
Fabbrocino 2010). It is based on the repeated computation of the SVD of the PSD matrix for a
number of records and on the construction of an averaged MAC vs. frequency plot which has
the aspect of a coherence function. Inspecting the sequence of MAC values at each frequency
line it is possible to identify the actual mode bandwidth to be used for modal parameter
extraction.
The algorithm has been implemented into a software package named LEONIDA (Rainieri &
Fabbrocino 2010) which is able to retrieve data from a measurement hardware, a remote
MySQL database or a file, thus allowing its integration into a fully automated monitoring
system. The software is able to carry out a proper identification process in a fully automated
way. Since it is somewhat time consuming, a faster modal tracking can be carried out by using
its results (namely, the mode shapes) for a spatial filtering of data. A reliable procedure for
modal tracking based on this concept (called AFDD-T) is described elsewhere (Rainieri 2008).
Such algorithm is able to follow the natural changes in modal parameters due to damage or
environmental effects. A new reference can be provided by LEONIDA periodically or on
demand (namely, after an extreme event).

135

Figure 1. The algorithm for automated modal parameter identification and tracking flowchart

These two algorithms can be combined in order to make an effective modal engine readily
available for SHM purposes. It is summarized in Figure 1, where the module for automated and
blind (no a-priori information is available) modal identification is followed by the modal
tracking procedure which takes advantage of the results of the first module in order to keep
computational efforts low. The overall procedure allows to overcome the typical drawbacks of
other automated modal parameter identification algorithms concerning flexibility, use of
threshold-based peak detection, discrimination between actual modal parameters and spurious
modes.
5 LONG TERM MONITORING
The Main Building of the School of Engineering in Naples (Fig. 2) has been equipped with a
structural health monitoring system some years ago. It has been conceived to combine the needs
of Structural Health Monitoring and of Seismic Early Warning (Rainieri et al. 2010a).


136

Figure 2. The School of Engineering Main Building in Naples
The main characteristics of such a system can be summarized as follows:
Integration (with SEWSs, of different sensors, of different models and data processing
procedures),
Redundancy (of sensors, of data transmission systems),
Scalability (applicability to small as well as to gigantic structures),
Durability (lifespan of the system comparable to that of the structure, with minor
maintenance interventions).
A modular architecture based on a remote relational database has been adopted.
A number of accelerometers have been installed at three different floors along the height of
the structure. They have been placed along the north-south direction and the east-west direction
in two opposite corners of the building and in two opposite corners near the stairs.
Data in operational conditions have been used to identify the fundamental modes of the
structure. The previously described procedure for automated modal parameter identification has
been integrated into the system.
Tracking of modal parameters over a year has pointed out a certain influence of temperature.
Results of monitoring of natural frequencies of the fundamental modes in operational conditions
are summarized in Table 1; effects of LAquila earthquake mainshock are described elsewhere
(Rainieri et al. 2009). During hot summer time, lower values of natural frequency for the first
three modes have been observed (in particular for the first and the third mode) compared to
those ones observed in winter.
Table 1. The School of Engineering (Naples): long term monitoring results.
Natural frequency [Hz]
Period Mode number
Mode Mean Standard deviation
I 0.92 0.9196 0.00979
II 0.99 0.9851 0.00808
Summer 2008 (7130 samples)

average temperature 25.2C III 1.29 1.2942 0.00836
I 0.94 0.9352 0.01011
II 0.99 0.9903 0.00736
Winter 2009 (13265 samples)

average temperature 11.5C III 1.31 1.3073 0.00750

The low values of standard deviation of natural frequencies reported in Table 1 point out that
the influence of environmental factors on this structure is relatively small (standard deviations

137
lower than 0.01 Hz) and that it is quite uniform for all modes. Nevertheless, environmental
effects were very useful to check reliability and robustness of the modal identification and
tracking procedures.
The effect of environmental factors, such as temperature, on modal parameters of the
structure requires further investigations. Development of a model to depurate modal parameter
estimates from the influence of environmental factors will make results readily applicable for
damage detection purposes.
6 SHORT TERM MONITORING IN THE EARLY EARTHQUAKE AFTERSHOCK
The automated procedure for modal parameter identification and tracking has been used, during
the seismic crisis in LAquila, for fast implementation of a modal-based SHM system for four
buildings belonging to Block B1 in the School of Guardia di Finanza in Coppito.

Figure 3. Plane view of tested building: nine buildings jointed among them and with stairs.
Block of buildings B1 (Fig. 3) is made by nine similar reinforced concrete moment frame
buildings distributed on three lines and connected by a number of separated reinforced concrete
stairs. Each building and stair structure appeared to be independent of the adjacent ones, since
seismic joints were provided in the superstructure. However, foundations were common to all
buildings and stairs of the block. In addition, as a consequence of ground shaking induced by
LAquila earthquake mainshock, pounding phenomena between buildings and stairs occurred.
The result was a certain degree of interaction among the different buildings and among
buildings and stairs so that natural frequencies of a given building or stair structure appeared in
the dynamic response in operational conditions of other close buildings in the block. Thus, the
main issues were related to the need of monitoring four building at the same time by a few
installed sensors on each structure and in presence of dynamic interaction effects and spurious
harmonics superimposed to the stochastic response. In particular, since the buildings were very
similar each other, as a result of their interaction response spectra were characterized by several
peaks in a narrow bandwidth. However, only a few of them were associated to actual structural
modes. The problem of output-only modal identification in presence of spurious harmonics and
dynamic interaction effects has been extensively described elsewhere (Rainieri et al. 2010b).
Results of manual modal identification according to a number of different frequency domain
and time domain methods have been then compared with those ones provided by the automated
modal identification procedure (Rainieri & Fabbrocino 2010).
Results reported in Table 2 point out that the algorithm for automated modal identification
was very effective and provided reliable results even in such a complex case study. In fact, it
was able to distinguish structural from spurious modes without any user intervention and with
only one false positive. The identified mode shapes have been then used as references for the
modal tracking (Rainieri 2008).


138
Table 2. Performance of the automated identification algorithm in the presence of dynamic interaction
effects and spurious harmonics.
Building ID Natural frequency [Hz] Structural mode? LEONIDAs performance
2.18 NO OK
2.58 YES OK
2.64 NO OK
2.88 YES OK
2.98 NO OK
3.52 YES OK
5.31 YES OK
2
7.09 YES OK
2.18 NO OK
2.58 YES OK
2.88 YES OK
2.99 NO OK
3.19 NO OK
3.52 YES OK
3
5.26 NO OK
2.20 NO OK
2.41 NO OK
2.67 YES OK
3.15 YES OK
5
3.96 YES OK
2.24 YES OK
2.41 NO OK
2.67 YES OK
2.81 NO FAILURE
3.20 NO OK
3.47 NO OK
3.96 YES OK
5.80 YES OK
6
7.54 YES OK

Figure 4. Block B1, Bldg #6 in the Guardia di Finanza School (LAquila Italy): short term monitoring.

The algorithm was also able to carry out an effective tracking of identified modes over time.
Even in presence of peaks in response spectra due to dynamic interaction effects and spurious
harmonics, the algorithm was able to carry out an effective tracking of identified modes over
time in the absence of transient input effects due to earthquake aftershocks. Slight variations of
natural frequencies over time (Fig. 4) were mainly and caused by ground motions occurred in

139
between two subsequent estimates (results of data processing of the structural response to
ground motions have been intentionally excluded from the plot). Results of modal tracking,
depurated from false estimates due to ground motion response, are summarized in Table 3.
Table 3. Block B1, Building n. 6 in the Guardia di Finanza School (Coppito, LAquila Italy): analysis
of short term monitoring results.
Building ID Mode
Average natural
frequency [Hz]
Standard deviation
[Hz]
Variation coefficient
[%]
I 2.58 0.04 1.6
II 2.88 0.03 1.0
III 3.52 0.03 0.9
IV 5.31 0.03 0.6
2
V 7.09 0.04 0.6
I 2.59 0.04 1.5
II 2.88 0.03 1.0 3
III 3.52 0.03 0.9
I 2.67 0.03 1.1
II 3.15 0.03 1.0 5
III 3.96 0.04 1.0
I 2.24 0.03 1.3
II 2.67 0.04 1.5
III 3.96 0.04 1.0
IV 5.80 0.04 0.7
6
V 7.54 0.04 0.5

The only drawback of the proposed procedure is related to the need of choosing a proper
sensor layout. In fact, it was found that if the estimated mode shapes between two modes are
very similar, the tracking procedure may sometimes fail. However, such wrong estimates can be
easily identified (the natural frequency estimates for the two modes having similar mode shapes
are equal at a certain iteration) and removed. Apart from the issues related to sensor placement,
the modal tracking algorithm was able to carry out an effective and reliable modal parameter
monitoring over time. Thus, the combination of automated modal parameter identification and
tracking represents an opportunity to implement a robust modal-based SHM system, even in
presence of dynamic interaction effects among adjacent structures.
7 CONCLUSIONS
In the present paper specific aspects related to implementation and operation of Structural
Health Monitoring (SHM) systems have been discussed. The main issues concerning sensing,
data acquisition and transmission, and data processing have been discussed.
After an extensive description of the main technological aspects concerning implementation
of SHM systems, attention has been focused on the assessment of the health conditions of
structures through modal-based damage detection procedures. One of the main issues relating to
their integration within fully automated SHM systems is the lack of reliable automated modal
parameter identification and tracking procedure. Thus, a procedure recently developed to this
aim has been described in its main aspects and applied to continuous monitoring of modal
parameters in operational conditions and in the early earthquake aftershock. In particular, it has
been shown how combination of effective automated modal identification and tracking
procedures can lead to a fast and easy implementation of effective vibration-based SHM
systems.
Application of the proposed procedure to data provided in operational conditions by the SHM
system of the School of Engineering Main Building in Naples has given the opportunity to
observe the effects of environmental factors on the modal parameters. Application to the Block
B1 of buildings belonging to the School of Guardia di Finanza in Coppito (LAquila - Italy), on
the other hand, has pointed out the robustness of the proposed algorithm even in presence of a
few installed sensors and spurious harmonics due to dynamic interaction effects. The only
drawback was related to the need of choosing an effective sensor layout when a few sensors are
used for modal tracking. Thus, an effective vibration-based structural health assessment can rely


140
on automated output-only modal identification procedures, even in seismically-prone areas if
the analysis of the structural response to an input ground motion is not of interest. They can be
confidently applied to follow degradation and damage phenomena, and to investigate the effect
of environmental factors on the structural behaviour of the investigated structure.
REFERENCES
Aktan, A.E., Tsikos, C.J., Catbas, F.N., Grimmelsman, K. & Barrish, R. 1999. Challenges and
Opportunities in Bridge Health Monitoring, Proceedings of the 2
nd
International Workshop on
Structural Health Monitoring, Stanford, CA, USA.
Aktan, A.E., Ciloglu, S.K., Grimmelsman, K.A., Pan, Q. & Catbas, F.N. 2005. Opportunities and
challenges in health monitoring of constructed systems by modal analysis, Proceedings of the 1
st

International Conference on Experimental Vibration Analysis for Civil Engineering Structures,
Bordeaux, France.
Azhari, F. & Banthia, N. 2009. Carbon Fiber/Nanotube Cement-Based Sensors for Structural Health
Monitoring, Proceedings of the 7
th
International Workshop on Structural Health Monitoring,
Stanford, CA, USA.
Brincker, R., Andersen, P. & Jacobsen, N.J. 2007. Automated Frequency Domain Decomposition for
Operational Modal Analysis, Proceedings of the 25
th
SEM International Modal Analysis Conference,
Orlando, FL, USA.
Devriendt, C., De Troyer, T., De Sitter, G. & Guillaume, P. 2008. Automated operational modal analysis
using transmissibility functions, Proceedings of ISMA 2008 Conference, Leuven, Belgium.
Doebling, S.W., Farrar, C.R., Prime, M.B. & Shevitz, D.W. 1996. Damage Identification and Health
Monitoring of Structural and Mechanical Systems from Changes in their Vibration Characteristics:
A Literature Review, Los Alamos National Laboratory, Technical Report LA-13070-MS, UC-900,
New Mexico 87545, USA.
Fabbrocino, G., Laorenza, C., Rainieri, C. & Santucci De Magistris, F. 2008. Seismic monitoring of
structural and geotechnical integrated systems, Proceedings of the 2
nd
Asia Pacific Workshop on
Structural Health Monitoring, Melbourne, Australia (peer reviewed).
Gul, M. & Catbas, F.N. 2009. Damage assessment Using a Novel Modified Time Series Analysis
Methodology, Proceedings of the 7
th
International Workshop on Structural Health Monitoring,
Stanford, CA, USA.
Habel, W.R., Schallert, M., Krebber, K., Hofmann, D., Dantan, N. & Nther, N. 2007. Fiber optic
sensors for long-term SHM in civil engineering applications, Proceedings of the 3
rd
International
Conference on Structural Health Monitoring of Intelligent Infrastructure, Vancouver, Canada.
Lanata, F. 2008. Statistical Algorithms for Damage Detection using Static and Dynamic Approaches,
Proceedings of the 4th European Workshop on Structural Health Monitoring, Cracow, Poland.
Li J., Zhang, Y. & Zhu, S. 2007. A wavelet-based structural damage assessment approach with
progressively downloaded sensor data, Smart Materials and Structures 17(2008).
Liu, C., Meng, X. & Yao, L. 2007. A Real-Time Kinematic GPS Positioning Based Structural Health
Monitoring System for the 32 km Donghai Bridge in China, Proceedings of the 6
th
International
Workshop on Structural Health Monitoring, Stanford, CA, USA.
Lynch, J.P. 2002. Decentralization of wireless monitoring and control technologies for smart civil
structures, Blume Earthquake Engineering Center, Technical Report #140, Stanford University
Stanford, CA, USA.
Lynch, J.P. & Law, K.H. 2002. Energy market-based control of linear civil structures, Proceedings of the
US-Korea Workshop on Smart Structural Systems, Pusan, Korea.
Lynch, J.P., Swartz, R.A. & Zimmerman, A.T. 2008. Distributed Data Processing Architectures for
Structural Health Monitoring Systems Implemented on Wireless Sensor Networks, Proceedings of the
4
th
European Workshop on Structural Health Monitoring, Cracow, Poland.
Magalhes, F., Cunha, A. & Caetano, E. 2009. Online automatic identification of the modal parameters of
a long span arch bridge, Mechanical Systems and Signal Processing 23: 316-329.
McConnell, K.G. & Reiley, W.F. 1987. Strain-gage instrumentation and data analysis, in A.S. Kobayashi
(ed.), Handbook on experimental mechanics, Prentice-Hall, pp. 79-116.
Mita, A., Inamura, T. & Yoshikawa, S. 2006. Structural Health Monitoring system for buildings with
automatic data management system, Proceedings of the 4
th
International Conference on Earthquake
Engineering, Taipei, Taiwan.
Mizuno, Y. & Fujino, Y. 2007. Wavelet decomposition approach for archiving and querying large
volume of Structural Health Monitoring data, Proceedings of the 3
rd
International Conference on
Structural Health Monitoring of Intelligent Infrastructure, Vancouver, Canada.
Mufti, A. 2001. Guidelines for Structural Health Monitoring, University of Manitoba, ISIS, Canada, pp.

141
1-127, ISBN 0-9689006-0-7.
Noh, H.Y., Lignos, D.G., Nair, K.K. & Kiremidjian, A.S. 2009. Application of Wavelet Based Damage
Sensitive Features for Structural Damage Diagnosis, Proceedings of the 7
th
International Workshop
on Structural Health Monitoring, Stanford, CA, USA.
Ou, J. & Li, H. 2009. Structural Health Monitoring in the Mainland of China: Review and Future Trends,
Proceedings of the 7
th
International Workshop on Structural Health Monitoring, Stanford, CA, USA.
Pitilakis, K., Manos, G., Raptakis, D., Makra, K., Manakou, M., Apostolidis, P. & Terzi, V. 2005.
Theoretical and experimental studies in Euroseistest, Proceedings of the Geotechnical Earthquake
Engineering Satellite Conference, Osaka, Japan.
Poncelet, F., Kerschen, G. & Golinval, J.C. 2008. In-orbit vibration testing of spacecraft structures,
Proceedings of ISMA 2008 Conference, Leuven, Belgium.
Rainieri, C. 2008. Operational Modal Analysis for Seismic Protection of Structures, Ph.D. Thesis,
University of Naples, Naples, Italy.
Rainieri, C. & Fabbrocino, G. 2010. Automated output-only dynamic identification of civil engineering
structures, Mechanical Systems and Signal Processing 24(3): 678-695.
Rainieri, C., Fabbrocino, G. & Cosenza, E. 2008. Integrated systems for Structural Health Monitoring:
worldwide applications and perspectives, Proceedings of the 4
th
European Workshop on Structural
Health Monitoring, Cracow, Poland.
Rainieri, C., Fabbrocino, G. & Cosenza, E. 2009. Real-time structural monitoring of the School of
Engineering Main Building at the University of Naples Federico II - Records of dynamic effects
induced by Abruzzo earthquake on April 6th, 2009, Reluis - StreGa Lab Technical Report, available
on-line at www.reluis.it.
Rainieri, C., Fabbrocino, G. & Cosenza, E. 2010a. Integrated seismic early warning and structural health
monitoring of critical civil infrastructures in seismically-prone areas, Structural Health Monitoring
An International Journal, accepted for publication.
Rainieri, C., Fabbrocino, G., Manfredi, G. & Dolce, M. 2010b. Operational Modal Analysis in presence
of dynamic interaction among adjacent buildings - an assessment, Proceedings of ISMA 2010
Conference, Leuven, Belgium.
Seim, C. & Giacomini, M.G. 2000. Instrumenting the Golden Gate Bridge to Record Seismic Behaviour
and to Deploy Rapid Inspection Response, Proceedings of the 12
th
World Conference on Earthquake
Engineering, Auckland, New Zealand.
Sica, S. 2003. Dynamic analysis of earth dams, Proceedings of the Workshop on Constitutive Modelling
and Analysis of Boundary Value Problems in Geotechnical Engineering, Naples, Italy.
Silkorsky, C. 1999. Development of a Health Monitoring System for Civil Structures using a Level IV
Non-Destructive Damage Evaluation Method, Proceedings of the 2
nd
International Workshop on
Structural Health Monitoring, Stanford, CA, USA.
Szyniszewski, S., Hamilton III, H.R. & Chih-Tsai, Y. 2008. Wireless Sensors in Prestressed Concrete
Piles Combining Driving Control with Long Term Monitoring, Proceedings of IABSE Conference
on Information and Communication Technology (ICT) for Bridges, Buildings and Construction
Practice, Helsinki, Finland.
Song, J. & Zhou, T. 2007. Working state monitoring of a pile-supported 18-storey frame-shearwall
structure, Proceedings of the 3
rd
International Conference on Structural Health Monitoring of
Intelligent Infrastructure, Vancouver, Canada.
Srensen, B.F., Lading, L., Sendrup, P., McGugan, M., Debel, C.P., Kristensen, O.J.D., Larsen, G.C.,
Hansen, A.M., Rheinlnder, J., Rusborg, J. & Vestergaard J.D. 2002. Fundamentals for remote
structural health monitoring of wind turbine blades - a preproject, Ris-R-1336(EN), Ris National
Laboratory, Ris, Denmark.
Steidl, J. & Nigbor, R.L. 2004. Network for Earthquake Engineering Simulation Experimental Arrays,
http://www.cosmos-eq.org:16080/Projects/GSMA/GSMA1/Paper/2.2_Steidl_et_al.pdf, Proceedings
of the International Workshop for Site Selection, Installation and Operation of Geotechnical Strong-
MotionArrays, Los Angeles, CA, USA.
Various Authors. 1999. Structural Health Monitoring, F.K. Chang (ed.), Proceedings of the 2nd
International Workshop on Structural Health Monitoring, Stanford, CA, USA.


142
G. Feltrin, R. Bischoff, J. Meyer & O. Saukh
Empa, Swiss Federal Laboratories for Materials Science and Technology, Duebendorf, Switzerland

Wireless sensor networks: A new tool for structural monitoring
1 MONITORING OF CIVIL STRUCTURES
Structural monitoring is a term that increasingly occurs within the civil engineering community
and which refers to a great variety of measurement systems implemented on full-scale civil
structures with the purpose to inform infrastructure operators about the actual state of their
structures. The current concern in western countries is driven by the increasing number of
structures that infrastructure operators have to manage and maintain as well as the increasing
number of structures that are approaching their design lifespan, and which require a decision
about their future use.
Assessing the condition of a structure is traditionally done by periodical visual inspection.
However, the aging of civil infrastructures demands for an increased inspection effort producing
considerable costs. Furthermore, periodical visual inspections do not provide up to date
information. This yields an increasing concern for rationalized, autonomous and continuous
inspection and monitoring concepts and technologies. Monitoring with physical sensors has the
potential to provide up to date information. Furthermore, it furnishes information about transient
processes like live loads, fatigue damage, vibrations, temperature, humidity etc. difficult to
assess by using traditional inspection methods.
The rapid progress in sensor, computing and communication technologies allowed to develop
structural monitoring systems that are sufficiently robust and reliable to be suitable for
autonomous medium and long term applications at allowable costs. This technological
breakthrough intensified the research on structural monitoring techniques with the goal to
improve the effectiveness of assessment procedures for structures, to identify the degradation of
structures or to assess their performance under real operation conditions.
A whole specialized field of structural monitoring, called structural health monitoring, was
established with the goal to develop techniques that are able to, possible autonomously, identify
damages in structures (Farrar & Worden 2007). Here, damage refers to changes of material,
geometric or system properties that affect adversely the performance of a structure. Influenced
by research and development efforts performed in the 1970s and 1980s for offshore platforms
and aerospace structures, the investigations were mainly focused on vibration-based damage
identification methods. A great variety of methods have been developed and tested for simple
structures in the laboratory. Field tests on full-scale structures, however, have demonstrated that
the effectiveness obtained under controlled conditions in the laboratory cannot be achieved
(Farrar et al. 1997; Rohrmann et al. 2000; Peeters & De Roeck 2001; Feltrin 2002). The
intrinsic complexity of civil structures, immanent uncertainties of the structural models,

143
uncontrollable environmental factors and noise diminish dramatically the likelihood of
successful damage identification in a stage that is beneficial for the infrastructure operator (Huth
et al. 2005). After two decades of intensive research, structural health monitoring has therefore
found little acceptance by infrastructure operators and remained essentially an academic
research topic.
Recently, structural monitoring has attracted the attention of scientists traditionally focused
on developing and implementing cost-efficient maintenance and management processes (Yanev
2003). In view of the recent developments in monitoring technology, which permits
deployments at allowable costs, they started to systematically investigate the benefits of
monitoring in terms of cost-effectiveness at all stage of bridges life-cycle management and
particularly as part of bridge management. The investigations are mainly focused on the
efficient integration of quantitative structural monitoring data into reliability assessment
frameworks and the updating of prediction models (Frangopol et al. 2008; Strauss et al. 2008).
The goal is not to supersede traditional inspections with monitoring but to complement and
optimize the inspection and assessment process with modern tools. The expectation is that the
additional quantitative information provided by monitoring will assist the efficient spending of
available budgets by permitting more focused intervention strategies. This monitoring approach,
which is less ambitious than the structural health monitoring approach based on damage
identification, has the advantage to be easier to integrate into existing bridge maintenance and
management processes and to be conceptually closer to the current engineering practice.
A very good overview about the state of the art of structural monitoring technology is
provided by the Encyclopedia of Structural Health Monitoring (Boller et al. 2009). In todays
daily civil engineering practice, however, the application of structural monitoring is still an
exception. Deployments are mainly limited to very large or severely deteriorated structures or
are performed for demonstration and research scopes. Important factors, which prevent the large
scale deployments of monitoring is the lack of incentives provided by codes or accepted
guidelines. Monitoring guidelines have been published by different national and international
organizations (Rcker et al. 2000; Mufti et al. 2001; Aktan et al. 2002; Bergmeister et al. 2003).
These guidelines, however, reproduce the state of the art and do not provide a guide for practice
how to design and implement a monitoring system for a focused objective on a specific bridge.
Furthermore, although less critical than decades ago, costs (instrumentation, deployment and
maintenance costs) are still an important issue since current monitoring technology and services
are generally quite expensive. The typical instrumentation costs per channel are in excess of
several thousands of Euros (deployment and maintenance costs are not included) which imply
high investment costs for service providers and high service costs for bridge operators even for
small size monitoring systems.
Monitoring with a wireless sensor network is a young and emerging technology that has the
potential to lower the cost and boost the structural monitoring market. The advantages of
wireless sensor networks over conventional wired technology are fast deployment, great
flexibility, easy scalability and self-organization. The objective of this paper is to provide a brief
overview of the technology and several specific challenges that occur when designing,
implementing and operating a WSN monitoring system. Furthermore, the second part of the
paper, describes the performance of a WSN monitoring system during a long-term deployment.
2 WIRELESS SENSOR NETWORKS
2.1 Introduction
A wireless sensor network (WSN) is a computer network consisting of many small,
intercommunicating computers equipped with one or several sensors (Culler et al. 2004). Each
small computer represents a node of the network and is commonly called sensor node. The
communication within the network is established using radio frequency transmission. All sensor
nodes are equipped with specific sensors tailored to their measurement tasks. One or several
sensor nodes act as root nodes and represent the data sink in the network.
The typical hardware components of a sensor node are the sensors, a signal conditioning unit,
an analog to digital conversion (ADC) module, a central processing unit (CPU) with random


144
access memory (RAM), a radio transceiver and the power supply (see Figure 1a). A physical
implementation of a sensor node is displayed in Figure 1b. Various hardware wireless sensor
networks platforms are commercially available today, which offer the basic functionality together
with specific sensing capabilities (Bischoff et al. 2009b). The diversity of platforms offers the
possibility to choose a platform which best fits the needs of a specific application.
a)
Sensors
Acquisition/
Conditioning Processing Communication
Antenna
Power supply
RF-
Tranceiver
CPU
Memory

b)

Figure 1. a) Hardware architecture of a sensor node. b) Physical implementation of a sensor node.
2.2 Power management
Since the sensor nodes have to be operated with batteries, the power supply is very limited.
Power saving is therefore of outmost importance in designing, implementing and operating
WSN based monitoring systems for medium and long term deployments, since too frequent
battery changes increase the maintenance costs severely compromising one of the main
advantages of WSN monitoring systems.
Energy consumption is reduced by using low power hardware (sensors, microcontrollers,
radio chips) for implementing sensor nodes that consume typically significantly less than 100
mW (Bischoff et al. 2009b). Another means to reduce energy consumption is to operate the
network in switched-off state for a significant amount of time by switching-on hardware
components only for the time period required to complete the scheduled tasks.
Communication is the most energy consuming task. Since the energy consumption increases
exponentially with the transmission distance (exponent greater than 2), power can be saved by
operating WSNs as multi-hop networks (Culler et al. 2004). By establishing communication
links only to neighbor nodes, thus reducing the transmission distance, the data are sent to the
data sink through several nodes, which act as relaying stations, receiving and forwarding data
from adjacent nodes.
Finally, power can be saved by processing the raw data in the node with the goal to reduce
the amount of data that need to be transmitted (sending of information instead of raw data).
When monitoring vibration based processes, which produce large samples of raw data, this
strategy is the most powerful energy saving method and is mandatory if a system lifetime of
several months should be achieved (Feltrin et al. 2006; Lynch et al. 2006).
2.3 Sensing
Several commercial WSN platforms offer specific low power data acquisition boards with
integrated MEMS sensors and signal conditioning circuits for sensing and acquiring
temperature, humidity, light intensity, gas pressure, or accelerations. MEMS sensors have
several advantages compared to conventional sensors: They are small, generally low power,
highly integrated and, if no high end accuracy and resolution is required, rather inexpensive.
Several commercial MEMS accelerometers have the characteristics to be suitable for
structural monitoring applications. MEMS accelerometers with an amplitude range of several g
have typically a sensitivity of 0.51 V/g and a noise density smaller than 0.1mg/ Hz . They

145
can be powered with a low DC voltage of 3 to 5 V, which can be provided with standard
batteries, and their power consumption is of the order of several mW.
a)

b)

Figure 2. a) Acceleration records obtained with a WSN equipped with a low cost MEMS sensor and a
conventional data acquisition device with a conventional piezoelectric accelerometer. b) Strain records
obtained with a WSN and a conventional data acquisition device.
Figure 2a displays a comparison of the performance of a 10 MEMS accelerometer and a
500 conventional piezoelectric accelerometer. The accelerometers were mounted on a shaker
that performed a sweep excitation between 2 and 50 Hz. The accelerations of the MEMS
accelerometer were recorded with a sensor node equipped with a 12 bit ADC and those of the
piezoelectric accelerometer with a 24bit high end data acquisition device. The agreement
displayed in Figure 2a is very good. In the investigated amplitude range both accelerometers are
essentially equivalent.
A sensor node can be designed to support also conventional sensors that are not low power. A
typical widely used sensor in structural monitoring is the electrical strain gage. Because of its
low resistance (typically 120 ), strain gages are rather power consuming (approx. 40 mW).
Furthermore, they require a specific signal conditioning for enhancing the small resistance
changes due to the elongation of the gage. Power saving can be achieved by providing the
signal conditioning with an on-off-switch and switching-on the sensing unit only when
recording (see section 2.6 and Bischoff et al. 2009a). Figure 2b displays strain records obtained
with a WSN and a conventional high end data acquisition device. The quality of the WSN
record is comparable to the record obtained with the conventional system. The average
difference between the two records has a magnitude of a few percent.
2.4 Data acquisition
A critical aspect of data acquisition is the analog to digital conversion (ADC) since it has an
important impact on the resolution of the digital data. Modern wired data acquisition devices are
typically equipped with 24 bit AD converters. In contrast, mainly for reducing the power
consumption, WSN platforms are provided with 8 or 12 bit AD converters. This represents a
quite stringent limitation that requires a careful balance between resolution and amplitude range.
At low amplitudes the quantization effect of the AD converter usually limits the resolution of
the whole data acquisition process. This effect manifests itself as noise and is displayed at small
amplitudes in the record displayed in Figure 2b (0 to 2 seconds). Contrary, at high amplitudes,
Figure 2b does not display a significant difference between the data obtained by the sensor node
and the high end data acquisition device.
A second aspect that has to be considered is time drift due to an inaccurate setting of the
sampling rate of the WSN platform. The inaccurate setting of the sampling rate is an effect of
the quantization of time due to the finite frequency of the quartz oscillator. The effective
sampling rate of the WSN ADC was tested by monitoring the change of the phase difference of
a sine signal between the WSN and conventional data acquisition. Figure 3a) displays the time
drifts computed by two sine signals with 1 and 5 Hz frequency. For both sine signals the time
drift is very similar and reaches 0.13 seconds after 120 seconds. The relative time drift of the


146
WSN data is therefore approximately 1.1, which is a very small figure but still has
appreciable effects on measurements that last for several minutes or have high frequency
oscillations. Since the time drift is linear, it can be easily corrected by an adequate post-
processing.
a)

b)

Figure 3. a) Time drift of a sensor node. b) Signal corruption by the duty cycle of the radio.

2.5 Duty cycle operation
Duty cycle operation is one of the main mechanisms for reducing power consumption in WSNs.
The duty cycle describes the fraction of the switched-on time of a hardware component within a
given period of time. A small duty cycle means that most of the time the hardware component is
switched-off or in sleep mode. The smaller the duty cycle, the less energy is consumed.
Since communication is very power consuming, operating the radio with a low duty-cycle is
a standard strategy to save power. However, because of the high power consumption, each time
the radio is switched-on, the power supply voltage drops by a small amount (e.g. 0.1 V) due to
the internal resistance of the power supply. If the sensors are connected to the same circuit that
powers the WSN platform, which is the standard situation in commercial WSN platforms, the
periodical voltage drop affects also the power supply of the accelerometer and, as a
consequence, produces a variation of the output signal, which reduces the quality of the sensor
signal. Figure 3b displays such a periodical acceleration signal noise. For ambient vibration
measurements that require a large amplification of the acceleration signal, the noise induced by
duty cycle operation is very likely to severely degrade the quality of the data.
The signal corruption can be eliminated by separating the power supply of the sensors from
the power supply of the WSN platform (Feltrin et al. 2009a) or by powering the sensors with a
voltage regulator circuit that keeps a constant output voltage, e.g. 3V, independently of the
current supply voltage of the batteries. Furthermore, if a particular sensor requires a supply
voltage that is greater than the supply voltage provided by the batteries, a voltage regulator
boosts the battery output voltage to the correct voltage level. The latter solution cancels also the
effect of the fading of supply voltage due to power consumption that is typical of batteries.
2.6 Switch-on effects on sensing
Switching-off hardware components that are not required is generally a good strategy to save
power. The more power consuming the hardware components is the greater are the advantages
to switch them off. Many sensors are nearly indifferent to switch-on operations, since they
require a short time to be ready for operation because of the inevitable electrical switch-on
transient state and perhaps initialization. MEMS accelerometers belong to this class of sensors.
Unfortunately, the electrical resistance strain gage, the most common sensor for measuring
mechanical strains is affected by the switch-on operation. Figure 4b displays the time series of
strains recorded with a strain gage on a steel bridge during the passage of a train (Bischoff et al.
2009a). The time series contains the expected strain cycles generated by the train axles passing

147
over the stringer. The time series, however, are biased by a significant time-dependent signal.
The bias was generated by switching-on the strain signal conditioning device shortly before the
measurement was started. Due to the resistance of the strain gage, the current flow heats the
strain gage increasing its temperature and consequently its resistance. The Wheatstone bridge
translates this resistance change in a decreasing voltage signal. Since the heat produced in the
strain gage flows into the surrounding metal, the temperature of the strain gage increases with
increasing time until eventually an equilibrium state is achieved.
Figure 4b) shows that the equilibrium state is achieved after approximately 25 seconds from
switching-on the strain gage board. Conventional monitoring systems do not show this bias
because they are always operated in a temperature equilibrium state since switching-on occurred
long time before the first measurement started and the devices are never switched-off between
two measurements.
a)

b)

Figure 4. a) Sensor node and strain gages mounted on a bridge girder. b) Strain record with switch-on
bias.
In this particular case, to avoid the bias the strain gage should therefore be switched-on 25
seconds in advance. This would require a triggering system that captures train arrivals long
before a train reaches the bridge. A solution that is not easy to achieve. The bias can be reduced
by adding a dummy gage in the Wheatstone bridge for achieving temperature compensation (2
gage system). Since the dummy gage is not bonded to the structure, it has to be designed to
provide the same thermal characteristics of the primary gage for achieving good temperature
compensation. This solution, however, is expensive since the temperature compensation must be
adapted for each specific application and complicates the deployment process.
2.7 Software
Contrary to conventional monitoring systems, which usually have a centralized system
configuration and data acquisition software, a WSN is a distributed data acquisition system.
Each sensor node executes a software that provides the functionality for performing many tasks.
This functionality comprehends
time synchronization of the network
scheduling and execution of the measurement tasks
signal conditioning and data acquisition for different sensors
temporary storage of the acquired data
data processing
self monitoring (e.g. supply voltage, communication link quality)
management of the data acquisition and processing configuration (e.g. changing the
sampling rate, reprogramming of data processing algorithms)
reception and forwarding of data packets
coordination and management of communication and networking.
The sensor node software of a WSN for civil structures that supports different sensors and data
processing algorithms is described in Feltrin et al. 2009b. Because of the very limited memory
resources, the software is usually tailored to the specific sensors and tasks of a sensor node in


148
order to keep its size as small as possible. Often, the software is written in NesC (Gay et al.
2003), an extension to the C language. The basic functionality like time synchronization, multi-
hop functionality etc. is provided by TinyOS (Levis et al. 2005), a widespread operating system
for WSNs that has been ported to many WSN platforms. TinyOS is highly tailored to the limited
node hardware . NesC as well as TinyOS are both Open Source projects.
a)

C6 C5 C4 C3 C2
C1
base station
root node C0
63.20
b)

Figure 5. a) Set-up of the WSN monitoring system. b) Stay cables of the Stork Bridge with mounted
sensor nodes.
3 PERFORMANCE ASSESSMENT WITH A FIELD DEPLOYMENT
3.1 Cable stays monitoring
The performance of a WSN monitoring system was studied with a stay cable tension monitoring
deployment. Cable tension is usually estimated by correlating the measured natural frequencies,
which are extracted from ambient vibration recordings, with natural frequencies predicted with
a cable model (Casas 1994; Feltrin et al. 2006). The natural frequencies are therefore the only
information needed to estimate the cable tension. The natural frequencies are extracted by
processing the raw data, ambient vibration records, in the nodes and transmitted to the net-work
sink, while the raw data can be discarded. From the initial amount of data, e.g. thousands of
samples, the data to be communicated is reduced to several natural frequencies. This approach
results in a drastic reduction of the power consumption.
3.2 Monitoring system
The WSN monitoring system has been deployed on a cable-stayed bridge (Stork Bridge) in
Winterthur. The network consists of 6 nodes (C1 to C6), which are mounted on 6 stay cables, a
root node (C0), which is located at the northern abutment under the bridge deck, and a relay
node (C7), which is mounted beside the bridge and has a line of sight to all other nodes. Figure
5a illustrates the set-up and Figure 5b displays the sensor nodes mounted on stay cables of the
Stork Bridge. The root node is attached to the base station, an industrial PC, powered from the
mains power supply. The base station establishes a communication link with a computer in
Empas laboratory over the cell phone network using standard Internet communication
protocols. The laboratory computer hosts the data base of the monitoring system and provides
the remote configuration tools.
Vibrations are measured with a LIS2L06 MEMS capacitive accelerometer from ST
Microelectronics. The accelerometer is integrated into an electronic circuit containing a low
pass filter with a cut-off frequency of 20 Hz and a signal amplifier with an amplification factor
of 50. The power consumption of the accelerometer and signal conditioning is approximately

149
5mW. The ambient vibrations were recorded with a sample rate of 50 Hz for 20.5 seconds and
contained 1024 samples. A typical time series of ambient vibrations is displayed in Figure 6a.
In addition to the ambient vibrations, temperature and humidity are periodically measured
with the single chip Sensirion SHT11 sensor. The sensors are mounted inside the enclosure to
investigate the time evolution of humidity as well as into an opening of the housing to measure
the ambient temperature and humidity. Power supply voltage, routing and link quality
information are also monitored. All data was acquired periodically with a time interval of 5
minutes.
3.3 Computation of natural frequencies
Typically, the estimation of natural frequencies via frequency spectrum is performed in three
steps. First, the Fourier transform of the acquired vibration data is computed with a FFT
algorithm. The second step computes the frequency spectrum with the real and imaginary of the
Fourier transform. The third step identifies the peaks of the spectrum that correspond to the
cable natural frequencies by using a peak detection algorithm.
Implementing this algorithm on a microcontroller with 10kB RAM is not a straightforward
exercise since a time series of 1024 16 bit samples takes 2kB of memory. Since the RAM stores
also the sensor node software, little memory is left for additional memory requirements of the
data processing algorithm. Fortunately, the FFT algorithm, which represents the most
demanding data processing step, can be implemented by using essentially only the memory of
the time series. The intermediate and final results are stored in the same memory, overwriting
the data from the previous stage.
Usually, the FFT algorithm is performed in floating point arithmetic. However, the
microcontroller performs the floating point operations very slowly since they have to be
emulated with integer operations. This overhead results in a greater code size and longer
execution time. Less memory demanding, much faster and less power consuming is to perform
the FFT algorithm with integer operations.
Evidently, an integer FFT implementation may introduce significant errors. A typical error
source is overflow. By scaling the recorded time series with a suitable factor, however, the
overflow can be avoided and the approximation error can be maintained within an acceptable
range. Figure 6b) compares the frequency spectra of the time series displayed in Figure 6a) that
are computed with a 16 bit integer approximation FFT and with the standard 32 bit floating
point FFT. As can be observed, the result of the integer approximation FFT matches very well
with the floating point FFT. The net execution time of the FFT was less than 0.55 seconds. The
FFT computation with floating point operations took approximately 10 times longer and namely
5.8 seconds.
a)

0 2 4 6 8 10
-4
-2
0
2
4
time [s]
a
c
c
e
l
e
r
a
t
i
o
n

[
m
g
]

b)

0 2 4 6 8 10 12
10
1
10
2
10
3
10
4
frequency [Hz]
s
p
e
c
t
r
a
l

a
m
p
l
i
t
u
d
e
f
1
f
2
f
3
f
4

integer FFT
floating point FFT

Figure 6. a) Ambient vibrations recorded on a stay cable. b) Fourier spectra computed with fixed and
floating point operations.
Computing the frequency spectrum (
2
norm) took 0.19 seconds and 8 natural frequencies
additional 0.21 seconds. Hence, the final result, the natural frequencies, was computed in less
than 1 second. Compared to the recording time of 20.5 seconds the data processing is very fast.


150
Figure 7 shows the time evolutions of the captured natural frequencies of cable C3 and C6. 5
natural frequencies could be regularly monitored for the cable C3 and 3 natural frequencies for
the cable C6. Natural frequencies with a magnitude greater than ~10 Hz are more difficult to
detect since the associated vibration modes are scarcely excited. The standard deviation of the
natural frequency estimations of cable C3 is less than 0.05 Hz and therefore approximately of
the same magnitude of the frequency resolution of the spectrum, which is 0.05 Hz. A more
pronounced scattering is displayed by natural frequency estimations of cable C6 that range
between 0.05 and 0.13 Hz for the first three natural frequencies. This effect is due to the
shortness of the cable C6 that limits the magnitude of the ambient vibrations. The 1h average of
the natural frequencies (12 estimations) of the cables C3 differs from reference measurements
by less than 2%. For cable C6 the difference between the average value and the reference
measurement is smaller than 4%.
a)

b)

Figure 7. a) Natural frequencies of cable C3. b) Natural frequencies of cable C6.
3.4 Availability
From the very first day on, the most challenging issue was to achieve system stability. The
software that integrated data acquisition, data processing, time synchronization, low duty cycle,
task scheduling etc. into one packet turned out to be very sensitive to many tiny details
regulating the various activities. Sensor nodes disappeared and spontaneously reappeared after
some time without any evident reason. When operating the network with TinyOS
1.x/Boomerang the routing tree was changing continuously. In this highly dynamic situation
quite often nodes failed to choose a parent node correctly and lost their link to the network.
System avail-ability could be improved by including watchdogs that identified abnormal states
and by periodic automatic rebooting of the nodes.
The system stability improved significantly after porting the software to TinyOS 2.x. This
progress is shown in Figure 8a that displays the hourly availability during 60 days of two test
periods. The availability is defined as the percentage of natural frequencies received by the
remote control unit with respect to the theoretical maximum.
Period 1, April-May 2007, displayed rarely an availability higher than 80% and several total
break downs of the whole network. Total failures of sensor nodes to deliver data occurred very
often. In the second period, January-February 2010, this pattern was not observed. The
availability was 100% for most of the time and was only briefly interrupted with partial break
downs that did not go below 80%. The mean availability of the two test periods was 53% and
99.5%, respectively.
Figure 8b displays the availabilities of different physical quantities during the crisis at day 33
of the second test period. The availability broke suddenly down at hour 4 and recovered also
suddenly, after 16 hours. The crisis was not due to a break down of the WSN but to a
malfunctioning of the base station. Problems with the base station were also the origin of the
crises at day 25 and 45.

151
a)

b)

Figure 8. a) Availabilities of two test periods. b) Availabilities of different physical quantities during the
crisis of day 33 of the second test period.
3.5 Energy Consumption
The average power consumption of the WSN node is determined by the time needed for sensing
and data processing, the amount of data to be transmitted, the duty cycle period, the power
management of the hardware and the network topology, since it determines the number of hops
for reaching the data sink. Sensors and signal conditioning boards were switched-off after
completion of the data acquisition. Furthermore, since the nodes were not equipped with a
voltage regulator, the radio was switched-on during data acquisition in order to avoid signal
corruption by the duty cycle. Figure 9a shows the battery voltage drop of the sensors node C1
and C7. The observed voltage curves are not decreasing monotonically but are oscillating
significantly. The good correlation between voltage and mid-term temperature variations, as
shown in Figure 9b, demonstrates that temperature changes originates the voltage oscillations.
The voltage drop of C1 was approximately 0.15 V in 60 days. Since a sensor node can be
operated correctly provided the supply voltage is higher than 2.4 V, a lifetime estimation based
on the observed voltage drop predicts a battery lifetime of approximately 240 days. The relay
node C7, which was not equipped with sensors, had half the voltage drop of sensor node C1 and
thus twice the lifetime of node C1. This figure indicates that roughly half of the energy of a
node is consumed by data acquisition, data processing and the extra radio-on time during data
acquisition.
a)

b)

Figure 9. a) Supply voltage drop during test period 2. b) Temperature variation during test period 2.


152
4 CONCLUSIONS
This 3 year of field test experience demonstrates that medium and long term data intensive
monitoring of civil structures with wireless networks is feasible and that the produced
information complies with the quality requirements in civil engineering. In the beginning,
operating the wireless network reliably over a period of months turned out to be non-trivial task.
The problems relied basically on balancing the requirements of a data intensive application with
the requirements of minimizing power consumption for achieving a sufficiently long battery
lifetime. Data reduction, a powerful method to save power and therefore a key factor for
achieving the targeted lifetime, destabilized seriously the WSN by disturbing the basic
operations like time synchronization and routing.
Nevertheless, with many hard- or software improvements, the problems could be
substantially solved. The availability of the most recent test phase was better than 99%.
Considering that the data loss of the most recent test phase was due to malfunctioning of the
link between the root node and the base station, the availability figure is very encouraging.
Despite the severe hard- and software limitations, relatively complex in-node data processing
could be performed and the accuracy of the generated information was better than 5%. This
figure is very close to comply with the quality requirements in civil engineering, which usually
do not require high precision information. The simultaneous application of several power saving
mechanisms, in particular in-node data processing, allows to easily achieve node lifetimes of
several months. Further significant improvements are still possible, since the duty cycle of the
radio was not optimized. The hardware limitations, however, imposes a tight specialization to
the monitoring task. This implies a detailed analysis and specification of the monitoring goals.
Although not all aspects were investigated in depth, the results obtained by this investigation
demonstrate that there are no fundamental obstacles preventing the application of long term
monitoring systems based on wireless sensor networks to civil structures. This technology is
very close to be mature for practical application. In the near future, progress in low power
hardware will increase the computational resources without increasing or even decreasing the
current power consumption. This progress will allow to perform more complex monitoring tasks
and in-node data processing with less compulsion to specialization.
ACKNOWLEDGMENTS
A significant part of this work was conducted within European (Sustainable Bridges, I-SSB)
and national research projects. The authors express their appreciation to the Swiss State
Secretariat for Education and Research, the European Commission, the Swiss Federal Roads
Office, and the directory board of Empa for the financial support.
REFERENCES
Aktan, A., Catbas, N., Grimmelsmann, K., & Pervispour, M. 2002. Development of a model health
monitoring guide for major bridges. Drexel Intelligent Infrastructure and Transportation Safety
Institute, Drexel University.
Bergmeister, K., Aktan, E., Bucher, C., Dorfmann, L., & Fehlin, E. 2003. Monitoring and safety
evaluation of existing concrete structures, state of the art report. fib, fdration international du bton,
Lausanne.
Bischoff, R., Meyer, J., Enochsson, O., Feltrin, G., & Elfgren, L. 2009a. Event-based strain monitoring
on a railway bridge with a wireless sensor network. 4th International Conference on Structural Health
Monitoring on Intelligent Infrastructure (SHMII-4), Zurich, Switzerland, 22-24 July 2009, on CD.
Bischoff, R., Meyer, J., & Feltrin, G. 2009b. Wireless sensor network platforms. In C. Boller, F. Chang,
and Y. Fujino (eds), Encyclopedia of structural health monitoring: 1229-1238. Chichester, UK: John
Wiley & Sons Ltd.
Boller, C., Chang, F., & Fujino, Y. 2009. Encyclopedia of structural health monitoring. Wiley,
Chichester, UK.

153
Farrar, C. R., Doebling, S. W., Cornwell, P. J., & Straser, E. G. 1997. Variability of modal parameters
measured on the alamosa canyon bridge. Proceedings of the 15th International Modal Analysis
Conference, Orlando, FL, February 3-6, 257-263.
Farrar, C. R., & Worden, K. 2007. An introduction to structural health monitoring. Philosophical
Transactions Of The Royal Society A-Mathematical Physical And Engineering Sciences, 365(1851):
303-315.
Feltrin, G. 2002. Temperature and damage effects on modal parameters of a reinforced concrete bridge.
Fifth European Conference on Structural Dynamics, EURODYN 2002, Munich, September 2-5, 2002,
373 - 378.
Feltrin, G., Meyer, J., & Bischoff, R. 2006. A wireless sensor network for force monitoring of cable
stays. IABMAS'06 - Third International Conference on Bridge Maintenance, Safety and Management,
Porto, July 16-19, 2006, on CD.
Feltrin, G., Meyer, J., Bischoff, R., & Motavalli, M. 2009a. Modular low-power wireless sensor network
for structural health monitoring. 4th International Conference on Structural Health Monitoring on
Intelligent Infrastructure (SHMII-4), Zurich, Switzerland, 22-24 July 2009, on CD.
Feltrin, G., Meyer, J., Bischoff, R., & Motavalli, M. 2009b. Wireless sensor networks for long term
monitoring of civil structures. Structure and Infrastructure Engineering: DOI:
10.1080/15732470903068573.
Frangopol, D. M., Strauss, A., & Kim, S. 2008. Bridge reliability assessment based on monitoring.
Journal Of Bridge Engineering, 13(3): 258-270.
Gay, D., Levis, P., von Behren, R., Welsh, M., Brewer, E., & Culler, D. 2003. The nesc language: A
holistic approach to networked embedded systems. Proceedings of the ACM SIGPLAN 2003
Conference on Programming Language Design and Implementation, 1-11.
Huth, O., Feltrin, G., Maeck, J., Kilic, N., & Motavalli, M. 2005. Damage identification using modal
data: Experiences on a prestressed concrete bridge. Journal of Structural Engineering-ASCE, 131(12):
1898-1910.
Levis, P., Madden, S., Polastre, J., Szewczyk, R., Whitehouse, K., Woo, A., Gay, D., Hill, J., Welsh, M.,
Brewer, E., & Culler, D. 2005. Tinyos: An operating system for wireless sensor networks. In (eds),
Ambient intelligenceNew York: Springer-Verlag.
Lynch, J. P., Yang, W., Loh, K. J., Jin Hak, Y., & Chung Bang, Y. 2006. Performance monitoring of the
geumdang bridge using a dense network of high-resolution wireless sensors. Smart Materials and
Structures, 15(6): 1561-75.
Mufti, A., Humar, J., Jalali, J., Newhook, J., & Rahman, S. 2001. Guidelines for structural health
monitoring, design manual no. 2. ISIS Canada Corporation, Winnipeg.
Peeters, B., & De Roeck, G. 2001. One-year monitoring of the z24-bridge: Environmental effects versus
damage events. Earthquake Engineering and Structural Dynamics, 30: 149171.
Rohrmann, R. G., Baessler, M., Said, S., Schmid, W., & Ruecker, W. F. 2000. Structural causes of
temperature affected modal data of civil structures obtained by long time monitoring. Proceedings of
the 18th international modal analysis conference, San Antonio, Texas, USA, 17.
Rcker, W., Rohrmann, R., Ruge, P., & Link, M. 2000. Merkblatt ber die automatisierte
dauerberwachung im ingenieurbau. Deutsche Gesellschaft fr Zerstrungsfreie Prfung e.V., Berlin.
Strauss, A., Frangopol, D. M., & Kim, S. 2008. Statistical, probabilistic and decision analysis aspects
related to the efficient use of structural monitoring systems. Beton- und Stahlbetonbau, 103(S1): 23-
28.
Yanev, B. 2003. Structural health monitoring as a bridge management tool. Proceedings of the 1st
International Conference on Structural Health Monitoring and Intelligent Infrastructure, Tokyo,
Japan, November, 13-15, 87-95.


154
G. Zuccaro & M.F. Leone
PLINIVS Study Centre, University of Naples Federico II, Naples, Italy
Structural Damage and Vulnerability Assessment for
Service Life Estimation through MEDEA tool
1 INTRODUCTION
1.1 Technology Retrofit and Life Cycle Thinking
In the last two decades, sustainability has become a top priority in design and construction, both
for new buildings and retrofit interventions. Only in recent years, however, issues related to a
reliable quantitative assessment of environmental impacts in the building process have been the
subject of extensive studies, aimed at identifying, along the entire life cycle, factors such as
material and energy intensity of products and components, emission of greenhouse gases and
other pollutants.

Figure 1. Building process and Life Cycle Thinking approach.
The aim is to define sustainability assessment tools that can integrate environmental (Life
Cycle Assessment), social (health and welfare) and economic (Life Cycle Cost) aspects.
The Life Cycle Thinking approach, upon which is focused the current European
environmental policy, allow to overcome purely ideological conception about sustainability
issues, moving towards a more rigorous scientific approach.
In fact, most of the recent regulations and codes confirm this direction, like ISO/TS 21931-
1:2006 (Sustainability in building construction - Framework for methods for assessment of
environmental performance of construction works), ISO 14040 (Life Cycle Assessment);
Directive EC 91:2002 (EPBD - Energy Performance of Buildings); the energy performance
assessment methods (CEN/TC89, TC156, TC169, TC228, TC247), the service life assessment

155
methods (Service Life Estimation of Buildings, ISO 15686-5:2008, ISO TC/59/SC14), the
declarations on products service life (Service Life Declarations, ISO TC59/SC14) and the
regulations upon dangerous substances (TC351).
Generally, the assessment tools used within the LCA analysis identify benchmarks for
different technical options, in order to guide decision making process and design choices. At
present however, while in the case of single products and systems evaluations is it possible to
refer to a set of common protocols and reliable tools, LCA methodologies and procedures for
whole building assessments are not as refined. This because of the high complexity due to the
interaction of several subsystems (structure, envelope, HVAC systems, etc.), each with different
implications in relation to environmental impacts in the life cycle.

Figure 2. Embodied energy of building components along the service life.
While in the case of structures, for example, major impacts occur during production and
construction phases (including all necessary operations from raw material extraction to
manufacturing in factory and work on construction site), envelope components and HVAC
systems produce the main impact during use and maintenance phases, mainly due to energy
consumption resulting from the thermo physical characteristics of materials and efficiency of
installations (Fig. 2). With reference to the different phases of the building process, different
aspects such as the choice of materials, the construction typology and the configuration of
technical elements can influence the environmental performance of the building during its
service life (Fig. 3).

Figure 3. Embodied energy of buildings for a 50 years service life: initial embodied energy 6,2%;
maintenance embodied energy 8,3%; operating energy 85,5%.
Several studies worldwide have confirmed the risks arising from a partial approach to the
theme of the LCA analysis of buildings, which could introduce elements of arbitrariness in the
final assessment because of the excessive simplification of the reference model. From this point
of view, the introduction of new tools and methodologies is needed, in order to support the
shared LCA models with specific evaluations related to the various building components. This
approach is especially needed in the case of technological retrofit interventions, where the


156
opportunities of preservation of parts of the existing building have to be seriously taken into
account, considering the environmental implications related to demolition and reconstruction.
Generally, retrofit actions are considered as inherently sustainable interventions, as they
can limit the impacts related to the disposal of substantial parts of the building (starting from the
structural elements), while avoiding the consumption of material and energy resources related to
their construction ex-novo. The benefits of the retrofit measures may however be nullified by
misjudgements about the expected service life of the building in relation to its structural
integrity and ability to ensure safety and reliability requirements. To affect this aspect there are
considerations about the degradation of the building components due to obsolescence or
physical structural damage occurred over time, as well as assessments regarding the
vulnerability to natural hazards in relation to construction typology and to expected damage
scenarios.
A proper analysis of vulnerability factors due to degrade conditions and to construction
features of buildings helps to guide the design choices for refurbishment interventions,
introducing assessments on the appropriateness of the retrofit action when the structural damage
or vulnerability level is too high.
The proposed evaluation model also includes the identification of possible mitigation
solutions according to the damage or vulnerability factors identified. For each solution it is
possible to analyse the aspects related to expected environmental and economic impacts which
can then be integrated into the LCA model. The aim is to favour structural retrofit technical
alternatives based on eco-efficient solutions that minimize the consumption of material and
energy resources in the life cycle, compatible with the envisaged refurbishment intervention.

Figure 4a, b. Comparison of Embodied Energy (GJeq) and Greenhouse Emissions (KgCO2eq) for
different R.C. construction technologies (Conventional concrete; precast concrete and UHPC Ultra
High Performance Concrete) considering a 100 years service life. 1. Pre-production and production; 2.
Transport; 3. Construction; 4. Use and Maintenance; 5. Disposal and recycling; 6. Avoided products.
1.2 Residual Service Life estimation and MEDEA assessment tool
The term "service life", according to ISO 15686 is referred to the period of time after
installation during which a building or its parts meets or exceeds the performance
requirements. During this period, buildings and infrastructures are able to withstand normal
conditions of use, if properly maintained.
The service life of a building depends on many factors, such as the design quality, the
durability of construction materials and systems components, the incorporated technology, the
location and climate, the type and intensity of use, and damage caused by human error and
natural hazards. These aspects influence the quickness of ageing of a building as well as the
amount of maintenance and repair required over its life cycle. In fact, although the building
performances gradually decline because of ageing, its service life can be optimized through
adequate and timely maintenance and repairs (Figure 5).
UHPC
Precast concrete
Ordinary concrete

157

Figure 5. Service Life and maintenance operations.
The estimation of Residual Service Life, defined as service life remaining at certain period
of consideration, is a relevant issue considering the great number of existing buildings showing
physical or functional obsolescence (in the EU only residential buildings are approximately 150
million, 32% of which was built before 1945, 40% between 1945-1975 and remaining 28% after
1975).
As stated before, the knowledge of the probable residual service life of building elements is a
key factor to define the opportunity of repair action, as well as need to replace components in
order to reduce energy consumptions or maintenance operations. In case of refurbishment
interventions, some of the main issues are the understanding of structural features, the potential
remaining useful life of the frame and the opportunities for re-loading the structures with the
refurbished components.
This kind of evaluation must take into account the vulnerability factors of the building and
the potential damage due to destructive events, which become crucial factors in order to define
the Residual Service Life of both structural and non-structural elements.
In order to produce an objective damage and vulnerability measuring tool, from which it
could be possible to derive the most frequent damage typologies in R.C. and masonry structures
as well as a possible damaging model for quick and guided safety assessment, PLINIVS Study
Centre, University of Naples Federico II, in collaboration with Civil Protection Department
designed a multimedia handbook called MEDEA (Manual for Earthquake Damage Evaluation
and safety Assessment). The assessment tool is based on the identification of possible
typological vulnerability factors of masonry and R.C. buildings, which are listed and combined
with potential collapse mechanisms to be activated under seismic excitation. The procedure
takes place from simple structural behaviour models, derived from earthquake observations, and
tested to evaluate the reliability of the model.
MEDEA provides the basis either for safety check analyses of the existing buildings or for
post-event structural safety assessment and economic damage evaluation. Mechanisms analysis,
mainly developed for the post event safety check surveyors training, allow to trace a simple
logic path in order to approach the evaluation of masonry and R.C. buildings safety check. The
procedure starts from the identification of the typological vulnerability factors to derive the
potential collapse mechanisms and their collapse multipliers and finally addresses the simplest
and cheapest strengthening techniques to reduce the original vulnerability. The procedure has
been introduced in the Guide Lines of the Regione Campania for the professionals in charge of
the safety check analyses and the buildings strengthening in application of the national
mitigation campaign introduced by the Ordinance of the Central Government n. 3362/03.
The flexible structure of MEDEA allows to define different uses for the multimedia tool, such
as:
a) a handbook for an homogeneous categorization of the structural elements in R.C. and
masonry buildings and of their relevant damage typologies; b) a teaching tool for the training of
technicians involved in damage and safety evaluations in the post earthquake management
(using standard, pre-defined assessment criteria); c) a supporting tool for more detailed safety
survey forms; d) a guided tool for macro-seismic evaluation on the field; e) a teaching resource
to be used at different levels in university courses in the field of seismic engineering or
emergency management.


158
Finally, with reference to the object of the present paper, MEDEA is proposed as an
assessment tool for the estimation of Residual Service Life of structural elements, to be used in
LCA or LCC studies, as well as to evaluate the opportunity of refurbishment actions on existing
buildings. The tool is based on the correlation between vulnerability factors and damage
mechanisms (visible or potential), which brings to the definition of a safety index,
corresponding to the expected Residual Service Life.
2 MEDEA EVALUATION TOOL
2.1 The evaluation of the structural damage and vulnerability through MEDEA
MEDEA tool is addressed to the technicians either involved in vulnerability assessment or
employed in post event seismic damage evaluation.
In both cases the univocity of the judgment of the surveyors is very important, either because
of the scientific aspects resulting from a standard damage measure (useful for the improving of
the vulnerability function on one hand, and for a bigger harmonization in defining macroseismic
fields on the other) or for the social aspects concerning the civil protection deriving from a fair
damage estimation and from a fair static safety evaluation.
The tool is structured so that the user can have a gradual knowledge of it. The different
sections of the multimedia didactical tool represent a guided training path able to create in the
users, step by step, a sensibility either to understand the typological characteristics of the
buildings or to evaluate vulnerability and damage, which are both very important aspects of the
final judgement of safety.
The CD-rom MEDEA has been made with particular care in designing a very simple and easy
interface, combining at the best the applications contained.
MEDEA is sub-divided into four main sections, following a principle of logical continuity
and strictly connected between them, thanks to functions of research, query and hypertext links.
The first section concerns a detailed dictionary of the main terms frequently used in technical
and scientific field. The glossary represents a crucial step to define an unified terminology either
for the reciprocals comprehension among the technicians or to improve the homogeneity in the
typological element classification and in the damage measuring.
The second section of the tool is an archive containing a rich anthology of pictures showing
different structural typologies and different levels and types of damages. The picture archive
aims to give an essential basic knowledge of the observed damages and typologies in order to
improve the capability of interpretation and judgment that the technicians must have during the
post-event inspections that will be tested then in the following sections of the tool.
The most important part of MEDEA is the third section, constituted by a detailed catalogue of
the main damages on structural and non-structural elements of R.C. and masonry buildings. The
multimedia catalogue represents a further step of fundamental importance in the training of the
technicians involved in the safety check of the buildings in the post event phase; in fact it
supplies to the user a guided path among the basic criteria of the structure behaviour in order to
develop interpretation skill on the observed damages. The damages classification of the
buildings structures is aided by their interpretation as possible collapse mechanisms. Therefore
the catalogue is organized on the base of logic rules to address at the most probable mechanism
identification. This section is, therefore, constituted by three sub-sections: Collapse Mechanisms
Abacus; Damages Abacus; Interactive Training Table.
The final part of MEDEA in constituted by a training section, in which some damaged
buildings, realized by Quick Time Virtual Reality technology, are shown. These allow the user
to perform training paths designed for safety analyses compiling the AeDes Form (safety form)
and eventually expressing evaluations on: the constructive typology (horizontal and vertical
structures) which a vulnerability class can be associated to; the damage level expressed both for
every single constructive element and for the whole building; the safety assessment of the
building; the possible provisional interventions to be adopted (Zuccaro et al. 2001, 2002).

159
2.2 MEDEA Form: Vulnerability Factors and Collapse Mechanisms
The MEDEA project originates to facilitate the buildings safety check assessment that the
technicians were called to perform in the post event phase. Now is proposed as a Quick
Assessment Procedure to identify the Potential Collapse Mechanisms (PCM) starting from the
evaluation of the Typological Vulnerability Factors (TVF) and Cracking Path (CP) of the
building. This approach has been developed in the MEDEA Form that includes a list of possible
relationship between TVF, CP and PCM.

Figure 6. Classification of the Global Mechanisms for masonry buildings.
The TVFs listed in MEDEA Form have been selected among those showing the greatest
influence in CP analyses of the damages observed in past seismic events in Italy, (Irpinia 80,
Abruzzo 84, Sicilia 90, Umbria-Marche 97, Molise 02, Abruzzo 09). In Figure 6, MEDEA
classification of the global mechanisms for masonry buildings is reported.
The collapse mechanisms are divided into Global and Local mechanisms. For the evaluation of
structural integrity of the building and the assessment of expected residual service life, only the
global mechanisms are here discussed, although local mechanisms can affect durability of non-
structural elements.
The Global Mechanisms are those mechanisms involving the structure as a whole and then
related to the evolution of the cracks in a sufficient number of elements such that a total
compromising of the static and dynamic equilibrium of the structural system is achieved; these
mechanisms have been subdivided as follows.
The Local mechanisms are related to marginal parts of the structure; their evolution, even if
determines the collapse of a single element, generally does not involve the whole structural
equilibrium.
The activation of the PCM is strictly related to structural lacks of the building. In Table 1 a
possible correlation between PCMs and the TVFs, involving constructive, mechanic and
resistance aspects are reported.
The MEDEA Form used during the damage survey after the San Giuliano di Puglia event in
2002 (Dolce et al. 2002, 2003), has allowed to collect information on the incipient CM observed
on the damaged masonry structures and to put forward an hypothesis on the global collapse
process of the building, achieved when several CM start and the sequence of which is strongly
correlated to the typological structural characteristic of the building, as well as the intrinsic
quality of the materials used.
The proposed analysis methodology suggests an useful path for the identification of the
vulnerability factors present on the structure and responsible of potential or incipient CMs. This
method addresses toward the identification of the strengthening priority pursuing, when a
complete upgrading of the structure is not possible, to eliminate at least the main structure
vulnerability factors responsible of PCMs. Furthermore, samples of the collapse multipliers of
the most frequents CMs are provided; these allow to derive the limit PGA capable to activate
the mechanism.


160
Table 1. Relationship between Typological Vulnerability Factors and Collapse Mechanisms.

The most frequent collapse mechanisms discussed in Zuccaro et al. this issue (as vertical
instability and overturning CMs: types 3, 4 and 5; Fig. 7a,b) have been studied (see ref.
MEDEA for detailed description of the CMs). In the old masonry buildings the walls are made,
in some cases, with two vertical layers without sufficient connection in between. Generally the
mechanism involves the external layer only. However, sometimes the internal wall interacts
with the external one speeding the collapse.

161

Figure 7a. Overturning Out of plane mechanisms.

Figure 7b. Instability Out of plane mechanisms.

Table 2. Vulnerability factors for masonry structures.
Absence of connection between orthogonal walls and/or tie-beams or stringcourse at different levels
Presence of stringcourses in breccia on masonries with double facings
Floors badly connected with the walls
Masonry of low-quality, reduced resistant area along one or both directions
High percentage of openings
Foundations inadequate to resist the vertical load increment due to the earthquake
Different consistencies of the foundation soils, presence of landslide or liquefaction
Presence of added buildings with different stiffness and/or with localized connections
Variation of the structural system at upper levels
Presence of a raising and/or a stiff and badly connected roof structure
Presence of staggered levels
Excessive distance between bracing walls
Pushing structure and/or absence of connection between the wall and the roof
Presence of lintels with reduced bending stiffness or with inadequate support length
Presence of lowered arches and/or inadequate support of lintel
Local reduction of the masonry section (presence of flues, niches, etc.)
Local discontinuities (filling of old openings, bad realization of masonry sewing, etc.)
Presence of ridge beam of considerable sizes
Presence of openings in the proximity of the roof ridge


162

Figure 8. Excerpt of potential damages to masonry structures linked to vulnerability factors.
Table 3. Vulnerability factors for R.C. structures.
Joints out of building code
Average distance between columns > 6 m
Average dimension of first floor columns < 25 cm
Slender columns
Squat columns (for the presence of ribbon windows, staggered floors, cranked beams)
irregularities in plan distribution (non compact or asymmetric plan, presence of strong eccentric
stiffening elements, presence of high concentrations of eccentric masses, presence of wide or heavily
loaded cantilevers)
irregularities in elevation (high variation of floor area, high variation of floor level distance, varied
overloads at different levels, high stiffness variations in vertical structure, significant super-
elevations, slender columns with heavy slabs)
Irregularities in infill panels disposition (absence or reduced presence of infill panels at the first
level, presence of soft planes, columns behind infill panels, presence of infill panels external to the
structural frame, presence of bow windows)
irregularities in the structural frame (absence of column at the intersection of two beams, presence of
staggered columns, presence of beams external to the column line, beams with reduced width
compared to the supporting column, frames in one direction and no connection beams, prevalence of
thickness beams and weak infill panels, frames with high beams on the perimeter with weak infill
panels and thickness beams)

163

Figure 9. Excerpt of vulnerability factors for R.C. structures.

2.3 Measures to eliminate vulnerability factors
The analysis of the damage surveyed in previous seismic events has shown that some
vulnerability factors are very diffused in the building heritage of the historic town centres either
in Italy or in other European countries.
In the following Table 5 the most frequents vulnerability factors and the relevant possible
interventions to reduce their negative effects are listed. This kinds of structural improvements
are particularly indicated when a full seismic upgrading is not possible and a priority program of
strengthening has to be traced.
As stated before, structural retrofit is the first issue in case of building refurbishment, both in
case of ordinary intervention and rehabilitation after a seismic event. To fully evaluate also
environmental impacts and cost factors in order to implement LCA analysis, each of the
proposed mitigation solutions has to be put in relation with material and energy consumption
required for its application.
Once identified the possible strengthening measures according to requirements of safety and
reliability, environmental and economic parameters become useful indicators for the choice of
design solution.


164
Table 4. Parametric cost of most common strengthening interventions for R.C and masonry buildings.
Strengthening
Cost
(/sm)
Application of FRP to R.C. structural elements 230
Application of FRCM to R.C. structural elements 220
Application of FRP to R.C. joists 125-150
Application of FRCM to R.C. slabs 100-120
Application of FRCM on infill panels 65-80
Application of FRCM on masonry walls 65-80
Iron ties insertion on masonry walls 5-7
Application of CAM system to masonry walls 100-120
Table 5. Diffuse vulnerability factors and relevant possible mitigation measures.
Vulnerability Factors Strengthening
Poor quality of the connections
between walls and between
walls and floors

Large distance between walls

Improvement of the connections among the walls through the use
of a ring beams at floor levels
Riveting of anchorage between the wall quoins and between walls
and floor.
Application of composite material bands to hoop the masonry at
floor level, or improvement of the connections between the
walls in the quoins
Small resistance area in one or
two directions (i.e. for high
percentage of openings or
small thickness of the walls)
and presence of floors of
notable mass
Jacketing with electro-welded mesh and gunite
Introduction of new shear masonry walls with a good embedment
to the existing walls
Jacketing with composite material
Treatments with epoxy resins
Deformable floors in his own
plan with poor connections

Presence of heavy and/or
thrusting roof and lack of
connections between
orthogonal walls and/or of
ties or ring beams
Increment of the floor stiffness with slabs of light concrete and
electro-welded mesh, cross ties, etc. with a good connections to
the masonry and to the adjacent fields of floor
Realization of ring beams with a good embedment to the masonry
Realization of ties to eliminate the thrust of the roof and/or
application of composite material bands to hoop the masonry at
level floor.

2.4 Safety index assessment
A procedure to evaluate a Safety Index I
A
to be associated with the surveyed building has
been proposed, based on the analysis of the evolution degree of the prevalent mechanism
recognized in the building.
Of course this index couldnt be assumed as an automatic judgement for the safety
assessment, but it offers to the operators an interesting parameter for heuristic evaluations.
The method is based on the analysis of the evolution level of the collapse mechanism
identified (or potential) as prominent for the building; to this purpose the surveyor assigns a
quotation to the state of progress of the mechanism: 1 = low, 2 = severe, 3 = very severe,
according to the damages surveyed.
The score assignment for each possible collapse mechanism is related to the vulnerability
factors or damages identified. A matrix highlights the compatibility between TVFs and CMs in
order to help the surveyor to proper evaluate the existing conditions.

165

Figure 10. CM/TVF compatibility matrix for masonry structures.

Figure 11. CM/TVF compatibility matrix for R.C. structures.
It is possible that the surveyor can detect the presence of damages or vulnerability factors
hard to link with any mechanism, In this case, these can be considered in the procedure only if
they show a compatibility between each other and there is uniqueness in the assignation of
MEDEA mechanism, with a minimum value of 1.
The procedure allows to identify the growing of collapse mechanisms, that can be considered
potential, where no visible damage is detected but one or more vulnerability factors are
recognized; at an initial or advanced stage, when structural integrity is compromised by degrade
conditions due to age, lack of maintenance or past destructive events.
This allows to define the prevalent mechanisms to which a synthetic value is assigned,
corresponding to the Safety Index I
A
of the building.
From the elaboration of the Safety Index is then possible to put in relation the existing
conditions of the building to the expected service life remaining if no structural rehabilitation
measures are provided, identifying at the same time the possible strengthening interventions
needed, to be assumed in LCA and LCC analyses.
In order to reduce the risks of personal interpretations of the surveyor and to provide a more
reliable safety analysis, an importance coefficient C
I
is assigned (Table 6), that is proportional
to the potential for each mechanism to bring to structural collapse.
The aim is also to distinguish local collapse from global collapse conditions and, in
the second case, the damages that can bring more rapidly to the structural failure of the
building.
In the case of masonry buildings, for instance, in plane global mechanisms often determine
a good response of structural elements, while the presence of out of plane mechanisms is
evidence of a much more instable equilibrium that can more rapidly evolve through structural
collapse.


166
Table 6. Importance coefficient for collapse mechanisms of masonry structures.
Code Mechanism C
I

M1 Shear on the wall for in plane actions 0.5
M2 Shear on the wall for in plane action localized only in upper parts 0.5
M3 Overturning of entire wall 1
M4 Partial overturning of the wall 1
M5 Vertical instability of the wall 1
M6 Bending crack of the wall 1
M7 Shift of horizontal plane 0.5
M8 Foundation failure 0.5
M9 Irregularity between adjacent structures 0.5
M10 Loss of bearing or shift of floor beams from support wall 0.5
M11 Failure of lintels 0.25
M12 Irregularity of the material, local weaknesses, etc. 0.25
M13 Overturning of tympanum wall 0.25
M14 Overturning of quoin upper part 0.25
M15 Overturning of attic supporting walls 0.25
M16 Rotation of arch or wall abutments 0.25

The building Safety Index I
A
evaluation procedure is then articulated through the following
steps:
a) Definition of the average level of progress associated to the i-th mechanism with reference
to the observed compatible damage (this operation is made for each collapse mechanism
identified by the surveyor and eventually for those automatically interpreted through an
expert evaluation of spurious damages or vulnerability factors:

i
n
j
ij
i
k
Ld
Ld

=
=
1
(1)
where ij Ld = level of severity of j damage in relation to i mechanism, with value variable
between 0 and 3; n = total damages number of MEDEA abacus; k
i
= number of damages or
vulnerability factors compatible with i mechanism.
b) Further calibration of the value obtained with (1), mediating it with the maximum value
between the scores associated with the damages compatible with i-th mechanism, obtaining:

2
max Ld Ld
Ld
i
i
+
= (2)
where ) ( max
i
Ld Max Ld = .

c) Definition of P
i
index, expressing the level of activation of i-th mechanism:

I i
i C Ld P = (3)
where C
I
= importance coefficient of the mechanism proportioned to the potentiality of that
mechanism to lead rapidly to the collapse of the structure (Table 6).

d) Normalization, after identifying as prevalent mechanism the one associated with the
higher P
i
index, in relation to the maximum ij Ld value (equal to 3), in order to obtain the Safety
Index I
A
of the building, comprised between 0 and 1:
3 max i P I
A
= (4)
where max i P = P
i
index associated with the prevalent mechanism recognized.

167
Finally a correlation between the Safety Index range and the global damage has been
investigated, defining media and deviation standard values. These results, compared to the real
safety judgements collecting on the fields, have allowed to find a bound value for the Safety
Index. Therefore a probabilistic correlation, in terms of percentage of unsafe buildings respect
to the total buildings showing a certain level of global damage, has been defined, as showed in
Figure 12.

Figure 12. Distribution of the Safety Index in relation with the global damage levels.
Of course, future development of the procedure, added to a larger amount of data set, will
allow to define a more confident threshold value for the Safety Index and more refined density
probability functions.
This perspective represents an interesting cue to reach quick safety evaluations to be used in
Residual Service Life estimation within LCA and LCC studies, as well as to define possible
alternatives for refurbishment interventions or mitigation action on buildings inside seismic
risk-prone areas.
3 CONCLUSIONS
The integration of support tools within LCA and LCC analysis is required in order to define
reliable environmental and economic evaluations according to specific design and construction
issues. The MEDEA method outlines a general approach to understand the structural
vulnerability factors and the behaviour of a building under seismic action. It can be used:
in the post-event phase: as tool to perform on the basis of the cracking path and of the
potential collapse mechanisms about to be activated the safety check of the buildings;
in the vulnerability assessments: as tool to identify vulnerability factors responsible of
potential activation of collapse mechanisms and to evaluate possible interventions of
improvement and upgrading as mitigation measure to reduce the effects of the vulnerability
factors, in order to parameterise the expected service life.
MEDEA not only allows to describe the main structural typologies but provide also an
implicit analysis of the possible behaviour of the structure related to the identification of
structural lacks and vulnerability factors of the building, as well as the potential collapse
mechanisms.
The tool provides a complete set of simplified numerical models at the limit state of collapse
for the main kinematics mechanisms. This will allow to define the structural vulnerability as the
probability to trigger mechanisms and to correlate these to a synthetic evaluation of
vulnerability and its expected service life.
The future goal of such a procedure is to propose a damage measure for structural elements
with the attribution of a Damage Index for every element based on the indication suggested in
the survey form and in the iconographic part.


168
REFERENCES
ISO 2001, ISO 15686-2:2001. Buildings and constructed assets. Service life planning. Part 2: Service life
prediction procedures.
ISO 2002, ISO/TR 14062:2002. Environmental management Integrating environmental aspect into
Product design and development.
ISO 2006, ISO/TS 21929-1:2006. Sustainability in building construction - Sustainability indicators -
Framework for development of indicators for buildings.
ISO 2006, ISO/TS 21931-1:2006. Sustainability in building construction Framework for methods for
assessment of environmental performance of construction works.
ISO 2007, ISO 21930:2007. Sustainability in building constructions Environmental declaration of
building products.
Campioli, A. Lavagna, M. 2007, Integrating Life Cycle Assessment in Building Environmental and
Energy Certification, Prooced. International Conference Sustainable Building South Europe, 7-8
June 2007, Turin. Torino: Celid.
Campioli, A. Lavagna, M. 2007. Life cycle design in building and construction sector, Proceed. 3rd
International Conference on Life Cycle Management. From analysis to implementation, Zurich,
Switzerland, 27-29 August 2007.
Fernandez, J. 2006. Material Architecture, emergent materials for innovative buildings and ecological
construction, Oxford: Elsevier Architectural Press.
Kawai, K. et al. 2005. Inventory data and case studies for environmental performance evaluation of
concrete structure construction, Journal of Advanced Concrete Technology, 3 (3): 435-456.
Kawai, K. et al. 2005. A proposal of concrete structure design methods considering environmental
performance, Journal of Advanced Concrete Technology 1 (3): 41-51.
McDonough, W. Braungart M. 2002. Cradle to Cradle: Remaking the Way We Make Things. New York:
NPP.
Neri, P. (ed). 2008. Verso la valutazione ambientale degli edifici: Life Cycle Assessment a supporto della
progettazione Eco-Sostenibile, Firenze: Alinea.
Angeletti, P. Bellina, A. Guagenti, E. Moretti, A. Petrini, V. 1988. Comparison between Vulnerability
Assessment and Damage Index, Some Results, Proceed. 9th World Conference on Earthquake
Engineering, 9WCEE, Tokyo-Kyoto.
Baggio, C. Bernardini, A. Colozza, R. Corazza, L. Della Bella, M. Di Pasquale, G. Dolce, M. Goretti, A.
Martinelli, A. Orsini, G. Papa, F. Zuccaro, G. 2000. Manuale di istruzioni per la compilazione della
Scheda di I livello di rilevamento del danno, pronto intervento ed agibilit nellemergenza post-
sismica (AeDES), Roma: Dipartimento della Protezione Civile.
Bernardini, A. Gori, R. Muneratti, E. Paggiarin, C. Parisi, O. Zuccaro, G. Vulnerability and probability of
collapse for classes of masonry buildings. In Faccioli, E. Pessina, V. (ed). 2000. The Catania Project:
earthquake damage scenarios for high risk area in the Mediterranean, Roma: Esagrafica.
Braga, F. Dolce, M. Fabrizi, C. Liberatore, D. 1986. Evaluation of a Conventionally Defined
Vulnerability of Buildings Based on Surveyed Damage Data, Proceed. 10th European Conference on
Earthquake Engineering, Lisbon.
DAyala, D. Spence, R.J.S. Oliveira, C.S. Papa, F. Zuccaro, G. 2000. The Performance Of Strengthened
Masonry Buildings In Recent European Earthquakes. Proceed. 9th World Conference on Earthquake
Engineering, 12WCEE, Auckland.
Dolce, M. Zuccaro, G. Papa, F. Masi, A. 2002. Remarks On The Seismic Damage In The Recent
Earthquakes In Europe. Proceed. 12th European Conference on Earthquake Engineering, London:
Elsevier Science Ltd.
Papa, F., Zuccaro, G. 2004. Un modello di valutazione dell'agibilit post-sismica attraverso la stima dei
meccanismi di collasso. Atti dellXI Congresso Nazionale Lingegneria Sismica in Italia, Genova.
Zuccaro, G. Papa, F. 1999. Inspection minutes process for the measurement of the seismic damage. In
Bernardini, A. (ed). Seismic Damage to masonry buildings, Rotterdam: Balkema.
Zuccaro, G. Papa, F. 2002. Multimedia handbook for seismic damage evaluation and post event
macroseismic assessment, 50th Anniversary of the European Seismological Commission (ESC):
XXVIII General Assembly, Genova..
Zuccaro, G. Papa, F. 2001. CD Multimediale MEDEA Manuale di Esercitazione sul Danno Ed Agibilit
per edifici ordinari in muratura. Roma: CAR Progetti.
Zuccaro, G. Cacace, F. 2006. Valutazione speditiva della Vulnerabilit per gli edifici strategici della
Regione Campania, Ingegneria Sismica 2.
Zuccaro, G. Cacace F., Rauci M., 2010. MEDEA: a multimedia and didactic handbook for structural
damage and vulnerability assessment, this issue.

Condition assessment
169
B. Khn
Verheyen Ingenieure GmbH & Co. KG, Bad Kreuznach, German
O. Hechler
ArcelorMittal, Sections S.A., Technical Advisory, ArcelorMittal, Esch-sur-Alzette, Luxembourg
Survey and Condition assessment
1 INTRODUCTION
During the last decades, focus has been put on inspection of the existing bridge stock and
related condition assessment procedures. In many countries bridge management systems have
been introduced. In parallel, recommendations for assessing the fatigue safety become in the
interest of many owners of old steel and iron bridges. Many of these bridges have their origin in
the second half of the 19
th
and turn of the 20
th
century. With the rapid development of new
materials and new calculation methods in these decades a lot of steel bridges were built, of
which many are still in use today.
The assessment of these bridges respectively the remaining fatigue life of its members is
difficult to estimate. Effectively, the assessment of fatigue safety was not included in the design
at the turn of the 19
th
to 20
th
century. Today, the old materials are not commonly known
anymore. Often, especially for road bridges, the load history is not reported. Further, changes in
the structure after repair or strengthening measures within the last hundred years may not be
indicated in the drawings. Information may got lost during the two world wars. In several cases
the loading of a bridge had to be increased drastically or even a new lane was added. In all these
cases, a re-assessment of fatigue is compulsory.
The recommendations presented in this chapter comprehend analyses methods for fatigue, old
material identification and rehabilitation. Once the fatigue damage is recognized, targeted repair
and strengthening measures can be decided. Consequently a clear definition of these tasks is
included by explaining the general applicability of the remedial measures to steel structures.
For the traditional assessment of the structure, the current fatigue resistance and the
remaining fatigue life, knowledge on the load histogram is required, e.g. for damage
accumulated from the past. If a crack is found or the load history is unknown, new methods are
to be applied. One potential method is the fracture mechanical approach (FMA). FMA is
neglecting the crack initiation phase, which represents a high percentage of the whole fatigue
life of a structure but focuses on the crack propagation phase only. Consequently with FMA the
remaining fatigue life of the structure is determined leading to a save assessment even though
the past load history is not known. Therefore FMA is additionally introduced in the
recommendation.
2 ASSESSMENT PROCEDURE
2.1 Limitation
The fatigue assessment of an existing structure results in a conclusion on the safety of the
structure subjected to cyclic loading for a specified remaining fatigue life. The application of the


170
recommendation presented herein is restricted to structures under normal environmental
conditions and temperatures between -40 and 150C. Assessment of structures exposed to fire is
not considered. Assessment under low cycle fatigue, as during seismic activities, is either not
included. Finally, the possibility to inspect the element to be assessed must be given.
2.2 Fatigue assessment procedure
For the fatigue assessment a step-by-step procedure is proposed. This procedure is based on a
general procedure developed by the Joint Committee for Structural Safety (JCSS 2001), which
has been enhanced with focus on existing steel bridges exposed to fatigue loading. If the
assessment by this procedure proves a sufficient remaining fatigue life in one phases, the
subsequent phases can be disregarded. The following phases for the assessment are proposed:
I. Preliminary Evaluation: Removal of existing doubts about safety of the
structure using fairly simple methods. Information from visual inspection,
including e.g. information from Bridge Management Systems (BMS) and own
inspection on site. The owner is informed by a first report.
II. Detailed investigation: The engineer may need the help of specialised
laboratories and experts for assistance. Information on the structure and
loadings are updated using specific tools as refined calculation models or more
realistic traffic loads. If the result is negative, further steps and remedial
measures are to be proposed. The outcome is reported in a second report to the
owner.
III. Expert investigation: A refined static model is used for probabilistic evaluation
and fracture mechanics for establishing final decisions. Measurements help to
obtain refined data from the structure and about loading. Advanced NDT may
be used in cross sections specified with the updated model. An expert report
informs the owner.
IV. Remedial measures: Retrofitting of the structure to achieve fitness for purpose
by using special measures such as intensified monitoring, reduction of loads,
change in use, strengthening, repair or rehabilitation. A final report summarises
the results of all working steps. All remedial measures, possible from the
technical point of view are proposed. The report will give all information,
which the owner of the structure needs for an economical decision about further
measures.
Often, the use of the step-by-step procedure leads to a significant extension of lifetime and to
postpone investments in new bridges. It also clarifies, if a bridge is safe without any further
measures or not sufficient anymore with the consequence of strengthening or demolishing. This
proposed stepwise procedure, see Fig.1, aims on the identification of the best strategy for the
optimal investment of life cycle costs.
For further information on each step of the fatigue assessment procedure reference is given to
(Khn et al. 2008).
3 STRUCTURAL AND MATERIAL INFORMATION
3.1 General
To get information on the resistance of a structure, both, field measurements and/or material
investigation are used to obtain the information directly from the structure itself. Data,
characterizing the old steel, design rules or connections between elements differ noticeably from
nowadays standards. That is why design, materials database, calculations and drawings as well
as additional experiences have to be studied first.
3.2 Identification of critical members
Civil engineering structures are mainly large and complex physical systems consisting of
many heterogeneous components. A detailed assessment of all components is economically and

171

Figure 1. Stepwise procedure for fatigue assessment (Khn et al. 2008)
technologically not feasible. Therefore, the assessment has to be restricted to a limited number
of components. Consequently, one major task prior to the assessment is the identification of the
critical details to be focused. Critical details of the bridge are details of structural components
having an outstanding damage risk (e.g. stress amplification). The general approach for the
identification of critical members for riveted steel bridges is given in Fig 2. For further
information reference is given to (Hechler et. al. 2005).


172

Yes
Structural analysis
Identification of
members in tension
Identification of members
in compression
Stability check OK
no failure expected
Determination of risk of breakdown using
fail-safe analysis methods for the members in tension
Does the failure of the member lead to the
breakdown resp. partial failure of the structure?
No
Fail-safe Critical member

Figure 2. Identification of critical members (Hensen 1992)

3.3 Material identification
The most important material characteristics are:
- Chemical characteristic values C, Si, Mn, P, S, N
- Yield strength R
eL
, R
eH
(f
y
)
- Tensile strength R
m
(f
u
)
- Fracture toughness expressed in K
Mat
(K
Ic
), J
Mat
(J
Ic
, J
crit
), K, K
th

which either can be determined in single material tests or, if a sufficiently large database is
available, expressed as a statistical values (mean value and standard deviation). The direct
determination of the material properties using material tests always provides the member-
specific and thus the exact material properties. However, material tests should only be used in
such cases, when statistically validated material characteristics, which are usually lower bound
values, lead to overemphasized conservative results. Therefore material tests to be performed on
samples taken from old steel bridges can be reduced to few cases (Langenberg 1995).
For riveted bridges of the 19
th
Century, wrought steel as well as mild steel were used.
Chemical analyses or sulphur prints can identify wrought steel, also known as puddle iron,
showing their characteristic lamellar microstructure consisting of ferrite matrix and slag layers.
Research from the years 1990ff proved, that fatigue assessment using S/N-curves as well as
fracture mechanic assessment are applicable.
Mild steel has a totally different microstructure, which can be characterized by sulphuric
segregation in the sulphur print. For more information, reference is given to e.g. (Helmerich
2005, Blanchard et al. 1997, Sustainable Bridges 2007).
Specimens, taken from riveted bridges, are distinguished on the basis of their original
production method by means of chemical and/or metallographic analysis. A scheme for this
characterization is presented in Table 3-2 of (Khn et al. 2008) covering steels produced by the
puddling process developed in 1784 or mild steels produced since 1855 by the Bessemer,

173
Thomas or Siemens-Martin process, see Table 1. However, the values given in this table have to
be understood as tendencies due to the high scattering of the chemical analyses as well as the
tension strengths. Other evaluation schemes are given in (Blanchard et al. 1997, Helmerich
2005, Mang et al. 1985).
A rapid development of different steel grades took place in the first three decades of the 20
th

century. Not all of them are considered in Table 1 (respectively Table 3-2 [Khn et al. 2008]).
Old steel structures may also contain of - through low-alloyed steel to high strength steel (not
included in Table 3-2) - the following steels developed: chromium steel (since ~1860), nickel
steel (since ~1908), high-carbon structural steel (since ~1923) and silicon steel (since ~1926)
(Helmerich 2005, Stranghner 2005). If any doubts on the steel grade of the investigated steel
structure are present, material tests are highly recommended especially with view on the
strength, toughness and weldability.
Furthermore, it has to be accounted for, that the quality of the steels themselves might be low,
especially during the years of World War I (1914-1918), the great depression (1929-1939) and
during and after World War II (1939-1950). Steel production had to be fast in these times, and
expensive alloys were not available. Assessing a structure, built during one of these periods,
material tests are strictly recommended.
Table 1. Information on commonly used producing processes
Cast iron Wrought steel
(Puddle steel)
Mild steel
(19
th
century)
Mild steel
(20
th
century)
Produced since 1709 (Abraham
Darby I.)
1784 (Henry Cort) Blast process
1855 Bessemer steel (Henry Bessemer)
1879 Thomas steel (Sidney G. Thomas)
Hearth process
1865 Siemens-Martin steel (Friedrich and
Wilhelm Siemens, Pierre and Emile
Martin)
Produced till today ~ 1900
~ 1900 Bessemer steel in steel structures
~ 1980 Thomas steel,
~ 1990 Siemens-Martin steel
Producing
process
Re-melting pig
iron by means
of coke and
cleaning
1. Puddling process:
Conversion of pig iron
into steel by stirring in a
pasty state carried out in
a reverberatory furnace
puddle balls
2. Mechanical slag
removal and welding
together of lumps of
steel by means of a
steam hammer or
squeezer
blooms
3. In the following years
further treatment by
heating and rolling in a
rolling mill
muck bars
Bessemer process
Air-refining process in which low-phosphorus
pig iron is converted into liquid steel in the
Bessemer converter consisting of an acid lining by
blowing air through the molten pig iron excess
carbon and impurities, such as silicon and
manganese, are oxidized acid Bessemer steel
Thomas process
Same process as Bessemer process but also for
high-phosphorus pig iron which is possible by a
basic lining based on dolomite (Thomas
converter) basic Thomas steel
Siemens-Martin process
Allothermic hearth process, in which pig iron with
or without steel scrap and/or iron ore is melted in
the Siemens regenerative furnace excess
carbon and impurities are burnt out basic
Siemens-Martin steel

As additional information, for modern steels from the middle of the 20
th
century the Thomas
process was replaced by the oxygen blowing process and further, to the end of the 20
th
century


174
the Siemens-Martin process was replaced successively by the electric arc process. The excellent
material properties of these steels are documented in e.g. (AMCS 2009).
Based on a statistical analysis (hypothesis testing) of the chemical and the microstructure
properties of specimens from riveted bridges it is concluded, that the obtained data for the
strength and toughness of wrought steel can be treated as a statistical homogenous population.
The statistical distribution of the material strength has been derived from an amount of 205 tests
at 0C and 283 tests at -30C from literature, see Table 2.
Table 2. Mech. and chemical properties of puddled iron and early mild steel (rimmed steel) based on tests
Puddled iron Rimmed steel Test, Standard
N/mm
2
N/mm
2

Comment
Tension test: EN 10002, part 1 Available tests ( > 600)
Yield strength R
eL
203 229
Ultimate strength R
m

Elongation min
max
6.0
26.5
22
42
Specimen:
min. diameter: 5 mm (B5)
min. thickness: 4 mm (E4)
Crack propagation tests, ASTM
Paris parameter C max C
min C
4.910
-17
(R=0.1)
5.710
-28
(R=0.5)
0.573910
-13

Paris parameter
m
max m
min m
9.3 (R=0.5)
3.8 (R=0.3)
3.299
Threshold of the cyclic stress
intensity factor K
th
[N/mm
3/2
]
13.49 (R=0.1)
6.36 (R= 0.5)
6.2 (R= 0.3)
Specimen:
SENB3, CT, modified CT
Proposed:
K
th
= 2 MPa m
1/2

C = 410
-13
,
m = 3 (upper limit)
Chemical analysis [%]
1)
[%] available tests
C
0.0032-0.15 0.026-0.20
Si
0.003-0.42 0.001-0.013
S
0.0034-0.018 0.063-0.176
P
0.011-0.39 0.009-0.136
N
0.0037-0.04
3)
0.011-0.022
Mn
0.054-0.11 0.036-0.52
1)
only wet analysis possible
3)
if N is below the soluble
limit (~ 0.014, aging effect
is neglect able)

The results of the statistical evaluation of old steels (except wrought steel) can be summarized
as follows:

+10 C R
eL,5%
= 229 N/mm 230 N/mm

0 C R
eL,5%
= 248 N/mm 240 N/mm
J
Mat, Modell, 5%
= 30 N/mm

-30 C R
eL,5%
= 257 N/mm 250 N/mm
J
Mat, Modell, 5%
= 17.5 N/mm 17 N/mm
For wrought steel a tension strength of R
eL,5%
= 203 N/mm 200 N/mm at +10 C was found.
3.4 S-N-curve determination using full-scale tests
The assessment procedure of phase I refers to S-N-curves for the detail to be evaluated. Riveted
details have been classified in the detail category 71 based on results of a great variety of full-
scale tests on original bridges and bridge elements from all over the world - this S-N-curve for
detail category 71 is shown in Figure 3. For the fatigue assessment of existing welded steel
structures, the detail categories given in Eurocode 3, part 1-9, should be used. Please note, crack

175
growth values from tests for specific structures resulted in higher values, corresponding to detail
category up to 90, investigation of the real elements can increase the estimated fatigue life
significantly.

Figure 3. SN curve proposed for fatigue assessment of riveted structures
Further, latest research results start the speculation about a differentiation of detail categories for
riveted members (Taras et al. 2009).
3.5 Inspection
The recommendations (Khn et al. 2008) concentrate only on fatigue. Consequently, the
information about inspection, measurement and remedial measures are limited on possible
damages caused by fatigue.
The older the bridges are, the more a bridge accumulates partial damages from service load
cycles. Depending on the type of structure different reasons may cause initial and fatigue
damages to be evaluated during inspection. Initial damages can be caused
- during fabrication, welding or riveting,
- due to unfavorable design with regard on fatigue (poor detailing),
- due to stresses and deformations unforeseen in design,
- or because the state of knowledge was too low.

During visual inspection almost all damages are detected. Thus visual inspections are essential.
In many countries, main inspections are performed all 5-6 years. Intermediate in the main
inspection interval, some countries perform simple inspections or inspection due to special
reasons. The bridge owners, such as state highway agencies or railway owners, fix the
requirements in inspection recommendations. During the last years, Non-Destructive Testing is
being more and more applied. The recommendations presented in this paper cover the
discussion on advantages and disadvantages for different applications. Table 3 shows available
NDT methods proposed for old riveted structures to be applied in different levels of the
assessment.

Fatigue limit: 29 N/mm
2
CAFL: 52 N/mm
2


176
Table 3. Non-destructive inspection tools for detection of fatigue damages
Method Damage to be detected Comment on effectiveness of the equipment
Visual inspection Surface cracks With help of magnifying glass and training
Magnetic particle test Surface cracks Only for magnetic materials
Dye penetration test Surface cracks Good alternative to magnetic particle test, also
for non-magnetic materials
Radiography Crack detection in sandwiched
element
Expert with permission required
Ultrasonic test Only the first layer can be assessed Not applicable for puddle iron
Eddy current technique Crack in rivet holes Not currently applied
Acoustic emission
technique
Active cracks Not applied for detection (insufficient research),
only for monitoring of detected cracks
3.6 Measurements
If a sufficient safety level cannot be shown by means of calculation, static and dynamic
measurements shall be performed. At this stage, the assessment is generally entering phase III
and aims to get detailed information from the structure. Although, with good experience and
guess, valuable measurements are already possible in phase II.
The objective of a measurement is always to gain information, either on the resistance or on
the loading of the structure, in order to reduce the uncertainties associated with the static
calculations made in design or in earlier fatigue assessment phases.
Measurements are appropriate for the fatigue assessment, if e.g.
- doubts about the acting static system are present,
- effects not known during design occurred,
- effects due to increased loads or additional lanes have to be assessed,
- Secondary stresses, which caused fatigue damage, need to be identified.
Table 3 gives an overview on physical dimensions to be measured and the related sensors to be
used.
Monitoring is a repeated collection of data of defined measuring points, to obtain information
about changes in the system or loading during a chosen time interval. In general, measurements
are used to obtain real information from structures; for fatigue assessment relevant strain
concentration regions with the minimum amount of sensors, e.g.:
- Strain distribution at high loaded cross sections,
- Critical elements, as anchors or braking trusses,
- Evaluation of the actual zero axis,
- Secondary stresses,
- Moments in fixed supports or restraint,
- Movement in bearings,
- Measurement of strains in theoretical zero elements,
- Local strain concentrations in connections assumed to be hinges/ joints.
3.7 Remedial measures
In the recommendations presented remedial measures concentrate only on fatigue damages.
Fatigue damage appears as a fatigue crack in details with the highest sensitivity to fatigue.
Design of remedial measures has to be done carefully, because the strengthening of a detail can
lead to changes in notch details or load re-distribution. Consequently remedial measures may
lead to additional fatigue sensitive details, if not smartly designed.
The typical fatigue details differ between welded and riveted structure. If the reason for a
fatigue crack is identified, e.g. by calculations or by intensifying monitoring, then only a proper
solution for remedial measure can be decided on. Remedial measures are:
- Reducing loads,
- Repair,
- Strengthening,
- Demolish structure.

177
Table 4. Measurement tasks and preferred sensors
Measurement task Preferred sensors Accuracy
Local strain distribution in
cross sections with assumed
high stress level or cross
sections with changes, caused
by damages
Strain gauges
Mechanical strain measurements
Fibre optical sensors (integral
sensors, Fabry-Perot sensors, Brag-
grid sensors)
2

0.5 1 depending on type of fibre
sensor and application
Actual static system System of strain gauges, Dynamic
measurement system

Influence of temperatures Temperature sensor (resistance based)
Termocouples (Semiconductors)
~0.1C (depending on the max.
temperature range)
Inclination Inclinometer Depending on the chosen max. angle (see
producer, ~0.1)
Displacements or settlement Potentiometer
Inductive measurement
Laser based system
Hydrostatic leveling / force
transducer
0.1%, depending on displacement
0.1%, depending on displacement
0.3mm + influence of measured distance
~0.1%, depending on max. force
Dynamic characteristics:
Acceleration Eigen
Accelerometers
Strain transducer
0.05 1% depending on force

A collection of typical causes for fatigue damages and a collection of details known to be
sensitive to fatigue are presented in the recommendation. Information about both, cracks in
bridges, given by consulters and results from full scale fatigue testing, are taken into account
and analyzed regarding their possible remedial measures. Fatigue damage causes for welded
structures refer to the database of the working group 5 from the International institute of
welding (IIW 2009). Fatigue causes for riveted structures are collected among the participating
institutes.
3.8 Case study
The recommendations contain an annex with a case study, in which the typical fatigue
assessment procedure is exercised.
The study deals with the assessment of a riveted single span bridge in Slovenia with two
equal trusses as main girders. In 2000, after over 100 years of service, the assessment has been
required as the bridge has reached the theoretically end of its design life. In this example,
calculation was done for the diagonal element of the main truss, which has been identified to
represent the most critical element regarding fatigue assessment. The first phase of the
assessment has been based on information on number of trains and transported tons in the past
and shows using the damage accumulation approach, that no remaining fatigue life is present.
Due to the fact, that the member chosen for the analysis is considered to be unsafe, further
measures are necessary. The proposal of these measures has additionally been included in the
example.
4 CONCLUSIONS
A step-by-step fatigue assessment procedure has been introduced, suitable to be applied by
practicing engineers for the evaluation of old steel structures respectively bridges, exposed to
dynamic loading. The proposed assessment procedure is divided into 4 phases. Consulting of
experts, non-destructive testing methods, measurements and analyses of the material are
assigned to different phases of the assessment. The procedure is designed to enable a practicing
engineer to carry out the first phase of the assessment on his own and, upon the results, give
advice to the owner. At the end of each phase, the owner has to take decisions based on a report
which shall help him to identify the most efficient solution for the further use of the bridge. The


178
proposed procedure is a milestone in knowledge transfer from scientific laboratory towards
practicing engineers.
ACKOWLEDGEMENTS
The authors would like to thank the members of working group A of the ECCS Technical
Committee 6 Fatigue for the agreement of publication and the technical collaboration for this
chapter. In particular thanks to Rosemarie Helmerich (BAM) und Alain Nussbaumer (EPFL),
who have significantly contributed to the content of this chapter.
REFERENCES
ArcelorMittal, Long Carbon Europe: Sections and merchant bars, Sales Programme, 2009-1,
Luxembourg
Blanchard, J., Bussell, M., Marsden, A., Lewis, D.: Appraisal of existing ferrous metal structures,
(Journal) Stahlbau 66 (1997), Heft 6, pp. 333-345, Ernst & Sohn Verlag, Germany.
Helmerich, R.: Alte Sthle und Stahlkonstruktionen Materialuntersuchungen, Ermdungsversuche an
originalen Brckentrgern und Messungen von 1990 bis 2003, BAM, Forschungsbericht 271, Berlin,
2005, ISBN 3-86509-362-0.
Hechler, O., Khn, B., Lsche, A., Sedlacek, G. 2005. Guidelines for monitoring of steel railway bridges.
WP5-19-T-050331-D-D5.2.1. Sustainable Bridges.Research project FP 6 - PLT 001653.
Hensen W.: Grundlagen fr die Beurteilung der Weiterverwendung alter Stahlbrcken. Dissertation
RWTH Aachen, Schriftenreihe des Lehrstuhls fr Stahlbau, 1992.
IIW, Working Group 5, http://iiw-wg5.cv.titech.ac.jp/reference.htm.
JCSS, Joint Committee on Structural Safety, Probabilistic assessment of existing structures, Rilem
publication, 2001.
Khn, B. Luki, M., Nussbaumer, A., Gnther, H.-P., Helmerich, R., Herion, S., Kolstein, M.H.,
Walbridge, S., Androic, B., Dijkstra, O., Bucak, .: Assessment of Existing Steel Structures:
Recommendations for Estimation of Remaining Fatigue Life, (Background documents in support to
the implementation, harmonization and further development of the Eurocodes), JRC Scientific and
Technical reports, Joint Report Prepared under the JRC ECCS cooperation agreement for the
evolution of Eurocode 3 (programme of CEN / TC 250), Editors: G. Sedlacek, F. Bijlaard, M.
Gradin, A. Pinto and S. Dimova, First Edition, February 2008, EUR 23252 EN, ISSN 1018-5593,
JRC 43401.
Langenberg, P.: Bruchmechanische Sicherheitsanalyse anrigefrdeter Bauteile im Stahlbau,
Dissertation, RWTH Aachen, Shaker Verlag, Band 14/95, 1995, ISBN 3-8265-1102-6.
Mang, F., Steidl, G., Bucak, .: Altstahl im Bauwesen, Schweien und Schneiden 37 (1985), Heft 1, S.
10-14.
Stranghner, N.: Werkstoffwahl im Stahlbrckenbau, Habilitation, TU Dresden, 2005.
Sustainable Bridges: Assessment for Future Traffic Demands and Longer Lives, Sixth Framework
Programme. 2003 2007. D1.2 European Railway Bridge Demography.
Taras, A., Greiner, R.: Development and Application of a Fatigue Class Catalogue for Riveted Bridge
Components, Research Report, TU Graz, ECCS TC 6 2009.

179
A. Taushanov
Department Structural Mechanics, Faculty of Structural Engineering, University of Architecture, Civil
Engineering and Geodesy, Bulgaria
Ch. Giarma
Laboratory of Building Construction and Building Physics, Civil Engineering Department, Aristotle
University of Thessaloniki, Greece

Concrete deterioration: Specification and assessment
1 INTRODUCTION
Reinforced concrete is a material that is widely used for the formation of buildings and
other civil engineering constructions bearing components. The wide use of reinforced concrete
globally is due to, among others, its high load bearing capacity, to its capability of providing
components of various shapes and forms, to the high level of safety it provides against seismic
loads as well as to the possibility of improving its performance and adapting its properties to its
service environment with the use of admixtures or with the application of certain technologies
(e.g. prestressed concrete). In fact, in some countries, such as Greece and Bulgaria, the
overwhelming majority of buildings bearing structure (buildings included) is made of
reinforced concrete. The reason for this preference lies mainly on local conditions, such as the
lack of local/ regional structural timber, the quick development of concrete industry, the need
for protection against the frequent and strong earthquakes in Greece and Bulgaria, which could
be fulfilled by the massive concrete structures, the design and construction of high residential
buildings even in the centres of urban areas, etc.
It is not only the vast use of reinforced concrete, but also its critical role in the constructions,
extending beyond the functionality and reaching the very essence of safety and structural
integrity, that set the demand for its timely and effective maintenance and repair. This need
becomes clearer when not only the structural, but also the social, economical and environmental
issues related to the subject are taken into consideration. As the building stock is ageing, the
building materials degradation (concrete included) increases and the demand for quick and
efficient interventions (regular maintenance, effective repair) becomes urgent.
The effective interventions on deteriorated concrete (as for any other material) presupposes a
series of actions, including the identification of the degradation type, the detection of the
problems origin, the analysis of the deterioration mechanisms and the assessment of the
destruction degree. It is only after this procedure that the right (in kind and in scale) measures
can be taken and the appropriate interventions can be schedules and performed. The field of
concrete deterioration is the central thematic axis of the present paper. More specifically, in the
first part of the paper, the main degradation mechanisms of reinforced concrete are presented in
their general context, organized in categories, depending on the nature of their action (physical,
chemical, biological/ biochemical). The second part of the paper is focused on the inspection of
deteriorated concrete, on the specification and the assessment of the concrete components
deterioration type and degree respectively. Several methods of identification and assessment are


180
reviewed, applicable at different stages of investigation and based on principles and
technologies ranging from simple and empirical to complex and technologically advanced.
2 DETERIORATION MECHANISMS OF CONCRETE STRUCTURES
2.1 Introductive discussion
Concrete, being a composite material (consisting of cement, water, aggregates, sand, etc.) is
susceptible to the degrading agents that attack its constituents separately. Additionally, the final
product of concrete, functioning as one body, is susceptible to mechanisms that attack exactly
these properties that ensure the cooperation and the successful merge of its constituent elements.
Adding the reinforcement bars to this rich in various elements mix, increases the number of
deteriorating agents and processes that attack reinforced concrete.
Obviously, the mass of concrete is, in principle, not threatened by the same mechanisms as
the reinforcement bars, e.g. the electrochemical corrosion degrading the steel embedded in
concrete is not attacking, as a process by itself, the mass of concrete. Nevertheless, there is a
strong interaction between the mechanisms attacking the mass of concrete and the reinforcement
bars when their progress in time is considered. For example, in the case of the electrochemical
corrosion mentioned before, for this corrosion to initiate and propagate, certain compounds must
have penetrated the mass of concrete cover; once the corrosion begins, the reaction products
occupy more volume than the initial elements leading this way to the formation of cracks in the
concrete cover (mass of concrete); the formation of cracks leads in turn to the facilitation of the
penetration of several elements that not only attack the mass of concrete, but also contribute to
the further corrosion of the steel bars. This example is rather illustrative and representative of
the complicated interrelationships among the deterioration mechanisms of reinforced concrete.
This is also the case when the classification of deterioration mechanisms into physical,
chemical, etc. is attempted. A mechanism chemical in its causes and nature can result in a
secondary deterioration type of physical character. Such a classification of possible reinforced
concrete deterioration mechanisms in physical, chemical and biological mechanisms and factors
is the structural basis of the following section.
At this point, it should be stressed that the focus is on the presentation of environmentally-
induced deterioration types. More specifically, reinforced concretes problems can be roughly
categorized into four sections (Rendell et al., 2002):
Internal design errors,
built-in problems (e.g. use of inappropriate aggregate),
construction defects and
environmental deterioration.
The focus of this paper is on the last section. For example, drying shrinkage of concrete,
which can lead to the formation of cracks, will not be addressed.
2.2 Physical mechanisms
2.2.1 Freeze-thaw cycles
Freeze-thaw deterioration is due to the formation of ice crystals within the concrete matrix.
Damage is due to internal pressures caused by the expansion that occurs when water is
converted into ice. This expansion results in internal forces to the solid matrix of the material;
these forces, when exceeding the tensile strength, lead to the formation of cracks. The damage
induced by frost action is usually a result of repeated freezing and thawing cycles, since the
expansion caused by re-freezing after thawing is bigger and the action of the phenomenon is
cumulative (Neville,1995). Not only the bigger pores, but also smaller pores of concrete
participate in the phenomenon (Vesikari et al., 2009). In several studies (Sun & Scherer, 2010)
it is cited that the damage during frost action is due to both hydraulic and crystallization
pressures. Deterioration caused by freeze-thaw cycles occurs in form of spalling, scaling and
cracking.
Some aggregates are themselves susceptible to frost action. This deterioration usually

181
manifests itself with the appearance of pop-outs on the concretes surface (Illston & Domone,
2002).
2.2.2 Salt crystallization
Salts are transferred in the mass of concrete through the movement of the water (capillary
conduction is the main, with regard to contribution, water transfer mechanism). The evaporation
of the water in the big pores leads to the development of crystals (and therefore of
crystallization pressures). Also in this case, cycles of dissolution and crystallization occur and
the damage appears with the form of surface scaling similar to the one induced by freeze-thaw
attack (Rendell et al., 2002). Generally, in building materials, the progress of this deterioration
mechanism depends on the ratio {big pores (vol. %)/small pores (vol.%)}.
2.2.3 Efflorescence
The transfer of soluble salts at the surface of the concrete where evaporation occurs leads to the
formation of salt deposits on this surface (efflorescence). Also, leaching can result in
efflorescence (Neville, 1995).
2.2.4 Abrasion
Abrasion to the surface of concrete is caused by solid particles carried by a fluid. More
specifically, water flowing in contact with surfaces of hydraulic structures and carrying solids
such as sand and gravel causes abrasion of the concrete surfaces. The abrasive forces developed
depend, among others, on the flow velocity. Abrasion is a characteristic pathology type of
concrete channels, sewers, pipes, etc. Abrasion leads to the removal of surface layers and to the
opening of the concrete matrix and the pore structure to further attack (Rendell et al., 2002).
2.2.5 High temperature impact (fire damage)
Several types of deterioration can be caused by fire: surface cracking, spalling, change of colors
to pinkish and at higher temperatures to white/grey and then to buff (indicating change of
phases in the cement paste and aggregates), etc. The changes occurring in concrete due to fire
result in loss of strength. The percentage of strength lost depends on various factors, e.g. the
mineralogical composition of the aggregates. In case that reinforcement bars are affected or
exposed to fire, the deterioration is enhanced.
2.3 Chemical mechanisms
2.3.1 Reinforcement corrosion
In some references, reinforcement corrosion is listed under the physical mechanisms of concrete
deterioration. This is due to the fact that reinforcement corrosion, leading to the formation of
products that occupy greater volume than the steel, creates internal stresses in the concrete
matrix. These stresses lead to concretes cracking. This mechanism of crack formation due to
internal stresses is by itself - in other words when the result for the concrete mass is under
consideration- of physical nature.
Nevertheless, the corrosion mechanism itself (electrochemical corrosion) and the processes
leading to the initiation and occurrence of corrosion of steel in the concrete components
(carbonation, chloride ingress) are of chemical nature. In this context, the reinforcement
corrosion is classified into the chemical deterioration mechanisms category.
In the following, a very short presentation of carbonation and chloride penetration into the
concrete cover is made. An analysis of the reinforcement corrosion itself is presented in
Vesikari et al.
2.3.1.1 Carbonation
In a few words, carbonation of concrete is the reaction of atmospheric CO
2
, that is diffused in
concrete, with cement paste constituents (cement hydration products, such as Ca(OH)
2
, hydrated
compounds like 3CaO.2SiO
2
.3H
2
, but also yet unhydrated constituents like 3CaO.SiO
2
,
2CaO.SiO
2
). In fact the whole process of carbonation is complicated. There are many
physicochemical processes that participate in the phenomenon (Papadakis et al. 1991). The


182
impact of carbonation on the concrete itself is not very important. The structural changes in the
cement paste caused by carbonation lead to the alteration of concretes porosity. It is accepted
that the density of concrete containing Portland cement increases due to carbonation (Steffens,
2000).
The main impact of concrete carbonation is related to the reinforcement corrosion.
Analytically, concrete creates an alkaline environment, in which a tight oxide layer is produced
around the steel, preventing corrosion from occurring. The carbonation of concrete leads to a
drastic reduction of the concrete pore waters pH value, resulting in the destruction of the
passive layer around the steel and in the initiation of corrosion. The progress of carbonation
depends on temperature, relative humidity, ambient CO
2
concentration, concrete quality, etc.
Many models for the calculation of carbonation have been developed (e.g. Papadakis et al.
1991; Saetta et al., 1993; Saetta et al., 1995; Steffens, 2000; Vesikari et al., 2009; Giarma,
2010).
2.3.1.2 Chloride ingress
Apart from carbonation, another chemical process that can lead to the same result (depassivation
of steel and initiation of reinforcement corrosion) is the penetration of chlorides in concrete.
Models calculating chloride penetration (chloride content) in concrete have been developed
(Vesikari et al., 2009). A basic input for these calculations is the critical chloride content. A
literature review on critical chloride content is presented in a recent study by Angst et al. (Angst
et al., 2009).
2.3.2 Sulfate attack
The presence of sulfate salts in concrete (as a result of their transfer into the material, usually in
buried components from the neighboring ground) leads to sulfate attack. More specifically,
sulfates attack the cement paste. Through the occurrence of chemical reactions, ettringite, which
is a product of bigger volume than the initial constituents (expansive reaction), is formed. The
deteriorated concrete has a whitish appearance and later cracks are developed. The rate of the
attack increases with the sulfate content and requires damp conditions. This type of deterioration
has been extensively studied. Concise reviews of the chemical processes and reactions can be
found in several sources (Rendell et al., 2002; Neville, 1995; Illston & Domone, 2002).
2.3.3 Thaumasite attack
This kind of attack also involves the action of sulfates, but it is characterized by distinct
differences from the previous category. It occurs in cold and wet environment, it involves
sulfates and carbonates (usually in the aggregates) and results in the formation of thaumasite.
This attack has serious consequences and can lead to serious damage.
2.3.4 Conversion of Ca+2ions to soluble and insoluble products
In the first case (e.g. action of acid rain) decalcification of concrete occurs, leading to the
opening of concrete matrix, and to the decrease of its pH. In the second case, the surface
porosity of concrete is sealed (Rendell et al., 2002). Of course, in the case of exposure to sea
water (2
nd
group), the cement paste is degenerated and weakened.
2.3.5 Alkali-silica reaction
It occurs due to the reaction of amorphous silicate of certain (reactive) aggregates with the
cements alkali. The products of this reaction occupy big volume and, therefore, lead to the
cracking of concrete.
2.4 Biological/ biochemical attack
It can be encountered in various environments, e.g. waste water treatment (Leemann et al.,
2010). Usually, the concrete surface is attacked.
2.5 Interactions
In the previous sections, some of the concrete deterioration mechanisms were presented. An
important aspect of these mechanisms development and action is their interaction. Indeed, in

183
most cases, the deterioration caused by the action of one degrading factor, facilitates the action
of another factor, leading this way to additional deterioration. There are some cases, in which
the action of a mechanism has an adverse impact on the action of another, e.g. carbonation
increases the surface porosity of concrete, in the preparation of which Portland cement has been
used.
Generally, in most cases, more than one deterioration mechanisms act simultaneously.
3 INSPECTION OF DETERIORATED CONCRETE AND ASSESSMENT OF THE
CONCRETE COMPONENTS DETERIORATION TYPE AND DEGREE
Although corrosion of reinforcement occurs in poor quality concrete throughout the world, more
severe problems are encountered in hot climates, particularly in coastal locations. Vital and
prestigious structures also suffer premature deterioration, as well oil refineries, power stations
and other industrial projects. (Shalaby, Daoud, 1990).
By identifying the causes of concrete deterioration in these important projects is possible to
avoid the repetition of such problems in future. Otherwise, after some years the structure
became extensively deteriorated to the extent that it had to be removed and replaced.
3.1 Diagnosis, on-site assessment and visual inspection
3.1.1 Structural Failure
Actual structural failure, or even structural cracking is only rarely encountered but it is
important to differentiate between cracking from structural and other causes. Such an
assessment should only be carried out by a structural engineer, but an initial inspection by a
materials engineer may highlight other causes of cracking. The materials engineer will know
when the advice of a structural engineer is required.
3.2 Corrosion of Steel
Steel reinforcement is normally chemically protected from corrosion by the alkaline nature of
the concrete. If this alkalinity is lost through carbonation or if chlorides are present which can
break down this immunity, then corrosion can occur. Obviously, when cover is low, the onset of
corrosion will be sooner.
The type of corrosion which occurs varies. Carbonation induced corrosion tends to affect large
areas of the bar causing a gradual loss of section over a relatively wide area. The corrosion
problem is obvious before serious damage can be done because the concrete cover will spall.
With chlorides, a different mechanism often occurs causing very localized severe loss of
section. This can occur without disruption of the cover concrete and almost total of section can
occur before problems become apparent at the surface. Where prestressed steel is used,
catastrophic failures have occurred with no prior warning on structures which had been load
tested shortly before the failure occurred.
3.3 Alkali Silica Reaction
Alkali silica reaction can occur in concretes made with aggregates containing reactive silica,
provided there is a sufficient supply of alkali (usually provided by the cement) and a supply of
moisture. The reaction product is a hygroscopic gel which takes up water and swells. This may
create internal stresses sufficient to crack the concrete.
One of the most frequently found aggregates in affected concrete is chert. This is a common
constituent of many gravel aggregates, but a number of other geological types may be reactive,
such as strained quartz in sands.
3.4 Frost Damage
Concrete of inadequate durability, if subjected to a wet environment and freezing can be
disrupted by a frost attack. Water enclosed in the pores of the wet concrete will expand on
freezing and the high internal stresses so created can disrupt the surface. The effects are


184
intensified by subsequent freeze/thaw action as minute cracks develop which, in turn, become
filled with water.
3.4.1 Shrinkable Aggregate
Some, mostly igneous, aggregates can contain inclusions of weathered material in the form of
clay minerals. These minerals, in common with the clays encountered in the ground, swell in the
presence of moisture and shrink as they dry out. They can cause excessive drying shrinkage of
the concrete and can cause a random crack pattern. The problem was first identified in Scotland,
where it is quite common, but has been observed in whole Great Britain. The cracking can pose
potential structural problems but is more likely to cause loss of durability and is frequently
associated with frost damage.
3.4.2 Chemical Attack
3.4.2.1 Sulfate Attack
Concrete buried in soils or groundwater containing high levels of sulfate salts, particularly in the
form of sodium, potassium or magnesium salts, may be subjected to sulfate attack under damp
conditions. Past experience has shown that true sulfate attack is rare in concrete, only occurring
with very low cement content concretes, with less than about 300 kg/m
3
of cement. As a guide,
levels of sulfate above about 4% of cement (expressed as SO
3
) may indicate the possibility of
sulfate attack, provided sufficient moisture is present. Sulfate attack requires prolonged
exposure to damp conditions.
3.4.2.2 Acid Attack
The most common sources are spillage from acid tanks, acidic groundwater and oxidation of
sewage effluents.
3.4.2.3 Other Contaminant
In addition to those specifically mentioned above, many potentially corrosive substances may
come into contact with reinforced concrete. The extent of their effect may depend on the type of
cement used.
Examples of such materials are certain alkalis, beer and wine (carbonic acid, lactic acid and
acetic acid), vegetable and fish oils, milk (lactic acid), lime, sugar and sulphides.
3.4.3 Fire Damage
Damage to concrete attributable to fire has been summarized in Technical Report No.15 of the
Concrete Society. Three principal types of alteration are usually responsible.
3.4.3.1 Cracking and microcracking in the surface zone
This is usually sub-parallel to the external surface and leads to flaking and breaking away of
surface layers. Cracks also commonly develop along aggregate surfaces presumably reflecting
the differences in coefficient of linear expansion between cement paste and aggregate. Larger
cracks can occur, particularly where reinforcement is affected by the increase in temperature.
3.4.3.2 Alteration of the phases in aggregate and paste
The main changes occurring in aggregate and paste relate to oxidation and dehydration. Loss of
moisture can be rapid and probably influences crack development. The paste generally changes
color and various color zones can develop. A change from buff or cream to pink tends to occur
at about 300oC and from pink to white-grey at about 600oC. Certain types of aggregate also
show these color changes which can sometimes be seen within individual aggregate particles.
The change from a normal to light paste color to pink is most marked. It occurs in some
limestone and some siliceous rocks - particularly certain flints and chert. It can also be found in
the feldspars of some granites and in various other rock types. It is likely that the temperature at
which the color changes occur varies somewhat from concrete to concrete and if accurate
temperature profiles are required, some calibrating experiments need to be carried out.

185
3.4.3.3 Dehydration of the cement hydrates
This can take place within the concrete at temperatures a little above 100oC. It is often possible
to detect a broad zone of slightly porous light buff paste which represents the dehydrated zone
between 100oC and 300oC. It can be important, in reinforced or pre-stressed concrete to
establish the maximum depth of the 100oC isotherm.
3.4.4 Poor Quality Construction
During construction lack of attention to proper quality control can produce concrete which may
be inferior in both durability and strength to that assumed by the designer. Particular factors in
this respect are compaction, curing conditions, low cement content, incorrect aggregate grading,
incorrect water cement ratio and inadequate cover to reinforcement.
3.4.4.1 Plastic Cracking
Cracks due to this phenomenon appear within the first two hours after placing and are of two
distinct types - plastic settlement and plastic shrinkage cracks.
3.4.4.2 Plastic Settlement Cracks
Typically found in columns, deep beams or walls. The problem tends to occur with high water
cement ratio concretes which have suffered from bleeding. The concrete literally "hangs up" on
the steel slumping between it, with cracks forming over the line of the steel. Caught early
enough, re-vibration of the concrete can repair the damage while the concrete is still plastic. The
plate shows a core taken through a plastic settlement crack in a car park. The damage that the
easy pathway for chlorides has caused can easily be seen.
3.4.4.3 Plastic Shrinkage Crack
More common in flat exposed slabs. This can occur anywhere where the rate of loss of moisture
due to evaporation exceeds the rate of bleeding. Not surprisingly it is more of a problem in hot,
dry climates, but can easily occur on hot days in flat slabs especially, where adequate attention
to protection and curing has not been given.
3.4.5 Thermal Cracking and Delayed Ettringite Formation
Thermal cracking of concrete can occur in large pours. Typically concrete can gain in
temperature about 14oC per 100 kg of cement in a cubic meter of concrete. In large pours this
sets up a thermal gradient, with the outer part of the concrete cooling more rapidly than the core.
This puts the outer skin in tension, and small cracks form. With the addition of subsequent
drying shrinkage, the cracks can become quite large.
During the hydration process, it is quite normal for ettringite to form inside the concrete. This
mineral is normally associated with sulfate attack, but in the context of a setting concrete is
quite normal. Any expansion resulting from its formation is taken up in the still plastic concrete.
(Grantham, Gray, 1999)
3.5 Assessment of concrete quality and cover of the steel
3.5.1 Visual inspection
Maintenance engineering should supply the research team with concrete blocks cut from the
structure during the demolition works. Cores should be cut for physical and mechanical
properties. Powder samples for chemical and mineralogical analysis should be obtained.
3.5.2 Covermeter
Adequate cover to the steel reinforcement in a structure is important to ensure that the steel is
maintained at a sufficient depth into the concrete so as to be well away from the effects of
carbonation or from aggressive chemicals. However, excessively deep cover has its own
problems; crack widths may be increased and the lever-arm decreased.
All covermeters are electromagnetic in operation. Electric currents in a coil winding in the
search head generate a magnetic field which propagates through the concrete and will interact
with any buried metal present, such as reinforcing steel. The interaction will be due to either or


186
both of two physical properties of the steel: it's magnetic permeability and it's electrical
conductivity. The interaction causes a secondary magnetic field to propagate back to the head
where it is detected by a second coil or, in some instruments, by modifying the primary field.
The signal received will increase with increasing bar size and decrease with increasing bar
distance. By making certain assumptions about the bar, and specifically by assuming that only
one bar is present within the primary magnetic field, the instrument can be calibrated to convert
signal strength to distance and hence to indicate the depth of cover.
If there is more than one bar within the range of the primary field the instrument will receive
a greater signal and indicate a shallower cover than the true cover. The skilled operator will
always carefully map out the position and orientation of the steel, breaking out some steel if
necessary, to ensure that accurate results are obtained.
Some manufacturers claim that the size of the reinforcing bar may be determined by the use
of spacer blocks and some inbuilt mathematical processing, or with some of the more
sophisticated machines by internal data processing only. Such methods work satisfactorily only
where a single bar is present within the range of the search head, but can be reasonably accurate
with some of the more modern devices.
British Standard requires that when measuring cover to a single bar under laboratory
conditions, the error in indicated cover should be no more than plus or minus 5 % or 2 mm
whichever is the greater.
The standard also lists a number of extraneous factors which are potential sources of error.
Those concerned with magnetic effects from the aggregates or the concrete matrix, and those
due to variations in cross sectional shape of the bars should not affect the modern covermeter,
but care must always be taken.
3.5.3 Assessment of concrete quality
Thick sections are cut from the cores and are inspected under the stereomicroscope to study the
condition of the concrete with regard to air voids, cracking and possible reactions between the
gravel and cement alkalies for samples.
Thin sections taken from different types of stones are examined using the polarizing
microscope. Sections should have different colors (often it varies from greenish through
brownish to reddish).
The existence of the iron oxide indicates the highly altered condition of the rock due to
serpentinization. Gravel may harm concrete in three ways:
1. The gravel may continue its alteration after its use in concrete. This serpentinization
process may be accompanied by an expansion of 25% causing volume instability and hence
cracking to the concrete matrix.
2. The high magnesium content of the serpentine minerals could cause migration of Mg
++

ions into the paste, damaging the cement reactions.
3. Under hydrous, alkaline conditions and in a calcium-rich environment, malic silicate such
as serpentinite may be converted into expanding layer silicates such as smectiles and
vermiculite.
The main cause of concrete deterioration in structures is the presence of gypsum in high
proportions in the concrete. Serpentine aggregate expand due to weathering and cause pop-outs
on concrete surface. The cause of the phenomenon can be attributed to the expansion of high
iron content brucite existing in some serpentine due to oxidation and carbonation process
thereby decomposing into coalingite Mg
10
Fe
2
(CO
3
)(OH)
24
.2(H
2
O). The crystal structure of
coalingite can be determined using single-crystal X-ray methods. The mineral is trigonal.
Investigations in the past include chemical analyses, optical, X-ray powder, infra-red
absorption, thermal decomposition studies etc. In more cases, X-ray single-crystal investigation
can obtain only limited information because of the poorness of the material.
3.5.4 Chemical testing
The chemical analysis shows the chloride and sulphate contents. The cause of many
unnecessary deterioration when the concrete is in no danger of sulphates as much as chloride
attacks. The high percent of SO
4
in the surface soil may be due to contamination from air born
fumes and gases existing in the refinery atmosphere and hence its existence was limited to a few
centimeters at the surface.

187
3.5.5 Chloride content and carbonation depth
The measurement of carbonation depth and of chloride ion concentration in the cover zone are
two standard procedures in assessing reinforced concrete deterioration, particularly in respect of
reinforcement corrosion problems. It reveals the level at which the concrete pH=9, which is
close to the boundary between concretes that are and are not sufficiently alkaline to provide
passivity for steel. Reinforcement will only corrode if there is sufficient moisture in the
concrete, and an appropriate elastomeric anti-carbonation coating can often significantly retard
further deterioration.
3.5.6 Phenolphthalein test
The concrete surface should be sprayed with an aqueous alcoholic solution of phenolphthalein.
The surface should only be lightly sprayed as transport of dissolved alkalis may obscure the
location of the true carbonation front. Where the concrete is no longer alkaline, the
phenolphthalein remains colorless, changing to pink where the concrete is alkaline. It is
important to note the speed of color change; a slow change may indicate partial carbonation.
The phenolphthalein test method is not suitable for concrete made with high-alumina cement.
3.5.7 On-site permeability measurement
A specific apparatus can be used for testing the relative permeability of materials, particularly
concrete structures. It allows to be sealed against the structure when a partial vacuum is
induced. The decay of the vacuum may be measured, and the rate of decay provides an index of
permeability. A liquid may be applied to the surface being tested and will allow identification of
cracks and fissures. In alternative embodiment, a fluid is introduced into a secondary chamber
defined by a secondary gasket and the head, and the rate of permeation of the fluid into the
structure provides a measure of permeability. The vacuum mounted test apparatus allows in-situ
testing without requiring the extraction of sample cores from the structure or making invasive
attachments to the structure, thus providing a quality control apparatus and method for rapid
testing of construction materials.
3.6 Laboratory testing
The transport properties of fluids in the cover layer are the main indicator of the durability of
reinforced concrete structures.
Many laboratory tests exist for measuring the transport properties of fluids in the cover layer,
because they are the main indicator of the durability of reinforced concrete structures. However,
there are relatively few tests which can be used on site.
It was developed a new non-destructive rapid test capable of measuring the air permeability
of in-situ concrete from Claisse, Ganjian and Adham from School of The Built Environment in
Coventry University. The method measures the movement of gas between different holes drilled
into the concrete and gives results for the permeability of the concrete and for the volume of
concrete which has been tested.
A pressure measuring method was developed to measure the pressure inside the concrete by
using a piezoresistive pressure transducer and a data logger. Four different techniques were
investigated and one preparation technique is recommended for in-situ use.
This new test is a modification of the original Figg test that was introduced in the 1970s.
There had been modifications to the original Figg test in order to improve its repeatability and
sensitivity, which were not taken into account while carrying out the investigation reported in
this paper. Therefore, the proposed test does not necessarily have improved repeatability and
sensitivity compared to the modified Figg tests. The primary objective of this investigation was
to obtain a coefficient of permeability from the original Figg test (Claisse, Ganjian, Adham,
1990).
3.6.1 Theoretical Analysis
The volume of concrete that is being tested is measured by the distance X at which the pressure
returns to atmospheric. To calculate this, the steady state is considered in which a constant
vacuum is applied to the centre hole.
The modeling is based on the Darcy equation for pressure driven flow:


188

where
F= flow rate; K= intrinsic permeability; E= viscosity of the fluid;
p= pressure at a distance x from the high pressure reservoir;
A= area across which the fluid is flowing.
In this test the permeating fluid is compressible and the observed flux F will therefore change
with pressure. The flow is therefore best expressed as molecular flow where N is the total flux
and dn/dt is the flow rate of the gas. Both N and dn/dt are approximately constant across the
sample (assuming a steady state within it).
where R=gas constant; T= temperature (K); and t= time from the start of the test.

The distance X at which the pressure returns to atmospheric was determined by applying a
constant pressure to the centre hole and measuring the pressure in the side hole. It was found
that the piezoresistive pressure transducers could be conveniently assembled directly onto the
hypodermic needles thus minimising "dead volume" which would have affected the readings.
4mm diameter side holes were used. In order to minimise the effect of the 20mm aggregate
these were drilled to 40mm deep with a 20mm deep void space. Before applying the vacuum
three readings of pressure were taken at 10 seconds intervals in order to zero the pressure
sensors.
3.6.2 Petrographic examination
3.6.2.1 Preliminary examination
The samples are examined with the binocular microscope as received and their dimensions and
main features are recorded. The features observed include the following:
- The presence and position of reinforcement.
- The extent to which reinforcement is corroded.
- The nature of the external surfaces of the concrete.
- The features and distribution of macro and fine cracks.
- The distribution and size range and type of the aggregate.
- The type and condition of the cement paste.
- Any superficial evidence of deleterious processes affecting the concrete.
3.6.2.2 Polished surfaces
A plate is cut, where possible, from each sample. This is typically about 20 mm thick and
usually provides as large a section of the sample as is possible. The plate is polished to give a
high quality surface that can be examined with a high quality binocular microscope or even with
the petrological microscope if necessary. The polished plate is used to assess the following:
- The size, shape and distribution of coarse and fine aggregate.
- The coherence, color, and porosity of the cement paste.
- The distribution, size, shape, and content of voids.
- The composition of the concrete in terms of the volume proportions of coarse
aggregate, fine aggregate, paste and void.
- The distribution of fine cracks and microcracks. Often the surface is stained with a
penetrative dye, so that these cracks can be seen. Microcrack frequency is measured
along lines of traverse across the surface.
- The relative abundance of rock types in the coarse aggregate is assessed.
3.6.2.3 Thin sections
A thin section is prepared for each sample as appropriate. The section is usually made from a
1 dn Kp dp
N = = -
A dt RTE dx
. .
K A dp
F =
E dx
.

189
plate cut at right angles to the external surface of the concrete, so that the outer 70 mm or so of
the concrete are included in the section. Sometimes it is more appropriate to make the section
from inner parts of the concrete. This might be appropriate where specific problems are being
investigated for example. The section normally measures about 50x70 mm.
In manufacturing the thin section a plate some 10 mm thick is cut from the sample. This is
impregnated with a penetrative resin containing a yellow fluorescent paint. The resin penetrates
into cracks, microcracks, and capillary pores in the sample. One side of the impregnated plate is
then polished and the plate is mounted on to a glass slide. The surplus sample is then removed
and the plate is ground and polished to give a final thickness of between 20 and 30 micrometres.
At all stages the cutting and grinding is carried out using an oil based coolant in order to prevent
further hydrating of the cement and excessive heating of the section. The thin section is covered
and then examined with a high quality petrological photomicroscope. The thin section supplies
the following types of information:
- Details of the rock types present in the coarse and fine aggregate and in particular
structures seen within those rocks.
- Details of the aggregate properties are measured such as the degree of strain in
quartz.
- The size, distribution and abundance of phases in the cement paste are assessed
including, for example, the occurrence of calcium hydroxide and the amount of
residual unhydrated clinker.
- The presence of cement replacement phases such as slag can usually be recognized
- The presence of high alumina cement can be detected and the type of cement clinker
can often be assessed.
- Any products of processes of deterioration of either the cement paste or the aggregate
can be recognized.
3.6.2.4 Broken surfaces
After the specially prepared surfaces and sections are completed, the remainder of the core is
examined with the binocular microscope. In particular, the pieces are broken to produce fresh
surfaces. These surfaces allow the contents of voids to be studied and the nature of aggregate
surfaces or crack surfaces to be investigated.
3.6.2.5 Composition
The composition of the sample is measured using either the polished slice or the thin section,
depending on the size of the sample and on details of the aggregate type and paste. The thin
section is preferable, for example where large quantities of dust are present. The volume
proportions are found by the method of point counting using a mechanical stage. The amount of
coarse aggregate can also be assessed by this method if a distinction can be made between
coarse and fine aggregate. The results obtained usually represent the sample reasonably, but
may not represent the concrete.
The amount of individual rock types present in the aggregate as a whole are assessed and the
saturated density of the sample is measured by the method of immersion in water using vacuum
impregnation to ensure saturation. From this information and the volume proportions, the
weight fractions of aggregate, cement and water can be calculated.
3.6.2.6 Water/cement ratio
The hydrating processes of cement paste vary significantly with the original water/cement ratio.
Concretes with a low water/cement ratio tend to leave substantial quantities of unhydrated
cement clinker and to develop only limited amounts of coarsely crystalline calcium hydroxide.
In particular, the extent to which calcium hydroxide is separated into layers on aggregate
surfaces and occurs in voids and on void surfaces varies with the original water/cement ratio.
The number and proportion of unhydrated cement clinker particles varies inversely with the
original water/cement ratio. Comparison with standard concretes made with known
water/cement ratios visually, and by measurement allows the water/cement ratio of the cement
paste to be assessed directly. The standard error attached to the estimation of water/cement ratio
by this means is considered to be approximately 0.03.


190
3.6.3 Compressive Strength Determination.
The compressive strength of a concrete has traditionally been used as a specification tool by
engineers to specify the quality of concrete delivered to a job. It has been used for many years
because it is a relatively simple property to measure, unlike chemical analysis which requires a
well equipped chemistry laboratory to determine whether an adequate cement content has been
used. On site, a simple slump test will normally be carried out to show whether the concrete
complies with the specified workability.
The concrete will normally be specified as a "designed mix", meaning that the engineer will
have asked for a concrete of a certain strength, either to ensure that he has adequate strength for
structural purposes, or, more usually, to ensure the overriding requirement that the concrete has
adequate durability. The first thing to realize when measuring concrete strength is that concrete
is not a uniform material and, when placed, it is even less so. Even with good quality control, a
spread of results of 16 N/mm
2
is likely with a Grade 40 concrete. This means that to be sure of
getting most (95%) of his results above 40 N/mm
2
, the concrete supplier will aim for a target
mean strength of some 48 N/mm
2
.
These days, BS 5328 allows the specifier to call for an "equivalent grade" concrete, meaning
that he can specify a strength grade that will give him a reasonable guarantee that he has
achieved his desired minimum cement content and/or maximum water/cement ratio. The
strength so specified will usually be higher than that which would be called for strength
purposes alone.
It must be appreciated that the results gained from concrete core samples will not be directly
comparable with those from the original cube tests. Cube samples are fully compacted and have
been stored under ideal curing conditions for 28 days, prior to test. Core samples, on the other
hand, have been taken from in-situ concrete, cured in the structure, often inadequately, with
perhaps less than perfect compaction. There are also settlement effects, with results on cores
from the bottom of a wall or column differing by 15-30% from those of cores taken near the top.
Not surprisingly, there are large differences in the strength measured on cores and on cubes
made from the original concrete, as supplied.
Core samples are the most common form of sample for this purpose, removed from the
structure by diamond drilling. Typically cores will be 100 mm in diameter, and should ideally
be at least three times the maximum aggregate size in diameter. The cores are usually visually
described and photographed, concentrating especially on compaction, distribution of aggregate,
presence of steel etc. and then trimmed to a length to diameter ratio approaching 1:1. Various
methods are available to ensure accuracy of the ends of the core; grinding to a perpendicular flat
surface, capping with high alumina cement mortar and capping with hot sulphur are all methods
which are used. The capped core is then crushed in a calibrated compression testing machine.
The resulting failure load is converted firstly to a cylinder strength and secondly to an
equivalent in-situ cube strength. This will generally be some 75% or so of the original cube
result, depending on where the core was taken, before allowing for any additional correction
due to compaction effects.
REFERENCES
Angst, U., Elsener, B., Larsen, C. K. & Vennesland, . 2009. Critical chloride content in reinforced
concrete A review. Cement and Concrete Research. 39: 1122-1138.
Claisse, P.A., Ganjian, E. & Adham T.A. 1990. A Vacuum-air Permeability Test for In-situ Assessment
of Cover Concrete. Coventry University.
Giarma, C. 2010.Numerical approach of the simultaneous moisture and heat transfer in concrete
building components under the influence of environmental conditions and estimation of the
carbonation depth. PhD Thesis (in Greek). Civil Engineering Department, Aristotle University of
Thessaloniki
Grantham, M.G., Gray, M.J. October 1999. Diagnosis, Inspection, Testing and Repair of Reinforces
Concrete Structures. M.G. Associates, pp. 1-75.
Illston, J. M. & Domone, P. L. J. 2001. Construction Materials: their Nature and Behavior. London:
E&FN Spon; New York: Chapman & Hall.

191
Leemann, A., Lothenbach, B. & Hoffmann, C. 2010. Biologically induced concrete deterioration in a
wastewater treatment plant assessed by combining microstructural analysis with thermodynamic
modelling. Cement and Concrete Research. 40: 1157-1164.
Neville, A.M. 1995. Properties of concrete. Harlow, Essex : Addison Wesley Longman Limited.
Papadakis, V. G., Vayenas, C. G. and Fardis, M. N. 1991. Fundamental modelling and Experimental
Investigation of Concrete Carbonation. ACI Materials Journal. 88: 363-373.
Rendell, F., Jauberthie, R. & Grantham, M. 2002. Deteriorated concrete: inspection and physicochemical
analysis. London: Thomas Telford.
Saetta, A.V., Screfler, B.A. and Vitaliani R.V. 1995. 2-D model for carbonation and moisture/heat flow
in porous materials. Cement and Concrete Research. 25: 1703-1712.
Saetta, A.V., Screfler, B.A. and Vitaliani R.V. 1993. The carbonation of concrete and the mechanism of
moisture, heat and carbon dioxide flow through porous materials. Cement and Concrete Research. 23:
761-772.
Shalaby, H.M., Daoud, O.K. 1990. Case Studies of Deterioration of Coastal Concrete Structures in Two
Oil Refineries in The Arabian Gulf Region. Cement and Concrete Research, Vol. 20. pp. 975-985.
Steffens, A. 2000. Modellierung von Karbonatisierung und Chloridbindung zur numerischen Analyse
der Korrosionsgefhrdung der Betonbewehrung, PhD thesis, Braunschweig.
Sun, Z. & Scherer, G.W. 2010. Measurement and simulation of dendritic growth of ice in cement paste.
Cement and Concrete Research, doi:10.1016/j.cemconres.2010.03.008
Vesikari, E., Giarma, C. & Bleiziffer, J. 2009. Degradation models of concrete structures. Proc. of the
Workshop of COST C25 Action Sustainability of Constructions: Integrated approach to Life- time
Structural Engineering, Timisoara, Romania, 23-24 October 2009, pp. 40-51.


192
T. Teruzzi
University of Applied Sciences of Southern Switzerland, Lugano-Trevano, Switzerland
Condition assessment and service-life prediction of reinforced
concrete structures: description and application examples of a
method developed for aged and new buildings
1 INTRODUCTION
For an effective lifetime management of buildings and civil infrastructures, durability
assessment methods are needed in order to verify the compliance of the expected service life
with the design value and, in case of non-conformity, to allow reasonable predictions about the
time-evolution of degradation. These predictions enable the owner to plan adequate
maintenance and repair actions.
In the case of concrete structures, durability and service-life are mainly affected by the
penetrability of the finished concrete cover layer, which in turn is influenced by the concrete
mix and also by the placing and curing conditions.
For the case of concrete building facades, whose most important degradation mechanism is
carbonation induced corrosion of the reinforcement bars, the penetrability directly determines
the velocity with which the front of the concrete carbonation reaction penetrates into the
structure.
Most degradation models developed to describe the progression of the carbonation front
assume that the carbonation depth increases proportionally to the square root of time. Basing on
this assumption, simple methods for service-life estimation can be set up. Owing to the
heterogeneity of concrete properties, service-life estimation has to be better approached from a
fully probabilistic point of view. Within this approach, service-life is defined as the time for
which the probability of reinforcement bars depassivation (failure probability) reaches some
limiting value. It is common practice to express the maximum acceptable probability by means
of a limiting reliability index. The actual reliability index can thus be used as a criterion for
evaluating the conditions of concrete building facades.
The failure probability can be calculated by means of a convolution integral, whose integrand
is given by the product of the probability density function of the carbonation depth, which is
time dependent, and the one of the concrete cover thickness. The former can be determined on
core samples taken out from the structures, whereas the latter can be determined non
destructively by means of a covermeter. The paper presents two case studies concerning aged
buildings, in which data on carbonation depth and concrete cover thickness were collected in
order to assess their condition and to estimate their residual service life.
On new structures the probability density function of the carbonation depth cannot be
determined directly. In this paper it is shown how it can be calculated basing on measurements
of the air permeability of the concrete cover and how these data can be used as input values for
an estimation method of the service life. A key element of the method is the relationship
existing between the carbonation rate and the air permeability of the concrete cover. The latter
has been measured non destructively with a Torrent permeability meter. The relation has been
explored and validated by means of statistical inference methods.

193
2 PROBABILISTIC SERVICE LIFE ESTIMATION METHOD
2.1 Probabilistic definition of service life
Service life is defined as the time period since the realization of a structure or of part of it,
during which the intended performance is achieved. With regard to durability and within a
probabilistic approach, this period of time ends when the probability of reaching some limit
state exceeds a predefined value.
According to reliability theory, if S(t) represents the time dependent solicitation acting on the
structure and R(t) is the resistance that the structure opposes against it, then the event expressing
failure can be specified as {failure} = {R(t) S(t) < 0}. The failure probability P
f
is so defined
as
{ } 0 ) ( ) ( ) ( < = t S t R P t P
f
(1)
and the service life
SL
t as
{ }
max ,
) ( : sup
f f SL
P t P t t s = (2)
The service life is dependent on the required reliability level
max , f
P . Reliability requirements
are generally expressed in terms of a statistical reliability index according to the equation,
) (
max ,
| u =
f
P , where is the distribution function of the standard normal distribution.
In the European standard EN 1990, a fixed minimum reliability index = 1.5 (corresponding
to a failure probability of 6.681%) is recommended for serviceability limit states and for a time
period of 50 years. Some authors propose to graduate the reliability requirements according to
the economic consequences deriving from a violation of the serviceability limit state. Rackwitz
(1999), for example, considers as a ranking parameter for the reliability index the ratio between
the costs related to risk minimization and those related to the rehabilitation actions.
Gehlen (2000) suggests that the reliability index has to be linked not only to the cost ratio
introduced by Rackwitz but also to the environmental exposition classes according to EN 206.
The reliability indexes suggested by Rackwitz and Gehlen are listed in Table 1.
Table 1. Required reliability index at the end of the service life as a function of the costs ratio and the
environmental exposition classes according to Rackwitz (1999) and Gehlen (2000)
Costs ratio Exposition classes Reliability index |
Low XC4, XD1, XS1, XS3, XD3 2.0
Normal XC2, XC3, XS2, XD2 1.5
High XC1 0.5
2.2 Computation of failure probability for carbonation induced corrosion
If the solicitation S and the resistance R are independent random variables with probability
density functions (pdfs) ) (s f
S
and ( )
R
f r (and cumulative distribution functions ) (s F
S
and
( )
R
F r ) respectively, then the probability of failure defined by equation (1) can be computed
using the following convolution integral
( ) ( ) ( ) ( ; ) ( ; )
s
f R S R S
P t f r f s dsdr F s t f s t ds
+ +

= =
} } }
. (3)
For carbonation induced corrosion, the solicitation, to which the structure is subjected, is
represented by the time dependent carbonation depth ) (t X
C
and the resistance is represented
by the concrete cover thickness
C
D . Here it is assumed that the serviceability limit state


194
corresponds to the onset of reinforcement depassivation. In this case, equation (3) can be written
as follows:
( )
} } }
+

+

= =
c c X c D c c c X
x
c D f
dx t x f x F d d dx t x f d f t P
C C C
C
C
) ; ( ) ( ) ( ) ; ( ) ( (4)
Knowledge of the probability distributions of the carbonation depth at time t and of the
concrete cover thickness is thus fundamental for computing the failure probability. The
distribution of the concrete cover can be derived from measurements performed
nondestructively by means of commercially available covermeters.
2.3 Distribution function of the time- dependent carbonation depth
Under the assumption that the front of the concrete carbonation reaction progresses
proportionally to the square root of time, t c t X
C
= ) ( (C is the carbonation rate factor), the
probability density function of the carbonation depth
C
X
at the time t can be written as:
) (
1
) ; (
t
x
f
t
t x f
c
C c X
C
= , (5)
where ) (c f
C
is the pdf of the carbonation rate factor. With expression (5) the task of finding
the time-dependent pdf of the carbonation depth is thus reduced to that of establishing the pdf of
the carbonation rate factor.
2.3.1 The case of aged structures
For the case of aged structures the probability density function of the carbonation rate factor can
be inferred from the individual carbonation rate factors
0
/ t x c
ci i
= , where
ci
x is the i-th
value of the carbonation depth at the time t
0
of the inspection, as obtained by means of the
phenolphthalein method on cores taken out from the structure.
For the assumed carbonation model, the mean value ) (t
C
X
and the standard deviation
) (t
C
X
o of the carbonation depth at time t are given by the following formulas:
) ( ) (
0
0
t t
t
t
t
C C
X X
= = (6)
) ( ) (
0
0
t t
t
t
t
C C
X X
= = o o (7)

If the concrete cover thickness
c
D and the carbonation depth ) (t X
c
can be approximately
considered independent normal variates, then the failure probability P
f
at time t is
)
) (
) (
( ) (
2 2
t
t
t P
C C
C C
X D
X D
f
o o

+
u = (8)
2.3.2 The case of new structures
In the case of new structures the method for calculating the failure probability just presented
cannot be applied, because at an early age, say 1-9 months, carbonation depth is barely
measurable. Alternative ways have therefore to be devised.

195
In the following, a method is presented, in which the distribution function of the carbonation
rate factor is estimated upon the basis of the air permeability kT of the concrete cover, which
can be measured directly on the structure according to the Torrent method (Torrent (1995)). In
Switzerland the Torrent method is a standard test for the assessment of the in situ air
permeability of concrete (Swiss standard SIA 262/1 (2003)).
The relationship existing between the carbonation rate factor and the air permeability has
been explored and validated by means of statistical inference methods (likelihood inference). To
this purpose, on site measurements of the air permeability and of the carbonation depth have
been performed on more than 50 points distributed over the facades of a 35 years old concrete
building located in the urban area of Lugano and differently exposed to the climatic agents. The
data used for the analysis are shown in Figure 1.

Figure 1. Carbonation rate against the air permeability of the concrete cover. The solid curve represents
the mean carbonation rate as a function of air permeability.
The main results of the statistical analysis can be summarized as follows (a detailed
description of the inference methods used are presented in Teruzzi (2009)):

1) the relationship linking the mean carbonation rate with the air permeability kT of the
investigated on-site concrete is
) / ( log
0 10
T k kT c = (9)
for = >
2 -16
0
0.01610 m T k kT and is 0 = c elsewhere (see Figure 1);

2) the carbonation rate factor is statistically distributed according to a Weibull distribution with
scale parameter ) 1 / 1 ( / ) ( ) ( + I = o kT c kT and shape parameter 949 . 1 = o ,
( )
|
|
.
|
\
|
|
|
.
|
\
|
|
|
.
|
\
|
=
o o

o c c
c f exp
1
(10)
Since the scale parameter depends upon the air permeability, expression (10) can also be
considered as the conditional density function ) ( kT c f
kT c
of the carbonation rate c given kT.
Substituting in (10) the corresponding parameters one gets the following function
( ) ( )
|
|
.
|
\
|
|
.
|
\
| + I
|
.
|
\
| + I
=
o o
o
o o o
1
1 / 1
exp
) 1 1 (
) (
c
c
c
c
c
kT c f
kT c
(11)


196
for
0
T k kT > and 0 elsewhere. The function is plotted in Figure 2 for some values of kT.

Figure 2. Conditional density functions of the carbonation rate
factor for kT = 0.10, 0.32, 1.0, and 10.010
-16
m
2
.

By making use of the total probability theorem, the probability density function of the
carbonation rate factor can be written as follows
0
( ) ( , ) ( ) ( )
C kT c kT
kT
f c f c kT f kT d kT
+
=
}
(12)
where ) (kT f
kT
is the pdf of the air permeability.
It is shown, for example, by Denari (2006) that the air permeability can be assumed to
follow a lognormal statistical distribution. Substitution of the appropriate functions for
kT c
f
and
kT
f into equation (12), followed by (numerical) integration, gives the pdf of the
carbonation rate factor and, hence, by virtue of equation (5), the pdf of the carbonation depth.
Under the assumption that expressions (9) and (10) are generally valid, measurements of the
concrete air permeability thus allow to determine the time-dependent pdf of the carbonation
depth. The failure probability can then be calculated according to formula (4).
2.4 Computation of service life
For given distributions of the concrete cover thickness and of the carbonation depth, the service
life
SL
t corresponds to the time t for which the probability of failure, as given by equation (4),
reaches its maximum acceptable value, ( )
max . f SL f
P t t P = = .
3 APPLICATION EXAMPLES
3.1.1 The case of aged structures
This section presents the results obtained for two complexes of concrete buildings out of the
many, whose condition has been assessed by applying the method presented in section 2.3.1.
The two building complexes, which are both scholastic facilities, are located in Gordola, a
village located close to the Lago Maggiore at about 200 m above sea-level, and in the urban area
of Lugano-Viganello at about 400 m above sea-level. (see Figure 3).
Both complexes consist of multiple buildings having facades of in situ cast in place concrete.
All the facades are fully exposed to the climatic agents, in particular to rain.
The condition of the investigated facades, expressed in terms of their reliability index| , are
presented in Table 2, jointly with the data characterising the distribution of the thickness of the

197
concrete cover and the depth of the carbonation reaction front. The service life has been
computed for % 275 . 2
max .
=
f
P ( = 2).

Figure 3. Concrete building complex of Gordola (left) and of Lugano-Viganello (right)
Table 2. Results of the inspections performed on the complexes of Gordola and Lugano-Viganello
Carbonation depth
[mm]
Concrete cover
[mm]
Complex Facade Age

[y] m* s* m* s*
Reliability
index
|
t
SL

[y]
Gym, North 13.7 5.5 25.5 3.7 1.8 21
Gym, East 22.0 8.9 25.2 4.3 0.3 8
Gym, South 26.3 8.4 26.5 5.6 0.0 6
Gym, West
24
21.2 10.8 29.8 6.5 0.7 8
Refectory, North 14.7 2.3 21.6 5.2 1.2 14
Refectory, East 13.2 2.2 18.6 4.7 1.0 12
G
o
r
d
o
l
a

Refectory, West
26
17.5 2.3 21.3 6.5 0.6 6
Carbonation depth
[mm]
Concrete cover
[mm]
Complex Facade Age

[y] m* s* m* s*
Reliability
index
|
t
SL

[y]
Building 1, North 4.3 4.2 31.3 4.8 4.2 53
Building 1, East 10.0 1.0 31.3 6.6 3.2 31
Building 1, South 10.7 5.0 32.5 3.8 3.5 22
Building 1, West
10
11.0 1.0 28.8 5.7 3.1 24
Building 2, North 4.3 1.5 42.4 2.5 13.1 >150
Building 2, East 12.0 6.1 42.1 2.4 4.6 35
Building 2, South 11.0 1.4 43.0 2.4 11.5 110
Building 2, West 16.3 8.7 43.6 2.0 3.1 20
Building 3, East 13.0 4.6 40.1 4.8 4.1 34
V
i
g
a
n
e
l
l
o

Building 3, West
12
7.3 2.3 33.5 5.7 4.3 70
* m: mean value/s: standard deviation

With regards to the complex of Gordola, except the northern facade of the gymnasium, the
overall condition of the investigated facades, with a reliability index ranging between 0.0 and
1.2, is largely insufficient. This is confirmed by the presence, on the facades, of a large amount
of homogeneously distributed points where cracking and spalling of the concrete cover is
observed. The insufficient condition of the facades of the gymnasium is mainly due to a high
carbonation rate factor (about 4.7 mm/y
0.5
), whereas the one of the facades of the refectory is a
direct consequence of the poor thickness of the concrete cover (13-18 mm). With a reliability
index | = 1.8, the condition of the northern facade of the gymnasium is good.


198
The state of all the investigated facades of the scholastic facility of Lugano-Viganello is
characterised by a reliability index | > 3. The good condition of the facades is due to the fact
that the older buildings of this facility are only 12 years old and, in particular, to the generous
thickness of the concrete cover (it is larger than about 37 mm on average). With respect to the
other buildings, which are less recent, those of Viganello, thanks to a revision of the Swiss
standard concerning the design of concrete structures in 1989, benefited from a more severe
requirement for the minimum thickness of the concrete layer protecting the reinforcing bars.
The knowledge of the estimated service life of a concrete structure is very useful when a
maintenance program has to be implemented. For example, it is possible to define three
different situations, to which may correspond the following intervention actions:
Situation 1: t
SL
t
0
> 50 years, actions: none,
Situation 2: 0 < t
SL
t
0
< 49 years, actions: application of a coating reducing or
preventing the ingress of degradation agents,
Situation 3: t
SL
t
0
s 0 years, actions: concrete restoration.
Basing on the estimated values of the (residual) service life, the owner and the facility
manager can thus define appropriate maintenance actions, which should also include periodical
visual and in-depth inspections in order to update the condition assessment data, thus managing
the inherent uncertainty of the method.
3.1.2 The case of new structures
Service life is computed by means of expression (4) and (2). Although there is no major
computation difficulty for any distribution of the concrete cover thickness, for the sake of
simplicity, in the following, service life is estimated for the case in which it is deterministic, that
is d
c
= d
c,0
. In this particular case, the expression for the failure probability becomes
}
+
|
|
.
|
\
|
= = =
O C
C C
d
c
c c X c c X f
t
d
F d F dx x f t P
,
0 ,
0 ,
1 ) ( 1 ) ( ) ( (13)
It is clear from equation (13), that the service life
SL
t corresponds to the time t for which the
term t d
o c
/
.
equals the ( )
max .
1
f
P - quantile of the cumulative distribution function
c
F of
the carbonation rate.
For example, basing on the results presented in 2.3.2, approximate values of the ( )
max .
1
f
P -
quantiles have been computed for
10
log ( )
0.4
kT
o = , 2 . 16
) ( log
10
=
kT
and 15.0 and for
% 275 . 2
max .
=
f
P ( = 2). The computation results are presented in Table 3. The value of 0.4
for the standard deviation of ) ( log
10
kT is observed to be typical for air permeability
measurements (Jacobs (2006)).
Table 3. Approximate values of the ( )
max .
1
f
P
- quantile computed for
4 . 0
) ( log
10
=
kT
o
,
2 . 16
) ( log
10
=
kT
, -16.0 and -15.0 and for

% 275 . 2
max .
=
f
P
(| = 2)
( )
max .
1
f
P - quantile of the cumulative distribution function F
c
[mm/y
0.5
]
) ( log
10
kT

) ( log
10
kT
o
-16.2 -16.0 -15.0
0.4 3.95 4.37 6.47

Basing on the quantiles presented in the previous table, service life values have been
calculated, for example, for a concrete cover thickness =
0 . c
d 30, 40 and 50 mm. The results are
shown in Table 4.

199
Already shortly after the end of the construction phase it is thus possible, through on site
measurements of the air permeability of the concrete cover, to estimate the expected service life
of a concrete building and, if necessary, to plan adequate maintenance or repair actions.
Table 4. Service life computed for
10
log ( )
16.2
kT
=
, -16.0 and -15.0 and for
d
c,0
= 30, 40 and 50 mm (
) ( log
10
kT
o is assumed to be 0.4).
Service life
SL
t [years]
) ( log
10
kT

Concrete cover
thickness [mm]
-16.2 -16.0 -15.0
30 58 47 21
40 103 84 38
50 160 131 60

The data of Table 4 indicate, for example, that for a 30 mm concrete cover the expected
service life exceeds the standard value of 50 years only if the geometric mean of the air
permeability of the concrete cover does not exceed about 1.010
-16
m
2
. The service life ranges
from 47 years up to 131 years for a 50 mm cover. For a geometric mean of 10.010
-16
m
2
service
life barely reaches 20 years. On the other hand, for the case of a 50 mm concrete cover the same
air permeability would be sufficient to guarantee an expected service life of 50 years. The
data of Table 4 outline the importance for durability of the balance between thickness and
compactness of the concrete cover.
Once the requirements about cover thickness and desired service life are specified, the method
just described can be used to determine upper bounds for the concrete cover air permeability.
These can be viewed as compliance criteria when assessing durability of a newly constructed
structure. For example, for monumental building structures, bridges and other civil engineering
structures, whose recommended service life is 100 years according to EN 1990, the method
gives a limiting value of about 0.610
-16
m
2
for a 40 mm cover.
4 CONCLUSIONS
Within a probabilistic framework service life is defined as the time at which the probability for a
structure to attain some serviceability limit state reaches a predefined maximum acceptable
value. For degradation of the structure due to carbonation induced corrosion of the
reinforcement bars, a suitable serviceability limit state is represented by the state of
depassivation of the reinforcement. The time-dependent failure probability can thus be
computed with a convolution integral, whose integrand is given by the product of the
probability density function of the time-dependent carbonation depth and that of the time-
invariant concrete cover thickness.
In order to model the progress of the carbonation reaction front it is assumed that the
carbonation depth increases proportionally to the square root of time. The constant of
proportionality is represented by the carbonation rate factor, which can be viewed as a random
variable. The pdf of the carbonation depth can thus be derived from that of the carbonation rate
factor.
For aged structures the distribution of the carbonation rate factor is obtained by means of the
phenolphthalein method on core sample taken out from the structure. Basing on the obtained
distribution the actual failure probability and the (residual) service live of the structure are then
calculated. The results can be used to evaluate the condition of the concrete structure.
For new structures the distribution of the carbonation rate factor can be derived from
measurements of the air permeability of the concrete cover with the Torrent permeability meter.
The data can be used to estimate the service life of the structure, thus allowing the owner to


200
check whether the durability requirements (expressed in terms of design service life) are
fulfilled.
REFERENCES

Denari, E. & Matre, M. & Conciatori, D. & Brhwiler, E. 2006. Air permeability measurements for the
assessment of the in situ permeability of the cover concrete. In Alexander (ed.), Concrete Repair,
Rehabilitation and Retrofitting. Taylor & Francis Group, London.
Ghelen, C. 2000. Probabilistische Lebensdauerbemessung von Stahlbetonbauwerken
Zuverlssigkeitsbetrachtungen zur wirksamen Vermeidung von Bewehrungskorrosion. Deutscher
Ausschuss fr Stahlbeton, Heft 510. Beuth Verlag, Berlin (in German).
Jacobs, F. 2006. Luftpermeabilitt als Kenngrsse fr die Qualitt des berdeckungsbetons von
Betonbauwerken. Swiss federal road office FEDRO. Report no. 604 (in German).
Rackwitz, R. 1999. Zuverlssigkeitsbetrachtungen bei Verlust der Dauerhaftigkeit von Bauteilen und
Bauwerken. In Kurzberichte aus der Bauforschung 40, No. 4: 297-301, Stuttgart (in German).
SIA 262/1. 2003. Concrete Constructions Supplementary Specifications. Swiss standard. Zurich: SIA
(in German and French).
Teruzzi, T. 2009. Estimating the service-life of concrete structures subjected to carbonation on the basis
of the air permeability of the concrete cover. Current State and Challenges for Sustainable
Development of Infrastructure; Proc. intern. EUROINFRA 2009 Conference, Helsinki, 15-16 October
2009
Torrent, R. & Frenzer, G. 1995. Methoden zur Messung und Beurteilung der Kennwerte des
berdeckungsbeton auf der Baustelle. Swiss federal road office FEDRO. Report no. 516 (in German).

Examples
201
G. Zuccaro
F. Cacace
M. Rauci
MEDEA: a multimedia and didactic handbook for structural
damage and vulnerability assessment LAquila Case Study
1 INTRODUCTION
The evaluation of the damage due to a seismic event on masonry or reinforced concrete
buildings is a central problem inside the scientific community and one of the main factors for a
post-event macroseismic assessment. Researchers attention is mainly focused on the
identification of criteria and methods able to guarantee an objective evaluation, in order to find
out the best correlation between the parameters measuring the seismic action and the damage.
At present many uncertainties and ambiguities exist either in the definition of the damage
levels to assume or in the identification of the damage by the operators working in the survey.
Moreover, during the post event emergency, the technicians always have to take decisions
concerning the real safety of the buildings.
To this aim, it has been realized (Zuccaro et al, 2002) a dedicated software, MEDEA (Manual
for Earthquake Damage Evaluation and safety Assessment), that offers interesting cues for an
objective evaluation of such damage. According to MEDEA, it could be possible to derive the
most frequent damage typologies in masonry and reinforced concrete structures and a possible
damaging model for quick and guided safety assessment. In fact, the damage assessment
protocol is supported by forms where every damage type is described with notes, iconographic
representations showing different damage levels, and possible links to their associated collapse
mechanisms. This helps to reduce the level of uncertainty in the assessment of safety during the
survey. It contains a technical glossary, a picture archive and a section describing seismic
damage analyses, which are also linked to compatible collapse mechanisms. It also contains a
section on virtual reality examples of damage surveys and safety evaluations.
MEDEA is therefore mainly addressed to the technicians either involved in vulnerability
assessment or employed in post event macroseismic evaluation. In both cases the uniqueness of
the judgment of the surveyors is very important, because of the scientific aspects resulting from
a standard damage measure (useful for the improving of the vulnerability function on one hand,
and for a bigger harmonization in defining macroseismic fields on the other) and for the social
aspects concerning the civil protection deriving from reliable damage estimation and static
safety evaluation.
In addition MEDEA has showed to be an interesting teaching tool for university courses in
the field of seismic engineering or emergency management.


202
2 MEDEA MULTIMEDIA TOOL
2.1 The logic structure of MEDEA and its gradual understanding path
The multimedia handbook MEDEA is structured so that the user can have a gradual
understanding way. In fact, the different sections represent a guided training path able to create
in the users, step by step, a sensibility either to understand the typological characteristics of the
buildings or to evaluate the damage, which are both very important aspects of the final
judgement of safety.
The CD-Rom has been made with particular care in designing a very simple and easy
interface, combining the applications contained to the best effect. MEDEA is divided into main
sections, following a principle of logical continuity. Each section is closely cross-referenced,
thanks to search functions, queries and hypertext links.
2.2 The preliminary sections: glossary and archive
The first section of MEDEA concerns a detailed dictionary of the main terms frequently used in
technical and scientific field. The glossary represents a crucial step to define an unified
terminology either for the reciprocals comprehension among the technicians or to improve the
homogeneity in the typological element classification and in the damage measuring. Every term
of the dictionary is associated to some pictures and graphics, to a descriptive text, and links to
other terms in the glossary. Moreover a search of terms alphabetically or by category is
possible. The terms are organized into five different categories: a) building and its structural
elements; b) structural seismic damage; c) site equipment in the emergency; d) provisional
interventions; e) environment.

Figure 1. MEDEA section frame: a) the glossary; b) the archive.

Examples
203
The second section of the tool is an archive containing a rich anthology of pictures showing
different structural typologies and different levels and types of damages. The picture archive
aims to give an essential basic knowledge of the observed damages and typologies in order to
improve the capability of interpretation and judgment that the technicians must have during the
post-event inspections that will be tested then in the following sections of the tool. A set of
information is joined to every picture of the archive; moreover query either on single
information field or on combined ones are possible. This option allows to select picture sub-set
with specific characteristics, that could result quite useful, for example, in macroseismic survey.
2.3 The evaluation of the seismic damage
The most important part of MEDEA is constituted by a detailed catalogue of the main damages
on structural and no structural elements of masonry buildings.
The multimedia catalogue represents a further step of fundamental importance in the training
of the technicians involved in the safety check of the buildings in the post event phase; in fact it
supplies to the user a guided path among the basic criteria of the structure behaviour in order to
develop interpretation skill on the observed damages.
The damages classification of the structures of the masonry buildings is aided by their
interpretation as possible collapse mechanisms. Therefore the catalogue is organised on the base
of logic rules to address at the most probable mechanism identification. This section is,
therefore, constituted by three sub-sections: Collapse Mechanisms Abacus; Damages Abacus;
Interactive Training Table.

The Collapse Mechanisms Abacus is an accurate work of synthesis, where the authors try to
classify the main recognisable collapse mechanisms for a standard structure. These mechanisms
are firstly classified, even for masonry or for reinforced concrete buildings, as:
Global mechanisms are those mechanisms involving the structures as a whole and then
related to the evolution of the cracks in a sufficient number of elements such that a total
compromising of the static and dynamic equilibrium of the structural system is achieved.
Local mechanisms: are concerned to marginal parts of the structure; their evolution, even if
determines the collapse of a single element, generally does not involve the whole structural
equilibrium.
For masonry structures, the global mechanisms have been subdivided as follows:
in plane: these mechanisms occurs when the walls of the masonry box, excited by in plan
actions in both versus, respond by showing the classical diagonal X cracks, consequent to
the formation of diagonal compressive beams. This mechanisms are due to poor tensile
strength of the masonry material;
out of plane: damage mechanisms that appear through an out of plane kinematism of one or
more walls of the masonry box that, under seismic actions, loses his own original toothing
connection between the walls of the facade and the orthogonal ones, possibly aided by the
action of thrusting floors and roofs;
other mechanisms: in this category are classified those mechanisms that couldnt directly
be recognized as in plane or out of plane, nevertheless are able to involve the building as a
whole, generating the total collapse of the structure (i.e. floor and roof beam unthreading,
irregularity between adjacent structures, etc.);
- while the local mechanisms have been classified as:
for localized dislocation: these mechanisms are those, for example, that arise for arch or
architrave failure, or in part of the structure characterized by different irregularities, often
connected to significant stiffness variations (i.e. non appropriate retrofitting such as R.C.
intervention in masonry structure, etc.). The phenomenon generally determines the
crumbling and the expulsion of the material in the neighbourhood of the involved part;
for thrusting elements: these mechanisms are determined by the action of single elements
that produce horizontal thrust on the supporting structures; good examples are the thrusting


204
elements of a roof or the vaults, the thrust action of which is not sufficiently balanced by
suitable devices.
On the base of this classification 16 different collapse mechanisms for the masonry structures
related to seismic damages and described by schemes, notes and photos have been classified.
These are shown in Figure 2(a-b).

Figure 2a. Abacus of the global collapse mechanisms for masonry structures.

Figure 2b. Abacus of the local collapse mechanisms for masonry structures.

For reinforced concrete structures, the global mechanisms have been subdivided as follows:
strong beam weak column: these mechanisms occurs in buildings characterized by high
strength beams against columns with appreciable lower resistance, in terms of their relative
ratio compared to the seismic action demand. This causes the arising of plastic hinges in
the columns (weaker) instead of the beams (stronger), with the subsequent concentration of
inelastic deformation in the pillars. The mechanisms could involve all or several storeys,
leading to ruinous collapses as pancake type.
weak beam strong column: these mechanisms occurs in buildings characterized by low
strength beams against columns with appreciable higher resistance, in terms of their
relative ratio compared to the seismic action demand. The development of plastic hinges in
the beams (weaker) prevents from the arising of hinges in the pillars (stronger), therefore
the inelastic deformation gather in the beams. The mechanisms could occur for bending in
the beams (more ductile), shear in the beams (less ductile) or iron bars unthreading;
weak nodes: these mechanisms occur when the structural frame is characterized by less
resistant nodes compared to the actions transferred by beams and columns. The mechanism
starts with the arising of diagonal cracks in the junctions up to notable sliding between
beam and column. The collapse of more junctions causes the frame lability, the loss of the
equilibrium and the subsequent collapse of the whole structure.
weak storey: these mechanisms occurs in buildings with a storey less strength than the
others, namely with a demand/resistance ratio significantly different at one level respect to
the others. It usually is due to the lack, at that level, of elements (i.e. infill panels) able to
give more strength against the horizontal actions. In these cases, seismic energy is mainly
dissipated through the rupture mechanism at that floor;
foundation subsidence: these mechanism occurs when seismic actions involve the
foundation level causing a vertical subsidence effects; in these cases the structures show
one direction diagonal cracks in the infill panels, tacking out of horizontal alignment in the

Examples
205
beams and cracks at the ends, or increasing vertical cracks along the height of the building.
In some cases, characterized by high stiffness structural frame and widespread subsidence
extension causing, the building could show a rigid rotation or sliding movements while the
local mechanisms have been classified as:
due to structural elements: in this category are classified those mechanism caused by
structural elements producing a localized action in a part of the structure, as the hammering
between adjacent buildings, the strut action at the column ends by the infill panels, the
floor collapse caused by a large displacements at a junction seat; usually these
circumstances are due to a structural deficiency;
due to the infill panels or partitions: these mechanisms are similar to the ones occurring to
masonry structures, connected to the poor tensile strength of the material; they could
produce in plane or out of plane ruptures,
due to the roof collapse: these mechanisms occurs for the collapse of weak masonry wall
supporting the roof or for the thrusting effect in sloping roof.

On the base of this classification the collapse mechanisms for R.C. structures related to
seismic damages and described by schemes, notes and photos have been classified, as shown in
figure 3 (a-b).

Figure 3a. Abacus of the global collapse mechanisms for R.C. structures.

Figure 3b. Abacus of the local collapse mechanisms for R.C. structures.
For R.C. structures, in addition to the previous ones, mechanisms for R.C. walls are
classified, taking into account the differences among different structural typologies: single walls
(solid or with various openings), double walls, walls/frame.


206

Figure 4. Logical structure of mechanism abacus for Masonry and R.C. structures

The kinematics models and the collapse multipliers of the most frequents CMs are
determined through specific functions. An example of partial wall overturning for masonry
structures is reported below. In the old masonry buildings the walls, in some cases are made
with two vertical layers without sufficient connection in between.
Generally the mechanism involves the external layer only. However sometimes the internal
wall interacts with the external one speeding the collapse. The example is referred to out of
plane mechanisms of one storey wall, considering either one and two layer wall.

Figure 5. Mechanism M6 Partial wall overturning.

Examples
207
A) ONE LAYER WALL
1 1 1 1 1 1 1
1 1 1 1 1
1
1 1 1 1
1
1
0
2 2
NH H P y F y W
H S y F H T
S
N x P x F
S
W
a V G
p a o p a V
+ + +
+ + + +
= o
(1)
B) TWO LAYERS WALL
( ) ( )
( ) | |
1 1 1 1 , 1 1 , 1 1 1 1 , 1 ,
1 1 01 1 1 1 1 1 , 1 ,
1 ,
1 , 1 , 1 ,
1 ,
1 ,
0
2 2
H N H N H P H P y F y W W
H S y F H T x F x P
s
N W x P
s
N W
B A B A a V G B A
p a a V pB B
B
B B pA A
A
A A
+ + + + + +
+ + + + + + +
= o
(2)
where:
i = (1,i th,.. ,n) number of storeys;
i
W Live load of the wall at i th floor;
vi
F Vertical Component of the arches thrust or on the vaults at the i th floor level;
oj
F Horizontal Component of the arches thrust or on the vaults at the i th floor level;
i
P Weight of the floor acting on the wall at the i th floor level on the basis of the influence
area;
i
T Maximum value of the action of a possible tie at the top of the wall at the i th floor level;
N Vertical live loads of the upper floors acting at the top of the wall, supposed centred on
the centroid of the section;
i
S Thickness of the wall at the i th floor level;
H
i
Height of the wall of the macro element at the i th floor with respect to his centre of
rotation;
ai
y Height of the arches or vaults thrust at the i th floor with respect to the centre of
rotation;
ai
x Horizontal distance of the arches or vaults thrust at the i th floor with respect to the
centre of rotation;
pi
x Horizontal distance of the weight of the i th floor with respect to the centre of rotation;
Gi
y Height of the centroid of the wall at i th floor with respect to the centre of rotation;
0
o Multiplier of the horizontal forces.


208

Figure 6. One layer wall.

Figure 7. Two layers wall.

The Damage Abacus provides, for each single element (vertical structures, horizontal
structures, stairs, non structural elements), a classification of the main damages that may be
found by an observed survey.
Every damage type is described by a specific form containing notes, iconographic review
showing different damage levels following a predefined scale. The basic idea is to create a tool
to assist macroseimic assessment and the surveyors to better compile the sections of the safety
form concerning a synthetic evaluation of the damage level for the structural component
analyzed (only for masonry structures, MEDEA provides 23 different types of vertical structure
damage, 13 different types of horizontal structure damage, 7 different types of stair damage, 6
different types of non structural damage).
Nevertheless, the structure of MEDEA tool allows to make also different kind of analyses,
such the possibility of defining the expected service life of building structures, in order to
integrate LCA (Life Cycle Assessment) and LCC (Life Cycle Costs) studies.
According to this specific point of view, the analysis of visible damages due to past
destructive events or potential damages due to vulnerability factors identified (Zuccaro &

Examples
209
Leone, 2010, this issue) is linked to the need of reliable information about the expectation of life
of structural components in case of retrofit or refurbishment intervention.
MEDEA tool can be used to evaluate the opportunity of maintenance or strengthening
interventions according to the damage and vulnerability parameters identified.
Possible correlation between the safety condition of the building and its expectation of life
are derivable either considering standard functions of resistance decay due to lack of
maintenance or by seismic intensity return period and expectation of damages for specific
building classes.
MEDEA can then be used either in the post-event phase, as tool to perform on the basis of the
cracking path and of the potential collapse mechanisms about to be activated the safety check of
the buildings either in the vulnerability assessments, as tool to identify vulnerability factors
responsible of potential activation of collapse mechanisms and to evaluate possible
interventions of improvement and upgrading as mitigation measure to reduce the effects of the
vulnerability factors, in order to parameterise the expected service life.

Figure 7. Excerpt of damage abacus for masonry structures
(refer to MEDEA multimedia handbook for the complete abacus).

Figure 8. Excerpt of damage abacus for vertical R.C. structures
(refer to MEDEA multimedia handbook for the complete abacus).
At last, an hypothesis of possible links to other damages on the same examined elements or
on those elements of the structure, that could be recognized as compatible with respect to a
specific collapse mechanism, is given. This aspect seems to be, certainly, the most interesting
one; in fact, though it doesnt pretend to be exhaustive and strictly irrefutable, it gives a lecture
key of the single damage in global terms respect to the response of the whole structural system.
This is performed through the analysis in diagnostics key of the whole cracking frame,
characterised by different damages, all compatible with respect to a single mechanism.
Therefore, the aim is to offer a tool able to drive the analysis of the structure through a step by


210
step process of associations among the exiting damages, in order to recognize them as congruent
to a possible global structural behaviour.
In this direction, the interactive training table, in this section of MEDEA, seems particularly
interesting for damage - mechanism correlation. In this table all the mechanisms and all vertical
and horizontal damages of the catalogue with their own codes are listed. The interactive table,
for every damage selected by the user, shows the possible mechanisms congruent to the chosen
damage. Therefore, the user could choose a mechanism to investigate and then he can select in
multi-choice mode the compatible damage typologies; the tool allows to verify in progress the
congruence between these damage typologies and the chosen mechanism.

Figure 9. The safety evaluation section, interactive training table.

Figure 10. An example of the damage description form and the interactive training table.
2.4 The training in safety evaluation
The final part of MEDEA in constituted by a training section, in which some damaged
buildings, realized by QTVR technology, are shown (figure 10). These examples allow the user
to perform training paths designed for safety analyses and therefore to express evaluations on
different aspects, such as:
the constructive typology (horizontal and vertical structures), which a vulnerability class
can be associated to;
the damage level expressed both for every single constructive element and for the whole
building;

Examples
211
the safety assessment of the building;
the possible provisional interventions to be adopted.
Moreover, the tool allows control on the judgement capability achieved by the surveyor.
3 AN APPLICATION TO THE ABRUZZO EARTHQUAKE (APRIL 06, 2009)
3.1 Case study results
On the base of the described approach to the analysis of seismic damage, a specific survey form
(MEDEA form) has been elaborated to be used in an on field data collection in the Comune of
LAquila (Abruzzo earthquake, April 06, 2009). Through this tool a set of data on damage and
collapse mechanisms have been collected in order to analyze the exiting correlation, as showed
in Figure 13.

Figure 11. Frequencies of the observed collapse mechanisms against to the global damage.
Moreover the MEDEA survey form has allowed to collect information about the in progress
collapse mechanisms on the damaged structures and finally to make a hypothesis about the
genesis of the total collapse. This could be reached through the growing up of different
mechanisms, whose sequence is significantly correlated to the typological-structural
characteristics of the building and to intrinsic material behaviour.
In order to define a synthetic parameter to be used in different kind of studies (LCA, LCC,
Safety Assessment), a further development of MEDEA tool consists in the definition of a
procedure to evaluate a Safety Index I
A
, based on the analysis of the evolution degree of the
prevalent mechanism recognized in the building. Of course this index couldnt be assumed as an
automatic judgement for the safety assessment, but it offers to the operators an interesting
parameter for heuristic evaluations.
Furthermore, a correlation between the Safety Index range and the global damage has been
investigated, defining media and deviation standard values. These results, compared to the real
safety judgements collecting on the fields, have allowed to found a bound value for the Safety
Index. Therefore a probabilistic correlation, in terms of percentage of unsafe buildings respect
to the total buildings showing a certain level of global damage, has been defined.
Future development of the procedure, added to a larger amount of data set, will allow to
define a more confident threshold value for the Safety Index and more refined density
probability functions.
This perspective represents an interesting cues to reach quick safety evaluations both on a
territory stricken by a seismic events, both in peace time as design supporting procedure.


212

Figure 12. Damage distribution for different type of mechanism.

Figure 13. MEDEA survey form.
4 CONCLUSIONS
The paper points out the importance of collecting data on damage as much detailed and
codified as possible, so that the different approaches to the vulnerability assessment proposed
by the scientific community could take advantage from such a kind of unambiguous and
homogenised survey. To this aim the authors propose a complete collection of photos, drawings
and detailed descriptions in order to improve the damage data collection and the understanding
of the safety condition of the buildings in the post event phase. The importance of an
homogeneous understanding of the terms used, the damages observed and the potential collapse

Examples
213
mechanisms is traced in the paper and a possible guided path of interpretation is provided.
Hence the multimedia handbook MEDEA for technicians training is presented; it represents an
important step toward the harmonisation of the post event data campaigns. Moreover MEDEA
provides the first attempt to categorise all possible collapse mechanisms of a masonry and
reinforced concrete structure under seismic action and associates at each collapse mechanisms
considered a set of possible damages compatible with it.
Of course, the adoption of MEDEA at international level requires a regionalization of the
procedure, since the difference in building construction techniques and materials used in
different countries.
REFERENCES
Angeletti, P. Bellina, A. Guagenti, E. Moretti, A. Petrini, V. 1988. Comparison between Vulnerability
Assessment and Damage Index, Some Results, Proceed. 9th World Conference on Earthquake
Engineering, 9WCEE, Tokyo-Kyoto.
Baggio, C. Bernardini, A. Colozza, R. Corazza, L. Della Bella, M. Di Pasquale, G. Dolce, M. Goretti, A.
Martinelli, A. Orsini, G. Papa, F. Zuccaro, G. 2000. Manuale di istruzioni per la compilazione della
Scheda di I livello di rilevamento del danno, pronto intervento ed agibilit nellemergenza post-
sismica (AeDES), Roma: Dipartimento della Protezione Civile.
Bernardini, A. Gori, R. Muneratti, E. Paggiarin, C. Parisi, O. Zuccaro, G. Vulnerability and probability of
collapse for classes of masonry buildings. In Faccioli, E. e Pessina, V. (ed). 2000. The Catania
Project: earthquake damage scenarios for high risk area in the Mediterranean, Roma: Esagrafica.
Braga, F. Dolce, M. Fabrizi, C. Liberatore, D. 1986. Evaluation of a Conventionally Defined
Vulnerability of Buildings Based on Surveyed Damage Data, Proceed. 10th European Conference on
Earthquake Engineering, Lisbon.
DAyala, D. Spence, R.J.S. Oliveira, C.S. Papa, F. Zuccaro, G. 2000. The Performance Of Strengthened
Masonry Buildings In Recent European Earthquakes. Proceed. 9th World Conference on Earthquake
Engineering, 12WCEE, Auckland.
Dolce, M. Zuccaro, G. Papa, F. Masi, A. 2002. Remarks On The Seismic Damage In The Recent
Earthquakes In Europe. Proceed. 12th European Conference on Earthquake Engineering, London:
Elsevier Science Ltd.
Papa, F., Zuccaro, G. 2004. Un modello di valutazione dell'agibilit post-sismica attraverso la stima dei
meccanismi di collasso. Atti dellXI Congresso Nazionale Lingegneria Sismica in Italia, Genova.
Zuccaro, G. Papa, F. 1999. Inspection minutes process for the measurement of the seismic damage. In
Bernardini, A. (ed). Seismic Damage to masonry buildings, Rotterdam: Balkema.
Zuccaro, G. Papa, F. 2002. Multimedia handbook for seismic damage evaluation and post event
macroseismic assessment, 50th Anniversary of the European Seismological Commission (ESC):
XXVIII General Assembly, Genova.


214
G. Hauf
O. Hechler
ArcelorMittal, Commercial Sections, Technical Advisory, Luxembourg
Conclusions
The overall objective of monitoring is to improve the knowledge of the real behaviour and to
survey the condition assessment of structures. Hereby monitoring is used for different purposes
like monitoring of structural performance, remaining life-time prediction, model identification
and correlation and long term structural health monitoring.
All these purposes require specific techniques which differ substantially from each other and
which impede the formulations of simple and generic recommendations covering each aspect of
monitoring.
The short introduction about monitoring techniques (Hechler & Hauf) shows that nowadays
monitoring sensor techniques assist to analyze the building condition, to identify existing or
even also starting damages on the structure and to start rehabilitation work at an early stage.
This report gives a short overview of existing monitoring objectives and concepts, their
potential and the technical application of sensors mainly used for building structures. The
development is still in progress; new techniques are regularly appearing, e.g. fibre-optic sensors
or wireless sensor techniques.
In conclusion sensor techniques are an adequate device for monitoring, although the user has
to assemble it carefully and prudentially. Monitoring concepts have to be designed to inspect the
critical locations on the structure. Also the huge amount of data produced by the sensors has to
be checked and filtered and the sensor information should always be counterchecked by
additional visual inspections.
The paper about MEDEA (Zuccaro, Cacace, Rauci) points out the importance of collecting
data on damage as much detailed and codified as possible, so that the different approaches to the
vulnerability assessment proposed by the scientific community could take advantage from such
a kind of unambiguous and homogenised survey. To this aim the authors propose a complete
collection of photos, drawings and detailed descriptions in order to improve the damage data
collection and the understanding of the safety condition of the buildings in the post event phase.
The importance of an homogeneous understanding of the terms used, the damages observed and
the potential collapse mechanisms is traced in the paper and a possible guided path of
interpretation is provided. Hence the multimedia handbook MEDEA for technicians training is
presented; it represents an important step toward the harmonisation of the post event data
campaigns. Moreover MEDEA provides the first attempt to categorise all possible collapse
mechanisms of a masonry and reinforced concrete structure under seismic action and associates
at each collapse mechanisms considered a set of possible damages compatible with it.
Of course, the adoption of MEDEA at international level requires a regionalization of the
procedure, since the difference in building construction techniques and materials used in
different countries.
New technologies in sensor techniques play more and more an importan role for the condition
assessment and survey of structreus (Feltrin, Bischoff, Meyer, Saukh). Wireless monitoring
systems offer new possibilities for the application of monitoring concepts. The 3 year of field
test experience demonstrates that medium and long term data intensive monitoring of civil
structures with wireless networks is feasible and that the produced information complies with

Conclusions
215
the quality requirements in civil engineering. In the beginning, operating the wireless network
reliably over a period of months turned out to be non-trivial task. The problems relied basically
on balancing the requirements of a data intensive application with the requirements of
minimizing power consumption for achieving a sufficiently long battery lifetime. Data
reduction, a powerful method to save power and therefore a key factor for achieving the targeted
lifetime, destabilized seriously the WSN by disturbing the basic operations like time
synchronization and routing.
Nevertheless, with many hard- or software improvements, the problems could be
substantially solved. The availability of the most recent test phase was better than 99%.
Considering that the data loss of the most recent test phase was due to malfunctioning of the
link between the root node and the base station, the availability figure is very encouraging.
Despite the severe hard- and software limitations, relatively complex in-node data processing
could be performed and the accuracy of the generated information was better than 5%. This
figure is very close to comply with the quality requirements in civil engineering, which usually
do not require high precision information. The simultaneous application of several power saving
mechanisms, in particular in-node data processing, allows to easily achieve node lifetimes of
several months. Further significant improvements are still possible, since the duty cycle of the
radio was not optimized. The hardware limitations, however, imposes a tight specialization to
the monitoring task. This implies a detailed analysis and specification of the monitoring goals.
Although not all aspects were investigated in depth, the results obtained by this investigation
demonstrate that there are no fundamental obstacles preventing the application of long term
monitoring systems based on wireless sensor networks to civil structures. This technology is
very close to be mature for practical application. In the near future, progress in low power
hardware will increase the computational resources without increasing or even decreasing the
current power consumption. This progress will allow performing more complex monitoring
tasks and in-node data processing with less compulsion to specialization.
The purpose of collecting datas with sensor techniques and monitoring systems is to assess
the structural condition or even the structural damage of constructions and buildings. Therefore
MEDEA (Manual for Earthquake Damage Evaluation and safety Assessment) was developed as
a multimedia and didactic handbook as one example for structural damage and vulnerability
assessment.
In the paper about monitoring of structural behaviour of buildings (Fabbrocino & Rainieri)
specific aspects related to implementation and operation of Structural Health Monitoring (SHM)
systems are discussed as well as the main issues concerning sensing, data acquisition and
transmission, and data processing.
After an extensive description of the main technological aspects concerning implementation
of SHM systems, attention is focused on the assessment of the health conditions of structures
through modal-based damage detection procedures. One of the main issues relating to their
integration within fully automated SHM systems is the lack of reliable automated modal
parameter identification and tracking procedure. Thus, a procedure recently developed to this
aim has been described in its main aspects and applied to continuous monitoring of modal
parameters in operational conditions and in the early earthquake aftershock. In particular, it is
shown how combination of effective automated modal identification and tracking procedures
can lead to a fast and easy implementation of effective vibration-based SHM systems.
Application of the proposed procedure to data provided in operational conditions by the SHM
system of the School of Engineering Main Building in Naples has given the opportunity to
observe the effects of environmental factors on the modal parameters. Application to the Block
B1 of buildings belonging to the School of Guardia di Finanza in Coppito (LAquila - Italy), on
the other hand, has pointed out the robustness of the proposed algorithm even in presence of a
few installed sensors and spurious harmonics due to dynamic interaction effects. The only
drawback was related to the need of choosing an effective sensor layout when a few sensors are
used for modal tracking. Thus, an effective vibration-based structural health assessment can rely
on automated output-only modal identification procedures, even in seismically-prone areas if
the analysis of the structural response to an input ground motion is not of interest. They can be
confidently applied to follow degradation and damage phenomena, and to investigate the effect
of environmental factors on the structural behaviour of the investigated structure.


216
In the contribution about the survey and condition assessment (Khn & Hechler) a step-by-
step fatigue assessment procedure is introduced, suitable to be applied by practicing engineers
for the evaluation of old steel structures respectively bridges, exposed to dynamic loading. The
proposed assessment procedure is divided into 4 phases. Consulting of experts, non-destructive
testing methods, measurements and analyses of the material are assigned to different phases of
the assessment. The procedure is designed to enable a practicing engineer to carry out the first
phase of the assessment on his own and, upon the results, give advice to the owner. At the end
of each phase, the owner has to take decisions based on a report which shall help him to identify
the most efficient solution for the further use of the bridge. The proposed procedure is a
milestone in knowledge transfer from scientific laboratory towards practicing engineers.
For the assessment of concrete structures the different and specific deterioration mechanism
of concrete components are presented in the contribution by Taushanov& Giarma. These
mechanisms are classified into physical, chemical and biological and are discussed mainly under
the prism of a short presentation of their driving potentials, their characteristics and the factors
influencing their initiation and development. Also, a short reference to their impact on concrete
components is made. In the second part of the paper, the analysis focuses on the inspection of
deteriorated concrete, on the specification and the assessment of the concrete components
deterioration type and degree respectively. Not only methods that can be applied on site for the
investigation and the assessment of the damage, but also laboratorial tests and methods are
referred to and discussed. The range of the methods presented varies from simple tests to
investigations employing complex and technologically advanced equipment.
After presenting the different deterioration mechanism of concrete the focus is now on the
assessment and service-life prediction of reinforced concrete structures (Teruzzi). Within a
probabilistic framework service life is defined as the time at which the probability for a
structure to attain some serviceability limit state reaches a predefined maximum acceptable
value. For degradation of the structure due to carbonation induced corrosion of the
reinforcement bars, a suitable serviceability limit state is represented by the state of
depassivation of the reinforcement. The time-dependent failure probability can thus be
computed with a convolution integral, whose integrand is given by the product of the
probability density function of the time-dependent carbonation depth and that of the time-
invariant concrete cover thickness.
In order to model the progress of the carbonation reaction front it is assumed that the
carbonation depth increases proportionally to the square root of time. The constant of
proportionality is represented by the carbonation rate factor, which can be viewed as a random
variable. The pdf of the carbonation depth can thus be derived from that of the carbonation rate
factor.
For aged structures the distribution of the carbonation rate factor is obtained by means of the
phenolphthalein method on core sample taken out from the structure. Basing on the obtained
distribution the actual failure probability and the (residual) service live of the structure are then
calculated. The results can be used to evaluate the condition of the concrete structure.
For new structures the distribution of the carbonation rate factor can be derived from
measurements of the air permeability of the concrete cover with the Torrent permeability meter.
The data can be used to estimate the service life of the structure, thus allowing the owner to
check whether the durability requirements (expressed in terms of design service life) are
fulfilled.
In the last contribution (Zuccaro & Leone) Structural Damage and Vulnerability Assessment
for Service Life Estimation through MEDEA tool is presented. The integration of support tools
within LCA and LCC analysis is required in order to define reliable environmental and
economic evaluations according to specific design and construction issues. The MEDEA
method outlines a general approach to understand the structural vulnerability factors and the
behaviour of a building under seismic action. It can be used:
in the post-event phase: as tool to perform on the basis of the cracking path and of the
potential collapse mechanisms about to be activated the safety check of the buildings;
in the vulnerability assessments: as tool to identify vulnerability factors responsible of
potential activation of collapse mechanisms and to evaluate possible interventions of
improvement and upgrading as mitigation measure to reduce the effects of the vulnerability
factors, in order to parameterise the expected service life.

Conclusions
217
MEDEA not only allows to describe the main structural typologies but provide also an
implicit analysis of the possible behaviour of the structure related to the identification of
structural lacks and vulnerability factors of the building, as well as the potential collapse
mechanisms.
The tool provides a complete set of simplified numerical models at the limit state of collapse
for the main kinematics mechanisms. This will allow defining the structural vulnerability as the
probability to trigger mechanisms and to correlate these to a synthetic evaluation of
vulnerability and its expected service life.
The future goal of such a procedure is to propose a damage measure for structural elements
with the attribution of a Damage Index for every element based on the indication suggested in
the survey form and in the iconographic part.

Chapter 4. Maintenance, repair and rehabilitation techniques and
planning

Chapter 4. Maintenance, repair and rehabilitation techniques and planning:
Introduction
221
O. Hechler
ArcelorMittal Commercial Sections, Technical Advisory, Esch-sur-Alzette, Luxembourg
Introduction
From a socio-economic viewpoint, novel engineering methods are needed to deal with existing
structures in an efficient way responding to the principles of sustainability. There is thus a great
need to establish corresponding guidance for existing structures which the structural engineer
can rely on.
In this chapter maintenance, repair and rehabilitation techniques and planning for structures
are addressed. As general introduction to this thematic management systems for maintenance,
repair and rehabilitation are presented. For this purpose, the objectives of management systems
management system for maintenance, repair and rehabilitation (MR&R) are referenced whcoh
are: making decisions and plans for the execution of MR&R projects based on the information
of the condition of structures, functional and social requirements and economic and ecological
constraints. In this context the demands on management system are again recalled in the context
of taking into account technical demands, functional performance, safety, economy, ecology
and other indicators of lifetime quality.
Focus of introductive paper is in addition put on computer-aided management system based
on extensive databases as appropriate tool to plan, organize and optimize the maintenance
strategy for structures and to conduct that strategy in a systematic way. Further guidelines for
predictive, life-cycle based and optimizing management practices are presented as exemplified
by the systems developed in Finland and in Germany.
In the second part of the chapter techniques of maintenance, repair and rehabilitation for
buildings are reviewed. The link between general criteria, that govern structural consolidation,
with levels of intervention are outlined and pinpointing on the most appropriate level of
intervention is done.
Further the levels of intervention - safeguard, repair, reinforcing and restructuring - including
suitable techniques are addressed and discussed in detail. Finally application examples for
various types of structures and reasons are presented to provide guidance in the choice of the
appropriate measures.
The second constribution to techniques of maintenance, repair and rehabilitation.for buildings
presents the Swiss codes for existing structures with special focus on its principles and
challenges. As this code is a pioneering initiative not only in Switzerland it can be seen as
template methodology for future structural engineering in the domain of existing structures. In
this contribution the objectives, principles and concepts of the Code as well as the issues of
updating, risk-based safety and proportionality of interventions are presented in detail.
In the third part of the chapter techniques of maintenance, repair and rehabilitation for
bridges are presented. In particular the life-cycle context is introduced and optimization
processes for determining the most effective and least expensive schedules for maintenance
interventions are addressed. For this purpose models that reflect the effects of essential and
preventive maintenance on the life-cycle system performance indicators are presented.
The last contribution aims to give general guidance in order to find a suitable strengthening
method for different parts of civil structures to owners, designers and contractors of bridges.
Superstructures of concrete, metallic bridges and subsoil are included and for each problem,
related strengthening methods are presented.
Further special focus is put on strengthening by bonded CFRP laminates which becomes
more and more important in the sustainability of old buildings and bridges. General


222
recommendation about an appropriate gluing procedure for CFRP-metal joints are given and
two possible surface preparation methods as well as some international testing procedures for
accelerated ageing tests are presented.
Another particular field of investigation, with regard to strengthening efforts on steel bridges,
has been the fatigue issue of orthotropic steel decks. Cracking hazards due to fatigue of the
orthotropic plate are systemized and classified and, as example, the reinforcement of the
deckplate by a steel-elastomer-sandwich (Sandwich Plate System SPS) is shown as one
method to enhance the durability of the deck-plate and of its welded connection.

General
223
E. Vesikari
A. Borrmann & K. Lukas
Technische Universitt Mnchen, Germany
Management systems for maintenance, repair and rehabilitation
1 INTRODUCTION
A management system of structures is a rational and systematic approach for organizing,
planning and executing all the activities related to maintenance, repair and rehabilitation of
structures. The overall activities include (Sderqvist & Vesikari, 2003):
Defining condition
Monitoring and rating structures
Finding and recommending optimum alternatives of maintenance, repair and
rehabilitation (MR&R) actions for structures
Identifying, predicting and prioritizing structures for MR&R actions or demolition
Allocating funds for MR&R actions
Maintaining an appropriate database of information.

The ultimate objective of a management system is to make the necessary decisions and plans
between the inspection of structures and the execution of MR&R projects, as schematically
presented in Figure 1. So, a management system should be able to answer the strategic
questions: Which structures should be repaired? Which MR&R methods should be used? When
to do the MR&R actions? How to combine the actions into projects? How to prioritize the
projects when the budget is limited? etc. All these questions should be answered taking into
account technical demands, functional performance, safety, economy, ecology and other
indicators of lifetime quality.
A predictive and optimizing management system is always based on a well-defined data
inventory. The data structure of the inventory must be consistent with the system needs and a
predefined division of structures into smaller parts. It should contain data on all levels of
structural hierarchy and allow the input of inspection and condition assessment data.
Potential users of a management system are owners of infrastructure and buildings, such as:
Governmental organizations
Municipal organizations
Private companies.

Public organizations and private companies are responsible for the maintenance of bridges,
tunnels, canals, dams, quays, lighthouses, industrial buildings, power plants, parking houses,
office buildings etc.. All these buildings/structures deteriorate with time and require
maintenance. The costs of maintenance, repair and rehabilitation to keep the condition of
buildings and infrastructure on a satisfactory level may be very high for the owners and
management methods for planning, operating and optimizing the MR&R activities are highly
required.


224

Figure 1. The role of a management system between the inspection of structures and execution of MR&R
projects.
The main requirements and expectations of a management system are usually the same,
regardless of the buildings/structures and their owners. Typical needs are the following:
Well organised condition assessment and data inventory for the structures
Guaranteed safety of structures
Safeguarding uninterrupted traffic, production and use of structures
Evaluation and minimization of MR&R costs
Life-cycle based optimization and correct timing of MR&R actions
Objective basis for long-term targets based on engineering, economic and ecological
grounds
Prioritisation of MR&R projects
Budget control

All organizations have some kind of practices to manage the MR&R activities. Only some of
them are systemized, computer-aided, predictable or optimizing. In many cases the MR&R
works are executed only when they are by chance observed necessary without any periodic
inspections or special pre-planning. In the following some guidelines for predictive, optimizing
and computer-aided management practices are presented.
1.1 Management system using LIFECON procedures and calculation methods
LIFECON was a European Union Project in 2001-2003. The objective of the LIFECON project
was to develop a European model of a predicted life-cycle management system for
infrastructures. A LIFECON type management system could be characterized with the
following attributes: predictive, integrated, optimizing, life-cycle based and probabilistic
(Sderqvist & Vesikari, 2003).
Several life cycle management tools based on the LIFECON principles and methodologies
were developed in Finland during the years 2002 2008 (Vesikari, 2008). The tools were
designed applicable to bridges, roads, building envelopes and concrete structures in nuclear
power plants. The structures, whether bridges, roads, buildings or nuclear power plants, were
divided into smaller parts which could be treated as homogenous with respect to materials,
structural features and environmental conditions. These structural parts are called modules
and they serve as basic structural units in the analysis and planning processes of the system. The
Inspection and condition
assessment of structures

Database Management system
Execution of MR&R
projects

General
225
structural databases which serve as initial data sources in the calculation processes are consistent
with the modular breakdown of structures.
Basically the LIFECON management system is generic i.e. independent on the structures to
which the system is applied. Only the degradation models and the MR&R actions are specific to
the structures for which the system is developed. The remaining part could be applied to any
structures in which predictable degradation occurs. The LIFECON management system is
described in detail in (Sderqvist & Vesikari, 2003).
1.1.1 Procedure of the life-cycle management system
Figure 2 shows the calculation process and the data flow of the life-cycle management system
schematically. In the upper part, the databases of structures, modules and MR&R systems are
presented. The databases which in some cases can be connected to exterior databases form a
considereble part of the system.

MR&R System ID
Material and structural data
Consumption and measuring data
Unit cost data
Environmental impact data
Degradation rate data
Retardation data (protection syst)
Reference service life data
Repair effect on condition data

Module ID
Structure ID
Surface area and measuring data
Environmental exposure data
Coefficient on degradation
Condition (inspection) data
Performed actions data (incl.
MR&R System ID, and year)
Data for automatic timing of actions
Manually timed actions
Damage data (inspection)
Structure ID
Construction data
Surface area and measuring data
Environmental exposure data
DATABASE OF
MR&R SYSTEMS
DATABASE OF
STRUCTURES
DATABASE OF
MODULES
DATA OF THE ACTIVE MODULE
Analysis of the protection
system
Timing
System ID of protection method
Predicted service life
Transition probabilities
Condition state distributions
Timing logic
Effects on structural degradation
Costs
Environmental impacts
Analysis of module
Timing
System ID
Predicted service life
Transition probabilities
Condition state distributions
(for several degradation types)
Timing logic
Costs
Environmental impacts

LC ANALYSIS TABLE
DATATABLE OF
SPECIFIED AND
TIMED ACTIONS
Macro
Module specific data from the structural and modular databases
Checking of the values (Min, Max check, default values)
Final values for parameters of LC analyses
Project ID
Structure ID
Module ID
MR&R System ID
Year of action
Quantity of action
Cost data
Environmental impact data
DATA FOR EVERY YEAR WITHIN THE DESIGN PERIOD
MR&R System ID
PROTECTION SYSTEMS
REPAIR SYSTEMS
RENOVATION SYSTEMS

Figure 2. Calculation process of the life cycle management tools.
Only one module in turn is active in the LC analysis processes. The data of the active module
is gathered from the databases. The checking system of initial data consists of a minimum-
maximum check and application of default values in case of missing data. Then the module
specific data are inserted in the LC analyse process.
The LC analyse process has been systemised so that the calculation routine is always the
same irrespective of the module at hand. Only the initial data change according to the active
module. As a result of the LC analyse process MR&R actions are defined and timed and the


226
data on action specific cost and environmental impact data are determined. The data on actions
are gathered by a special macro program to the data table of specified and timed actions.
In some of the management tools the design process goes on to Project Design and Annual
Resources Design. These phases entail sorting, filtering and supplementing of data for the
datatable of the specified and timed MR&R actions.
1.1.2 Combined Life Cycle Analysis
The core of the management system consists of a combined condition, cost and environmental
impact analysis. The condition analysis is produced automatically based on degradation models
and predefined limit states of condition. The condition analysis is stochastic (based on the
Markov Chain method) and is capable of predicting the probability of the structure to be at any
of the condition states during the treated design period. The analysis comprises also an
automatic condition guarding system which is able to trigger MR&R actions whenever the
predefined limit state of condition is exceeded.

Condition
analysis
based on
Markov Chain
+
Condition
guarding
system

Life cycle
cost
analysis

Environ-
mental
impact
analysis

2007
2008
2009
2010
2011
.
.
2050
2051
.
.
.
.
2101
2102
.
.
Design period
MR&R actions

Figure 3. Combined life-cycle analysis.
The Markov Chain based condition analysis and the combined life-cycle analysis is described
in more detail elsewhere (Vesikari, 2008).
1.1.3 Making a life-cycle plan
In the life-cycle management systems the preliminary life-cycle plan is been made automatically
(Vesikari & al., 2009). The designer only selects the structure, design period and the discount
rate. All the other information required for making the plan is stored in the database. The
designer can view the database data and also change them in some cases interactively (such as
the inspection data). After starting the automatic design process the program goes through all
the modules of the selected structure and specifies and schedules the necessary MR&R actions
for each module at hand. The timing of actions is based on the degradation models and the
automatic condition guarding system. As a result of the calculation process the Results form is
opened and the designer can do the following (Figure 4):
View the component specific results of planning including
o MR&R action profiles
o Degradation curves (several types of degradation can be taken into account)
View the structure specific results of planning
o Project data summary (in the program Bridgelife the MR&R actions are
readily organized into projects)
o LCC data
o LCA data
Do manual changes to the preliminary life-cycle plan and
Print the results of the life-cycle planning on paper or to store them in another file.

General
227

Figure 4. Results display of Bridgelife.
Figure 5 shows a typical life-cycle costs display. The cumulative life cycle costs of the
selected structure from the selected design period are presented. Also the average annual costs
are presented. The MR&R costs and user costs are shown as real costs and present value costs.

Figure 5. Life-cycle costs display.
The automatically prepared life-cycle plan can be changed manually. MR&R actions can be
removed, changed (specification/timing) or added manually. After manual specifications the
life-cycle plan for that module is recalculated. The automatic condition guarding system of the
program may be kept in force during the recalculation in spite of the manual specifications. So
the program automatically shifts the timing of non-manually defined actions and secures that the
pre-set limit states of degradation are never exceeded.


228
1.2 Predictive management system based on BIM
1.2.1 Introduction
A software system for the predictive life-cycle management of reinforced concrete buildings is
being developed at Technical University of Munich (TUM). The system can be used for any
kind of buildings, including houses and bridges. Its core is formed by a building information
model (BIM) which not only comprises the 3D geometry of the building but also all data
relevant for performing the life-cycle analysis. A hierarchical subdivision of structures into
several levels allows a detailed allocation of information (Lukas &. al., 2009).
1.2.2 3D Building Model
In conventional life-cycle management systems the data concerning the building, e.g. inspection
data, is allocated textually. This approach is very non-transparent and therefore error-prone as
the inspection planner has to assign the data mentally to their real locations and building
components.
Therefore the TUM approach uses a 3D building model as centre of all data acquisition and
data retention activities. All non-geometric information as
material properties,
environmental loads,
inspections,
monitoring data,
condition changes and
repair actions

is stored in reference to the geometry model. In addition also photos taken at inspection or files
containing inspection results can be attached to the geometric model at the correct location. This
makes the allocation of information much easier and guarantees a good overview over the
buildings condition. The danger of adding information to the wrong building element is
minimized.
The building information model is structured vertically into five levels of detail (LoD) as
shown in Figure 6 (Schiel & Mayer, 2007). This structure is necessary to make the optimum
use of fully-probabilistic deterioration models that are developed.
The first level represents a whole building. There can be multiple entities on this level under
an imaginary level 0 meaning all buildings under the administration of the public or private user
of the management system.
The second level delineates modules. These are groups of building elements belonging together
either from organizational or from functional point of view.
The building elements themselves are stored on level 3. They can be further divided in sub-
element parts on level 4 and 5.
Level 4 describes element parts. These are parts of building elements that are subject of
different environmental stresses than the rest of the building element. For example, wall bases
under salty splash water conditions normally contain higher concentrations of chlorides than the
upper wall segment. Thus the wall bases should be considered separately from the rest of the
wall.
On level 5 hotspots are managed. A hotspot is a place with extraordinarily high
environmental loadings and low material resistance or a place where already damage was
observed. Thus a hotspot can be set by the engineer or bridge-owner during planning (e.g.
jointings) or can be added later when changes in the environmental stresses or damages occur.
As the management system should be useable for such different infrastructural building types
as bridges, parking garages, etc. the exact building structure inside those five levels can be
configured by the user according to his specific needs.

General
229

Figure 6. Hierarchical structure of the building information model (Schiel & Mayer, 2007).
1.2.3 Architecture
The architecture of the life-cycle management system is shown in Figure 7. The system is
structured into five modules grouped around a central database.

Figure 7. Architecture of the predictive life-cycle management system (Schiel & Mayer, 2007).
1.2.4 Database
The database is the core of the system. Herein, all relevant data concerning a building is stored.
This comprises the geometry, all non-geometric data, the material properties, inspection data,
environmental loads and so on. The geometry is stored in form of a boundary representation
model (B-Rep). All the non-geometric data is linked to geometric features, e.g. to surfaces.
1.2.5 Acquisition Module
In the acquisition module the user can add a new building to the database. In a first step he
uploads a 3D geometry constructed in an external CAD program. Inside the acquisition module
he can structure the building according to the five levels mentioned before and add additional
data such as material properties, construction dates, environmental loads and so on. He thereby

Repair Module Prognosis Module
Assessment
Module
Database
Acquisition
Module
Condition Acquisition
Module
Level 2 - Modules
Level 3 - Elements
Level 4 - Element Parts
Level 5 - HotSpots
foundation, pylon, superstructure,
Level 1 - Buildings
bridge A, bridge B, ...
multi-story car park A, ...
bridgehead, ... 1st floor, 2nd floor,
pavement, pylon, ...
wall, slab, binding girder, ...
pylon base, pylon shaft, ...
wall base, ...
pylon base with cracks, ...
wall base coated, ...


230
is assisted by a three dimensional representation of the buildings geometry implemented with
the Java 3D library as shown in Figure 8.

Figure 8. Graphical user interface of the Acquisition Module
1.2.6 Condition Acquisition Module
In the condition acquisition module inspection data is added to the building information model.
The inspection results are preprocessed and statistically evaluated so the user only has to enter a
distribution type and the respective distribution parameters.

Figure 9. Graphical user interface of the Condition Acquisition Module

General
231

This information will be needed for the successional prognosis computation. In addition also
photos and original files created by inspection sensors can be attached to the 3D model. Similar
as in the acquisition module (see Figure 8) a 3D representation of the geometry assists the
correct localization of the data.

1.2.7 Prognosis Module
In the prognosis module the condition prognosis over the buildings lifetime is computed. Here
two different cases have to be considered: For those deterioration processes for which already
quantified deterioration models exist (at the moment depassivation of reinforcement due to
carbonation or chloride ingress) the software package STRUREL (RCP, 2007) is used. In all
other cases Markovian Chains are used as place holders that can later be replaced by fully
probabilistic deterioration models.
STRUREL has originally been developed to perform probabilistic computations for structural
analysis. But as the limit functions can be defined freely by the user, the program can be used
for reliability studies in general.
The communication between the management system and STRUREL is done via files. The
management system exports the deterioration model containing distribution functions for all
parameters needed for the computation. STRUREL returns the condition state of the structure
expressed into a structural reliability and/or a probability of failure over time. These results are
imported into the database.
By incorporation of inspection data which have been obtained from the structure with non-
destructive inspection techniques (e.g. chloride profiles, carbonation depths) it is possible to
specify the prior service life design.
1.2.8 Assessment Module
In the assessment module the condition state of the building is visualized. The user can choose
the level of detail of the visualization. Only the condition indices of the lowest level (surfaces)
are stored in the database, for higher levels (building elements, modules, the entire building) the
condition states (e.g. reliabilities) are aggregated with specific algorithms (run-time based).
The smallest unit within the acquisition of structures is a single surface. For instance one
surface can be one side of the six sides of a column. Therefore one element comprises several
surfaces and one building comprises several elements again and so on. To generate a condition
state on building level (which is called the Building Condition Index BCI), the condition states
of all corresponding elements are needed (Condition Index on element level CI
element
).
To weight different elements to their static and safety relevance it is necessary to introduce
weight-factors which allow a functional assessment of each type of element.

Figure 10: Visualisation of the condition in the Assessment Module


232
1.2.9 Repair Module
Whenever repair actions are taken they have to be recorded in the PLMS. This is done using the
repair module. The repair module provides a catalogue of repair measures for the user to choose
from.
A maintenance measure either resets the condition to a higher level (e.g. in case of
replacement) or slows down the deterioration process (e.g. coating of the surface). In both cases
a new condition prognosis has to be computed starting in the year of the maintenance. The new
prognosis again is fed into the database.
2 DISCUSSION AND CONCLUSION
2.1 General architecture of the recommended management and planning system
The concept of a holistic design platform is recommended for the management and planning of
MR&R activities in the future. The design platform would contain a user interface to databases
of various structural objects and all necessary tools for maintenance and planning. The intended
framework would cover appropriate tools for at least the following purposes:
Optimization and decision making support for selection between design alternatives
or maintenance strategies,
Design tools for life-cycle assessment (environmental impacts), life-cycle costs and
estimated service life,
Design tools for evaluation of the structural performance and safety of structures,
Tools for receiving monitoring and measurement data and for evaluation of structural
condition, capacity, durability or maintainability based on that data,
Tools for management of existing structures i.e. for planning, organizing and
optimizing MR&R over a design period.

In particular, the platform would allow connection of design and analysis tools to a central
database in a plug-and-play fashion. The structure of data is based on BIM and specially
designed for life-cycle design and management. The platform guarantees the data compatibility
of the design tools with the central database and among each other. The platform would also
help to access other applications and data services that may possibly be needed in the design
processes, such as climate information, geological information, external databases, applications
for structural analyses etc.
The design and management tools implanted on a common platform and connected to a
central database will serve as an easy-to-use package for practical designers and managers to
work with. By this, the platform would efficiently promote to sustainable design and
management of buildings and infrastructure that nowadays are in many respects lacking both
resources and systematic methodology.
The platform with appropriate applications would improve the information management and
the quality of design especially from the viewpoint of economy and ecology. The optimization
included in the design may cause savings in both financial and ecological costs. The safety of
structures may also be improved by better information management.
Figure 11 shows a scheme of the life-cycle design and management platform. The
recommended design and management tools on the platform would be the following:
Life-cycle assessment (LCA) of structures. The tool outputs environmental impact
data based on the information entered by the user on structures and material
quantities. The results are given in several impact categories, such as abiotic
depletion, acidification, eutrophication, global warming, etc. A prototype of such a
tool has already been made (ETSI, 2009).
Evaluation of life-cycle costs (LCC). The program integrates all the costs resulted
from actions during the intended life-span of a structure including the construction,
operation, repair works and the demolishing. A prototype of such a tool already
exists (ETSI, 2009).

General
233
Optimisation
Module
LCA
Module
LCC
Module
Servicelife
Design
Module
Condition
Assessment
Module
Structural
Safety Module
Life-cycle
Design
Module Monitoring &
Simulation
Module
Data Input
Database
FEM
Statistics
Climatic
Data
Geology
Data

Figure 11. Platform for life-cycle design and management.
Estimation of service life. The service life is estimated as a function of material,
structural design parameters, exposure classification, maintenance level and the
maximum allowable degradation. The predicted service life is given based on various
safety levels taking into account the scatter of service life. Examples of such
calculators of predicted service life already exist.
Evaluation of the structural condition, performance and safety. The effects of
degradation in materials and components on the structural safety and performance are
evaluated using the Finite Element Method (FEM)
Monitoring and simulation. Reception of monitoring and measurement data from
structures and evaluation of structural condition, capacity, durability or
maintainability based on that data.
Life-cycle design and management of existing structures i.e. planning, organizing and
optimizing MR&R activities over a long design period. The module predicts the
condition of structures taking into account the MR&R actions during that time and
calculates the costs and environmental impacts from the MR&R actions.

The life-cycle design and management tool would form a central part in the system of
planning and organizing the annual MR&R projects. The LIFECON concepts and
methodologies presented in Chapter 1.1 are recommended although other methods are also
possible. The system should be preferably predictive, life-cycle based and optimizing with
multiple attributes. The attributes for optimization would contain safety, performance, costs,
environmental impacts and cultural aspects.

REFERENCES
Sderqvist, M-K. and Vesikari, E. 2003. Generic technical handbook for a predictive life-cycle
management system of concrete structures (LMS), Lifecon GIRD-CT-2000-00378 Lifecon
Deliverable D1.1, final report. 170 p. http://lifecon.vtt.fi/.
Vesikari E. 2008. Life Cycle Management Tools using LIFECON Procedures and Calculation Methods.
Proc Int Sem on Sustainability of Constructions. Integrated Approach to Lifetime Structural
Engineering. Oct 6-7 2008, Dresden. 10p.
Vesikari E., A. Borrmann & K. Lukas. 2009. Latest Software Development for Life Cycle management
of Bridges in Europe. Proc Int Sem on Sustainability of Constructions. Integrated Approach to
Lifetime Structural Engineering. Oct 23-24 2009, Timisoara. 14p.
ETSI, 2009. Salokangas, L. (Ed.). Bridge Life Cycle Optimisation. ETSI Project (Stage 2). Helsinki


234
University of Technology. Faculty of Engineering and Architecture. Department of Structural
Engineering and Building Technology. TKK Structural Engineering and Building Technology
Publications B TKK-R-BE3. 98 p.: http://brolcc.byv.kth.se/etsi.
K. Lukas, A. Borrmann, M. Zintel, T. Mayer and E. Rank, 2009: Developing a Life-Cycle Management
System for Reinforced Concrete Buildings based on Fully-Probabilistic Deterioration Models. Proc.
of the 12th Int. Conf. on Civil, Structural and Environmental Engineering Computing.
Schiel, P, T.F. Mayer, 2007. Schlussberichte zur ersten Phase des DAfStb/BMBF
Verbundforschungsvorhabens "Nachhaltig Bauen mit Beton". Deutscher Ausschuss fr Stahlbeton
Heft 572, Beuth Verlag, Berlin, Germany.
RCP, 2007. Documentation of STRUREL, RCP, Munich, Germany.

Buildings
235
V. Ungureanu
1,2
, A. Dogariu
1
, D. Dubina
1,2
, A. Ciutina
1

1
Politehnica University of Timisoara, Romania
2
Romanian Academy, Timisoara Branch, Romania
R. Landolfo, F.M. Mazzolani, F. Portioli
Maintenance, repair and rehabilitation of buildings
1 GENERAL APPROACH OF STRUCTURAL INTERVENTION
The general criteria that govern structural consolidation are linked with the preliminary
definition of the level of intervention. Pinpointing the most appropriate level of intervention
depends on multiple variables concerning: the importance of the building, the destination of the
building after consolidation, the technological systems engaged, the degree of safety necessary,
and last but not least the available funds.
A classification of the level of importance can be put into relation with the fundamental
stages of the logical and chronological process which has the final purpose of recovering of the
static efficiency of the building and its preservation in time.
The levels of intervention are (Mazzolani, 1990):
o Safeguard;
o Repair;
o Reinforcing;
o Restructuring.
SAFEGUARD implies a set of measures meant to assure the safety of the building in the
transitory phase, if important damage occurs after an earthquake.
REPAIR implies measures that aim to restoring the initial structural efficiency, prior to
earthquake action.
REINFORCING and RESTRUCTURING, do not imply, as opposed to the first two levels,
the existence of damage that compromises the safety of the building, but are necessary when
functional reorientation of the building and its alignment with new standards of strength and
stability are desired.
1.1 Safeguard
Safeguard represents the first level of intervention that consists of provisional measures meant
to assure an adequate level of safety in the transition phase. This precedes any subsequent
interventions that have a definitive character.
It refers to the situations where quick recovery interventions are needed until partial or total
consolidation of the building is done. It is used both for the protection of the site and the
prevention of collapse.
Typical cases when safeguard measures are needed are the emergency cases that follow an
earthquake, when the lack of material requires a quick and simple intervention with maximum
efficiency.
For these cases steel elements are recommended because of the ease of application, their
prefabricated characteristic, ease in transportation and assemblage regardless of the atmospheric


236
conditions and last but not least the economical advantage that derives from the recycling and
possibility or re-using steel.
1.1.1 Field of application
This intervention level finds its utility in the case of:
o temporary support of a building during the reconstruction of a new building placed
between two other buildings (Figure 1a);
o steel structures that support the faade of a building during the demolition of a part of this
building for replacement (Figure 1b);
o temporary support for building faades in the consolidation phase after a earthquake
(Figure 1c);
o provisional roofing during consolidation works (Figure 1d and 1e) (Mazzolani &
Mandara, 1993).

(a)

(d)

(b) (c) (e)
Figure 1. Safeguard application (Mazzolani & Formisano, 2005)
1.1.2 Technical details of the intervention
Tubular steel scaffoldings are used as ad hoc solutions that present good flexibility and
optimal results.
Heavy steel structures (welded sections or rolled sections with bolted connections) and light
steel structures (bolted tube sections) proved very effective in passive or active safeguard
measures.
The character of the safeguard measures must satisfy the following conditions (Mazzolani,
1990):
o speed of execution;
o flexibility of the constructive systems;
o adaptability to narrow and hardly accessible work spaces;
o reversibility of the intervention.
In this phase traditional materials as masonry and especially timber still represent a viable
alternative.
Provisional works for support and protection have a temporary character and belong to the
framework of safeguard interventions. Provisional works are used even during the execution of
final consolidation works (to protect the building during replacement of the covering), or during
the erection on a new building (scaffolding, passageways).
Until a few years ago these types of works have been done with timber elements, for example
passageways and support struts etc.

Buildings
237
In the framework of modern solutions, prefabricated steel, due to the current technology can
be easily implemented ad-hoc due to its great flexibility and diversity of shapes and sizes.
Their advantages consist in the ease of removal and possibility of re-using. Thanks to the great
number of possibilities offered by the constructive system, tubular scaffolding can be used in
any typical consolidation works. Reticular structures with braces can be built as real supporting
structures scaffolding with excellent set up and dismantling time, offering a viable alternative
to brick masonry contraforts type structures.
One must not forget about works done as a precaution, capable to prevent landslides or any
other phenomena, not entirely controlled, that tend to cause degradation of the existing
structure.
1.1.3 Examples of application
In the field of safeguard, passive interventions, structural masonry elements are used as
contraforts to support the faades of buildings, especially in Mediterranean countries (Figure 2).
This system has the disadvantage of high weight of the masonry supports that require their own
foundation, and in many cases the support structure fails before the consolidated building.

(a) (b)
Figure 2. Contraforts application (Mazzolani & Formisano, 2005)
Timber struts of different types and shapes are inserted into the masonry structures in order to
achieve an even distribution of vertical loads before the consolidation works (Figure 3a) or to
assure lateral support and access facilities (Figure 3b). Bamboo structures and scaffoldings are
widely spread in Asia as service systems during refurbishment operations (Figure 3c). There is
the advantage of the reduced weight with satisfactory strength, ease of manipulation and good
workability, flexibility and adaptability of the material in all situations, and the reduced price
makes of timber a very frequent choice in this field.

(a) (b) (c)
Figure 3. Wooden safeguard works (Mazzolani & Formisano, 2005)


238
An interesting solution proposed for supporting the Pisa tower in order to avoid collapse is
the use of a steel ring which surrounds the tower, as presented in Figure 3.
Steel elements both for passive intervention and active ones proved efficient in many cases.
Consolidation of Palazzo Carigliano in Turin (Figure 5). A provisional support system
consisting of hot-rolled profiles was designed to support the stone columns at the entrance hall
during the consolidation operations of replacing the old floors with new ones. Steel profiles and
corrugated steel sheets acted as formwork for the concrete cast slab (Mazzolani, 1990).

Figure 4. Supporting rings for Pisa Tower (Mazzolani & Formisano, 2005)

Figure 5. Palazzo Carigliano (Mazzolani, 1990)

Rebuilding of the Waring and Gillow store in London (Figure 6). Steel columns and girders
acted as a bracing system used to preserve the old faade. Once assembled, these elements
created an ad-hoc structure that supported the faade and a 5 m area behind it during the
demolition of the remaining parts of the building. After interior reconstruction, in conjunction
with current functional lay-out, the steel elements were gradually removed.
Restructuring of the old Muller theatre in Darmstadt (Figure 7). Horizontal steel rings fixed to
the perimeter walls were used as temporary bracing that kept the walls of this building together
while it was transformed into State-record Office.
Restructuring a building in the centre of Lisbon (Figure 8). A major fire led to the emptying of
a large masonry building in the centre of Lisbon. The remaining faades and the internal core
were temporary supported by means of steelwork and perimeter ties prior to the restoration and
consolidation operations.

Buildings
239

(a) (b)
Figure 6. Waring and Gillow store (Mazzolani & Formisano, 2005)

Figure 7. Darmstadt theatre
(Mazzolani & Formisano, 2005)
Figure 8. Lisbon building

Max Plank Institute in Rome (Figure 9). A complex consolidation operation for an old building
in Rome required the temporary support of the faades by means of steel structures especially
designed for this purpose before the degutting operations began (Marimpietri et al., 2004).

(a) (b)
Figure 9. Max Plank Institute (Mazzolani & Formisano, 2005)


240
1.2 Repairing
Repairing follows in chronological order after provisional safeguard and involves a series of
operations to restore the buildings to its former structural efficiency. In essence it is about
restoring the status quo ante of the retrofitting elements (Mazzolani, 1990).
The repair takes place when the safety of the building is compromised by damages sustained
due to earthquakes, atmospheric agents, or simple aging. As opposed to safeguard, repairing is a
definitive operation and is normally used when the degradation of the structure can be easily
diagnosed and comes from long term use or old age and doesnt require urgent intervention.
This is why there are cases when the economic aspect comes before safety aspects, when simple
restoration of the structural performance is carried out without any strengthening of the
construction. There is the possibility of making a simple repair even in the case of severe
earthquakes when the lack of funds doesnt allow a large scale consolidation.
From the point of view of safety, repairing offers a simple restoration of structural
performance without introducing supplemental elements or strengthening of the building.
1.2.1 Fields of use and examples of application
In repairing there are a series of consolidation techniques based on the use of steelwork to
improve the behaviour of all types of structures from steel to masonry and reinforced concrete
or timber (Mazzolani, 1990). Beside steel there can also be used for repairing other traditional
materials like masonry, reinforced concrete and timber. Interventions on buildings affected by
atmospheric agents and the passing of time are presented in Figure 10, Figure 11, Figure 12,
Figure 13.
The most commonly used materials in repairing operations are steel and more recently fibre
polymers, which are considered modern and high performance materials.
Structural steel elements offer, by means of prefabricated technology, ad-hoc solutions
designed to achieve optimal results and tailored to meet specific requirements (Mazzolani,
1994).

Figure 10. Masonry building damaged in time
Figure 11. Old wooden floor (Mazzolani &
Formisano, 2005)

Figure 12. Damaged wooden structure (Mazzolani
& Formisano, 2005)
Figure 13. Corrosion phenomena in steel structure

Buildings
241
The specific requirements mentioned above can be detailed in the case of steel as follows:
o reduced weight that allows for easy transport, easy handling and erection which are
important in conditions that require manoeuvring in the narrow spaces of historical centres
of old towns;
o reversibility, thanks to the use of bolted joints, that allow the reuse of the structure after
dismantling;
o speedy erection, a very important characteristic in the case of emergency repairs when
damage spreads quickly;
o economically convenient, thanks to the possibility of re-use;
o modern feature clearly identified with the additional advantage of the use of technologies
and materials that can be removed at any time without damaging the building.
Recently, the use of fibre reinforced polymers (FRP) has greatly increased in the repair of
existing buildings. This material is one of the new materials, of great innovation, that are studied
in recent times.
FRP are used as reinforcement bars for concrete or masonry structures or as strips for the
strengthening of certain structural elements.
The most important characteristics of FRP are:
o excellent strength/self-weight factor, approximately 40-50 times higher than steel;
o flexibility, easy to shape and use;
o corrosion-free;
o highly resistant to fatigue;
o reduced weight, easy transportation and use and does not bring additional load to the
building;
o short time of application;
o durability no special maintenance is required and can withstand aggressive environments;
o satisfactory fire resistance;
o aesthetics can be of different colours and textures.
The development of FRP is mainly justified by the absence of the corrosion phenomena that
affect both steel and reinforcement in the case of reinforced concrete structures. The
disadvantages of this material are its fairly high price, lack of ductility and insufficient research
of its behaviour in time. In the matter of ductility, those who promote FRP support the existence
of a global ductility, in the absence of the material ductility.
Masonry structures
Vertical structures made of masonry, like columns or walls can be repaired using one of the
methods that will be described below (Mazzolani & Ivanyi, 2002). Repair methods aim to
improving the behaviour of the element by restoring its load bearing capacity to gravitational
loads but also to horizontal loads that come from the failure of foundations, geometrical
asymmetry and seismic action. These methods in this stage of the consolidation do not aim to
improve the dynamic behaviour but only to achieve local repair of certain elements. In this
paragraph only masonry structural elements will be discussed, the partitioning elements being
neglected.
Improving the bearing capacity to vertical loads
With the help of steel elements:
o Confinement of the damaged masonry column by vertical angles and horizontal plates or by
a combination between them in the case of columns with rectangular cross-section, and with
rings in the case of circular cross-sections of the columns (Figure 14).The purpose of this
method is to achieve a tri-axial state of stress by preventing lateral deformation, which
results in a significant increase in resistance to compression and in the ultimate ductility of
the consolidated element. The steel system prevents the instability phenomena of the
masonry column and reduces cracks that occur parallel to the force direction. Steel elements
must be fitted with a precompression force that increases the effectiveness of the method.
An old method of achieving this precompression was to apply the steel rings at a high


242
temperature. Nowadays the use of circular rings tightened with bolts leads to the same
result.
o Inserting a new steel element along or inside the consolidated element, that can be either a
column or a wall (Figure 15). The new inserted elements take up completely or partially the
loads that belong to the vertical element. All types of steel, hot- or cold-rolled steel products
can be applied successfully for this method.
o Inserting a steel frame, inside or around the opening of a wall, to restore the initial strength
of the wall (Figure 16).This frame takes up vertical loads (acting as a lintel) and gives the
masonry panel higher resistance to horizontal loads, achieving a unitary behaviour.

Using fibre reinforced polymers:
Repairs can be done by means of polymeric materials used as reinforcement bars or as strips.
The most used fibre reinforced polymers are CFRP Carbon Fiber Reinforced Polymers
which comes in strips, glued by epoxy resins on the surface of the consolidated element. For the
use as reinforcement bars, they are called NSMR Near Surface Mounted Reinforced. NSRM
are placed in channels that are cut into the consolidated element and are covered with concrete.

Figure 14. Consolidation of a masonry column
Figure 15. Consolidation of a masonry wall

Figure 16. Steel frame used for a wall opening
Figure 17. Repair of a reinforced concrete column
by means of steel angles and ties (Mazzolani &
Formisano, 2005)
Improving load bearing for horizontal loads
o Securing the faade walls with horizontal steel beams connected to transversal walls by
means of ties that on the same time confine each floor in an attempt to tie the walls together
so to achieve rigid body behaviour of the floor.

Buildings
243
o Securing the corners of the building by means of steel columns connected by a system of
beams and ties achieving the unification of the walls and making the transfer of horizontal
loads to the walls that are parallel to the direction of the possible earthquake. Thus, a
confinement of the building is obtained and the rigid box behaviour of the building is
achieved.
Reinforced concrete buildings
In recent times, most of civil buildings, both apartment buildings and office buildings, state
institutions, schools, hospitals, have been constructed out of reinforced concrete. A
classification function of the resistance structure divides reinforced concrete buildings into two
categories: frame structures (with beams and columns as main resisting elements) or structures
with reinforced concrete bearing walls diaphragms. In the recent past, a great accent has been
put on the development of prefabricated reinforced concrete structures, and, nowadays, a great
number of buildings exist that have been built by means of this technology.
Frame structures
With the help of steel elements
The methods used in the case of reinforced concrete elements are similar to those used for
masonry and are based on the same principles, that is by achieving a tri-axial state of stress,
meaning confinement of the element, and preventing it from falling apart. There are multiple
measures to repair columns that have the effect of restoring the initial resistance to horizontal
action but also to vertical actions. Some of these measures are mentioned below (Mazzolani,
1992):
o Reinforced concrete columns can be consolidated with steel angles connected by steel
profiles or steel ties (Figure 17) or by welded plates, thus achieving a higher degree of
confinement (Figure 18). In the case of circular columns, steel rings bolted together are
used. In this case steel elements can be concealed by covering with a concrete layer.
o When the intervention must be visible, steel elements can be bolted to the existing column,
and the steel ties that are used in connecting the steel profiles (U) go through the reinforced
concrete (Figure 19a). Cold-formed steel profiles with bolted connection placed around the
column give the same result (Figure 19b).

(a)
(b)
Figure 18. Steel angles used at consolidation of
RC columns (Mazzolani & Formisano, 2005)
Figure 19. Increasing RC column section with hot-
rolled profiles (a) and cold-formed profiles (b)

The beams can be repaired in one of the following ways presented below. The plating of beams
with steel angles and batten plates is used in order to improve the resistance to bending and
shearing of the beam and also of the connection between columns and RC beams. Welding,


244
bolting or gluing by epoxy resins are used to ensure the connection between the steel elements
and the consolidated RC element. The method is similar to the one used for columns or elements
subjected to compression, but an increase in flexural behaviour is also desired. The flexural
resistance modulus is significantly increased by enlarging the section and attachment of
longitudinal steel elements to the extreme fibres of the beam. Moreover, horizontal steel plates
have a significant role in improving shear performance. Additional elements can be compared,
from the point of view of their function, with longitudinal reinforcement of the horizontal
elements along the beam and the vertical transversal elements with stirrups.
Vertical ties that go across the upper part of a reinforced concrete slab and bolted to a bottom
steel plate act like additional stirrups for the existing RC beam. The way this system works is
similar to the system described before.
Cold-formed steel profiles or plates can be placed at the lower part of the RC beam. Steel
profiles are fixed by means of bolts or epoxy resin or by adhesive mortar (Figure 20). The
gluing process has the advantage that unlike bolted connections it doesnt introduce stress
concentration in the concrete element.

(a)
(b)
Figure 20. Steel plates or profiles fixed at the bottom:
(a) or laterally (b) on the beam by means of bolts or epoxy resins (Mazzolani & Formisano, 2005)
With the help of fibre reinforced polymer
One of the modern means of repair consists in the use of FRP in order to strengthen the RC
beams (Figure 21) and columns (Figure 22). Prior to the application of the FRP, the support
layer needs to be prepared, with sand or water under pressure, by degreasing and drying of the
surface, and the use of special adhesives with epoxy resins. This method is easy to use and has
the advantage of quickness of application. The application of the FRP can also be done by
gluing the strip on a pretension element or achieving the pretension with the help of the strip. In
this case the end zones of the strips require special attention.

Figure 21. Repairing intervention on a RC beam
using CFRP (Mazzolani & Formisano, 2005)
Figure 22. Repairing intervention on a RC
column using CFRP (Mazzolani & Formisano,
2005)

Buildings
245
In the case of columns the aim is to obtain a tri-axial state of stress.
Prefabricated FRP angles that are different in shape and obtaining process than the normal
strips are often used for beams. The use of these elements requires sufficient anchorage of the
FRP in the compressed zone of the element.
Steel structures
Steel structures can be divided, from the consolidation point of view, into new and old steel
structures. The difference between these two types consists in the different construction method
and profiles used. Two different means of approach can be distinguished.
In the case of structures built with materials and technologies that are still used nowadays, the
consolidation technology focuses on integrating new steel elements. Methods with a high degree
of generality and applicability can be established due to the modular characteristics of steel and
to the existence of standard profiles.
One of the following methods can be used for repairing new steel structures:
o The increase of the member cross-section by welding or bolting new structural steel
elements results in an improvement of the strength and rigidity characteristics (Figure 23).
New elements can either reinforce or increase the area of the flanges or of the web in
function of whether an increase of resistance to bending or to shear is desired.
o Stiffening the structural nodes by introducing steel profiles, plates or stiffeners.

Figure 23. Repairing of steel cross-sections by welding or bolting new elements (Mazzolani &
Formisano, 2005)
Floor structures
Floors have mainly vertical load bearing capacity but the rigid diaphragm behaviour is a
prerequisite in the case of any consolidation work. So as a consolidation must provide, from the
anti-seismic point of view the diaphragm behaviour. The floor types that are usually subjected
to consolidation are the following (Mazzolani & Mandara, 1991):
o Floors with wooden beams;
o Floors with steel beams;
o Floors with reinforced beams and tiles.
Wooden floors can suffer degradation and deterioration with the passing of time: rotting
caused by atmospheric agents, loosening of the supports due to vibrations, weakening of the
link between the floor and other structural elements etc. As an alternative to completely
replacing the wooden floors, one of the two methods described below can be used:


246
a) Working from the bottom upwards, repairing each wooden beam in order to increase their
strength and reduce their deformations. Following this procedure, the following methods are
used:
o Steel beams are placed in the middle of each wooden beam (Figure 24);
o Placing a pair of steel profiles (double T, U, angles, cold-formed) on each side of the
wooden beam (Figure 24); in some cases they must be integrated by means of steel sheets or
ferro-cement coating in order to support secondary elements (Figure 25).

Figure 24. Repairing a wooden floor from the bottom upwards (Mazzolani & Formisano, 2005)

Figure 25. Example of intervention on a wooden floor (Mazzolani & Formisano, 2005)

b) The solution of working from the top downwards is chosen when the wooden beams are in a
good condition and they may remain visible. This solution can be used in the following ways:
o A steel beam is attached at the upper part of the wooden beam, linked by an appropriate
connection system (connectors) to ensure that the two elements work together (Figure 26).
If the connectors are correctly dimensioned so as to prevent slipping, a mixed steel-wood
system is achieved that has improved strength and rigidity;
o A more complex alternative to this solution is by creating a multiple composite system: the
reinforced concrete slab together with the profiled steel sheet work together with the steel

Buildings
247
beam by means of welded stud shear connectors. In the same time, it is assured that the steel
beam works together with the wooden beam. Thus, a composite steel-concrete-wood system
is achieved (Figure 27).

Figure 26. Repairing wooden floors by steel beams and connectors that
create a composite steel-wood system (Mazzolani & Formisano, 2005)

Figure 27. Repairing a wooden floor by using a multi-composite system (Mazzolani & Formisano, 2005)

In both cases presented, the achieving of a satisfactory collaboration between the steel and the
existing wooden element is difficult because a connection with enough strength and rigidity to
shear is difficult to achieve. That is why the finding of an appropriate solution for the
connection is very important.
Modern methods propose a system that uses glued steel elements (bars or plates) completely
hidden in the wood. Although these new systems are highly expensive, they present a
significant advantage: good rigidity of the connection, ductile design of the joints due to the
yield of steel, protection of the connection against aggressive factors, better architectural look.
Wooden floors are also characterized by weak horizontal rigid diaphragm behaviour, a
fundamental requirement in modern anti-seismic design. This behaviour can be improved by
installing steel braces or profiled steel sheeting, or by casting a concrete slab that significantly


248
increases the rigidity of the floor. However this procedure requires that the floor should be in a
good condition and correctly dimensioned.
The floors with steel beams frequently used at the beginning of the 20th century, are
composed of I-shaped steel beams together with brick or tile vaults with concrete. Due to the
degradation of the restraint condition, an increase in rigidity is often required. This can be done
by using one of the techniques bellow:
o Working from the bottom upwards, the section modulus can be improved by welding to
their bottom flange steel sections like plates, box sections, upside down T, double T etc.
(Figure 28). In addition, cold formed profiles and omega profiles, can be attached by means
of bolted connections.
o Working from the top downwards, a reinforced concrete slab can be connected by means of
appropriate connectors (stub, angles, T etc.) so that it should collaborate with the existing
floor slab. Additional steel elements can also be placed at the top part of the steel beam
(Figure 28) or the section modulus can be increased by means of CFRP strips.

Figure 28. Increasing the steel beam resistance by welding additional elements

Composite slabs or tile floors can be repaired using one of the methods bellow:
o Repairing the RC beams by adding new reinforcement and special concrete after the
degraded parts have been removed;
o Gluing steel plates or FRP at the bottom part of the RC beams (Figure 29);
o Inserting steel I profiles between the RC beams.

Figure 29. Repairing by attaching steel plates (Mazzolani & Formisano, 2005)
Roofing structures
Roofing of masonry structures is generally made of timber trusses which are often deteriorated
due to the permanent contact with atmospheric factors. If the wooden elements are in an
acceptable condition, a repairing solution can be found.
The repair operation consists in covering the wooden elements with steel plates or cold-
formed profiles with bolted connection or glued with resins.
The use of CFRP, in the form of strips or reinforcement, continues spreading, although it has
no solid economical justification.

Buildings
249
If the wooden structure is in an advanced state of deterioration, the optimum solution is to
replace the whole structure with new steel trusses integrated by trapezoidal sheeting and
concrete casting. This system was used in many cases for the replacement of roofing structures
of churches damaged by earthquakes (Figure 31 a, b).

Figure 30. Repairing a wooden roof structure by using steel plates (Mazzolani & Formisano, 2005)

Figure 31. New steel roof structure of a restored church (Mazzolani & Formisano, 2005)
1.3 Reinforcing
It is the next step after repair, and it does not necessary imply a state of emergency of the
building, but rather has the purpose of updating the bearing capacity and giving the structure the
capacity to face new functional demands (new loads resulted from the change of functionality)
or the change of position (as inclusion in a more severe earthquake risk zone) (Mazzolani,
1990).
The consolidation does not produce any significant change in the structural scheme but brings
new structural elements that need to be integrated with existing elements without changing or
significantly altering the mass and rigidity distribution of the structure.
Unlike a simple repairing intervention, the consolidation can be made at different
performance levels that depend on the level of safety that is desired to be achieved. This makes
the consolidation operations fall into two categories from a seismic point of view:
o simple improving;
o seismic upgrading and adaptation.
Consolidation for improvement includes interventions on the whole building or just a part of
it, with the aim of achieving a higher degree of safety without excessively modifying the static
scheme and the global behaviour of the building. Improvement interventions repair elements in
order to prevent local deficiencies that appear because of inadequate structure designs or
execution (Figure 32 and Figure 33).


250
Consolidation for adaptation consists in a whole complex of operations necessary in order to
assure that the structure has the required seismic capacity as presented in anti-seismic codes.
This can also include the change of the static scheme with complete modification of the global
behaviour (Figure 34).

Figure 32. Improving operations on a beam-to-
column connection by adding welded parts
Figure 33. Improvement operations on a RC node
by means of steel angles, plates and battens

Figure 34. Metallic braces systems for seismic conformance of reinforce concrete frame

Buildings
251
1.3.1 Fields of use
Improvement is applied in the following situations:
o Buildings under the condition of load increase due to the change of destination of the
building, which leads to an increase in live load;
o Existing buildings that were included in a new seismic zone or for which the conditions
of determining the seismic loads were changed, leading to an increase of seismic load.
For monumental buildings seismic conformation, especially in the case of big churches, there
raise a big number of problems due to large free spaces and painting on the walls that could
suffer and loose their artistic value if extended consolidation works are executed.
For all these situations, the design codes foresee distinct conditions for structural
improvement or for a rigorous anti-seismic conformation.
This last provision can be applied in the following cases (Mazzolani, 1994):
o increase of loads and functional demand changes as a result of the change of
destination;
o for storey adding or extending, which results in the increase of volume and area;
o when the necessary changes for consolidation essentially modifies the static scheme of
the existing building or the global behaviour.
The conformance can be avoided only if the loads are not modified or if the consolidation
operations are made for monumental buildings. In these cases it is enough to limit to the
amelioration interventions, which can include techniques with a reduced impact on the existing
building.
1.3.2 Technical details of intervention
In order to obtain different levels of strengthening, from the simple improvement of elements
behaviour to the modification of global response, the same techniques can be used as for repairs,
but with a higher degree of consistence and a much larger generalisation.
The solutions based on steel are used widely to obtain an improvement in static behaviour of
brick masonry or reinforced concrete structures. Lately, improvement solutions based on FRP
have been used with good results for brick masonry or reinforced concrete structures.
Metallic braces systems are widely used for anti-seismic conformance for the structures
mentioned before (Figure 37 and Figure 38) (Mazzolani, 1992). These systems are supposed to
optimise the structural response by increase of strength and rigidity of the structural assembly
for horizontal forces. Systems based on centric or eccentric braces can be used. The use of
centric braces (St. Andrews cross, as Figure 35 and Figure 36 shown) offers a significant
increase of rigidity and a negligible weight increase. For some cases, less rigid solutions are
suitable, like eccentric braces which offer considerable increase in ductility. One of the
advantages of the system is the fact that no new independent foundations are required, but a re-
evaluation of the existing ones for the new global behaviour of the structure. Special attention
must be given to the connection of the concrete element with the metallic system because the
connections are the vulnerable element during an earthquake.

Figure 35. Electrical power plant, Hungary
Figure 36. Church in Avelino (Italy) (Mazzolani &
Formisano, 2005)


252

Figure 37. Block of flats in Santa Monica
(California, SUA) (Mazzolani & Formisano, 2005)
Figure 38. Building in Thessaloniki (Greece)

In the category of systems based on braces also fall new systems that use modern
technologies: steel buckling restrained bracing BRB (Figure 40), braces that are made of
shape memory alloy bracing SMA-B (Figure 41), or low yield steel or pure aluminium
stiffened panels (Figure 42) (Experiments done in Bagnoli, Naples, Italy ILVA Project)
(Mazzolani et al., 2003).

Figure 39. Reinforced concrete structure consolidated
with EBF (Mazzolani & Formisano, 2005)
Figure 40. Reinforced concrete structure
consolidated with BRB (Mazzolani &
Formisano, 2005)

Figure 41. Reinforced concrete structure strengthened
with SMA braces (Mazzolani & Formisano, 2005)
Figure 42. Aluminum stiffened panels

The accepted technologies for consolidations can be schematically represented in the matrix
described bellow. It can be observed that not all the materials are suitable to be used for the

Buildings
253
consolidation of structural types defined by the material used. The use of materials different
from the base material can create the so-called mixed technologies based on the composite
action of different materials, which form new mixed materials (Table 1).

Table 1. The structure material relation (Mazzolani & Formisano, 2005)
New composite materials
Materials used for consolidation
Structure to be consolidated
Metal Concrete Masonry Wood FRP
Metal XX X
Concrete XX X X
Masonry XX X X X X
T
y
p
e

FRP XX X X

There can be noticed, from Table 1, the impossibility to combine all the materials, but the
composite action of two different materials can result in a behaviour superior to individual
materials.
Not all the possibilities presented before fulfil one of fundamental demands in modern
consolidation philosophy, i.e. the reversibility. This requirement becomes very important in the
case of monumental buildings. If the designer decides to use irreversible techniques, it is very
important to take into account the following aspects:
o Compatibility mechanical properties for the material used for consolidation in relation
with the structure (strength, deformability, thermal expansion coefficient etc.);
o Durability refers to the new material characteristics that must be compared with the
existing traditional materials.
Sigma Coating building from Agnano, Naples. The change of destination for an industrial hall
required the change of capacity due to the increase of live load from 2 kN/m
2
to 20 kN/m
2
. The
existing structure was composed of reinforced concrete, two storeys, and frames with isolated
foundations under the columns. The consolidation consists in (Mazzolani & Mandara, 1991):
o the transformation of the foundations from isolated to continuous reinforced concrete
foundations (Figure 43);
o the increase of the axial force capacity for columns by disposing supplementary cold-
formed profiles;
o the increase of the bending moment capacity for beams by disposing a 20 cm concrete
slab at the top and achieving the composite action, and binding at the bottom of a
metallic plate, connected to the top through metallic ties (Figure 44).

Figure 43. Consolidation of an isolated reinforced
concrete foundation (Mazzolani & Formisano, 2005)
Figure 44. Consolidation of a reinforced concrete
beam (Mazzolani & Formisano, 2005)

Reinforced concrete building at University of California, Berkeley (USA). After the Loma
Prieta earthquake, in 1989, California, many buildings from Berkeley were strengthened by
disposing metallic braces systems (Mazzolani & Ivanyi, 2002). The main university building is


254
a representative example of metallic braces that form a network disposed over two storeys. This
network increases the strength of the building and gives a new architectural aspect to the old
faade (Figure 45). The same intervention system, based on the same structural principles, was
applied to an apartment building for students (Figure 46) and for a car parking building (Figure
47).

Figure 45. The main university building Berkley,
California (Mazzolani & Formisano, 2005)
Figure 46. Apartment building for students,
Berkeley (Mazzolani & Formisano, 2005)

Figure 47. Car parking building in Berkeley
(California) and detail of the base hinge
Figure 48. A steel mill in Bagnoli (Naples)

ILVA continuous casting mill in Bagnoli, Naples. The earthquake from November 1980
happened when the structure was in the erection phase (Figure 48), the structure being designed
in accordance to old design codes.

Figure 49. The connecting structures at the foundation level (Mazzolani & Formisano, 2005)

Buildings
255
After this event, the location was included into a new seismic zone, so that the structure had
to be modified in order to fulfil the new demands and it was compulsory to upgrade the
structure. Following the calculus, it was established that the structure have not sufficient
strength on longitudinal direction. Due to columns design, built in for transversal direction and
pinned on the longitudinal one, it was possible to limit the intervention only to the longitudinal
direction. The necessary structural intervention consists in connecting the columns bases
through a metallic system and in increasing the capacity for braces (Figure 49).
1.3.4 Performance Based Seismic Evaluation of a Non-Seismic Masonry Building of Metal
Sheathed Walls
Recent earthquakes showed a poor performance of masonry structures subjected to seismic
action. Masonry buildings are characterized by lack of resistance, particularly the tensile stress
and shear strength, small deformation capacity and low ductility. Consequently old non-seismic
masonry buildings located in seismic zones, which usually are very rigid, are characterized by
high need for retrofitting.
Damage patterns for masonry buildings recorded during earthquakes can be classified in one
of the following four categories:
Out-of-plane damage or collapse of walls;
In-plane shear or flexural cracking of walls;
Loss of anchorage of walls to floor or roof diaphragms;
Damage or collapse of corners.
Out-of-plane failure modes, e.g. falling down, can be a result of: load capacity exceeded due
to inertial seismic forces, excessive deflection imposed on walls from diaphragm action, lack of
anchorage, poor possibility of transferring deflection and inertial forces to horizontal elements.
In-plane damage can be a result of: diagonal cracking through masonry units due to excessive
principal stress (tensile stress), shear sliding along bed joints, excessive toe compressive stress
causing crushing (sliding shear), or tensile cracking normal to bed joints resulting in rocking
(e.g. bending).
The interaction of in-plane and out-of-plane forces has as consequence the failure of corners.
An innovative intervention technique for retrofitting masonry walls using metallic sheathing
was proposed at the Politehnica University of Timisoara (Dogariu et al., 2007) and evaluated
in terms of strength and ductility performance criteria. The innovative strengthening solution for
masonry walls developed in PROHITECH (FP6 INCO-CT-2004-509119/2004) is presented in
detail in (Dogariu et al., 2007; Dubina et al. 2007). The technique consists in fixing a mild
carbon steel or aluminium plate either on both sides or on one side of the masonry wall, using
prestress steel ties, or using chemical anchors (see Figure 50). The proposed retrofitting
technique, described was proved to be suitable for a Performance Based Seismic Assessment. In
order to apply this concept, acceptance criteria are needed, together with numerical simulation
to sustain and extend the experimental observations. This technique was applied later, in a case
study, to retrofit a historical masonry building.

CHEMICAL ANCHOR
METALL SHEATING
MASONRY WALL
PRESTRESSED TIE
METALL SHEATING
MASONRY WALL

Figure 50. Proposed solution

Such a type of solution can be successfully applied in case of masonry wall structures, but it
is not appropriate in case of masonry vaults and arches.
This retrofitting technique attempts to respond at all the major desiderates previously
mentioned. Combining metal sheeting, which is resistant and ductile, with masonry, by means
of a proper connecting technology, seems to be a suitable solution. The use of dry connection


256
enables easy for removal of metallic elements. Prestressed tie connectors provide a very
significant confinement effect to masonry. Additionally, the solution offers the advantage of
high mechanical properties, e.g. strength and ductility, without changing too much the initial
rigidity.
In order to characterise the behaviour and performance of metal sheathed masonry walls
experimental tests have been performed on 25x1500x1500 mm panels. Such specimens
represent an entire shear wall or a critical pier zone between wall openings, and enable to
validate the applied technique and to establish the performance levels. The summary of the
experimental program carried out in the Laboratory of CEMSIG Research Centre
(http://www.ct.upt.ro) at the Politehnica University of Timisoara is presented in Table 2.

Table 2. Tests on large specimens
Reference masonry wall test REF
Chemical
anchor
SSP-CA
Steel shear panel
Prestressed ties SSP-PT
Chemical
anchor
ASP-
CA
Aluminium shear
panel
Prestressed ties ASP-PT
M
o
n
o
t
o
n
i
c

&

C
y
c
l
i
c

Steel wire mesh SWM

Figure 51 and Figure 52 show, as an example, the results obtained for a specimen sheathed on
one sides of steel plates connected by chemical anchors (more details in Dogariu et al., 2007).
Diagonal failure mode was observed for all specimens, both under monotonic and cyclic loading
and the tri-linear parametrical behaviour curves in terms of load displacement at elastic limit,
maximum and ultimate failure points for metallic, steel and aluminium, are presented in Figures
51 and 52.

Figure 51. Parametric curve and failure mode for masonry panel sheathed with
steel plates on one side, connected thru chemical anchors (monotonic loading)

Due to large in-plane stiffness of masonry walls, the strengthening solution does not avoid
completely damage to masonry. A limited amount of damage to masonry has to be allowed in
order to take benefit from ductility of the metal used for sheathing. It can be observed that,
despite strengthening, the masonry panel cracks at almost the same force and displacement
values as pure masonry reference panel. The mixed masonry-metallic plate work is activated
only after masonry cracking. This can be observed also by the fact the initial stiffness of both
strengthened and reference panels dont change too much. This is an advantage for global
behaviour of retrofitted building. These observations emphasize the system doesnt modify the
life safety performance level. The major advantage of this techniques, experimentally
confirmed, is the important increase, of 3-5 times, of the ultimate displacement accompanied by

Buildings
257
increase of load bearing capacity of the retrofitting masonry panels in comparison with the
reference unreinforced panel.

-575
-475
-375
-275
-175
-75
25
125
225
325
425
525
-6 -4 -2 0 2 4 6 8
SSP-CA-1

Figure 52. Experimental behaviour: parametric curve and failure mode for masonry panel
sheathed with steel plates on one side, connected thru chemical anchors (cyclic loading)

Choice of a specific strengthening technique for an existing building is a multi-criteria
problem, involving structural, technical, cultural and social, economical and sustainability
aspects. The designer has always several solutions at his disposal. Finally it needs to select one
of them which match better with assembly of validation criteria. In fact, the solution will
represent always a rational compromise among different criteria, because a single criterion
based optimisation leads, in general, to an unacceptable choice.
Modern retrofitting strategies insist in use of mixed and reversible technologies. The use of
mixed material based technology enables for optimisation of the performance of retrofitted
structure. The reversibility is very important because it offers the possibility to remove a
solution when more advanced technology (or more financial support) will be available. This
aspect cannot be neglected in case of buildings of cultural or historical value.
Dogariu et al. (2007) proposed a very complex Decisional Matrix for selection and validation
of rehabilitation method. Even if at level of single element the solution have shown god
experimental results (Dogariu et al., 2007; Dubina et al., 2007), to validate this technique, the
application on a real masonry building is required.

Figure 53. Plan view of the first floor

The masonry building subjected to retrofitting is located in Toscana region, Italy, designed
according only to geometrical considerations (as typical at the beginning of the XX century);


258
this building was selected as reference benchmark structure for the performance analyses of the
steel intervention techniques in the frame of STEELRETRO Project (RFSR-CT-2007-00050)
(Nardini, 2008). Figure 53 and Figure 54 present the plan view of the first floor and the faade
and vertical section.
The reference building respect all the main features of traditional masonry building, ground
floor plus two floors, symmetrical in plan and elevation with small and well positioned
openings, with an almost cubical shape of 15m width, long and height. The bearing wall
thickness varies from 350 mm to 650 mm and is made from stone masonry.

Figure 54. Plan view of the north faade and vertical section

The material properties adopted for the structural modelling of the masonry benchmark are
drawn by literature (O.P.C.M. 3431/2005).
In order to apply the metal sheathing retrofitting to the walls, it was assumed the measures
necessary to provide the diaphragm effect of floors and roof, the integrity of the wall junctions
have been already done.
The building faade was reinforced on the entire height of the building as shown in Figure
55, and all the internal transversal shear walls from the ground floor (see Figure 55). Other
possible location of sheathing would be at the corners or the entire ground floor. Beside
structural aspects, selection of intervention solution must consider the costs and time, as well as
the aesthetically reasons. The possibility to keep using the building even partially during
intervention is also very important.
The applied techniques attempt to be minimal one and avoid affecting internal walls to not
disturb the occupancy of the building.

Figure 55. Reinforcing areas

Buildings
259

A numerical experimentation procedure to analyses and evaluate the behaviour of the
masonry structures retrofitted by metallic plates on base of performance criteria was proposed.
In the first step of the procedure, a stable and robust FE Model able to replicate the experimental
observed failure mode and global behaviour was build.
In second step, numerical simulations of wall panel un- and reinforced masonry considering
the real mechanical characteristic was done; the main advantages and benefits of the
strengthening solution and acceptance criteria for the retrofitted elements have been obtained.
In the third step, equivalent materials that replicate the numerical results need to be
determined. This approach allows for performing global analysis on real faade or entire
buildings and asses the damage at a certain seismic demand using a non-linear evaluation
method, based on the acceptance criteria previously established.
In the fourth step, the validation, the most critical areas of the building must be selected to
verify the local behaviour introducing relevant continuity conditions and using the calibrated
model in step one.
The retrofitting solution has showed a god behaviour being able to preserve the initial
capacity simultaneous by allowing for considerable ultimate displacement of approximate 0.7%
drift ratio, which corresponds to collapse prevention level of the building.
At the end, if applies the Decisional Matrix, as Table 3 shows, to scoring the intervention
technique, the result is 8.33. So according to this matrix, even better intervention techniques are
possibly, the one it was applied in present study is good enough.

Table 3. Decisional matrix
Structural aspects L M H Mark
Capability to achieve requested performance objective (after building
evaluation!)
X
Compatibility with the actual structural system (no need of
complementary strengthening or confinement measures)
X
Adaptability to change of actions seismic typology (near field, far
field, T<>Tic, etc)
X
Adaptability to change of building typology X
9
Technical aspects L M H Mark
Reversibility of intervention X
Durability X
Operational X
Functionally and aesthetically compatible and complementary to the
existing building
X
Sustainability X
Technical capability X
Technical support (Codification, Recommendations, Technical rules) X
Availability of material/device X
Quality control X
8
Economical aspects L M H Mark
Costs (Material/Fabrication, Transportation, Erection, Installation,
Maintenance, Preparatory works)
X 8
Legend: L = low, M = medium, H = high; Mark L (5-6), M (7-8), H (9-10)
1.3.5 Shear retrofit of clay brick masonry walls with FRP composites
Another experimental program, carried out at the Politehnica University of Timisoara (Stoian
et al., 2003), is mentioned briefly, showing the beneficial effects of applying FRP overlays. The
experimental program relates on masonry walls where the first walls were tested in as-built
condition up to failure and then retrofitted on one side with FRP overlay (glass and carbon
fibers, different orientations of fibers, different resins). The test set up is shown in Figure 56.
The experimental specimens were 150 cm wide and 150 cm high, built from solid clay bricks
with dimensions of 6.3 24.0 11.5 cm and unit strength of 9.0 10.0 N/mm
2
and cement
based mortar with strength of 13 16 N/mm
2
. At the top and at the bottom of the wall it was
placed a reinforced concrete beam with dimensions 50 x 150 x 25 cm.


260
The first system was applied in the form of 15 cm wide CFRP sheets in vertical direction.
The second system, another type of CFRP, covered the entire surface of the wall, in one of the
cases in vertical direction and in other one in horizontal direction. The third system, covered
with GFRP, was applied in vertical direction only. The systems characteristics used in the tests
are presented in Table 4.

Figure 56. The specimen test set-up

The walls were tested in a special device, composed of a pair of L-shaped solid steel elements
attached to the massive reinforced concrete block, which was part of the wall at the top and the
bottom. The forces have been applied with hydraulic jacks. The vertical force was applied on
the top of the specimen, acting through the reinforced concrete bond beam. The horizontal
(shear) force was applied through a series of steel bolts embedded in the reinforced concrete
block and mounted to the L-shaped steel elements at the top as well as at the bottom.

Table 4. Characteristics of the composite systems used
System
(Tested
specimens name)
Components
Tensile
Strength
[N/mm
2
]
Tensile
Modulus
[N/mm
2
]
Strain at
Failure
[]
SikaWrap 230C Fabric 4100 231000 17
System 1
(RM1)
SikaDur 330 Resin 30 3800 -
SikaWrap Hex - 103C Fabric 3900 231000 15
System 2
(RM3/RM4/RM5)
SikaDur 306 Resin
55 2100 -
SikaWrap Hex - 100G Fabric 2250 72400 37
System 3
(RC3)
SikaDur 306 Resin
55 2100 -

The specimens were tested in as-built condition up to failure and then retrofitted on one side
with FRP composite layers and retested afterwards. The recorded data was the horizontal load,
the horizontal and vertical displacement, the strain in the composite and the specimen failure
modes. In Figures 59 to 65, the results of the tested specimens are presented together with the
load-displacement curves. Increase of strength and especially ductility is evident.

Buildings
261

0
20
40
60
80
100
120
140
160
180
200
0 2 4 6 8 10 12 14 16 18 20
HORIZONTAL DISPLACEMENT [mm]
H
O
R
I
Z
O
N
T
A
L

L
O
A
D

H

[
k
N
]
UM1L
UM1R
RM1L
RM1R

Figure 57. RM1 retrofitted wall test Figure 58. Load-displacement diagram of UM1 and RM1
(vertical applied CFRP) (L left side, R right side)

0
50
100
150
200
250
300
350
400
0 2 4 6 8 10 12 14 16 18 20
H
O
R
I
Z
O
N
T
A
L

L
O
A
D

H

[
k
N
]
UM2L
UM2R
RM3L
RM3R

Figure 59. The tested RM3 wall Figure 60. Load-displacement diagram of UM2 and RM3
(vertical applied CFRP) (L left side, R right side)

0
50
100
150
200
250
300
350
0 4 8 12 16 20 24 28 32 36 40
H
O
R
I
Z
O
N
T
A
L

L
O
A
D

H

[
k
N
]
UM4L
UM4R
RM5L
RM5R

(vertical applied GFRP) (L left side, R right side)

0
50
100
150
200
250
300
0 2 4 6 8 10 12 14 16 18 20
H
O
R
I
Z
O
N
T
A
L

L
O
A
D

[
k
N
]
UM5L
UM5R
RM6L
RM6R

(horizontal applied CFRP) (L left side, R right side)


262
1.3.6 Performance Based Evaluation of a RC Frame strengthened with BRB Steel Braces
The research presented here was focused on the evaluation of a reinforced concrete building
designed for gravity loads, retrofitted with Buckling Restrained Braced systems (BRB), a
building selected as benchmark case study within RSFR-CT-2007-00050 STEEL RETRO
project. Numerical and experimental investigations were carried out at P.U. Timisoara, within
CEMSIG Research Centre (http://cemsig.ct.upt.ro) in order to study and realize this intervention
(Bordea, 2010). Due to high seismic vulnerability of this building, it was decided for a global
retrofitting by means of steel Buckling Restrained Brace (BRB) system, considered to be more
efficient than inverted braces (CBF).
Thus, a Performance Based Evaluation of the RC frame building was applied before and after
retrofitting. The building used as case study is a historical building built during the first half of
20 century, designed to resist mainly gravity loads and wind. Nonlinear static and dynamic
analyses were performed using the relevant EC8 elastic spectrum adapted to the soil
characteristics of the building location. In order to evaluate the seismic reduction factor q for
retrofitted frame an Incremental Dynamic Analysis (IDA) was performed. To validate the
results of IDA a portal frame was isolated from the building frame, and replicate in a full-scale
specimen to be tested with and without BRBs under monotonic and cyclic loadings.
The dimensions of the 3 story building are 23.4 x 18.4 m, in plane, and 11.95 m height
(Figure 65). Common construction materials used in that period were the concrete with
characteristic compressive strength, f
ck
=20N/mm
2
, and re-bars with a characteristic yield
strength, f
sk
=230N/mm
2
. The detailing of the reinforcement is also characteristic for the design
practice of the time, with poor anchorage length of the rebars at the external beam-column joint,
the use of plane rebars (not ribbed), inclined reinforcement for shear force resistance and large
spaced stirrups (15 cm for columns and 25 cm for beams) in potential plastic zones.

Figure 65. RC building model and BRB system distribution RC portal frame location

In the first step, the research was focused on the numerical evaluation of the existing
reinforced concrete building designed for gravity loads only. The existing RC building was
verified for ULS conditions. Details are presented by Dubina et al. (2010).
The preliminary numerical results have shown that the structure is vulnerable and does not
meet the seismic requirements. Therefore, it was decided to retrofit the structure by means of a
BRB system. Additionally, a local confinement of the concrete elements by means of FRP was
envisaged. The results have shown that the initial structure MRF and the initial structure with
local retrofitting MRF+FRP have a limited ductility. The benefit of local confinement is
reduced. When the global retrofitting is accomplished MRF+BRB, the behaviour is much
improved. The stiffness and strength increase. The numerical results proved the possibility of
seismical improvement of existing RC building frames designed for gravity load by retrofitting
with BRB systems, but also the necessity of local retrofitting of RC elements.
Following the results from the nonlinear static and dynamic analysis, a RC portal frame was
isolated from the case study building (Figure 65), on the purpose of experimental validation of
the BRB retrofitting technique. The experimental program comprised 4 RC frames of 3.2m
height and 4.5m span: 2 RC frames without any retrofitting and 2 retrofitted with BRB, tested
under monotonic and cyclic loads respectively. The testing set-up, the loading system and the
specimen installed in testing rig is presented in Figure 66.

Buildings
263

Figure 66. RC portal frame and BRB system (MRF+BRB)

All details regarding the concrete and steel reinforcement of portal frame are described in
Bordea (2010).
Monotonic tests were also conducted on the frame in order to evaluate the yield point. The
results from monotonic tests are also used as reference values when comparing to the cyclic
tests.
After retrofitting, the ductility of the structure was very much improved and the failure was
caused by the failure of the steel brace in tension. The workability of the system with pre-
stressed ties was also tested and showed a very good behaviour (Dubina & Bordea, 2010). The
results recommend the application of this connecting system for such interventions.
Apart from the evaluation of the effectiveness of BRB retrofitting technique, the experimental
program also aimed at evaluating the reduction factor q.
1.4 Restructuring
The most general level of consolidation is the restructuring, which is the partial or total
modification of volume distribution and implies a radical change of the initial static scheme.
This can be done when the destination change is needed, the creation of new functional spaces,
new volumes, or when the imposed capacity level is impossible to be attained without a
substantial resistance skeleton change (Figure 67 and Figure 68).
Restructuring interventions:
o interventions that substitute the internal part of a building with new structures of
different types;
o vertical or horizontal additions of new volumes;
o introduction of new structures or structural elements into the existing structure;
o replacing of heavy elements with other lighter elements.
1.4.1 Historic context - past situation and actual tendencies
Cultural debates on forms and restructuring criteria of buildings are two centuries old and they
started in the same time with the systematic study of materials and their use for constructions.
At the origin of divergences are not only the structural aspects but also and mainly cultural and
architectural issues. First of all, an ambiguity derives from the word restructuring itself Lat.
re-sisto which means to build, to rise, to do, where the prefix re can mean re-
systematisation, recovery, status quo ante, and even revision, transformation, reinterpretation
intent to rise, in the past centuries conception, the necessity of adaptation to new static and
functional demands.
When it comes to the preservation of monumental values, as witnesses of past historic times,
the procedure of restructuring, the creativity and the artistic and architectural gumption must be
eliminated, in order to preserve the initial philosophy and background, excluding any
modification possibility. The intervention must be centred upon restoring and optimising the
resistance structure, as a simple static consolidation.


264
Nowadays, there is an attempt to refunctionalize constructions in such a way that the
intervention should offer to the past a new vitality that increases the intrinsic value. The new
problems are how to obtain and capitalize new spaces without hiding the existing elements. In
this approach the opposite and contrasting forms and colours may be used, as well as different
shapes, in order to put the old values in a new light. The rehabilitation work in this case can be
seen as an architectonic matter.

Figure 67. A new steel structure inserted in the
original lay-out of a RC building, Cant, Como,
Italy (Mazzolani & Formisano, 2005)
Figure 68. Insertion of an additional floor in an
existing steel building, Amstelveen, The
Netherlands (Mazzolani & Formisano, 2005)

After the Second World War, the general principles of rehabilitation have been stated in the
following Charta from: Paris (1957); Venice (1964); Italy (1972). These documents allow for
restoration works an individual character and the use of independent structures. This fact
justifies all the restructuring interventions.
Restructuring interventions typologies
Depending on the intervention level required, the consolidation may be classified in two distinct
categories:
o Conservation;
o Remediation.
Conservation intervention
This type of intervention is generally required when no modification of the destination is needed
or the modification doesnt change the internal spaces or the existing volumes. This applies
when architectonical reasons dont permit a substantial change of the original scheme, or the
presence of art work as paintings, statues obliges to keep the initial shape untouched. This is
why this intervention is named conservation and finds its application in the case of buildings
with highly artistically or monumental values. The purpose of the intervention is to well
preserve and offer a good resistance and functionality to the building that is in an advanced
stage of damage.

Buildings
265
Techniques that can be applied are related to elements consolidation, partial or total
consolidation of structural components with or without upgrading the initial performances, but
by keeping unchanged the static scheme and volumes of the structures. Thus, these techniques
must harmonize with the existing structure without negatively affecting the architectonic
balance. Steel can find a good justification in these cases, being the optimal choice, mainly for
technical reasons, not just economical. Since metallic materials shows high strength, it reduces
the weight and dimensions, so that the dry connection technique and the easy erection
integrate well into the existing assembly. The intervention technique interferes as minimum as
possible with the occupancy of the building.
Remediation intervention
It is needed when a new destination, different from the initial one, is desired. It implies
substantial modifications of the static scheme and the interior volumes. Due to this major
modification of the static scheme, the load paths, and the interior partitioning, the load bearing
structure can suffer important revisions. Sometimes this intervention may be confused with
restoration intervention.
The main features of ameliorating interventions divide them into specific operations
characterized by certain philosophies, methodologies and operations.
These intervention methods are classified into (Mazzolani, 1990):
o Degutting;
o Insertion;
o Addition;
o Lightening.
The first two imply the alteration of the initial structure, the load paths, the internal spaces, by
building new structures, possibly statically independent, inside the existing building (insertion)
or by total or partial substitution and replacement of an existing part of the structure by a lighter
new one (degutting).
Addition modifies the external volumes by increasing them, when new spaces are needed for
new functional demands.
Lightening, in opposition with addition, tries to reduce the volume by eliminating some
elements, assemblies, and parts in order to create new spaces or to reduce some supplementary
loads. Lightening replaces heavy RC, masonry or wood elements by lightweight steel elements.
In opposition with the conservation intervention, in the cases of ameliorating interventions,
the existing and the new elements coexist and emphasise the architectonic value of the existing
building. The new elements mainly have the role of load bearing structure. Thus, the contrast
between old building and new materials and technologies used for consolidation is obvious.
Once again, steel perfectly meets these demands.
1.4.2 Degutting
It consists in the total or partial substitution of the internal structure of a building for the
purpose of changing its destination and structural concept.
This is used when, for architectural and urban reasons, the preservation of the faades is
needed or a much too drastic change of the structure is needed in order to meet new structural
and functional demands.
The objects of degutting usually are buildings situated in the historic town-centres, when
the preservation of the faades is desired for image reasons (as in the case of banks). This is
preferred when the consolidation of the existing structure would be much too difficult or too
expensive. In this case, a new independent structure, with its own foundations, is the support of
the remaining faades, now without any static role, but preserved for aesthetic reasons (Figure
69).
Modern philosophy imposes that the structure be made with modern performing techniques,
new materials, which are, again, good arguments for the use of steel resistant, performing,
adaptable to new technologies and flexible to any structural scheme that needs quick execution
or high manoeuvrability.
Because of the reduced dimensions of the steel elements, the maximisation of space
availability and thus of the functionality is obtained.


266

Figure 69. Example of degutting (Zurich, Switzerland) (Mazzolani & Formisano, 2005)
1.4.3 Insertion
All operations and interventions that imply the adding of structural elements inside the existing
structure fall into this category.
A characteristic of this intervention is the maintaining of the added elements in plain sight, so
that they should have their own artistic expression. These operations must respect the
architectural identity of the existing buildings.
Insertion is usually done when an improvement of the interior functionality is desired, thus
giving birth to new usable spaces, or it can be conceived for the static optimisation of the
existing structure, giving part of the structural load to the inserted structure (Figure 70).
Many structural and architectural solutions can be used for this type of intervention:
o introducing an intermediate floor with the purpose of improving and enlarging the free
surface inside an interior volume;
o introducing an elevator shaft in order to improve the building feasibility and to satisfy
new and more strict safety codes;

Figure 70. Example of insertion (Genoa, Italy) (Mazzolani & Formisano, 2005)

o structures for sustaining the roof or the upper floors, built with own foundations in
order not to bring excessive loads to the existing structure;

Buildings
267
o self bearing skeleton destined for museums or showroom equipment.
The use of steel is mandatory in the case of insertion interventions because of the need to
preserve the character of the existing structure that requires the addition of slender, light, easily
removable and, in some cases, reversible structures.
That is why insertion of non-autonomous structures that unload on the existing structure
requires the use of lightweight materials that do not overload the structure and have a reversible
character, important when the provisional steel structure must be taken out.
1.4.4 Addition
It implies the building of another level on top of the existing ones, or near the existing building,
destined to change the initial global volumes. This modification can be done on the vertical or
on the horizontal direction.
This intervention is needed when new functional expectancies demand to create new spaces,
the function of which is in correlation with the destination of the existing structure.
When new levels are built on top of the building, the addition is called vertical addition. This
is one of the most delicate cases of addition, it depends on the horizontal and vertical
configuration of the new added masses and requires a careful examination of the static
conditions of the existing structure in order to decide whether a consolidation is needed or not.
From this point of view, the evaluation of the degree of safety can be very complex and
difficult, especially in the case of structures built in the past century, the structural skeleton of
which is comprised of masonry bearing walls and for which a highly accurate analysis must be
made in order to find out all the morphological and resistance aspects of the existing structure.
Vertical addition
The addition of a level (Figure 71a) can be done if there is a surplus of resistance of the
structure, or the choice of consolidation is made to give adequate capacity to the structure in
order to withstand the new loads. This intervention is a very delicate one, given both the
extending character of the building, which implies all structural elements, and because it is
difficult to predict the actual behaviour of the structure with the new floors. The problem
becomes even more pressing in seismic areas, where the evaluation of the global behaviour of
the building under dynamic load is significantly changed by the presence of new inertial masses
at the top of the building.

(a) (b)
Figure 71. Example of vertical addition (a. Canada) and horizontal addition (b. La Villette, Paris)

In this case it is advisable that the added structure should not interfere with the existing
structure in order not to compromise its static equilibrium; the support structure for the vertical
addition to be independent of the main structure and to transfer additional loads directly to the
foundations by means of vertical bearing elements positioned inside or outside of the structure.
This problem brings about the choice of the technological plan, in opposition with the choice of
light materials with maximum structural and mechanic performance so as to modify as
favourably as possible the behaviour of the structure. Once again, steel is an appropriate choice
capable of offering a good and rational solution. When the existing structure is capable of taking


268
extra loads, the reduced weight of steel materials helps sparing the resistance reserve of the
structure maximum. Even when the new floors are added on a static independent structure, the
use of steel will prove correct by reducing the inertial masses concentration at the top.
Horizontal addition
Finally, when new volumes are added alongside the existing building (Figure 71b), in this case a
horizontal addition, the intervention assumes in most cases the characteristics of the existing
structure. In this case the problems arise more from the aesthetic point of view than the
structural one, because of this need to attune very different architectural languages. Even in
this case the use of steel was appropriate as many interventions of this type proved.
1.4.5 Weight reduction
For clear reasons, the weight reduction and storey adding are opposite, and it is possible to
foresee total or partial demolitions of one or more storeys of the building. The goal is to reduce
the tension state from the existing structural elements. Actually the term discharge can acquire
a more ample meaning, generally any operation that somehow reduces the self weight of the
building in order to improve its functional aspect. In this category fall the substitution
operations, the roof replacement (Figure 72), and the replacement of coverings and even of all
structural elements type.
For example, a partial emptying regarding locally the structural resisting case of the building
by replacing the structural elements by lighter ones can fall into the weight reduction strategy.
This procedure has a great favourable impact on constructions located in seismic areas, by
reducing the structural masses from the upper storeys and regulating masses in plane and by
height, thus reducing in this way the torsion effects.

(a) (b)
Figure 72. Examples of weight reduction by roof replacement: (a) The Rivoli Museum, Torino, Italia;
(b) Mongiana Factory, Italia (Mazzolani & Formisano, 2005)

This type of interventions, which imply the replacement of structural elements and not their
removal from the building are practical when the old and heavy wooden roof carpentry is
replaced with a new and light structure. Once again, steel proves to be suitable due to the low
weight structures which can replace carpentry, roofs, stairs and envelopes, obtaining not just a
lighter structure but a more efficient structure from the structural point of view.
It is particularly underlined the economical possibility to profit from the weight reduction
which can be done by simply replacing the heavy or non efficient elements, replacing the
carpentry with steel work, thus increasing the bearing capacity of the building: a consequent
weight reduction can eliminate the necessity to consolidate other elements, which do not meet
the requirements for the existing situation (vertical structure), thus saving time and money.
1.4.6 Domains of use and technical details
Restoration applies rationally in one of the following situations:
o when the modification of the structure destination and of the interior structure needs adding
and the introduction of new spaces, volumes and areas;

Buildings
269
o when the requirements imposed by the design code implies the modification of the
resistance structure;
o in the case of highly structures for which the simple consolidation operations are not
enough.
The choice of a method of consolidation must take into account the following aspects:
o if the object of the consolidation is a historical building, it is required to use reversible
technologies;
o modern restoration theories and the preservation of existing buildings by integrating
reversible works and with a clear individuality represent basic criteria for every
intervention;
o steel and the applied technologies in steel works offer the answer to these demands by the
modern character of reversibility and in particular by its ability to harmonize with old
materials, thus building unitary structural systems.
Degutting (The Law Court Palace from Ancona, Italy)
The restoration of the Court Palace in Ancona is an emblematic example of emptying. The
building was emptied and restored, and in the same time the masonry faades were maintained,
preserving the new-Renaissance style of the building. Four towers made of reinforced concrete
of 9 by 9 m, including the stair case, the elevator shaft and the utilities, were placed at the
corners of the area, and were limited by the remaining faades.
The structural role of these towers was to overtake the vertical loads from the roof and from
the suspended floors, as well as to offer the necessary resistance in the event of an earthquake
(Figure 73). The system of four suspended floors is composed of steel beams and concrete slabs
reinforced with corrugated sheet. The floors are placed in four zones of 9x20m between the four
towers. In Figure 74 it can be observed in detail the building roof made of a metal beams system
(Mazzolani, 1992).

Figure 73. The tower and the suspended floors from
the Court in Ancona, Italy (Mazzolani & Formisano,
2005)
Figure 74. Trellis girders from the upper part of
the Court in Ancona, Italy (Mazzolani &
Formisano, 2005)


270
Insertion (Ducal Palace from Genoa, Italy)
During the past centuries the buildings have suffered many changes in destination, this fact
leading to an advanced degradation of the resistance structures. Due to the fires of 1591 and
1977 and to the bombing of 1944, structural restoration was needed. The use of steel was
justified by the need to differentiate the new and the old parts of the building so as to give a
modern aspect to the building (Mazzolani, 1992).
Particularly, insertion work was done as follows:
o new suspended ramp in order to create a connection between Loggia degli Abati and
Torre di Palazzo (Figure 75);
o new amphitheatre made of curved steel beams;
o articulated system from service stair case and the Intermediary mezzanine (Figure 76a, b).
The ramp was suspended from the upper part of the palace roof, which was made of steel
trellis girders. Each of the works is conceived in such a way as to interact with the structure as
little as possible.
Horizontal extension (The economic Science and Marketing Faculty in Torino, Italy)
This work represents a lateral extension example of an existing building, a former retirement
home for the elders (Figure 77a). In order to create the mezzanine and the staircase inside the
existing building, metallic confection was required, with respect to the reversibility demands of
the consolidation works. Moreover, new spaces were created and new volumes for the new
amphitheatre. The new building is characterized by a modulated faade (Figure 77b) made of
panels which keep the same window configuration as the existing building. The principal
problem that arises is the architectural harmonisation.

Figure 75. Suspended ramp from the Ducal
Palace in Genoa, Italy (Mazzolani &
Formisano, 2005)
Figure 76. (a) Amphitheatre; (b) staircase and
mezzanine from the Ducal Palace in Genoa, Italy

Buildings
271

(a) (b)
Figure 77. The Economic Science and Marketing Faculty in Torino, Italy: (a) before the intervention,
(b) after the intervention (Mazzolani & Ivanyi, 2002)
Vertical extension (Country Club in Briatico, Catanzaro, Italy)
The existing building is an ancient sugar factory (XIII century), a relevant example of industrial
archaeology. The restoration has the purpose to change the structure according to its new
destination for social activities and exhibitions.
The building has suffered important damage due to the last earthquake and soil erosion
because of the sea (Figure 78a). Due to the aforementioned conditions, the reconstruction
solution with old methods was excluded. A new layer was created over the existing masonry
walls, layer which transmits to the masonry underneath only the vertical loads, due to a special
supporting system (Figure 78b).

(a) (b)
Figure 78. Briatico building (a) before restoration (b) after restoration (Mazzolani & Formisano, 2005)
Vertical extension (Building in Toronto, Canada)
The original building had 6 storeys and was made of reinforced concrete frames (Figure 79a). It
was designed to be extended by four more storeys, also made of reinforced concrete.

(a) (b)
Figure 79. Building in Toronto, Canada (a) original building (b) building with added storeys (Mazzolani
& Formisano, 2005)


272
After a few years the storey adding began, but steel was used instead of reinforced concrete.
Due to the advantages offered by steel, instead of four storeys there were added eight storeys,
the building becoming a construction with fourteen storeys (Figure 79b).
Weight reduction (Cultural Center in Succivo, Caserta, Italy)
The former Carabinieri Barrack of Succivo was transformed into an antique shop and a Cultural
Centre. The change of the building destination, together with the static improvement demands,
required the creation of additional new spaces with a contemporary lightening of the structural
conditions. In order to fulfil both static and functional requirements, the original roof structure
has been replaced by a new metallic attic floor (Figure 80), composed of a series of Vierendeel
trusses, with four vertical elements, tall enough as to accommodate a new suitable space (Figure
81). Despite a little increase of the overall building volume, due to a slight increase of the height
of the roofing level, a sensible weight reduction has been obtained thanks to the use of light
gauge steelworks.

Figure 80. New roofing of Cultural Centre
Succivo, Italia (Mazzolani & Formisano, 2005)
Figure 81. New spaces created in penthouse
2 METAL BASED CONSOLIDATION TECHNIQUES OF EXISTING BUILDINGS
2.1 Advantages of metallic materials
Because of the wide range of steel types that can be found on the market, hot-rolled or cold-
rolled profiles, steel sheeting, tube sections, I, H sections etc., and of the wide range of
mechanical characteristics that it can possess, steel can be used with a high degree of flexibility
and it is able to solve almost any consolidation issue.
The possibilities are thus numerous and can be extended to a wide range of operations from a
simple intervention of consolidating a single structural element to restoring the whole structural
assembly taking into account the anti-seismic behaviour of the building.
In seismic areas, the problem of static rehabilitation of buildings becomes an increasingly
delicate matter because of the need to offer either new or existing buildings sufficient strength
in the case of seismic motion. In the same time, the problem of fast and efficient rehabilitation
of buildings affected by earthquakes arises in the process of facilitating the social-urban
rebuilding of the affected zones. From this point of view steel products fully satisfy the need to
create in a short amount of time provisional buildings which can be easily removed, thus
making room for the final consolidation of the buildings.
An optimal solution to accommodate these needs is the use of steel structures that have a
modern and reversible character.
Steel has clear advantages given by his structural characteristics, like (Mazzolani & Mandara,
1991):
- clear shape definition;
- prefabrication;
- reversibility;
- high mechanical strength;

Buildings
273
- mechanical isotropy;
- reduced dimension and weight;
- ease of transportation;
- ease of application;
- ease of manoeuvring in tight spaces;
- workability;
- commercial availability;
- prefabricated steel of different shapes, dimensions, mechanical characteristics;
- recyclable.
Details on these characteristics (Mazzolani & Mandara, 1991):
Prefabrication a major part of the labour is done in plants, thus reducing the on-site labour to
an assembly process. This leads to the optimization of the productivity and making on-site
labour in narrow spaces easier. Steel elements can be prefabricated at optimal dimensions, thus
facilitating transport and manoeuvrability. The severe tolerances imposed are easily respected
by prefabrication in plant and the on-site assembly done with bolted connections becomes very
easy.
Reversibility a characteristic of steel available mainly in the case of bolted connection. This
advantage can be exploited in different types of interventions, like: provisional works for
protection, bearings, made from tubular systems connected by mechanical fasteners for
passageways etc.
The reduced weight of elements is of great importance especially when the new structure
interacts with the old structure. The reduced weight in relation with high strength minimize the
supplemental load of the existing walls of the old structure, which is very likely to be
deteriorated or severely compromised by modifications during their lifetime by atmospheric
agents or by earthquakes. Reduced weight also implies fewer problems with transportation and
manoeuvring, the different elements can be set in place on site using light machinery or even
manually.
Reduced structural dimensions help in the preservation of the existing building when the static
function of it was eliminated or integrated into the new structure. That is why in most cases it is
recommended that the new structure be left visible.
Ease of application makes steel an ideal instrument in all cases especially when there is a short
time available for execution of the salvage or of interventions. This quality is most useful in
emergency cases that follow a calamity and require bracing.
Great commercial variety of products offers flexibility in meeting any design requirements or
execution demands. Steel products can be found as hot-rolled profiles (I, T, L, U etc.), cold-
formed profiles, pipes, tubes, profiled sheeting, plates, strips etc. Steel offers strength and
ductility and can be used in any procedure to replace old damaged elements, which are no
longer capable of carrying out their static function, and in any procedure to integrate new
elements in the existing structures and in complete reconstruction procedures.
All the above presented qualities makes steel an optimum material for all types of
consolidation works of different elements and assemblies, especially those in seismic areas.
2.2 Modern metallic materials used in consolidation
Modern metallic materials have not yet found an ample application in the field of civil
engineering and especially in the consolidation domain. The lack of application on a large scale
of these products can be justified on one hand by the lack of clear theoretical reference
regarding their use in the consolidation field and on the other hand by the lack of a code that
validates the use of these materials in structural applications. Despite having a high cost
compared to regular steel materials such as stainless steel, copper, aluminium alloys, titanium
alloys or shape memory alloys, they have a whole range of advantages due to their
characteristics (Mazzolani & Mandara, 1997; 2001). A comparison of their characteristics with
respect to steel is presented in the Table 5.
The advantages of these materials can be summarized as follows:
- high corrosion resistance;
- high weight-strength ratio;


274
- good ductility;
- ease of application and manufacture;
- aesthetics appearance;
- reversibility and availability on the market.

Table 5. Main characteristic of metallic materials (Mazzolani & Mandara, 1997; 2001)
Material

(g/cm
3
)
E
(kN/mm
2
)
f
0,2
(N/mm
2
)
f
t
(N/mm
2
)
c
t
x100
(A
5
)
o x 10
6
(
o
C
-1
)
Mild carbon steel 7.85 210 235365 360690 1028 1215
Low yield steel 7.85 200 90100 300350 4060 1215
Stainless steel 7.8 196 200650 4001000 1040 1719
Aluminium alloys 2.7 6573 20360 50410 230 2425
Titanium alloys 4.5 106 200975 3001100 830 67
Copper 8.28.9 88118 70400 170720 650 18
SMA Ni-Ti 6.5 2875 100560 750960 15.5 6.611

A great advantage of these materials (except for low yield steel) compared to normal steel is
the high corrosion resistance which makes it possible for these materials to be used in
aggressive, high humidity environments leading to reduced maintenance costs. The choice of
whether to use these materials is not only done by taking into account mechanical characteristics
like strength, deformation capacity or ductility, but rather from a more general point of view
regarding their physical and chemical compatibility with existing materials, resistance to
corrosion or the possibility to obtain particular shapes with aesthetic value.
For these materials as opposed to normal steel whose mechanic behaviour is reduced to a
simple elastic-plastic model, their behaviour curve oc is a continuous one without yielding
branch. It is necessary to adopt a more complex nonlinear behaviour which leads to the
conclusion that these materials do not always exhibit the same ductility as soft steel and the
evaluation of their behaviour at ultimate limit state must be done by a strength calculus at a
certain state of deformation.
3 BASE ISOLATION FOR SEISMIC RETROFITTING OF STRUCTURES
Base Isolation has demonstrated to be a very efficient means for protecting buildings against
destructive earthquakes. The application of base isolation to existing monumental buildings has
been made quite extensively all over the world.
For the purpose of a superior performance of a building designed according to traditional
methods, an isolated structure must comply with the following demands (Skinner, Robinson,
McVerry, 1993):
- the improvement of general safety conditions, in order to reduce to the minimum the
structural damage in the event of a major earthquake;
- the reduction of seismic forces considered by the standard in the structural design,
relatively to a medium intensity earthquake;
- the possibility to adapt the constructive systems with the help of low ductility materials,
which are characterised by a small reserve of plastic deformability;
- seismic adaptation of existing buildings.
Even if the first requirement is not fulfilled by traditional strategies, the advantage of isolated
structures can be seen for the structures of maximum importance for which no damage is
allowed. For monumental buildings that are part of the artistic heritage and for buildings with
intrinsic economical value, the seismic base isolation is fully justified. On the other hand, even
in a traditional approach, the base isolation system shows an important advantage, that is to
allow the use of reduced seismic forces for the structural design, as presented in the second
requirement.
The third requirement permits the rational and efficient use of non-ductile but cheap
materials, with great economical benefits.

Buildings
275
The fourth requirement underlines an important and actual problem for all the engineering
community, namely the preservation of historical buildings and monuments. Seismic adaptation
using base isolation methods does not necessarily require a major intervention on the existing
resistance structure with the obvious benefits of not modifying or altering in any way the
architectural value of the edifice. Thus, the seismic base isolation is not only considered a
particular method to save certain monumental edifices, but also a viable economical alternative
to traditional methods that use conventional materials and to successive maintenance operations.
In this regard, we can underline that seismic base isolation is a very good answer for the
consolidation of a wide range of structures for which the application of traditional methods
would be impossible.
The isolation technique, in the case of existing buildings, must ensure, in contrast with
traditional methods, the substantial reduction of seismic loads by reducing seismic energy, thus
opening new perspectives on preserving built heritage.
3.1 Design strategies for isolated systems
As presented before, the basic working concept of these systems is the following (Kelly, 2001):
- modifies the seismic response of the structure;
- it increases the quantity of dissipated energy.
Starting from this, the isolating systems can be classified into the following distinct
categories:
- Base isolated structural systems. The purpose of this type of systems is the reduction of
seismic loading by means of increasing the own vibration period. The most used
devices are elastomeric devices, lead based devices.
- Partial base isolated systems. Into this category fall the systems which insolate only a
certain part of the structure. A typical example of applying this system can be found in
the field of bridges where the support elements are isolated.
Generalised passive control devices: under this name can be found a series of applicative
devices which dissipate energy by hysteretic or viscous means. Oleodynamic devices also fall
into this category.
It is important to identify some possible types of isolation and the isolator or dissipators
position in the structure. After a primary analysis, it can be observed that these devices are
placed in the points corresponding to the maximum values for displacements generated by the
inertial horizontal actions.
Seismic isolation systems can be divided into three big conceptual categories according to
their working principles:
- Period elongation systems (PE) the reduction of the seismic forces is mainly done by
means of horizontal flexible elastic supports.
- Force barrier systems (FB) have a plastic rigid behaviour or even elastic nonlinear with
zero or very low consolidation. The systems are characterised by a well defined step
force which prevents the transmission of forces greater than it to the superstructure.
- Period elongation / Energy dissipation systems (PE/ED) The reduction of seismic
forces is achieved by the increasing of the own period as well as by energy dissipation.
They are made of horizontal deformable supports which act as a nonlinear hysteretic
system with high energy dissipation capacity.
In order to select the right system, a series of factors have to be taken into account:
- The level of the own frequencies of earthquakes which are expected to occur on the
given location. If there are expected components of great amplitudes in the domain of
low frequencies, the base isolation and own period increasing systems have to be
eliminated, the energy dissipating systems being a better option.
- The building type and structural system. Normally, the base isolation applies to rigid
low and medium rise buildings whereas, the energy dissipaters are recommended for
flexible frame structures. Specific structural configurations can require ad hoc
decisions.


276
- The necessity of content protection from high frequencies of vibration or in order to
avoid peoples panic inside an overcrowded building. In these cases, the seismic
isolation is the best suited solution.
In case of consolidation, supplementary difficulties arise because of the real on-site situation
(geometrical limitation, structural system type, actual structural resistance, etc) which reduce the
practical possibilities of the system set up.
General design criteria are set to obtain a correct behaviour of the overall structural system:
- to build structural networks underneath and over the isolation system of a building;
- to install ultimate displacement limiting systems;
- the devices to be easily inspectable and replaceable;
- to ensure the compatibility with vertical service loadings (vertical rigidity);
- to prevent torsion effects;
- to ensure the compatibility between structural and non-structural connections with
design displacements;
By contrast, the demands of designing a new building and the demands of seismic
rehabilitation of buildings are the following:
- choice of desired performance level;
- defining the expected behaviour of the building in the event of a design code seism in
terms of damage limitations;
- seismic risk: determination of seismic motion and other location related risks;
The construction characteristics:
- basic characteristics determination of the building and resistance capacity of the
existing building at an earthquake;
- rehabilitation methods: the choice of simplified or systematic method;
- rehabilitation strategies: the choice of the basic strategy, e.g. supplementary elements
able to undertake lateral loads;
- general and design analysis: the specification of force type actions and deformations for
which the given components of a structure and the setting of the minimum
interconnection criterion for structural elements must be evaluated;
- analysis and design procedures: for systematic rehabilitation approximations, the
selection between linear static, linear dynamic, nonlinear static and nonlinear dynamic.
3.2 Reduction devices
These devices have the purpose to reduce the negative effects introduced into the structure by
the action of the seismic force. The fundamental principle provides that this type of isolator
devices permit the ground motion without transmitting it to the structure. In the real situation,
there is no need of contact between the building and the supporting base. The functioning
principle is illustrated in the Figure 82. Also, the manners in which a rigid and a flexible
structure respond to seismic action are presented in Figure 83.

GROUND MOVES
STRUCTURE STAYS STILL

Figure 82. Principles of based isolation systems (Kelly, 2001)

Buildings
277

Ground Acceleration
Ground
Displacement
Zero Acceleration
No
Displacement
RIGID STRUCTURE FLEXIBLE STRUCTURE
R
I
G
I
D
D
i
s
p
l
a
c
e
m
e
n
t
A
c
c
e
l
e
r
a
t
i
o
n
D
i
s
p
l
a
c
e
m
e
n
t
A
c
c
e
l
e
r
a
t
i
o
n
F
L
E
X
I
B
L
E

Figure 83. The comparison between the behaviour of a rigid and a flexible structure (Kelly, 2001)

Over the time, many types of systems and isolating devices have been proposed. Many of
them were put into practice, others were just proposals and others were more in the state of
concept and impossible to achieve in practice. Further on, the principal types of systems
available on the market will be presented.
3.2.1 Sliding Systems sliding or friction isolators
This type of systems has a simple setup based on a clear and easy-to-use concept. Besides this,
the working principle is backed up by a facile theoretical implementation. One layer with a
certain friction coefficient will limit the acceleration and the forces transmitted into the system
(building) at a prescribed value, equal to the ratio of the friction coefficient and the weight of
the building.
A pure system does not have the displacements locked and they do not prevent the comeback
of the structure to the state before the seismic action. This flaw can be removed by the
combination with other devices which are able to exert comeback forces or by using spherical
sliding surfaces (see Figure 84).

Figure 84. Scheme of sliding devices (Kelly, 2001) Figure 85. Scheme of friction damper
3.2.2 Elastomeric supports (rubber)
These devices are made of multiple layers from thin natural or artificial rubber tied together by
metal plates (see Figure 86). The metal plates prevent the swelling and excessive deformation of
the device under vertical loads which can lead to stability loss for these elements. Under vertical
loads, they show reduced deformations, while, in the case of horizontal loading, they are very
flexible. Plane elastomeric supports assure flexibility but not enough damping and they will
deflect under service loads. In order to remove this flaw, a lead core is inserted in these devices
(see Figure 86).


278

Figure 86. Scheme of elastomeric supports (TI 809-04: 1998)
3.2.3 Springs
They are specific devices based on metallic springs (see Figure 87). This family of devices is
not widely used and they are generally used for the isolation of mechanical machines. The main
disadvantage of this type of isolators is the increased flexibility, horizontal and vertical. The
vertical flexibility leads to unfavourable dynamic responses. The decreased damping and
excessive displacement under service loads make the independent use of this kind of insulators
improper.

Figure 87. Steel damper (Obayashi, 2002)
3.2.4 Roller and spherical supports - isolators though rolling
They include cylinder rollers as well as spherical devices. As springs, they are mostly used in
mechanical engineering applications. Depending on the support material, they can offer a good
resistance to displacements and sufficient capacity for service loadings.
3.2.5 Flexible storeys and short pendulums
The flexibility is ensured by short pendulum columns (pinned at both ends) which allow
displacements or transform the storey in a soft storey. This system offers flexibility, but does
not ensure damping and resistance under service loads, and that is why they are used together
with other systems.
Dissipating devices
Some of the devices presented before offer flexibility but does not offer sufficient damping
and therefore supplementary devices are introduced into the system:
o Viscous dissipaters (oil dissipaters) these devices offer a good damping but they do not
have any resistance to service loads. They do not have elastic rigidity and therefore they
introduce into system only a small amount of energy;
o Devices based on steel yielding set in such a way as to yield to various types of efforts, as
bending, torsion, etc. They offer rigidity as well as a good damping and dissipation (see
Figure 88). Devices based on lead yielding act to shear and offer rigidity and damping.

Buildings
279

(a)
(b)
(c)
(d)
Figure 88. Steel dissipating devices (a) torsional beam device; (b), (c) flexural beam device;
(d) flexural plate device (TI 809-04: 1998)

Devices based on extruded lead (see Figure 89), in which lead is forced to pass through an
orifice, they add rigidity and damping to the system.
All these devices, except for the viscous dissipaters, are displacement dependent and
therefore they offer maximum force for maximum displacement. The viscous dissipaters are
velocity dependent, being able to offer a maximum force for zero displacement. Therefore they
can be more appropriate than other devices in the case of application on rigid buildings (e.g.
masonry buildings). A more accurate classification can be found in the Table 6.

(a) (b)
Figure 89. Lead damper (a) and lead extrusion (Obayashi, 2002), (b) devices (TI 809-04: 1998)


280
Table 6. Advanced devices for structural control
Natural Rubber Bearings
Lead Rubber Bearings Laminated Rubber Bearings
High-Damping Rubber Bearings
Elastic Slide Bearings
Isolators
Slide Bearings
Rigid Slide Bearings
Bar Type
Loop Type
Portal Type
Plate Type
Steel Dampers
Ring Type
Lead Dampers
Hysteretic-type Dampers
Friction Dampers
Oil Dampers
Dampers
Velocity-type Dampers
Viscous Dampers
REFERENCES
Bordea, S.: Dual frame systems of buckling restrained braces PhD Thesis, The Politehnica University
Timisoara, Ed. Politehnica Timisoara, Romania, 2010.
Dogariu A.: Study on seismic retrofitting techniques based on metallic materials of RC and/or masonry
buildings PhD Thesis, The Politehnica University Timisoara, 2009, Ed. Politehnica Timisoara,
Romania, 2009.
Dogariu A., Stratan A., Dubina D., Gyorgy-Nagy T., Daescu C., Stoian V.: Strengthening of masonry
walls by innovative metal based techniques COST 26 Urban Habitat Construction Under
Catastrophic Events Proceedings of Workshop in Prague, 30-31 March 2007, pp.201-210.
Dogariu A., Dubina D. & Campitiello F., De Matteis G.: FEM Modelling Masonry Shear Walls
strenghtened with metal sheating - datasheet no. 2-21 Urban Habitat Construction under Catastrophic
Events - COST 26 - Editors: Mazzolani et al., pp. 235-240, 2008.
Dubina, D., Stratan, A., Bordea, S.: Seismic Retrofit of RC frames with hysteretic bracing. Steel and
Composite Structures, Ed. Y.C. Wang and C.K. Choi, Taylor and Francis, London, UK, 2007, pp. 833
840.
Dubina D., Dogariu A., Stratan A., Stoian V., Nagy-Gyorgy T., Dan D., Daescu C.: Masonry wall
strengthening with innovative metal based techniques. Proceedings of the International Conference on
Steel and Composite Structures ICSCS07, Manchester, UK, 2007, pp. 1071-1077.
Dubina, D., Bordea, S., Stratan, A.: Performance Based Evaluation of a RC Frame strengthened with
BRB Steel Braces. Proceedings of the International Conference PROHITECH, Roma, Italia, pp. 1009-
1014, 2009.
EN 1998-1, 2004, Eurocode 8: Design of structures for earthquake resistance. Part 1: General rules,
seismic actions and rules for buildings CEN - European Committee for Standardization
FEMA 356, 2000, Prestandard and commentary for the seismic rehabilitation of buildings, Federal
Emergency Management Agency, Washington (DC).
Fulop L., Sippola M.: Constructive and performance analysis of the retrofit systems for vertical masonry
elements VTT-R-11115-07 SteelRetro Task 2.2 report, 2008.
Kelly T.E.: Base isolation of structure Design guidelines. Holmes Consulting Group Ltd., 2001.
Marimpietri F.M., Mezzi M., Parducci A.: Constraints and construction procedures to insert a new
building in a historical centre, Proceedings of the 8th Int. Conf. on Modern Building Materials,
Structures and Techniques, Vilnius, Lithuania, 19-21 May 2004.
Mazzolani F.M.: Refurbishment. Arbed, Luxemburg, 1990.
Mazzolani F.M., Mandara A.: Lacciaio nel consolidamento. Assa Editore, Milano, Italy, 1991.
Mazzolani F.M.: The use of steel in refurbishment. Proceedings of the 1
st
World Conference on
Constructional Steel Design, Acapulco, 1992.
Mazzolani F.M., Mandara A.: Structural preservation of the Centro Storico of Naples. IABSE,
Structural Engineering International, January, 1993.
Mazzolani F.M.: Strengthening of Structures. Lecture 16.1, ESDEP, 1994.
Mazzolani F.M., Mandara A.: Materiali metallici innovativi nel recupero: Attualita e prospettive future di
impiego. Proceedings of the 5
th
National Congress ASS.I.R.C.CO., Orvieto, 1997.

Buildings
281
Mazzolani F.M., Mandara A.: Restoration of monumental constructions by innovative metals.
Proceedings of 3
rd
Congreso Internacional Ciencia y Tecnologia Aplicada a la Proteccion del
Patrimonio Cultural en la Cuence Mediterranea, Alcala De Henares, Spain, 2001.
Mazzolani F.M., Ivanyi, M. (editors): Refurbishment of buildings and bridges (CISM Course). Springer -
Verlag Wien - New York, 2002.
Mazzolani F.M., Calderoni B., Della Corte G., De Matteis G., Faggiano B., Panico S., Landolfo R.: Full-
scale testing of different seismic upgrading metal techniques on an existing RC building. Proceedings
of the Fourth International Conference STESSA 03, Naples, Italy, 2003.
Mazzolani F.M.: General Principles and Design Criteria for Consolidation and Rehabilitation Chapter
1, FP6-2002-INCO-MPC-101 Prohitech WP1, Doc. 01.01.01.01, Deliverables I (www.prohitech.com)
2005.
Mazzolani F.M., Formisano A.: Consolidation levels Chapter 2. FP6-2002-INCO-MPC-101 Prohitech,
Doc. No. 01.01.02.01, Deliverables I (www.prohitech.com) 2005.
Mazzolani, F.M.: Seismic upgrading of RC buildings by advanced techniques. The ILVA-IDEM
Research Project, Polimerica International Scientifical Publisher (www.polimerica.com), 2006.
Nardini L., Braconi A., Osta A. & Salvatore W.: RFCS STEEL solutions for seismic RETROfit and
upgrade of existing constructions Definition of the masonry benchmark building for the execution of
comparative performance analyses between steel intervention techniques. (intranet report RFSR-CT-
2007-00050), 2008.
Obayashi Corporation: Friction Damper with Coned disc springs, 2001 & Base Isolation Technology,
2002.
O.P.C.M. 3431/2005 Technical Italian Standards for Design, Seismic Assessment and Retrofitting of
Buildings.
Skinner R.I., Robinson W.H., McVerry G.H.: An introduction to seismic isolation. John Wiley & Sons,
1993.
Stoian V., Nagy-Gyrgy T., Dan D., Gergely J.: Retrofitting the Shear Capacity of the Masonry Walls
Using CFRP Composite Overlays, The SE 40EE International Conference in Earthquake Engineering,
Skopje, Macedonia, 2003.
US Army Corps of Engineers: TI 809-04 Technical Instructions: Seimic Design for Buildings, 1998.
Workpackage 6: Set-up of advanced reversible mixed technologies for seismic protection Report. FP6-
2002-INCO-MPC-101 Prohitech (www.prohitech.com).
***, PROHITECH- Earthquake Protection of Historical Buildings by Reversible Mixed Technologies,
(www.prohitech.com), Sixth Framework Programme FP6-2002-INCO-MPC-1 2003.
***, STEELRETRO: Steel Solutions for Seismic Retrofit and Upgrade of Existing Constructions.
Research Programme of the Research Fund for Coal and Steel, Project RFSR-PR-06054, 2006.
***, COST 26 Urban Habitat Construction under Catastrophic Events (http://www.civ.uth.gr/cost-
c26/), European Cooperation in Science and Technology.


282
Eugen Brhwiler
EPFL, Institute of Structural Engineering, Lausanne, Switzerland
Thomas Vogel
ETH Zrich, Institute of Structural Engineering, Zrich, Switzerland
Thomas Lang
Swiss Federal Office of Transportation, Bern, Switzerland
Paul Lchinger
Dr. Lchinger + Meyer Bauingenieure AG, Zrich, Switzerland
Swiss codes for existing structures Principles and challenges
1 INTRODUCTION
When dealing with existing structures, most structural engineers apply the codes valid for the
design of new structures. This is a problematic approach since the codes for new structures are
in principle not or only analogously applicable to existing structures.
The professional approach to existing structures is based on an inherent methodology that
essentially includes collecting detailed actual information since the structure already exists. The
controlling parameters are determined more precisely, and for example, the structural safety of
an existing structure is proven using so-called updated values for actions and resistance.
In this way, it can often be shown that an existing structure may be subjected to higher
solicitation while meeting the safety requirements. Such an approach is needed to avoid rather
cost-intensive or even unnecessary maintenance interventions (which are often the result of
insufficient know-how and information about the existing structure).
Over the last 20 years, a methodology inherent to existing structures has evolved and already
been successfully applied. However, it has not yet been really adopted in practice by the
majority of structural engineers. This is explained by the fact that there are no codes available
which the engineer can rely on.
For this reason, the Swiss Society of Engineers and Architects (SIA) launched a pioneering
project in 2005 to develop a series of codes for existing structures. In a country with a rather
well-developed infrastructure such as Switzerland, the establishment of this series of codes is a
real need arising from the fact that more than half of all current and future structural engineering
activities are and will be related to existing structures.
2 PROJECT OBJECTIVE
The objective of the project consists of compiling a series of user-friendly codes dealing with all
aspects of existing structures. In particular, the following typical challenges facing the structural
engineer are addressed:
- Higher live loads (such as traffic loads or live loads in buildings) are anticipated for which
an existing structure has not been initially designed. The structural engineer has to prove

Buildings
283
that these can be carried by the existing structure without asking for costly strengthening
interventions.
- In case the structural safety for higher live loads can be verified, the fatigue safety and the
remaining fatigue life (of fatigue vulnerable structures such as bridges) and the
serviceability become predominant issues requiring advanced analysis methods.
- Risk and safety issues of structures subjected to accidental actions due to natural or man-
made hazards (e.g., earthquakes or impacts) need to be addressed. These issues often have
not been considered when the existing structure was designed.
- Durability of structures showing major damage and deterioration needs to be re-established
and improved in an efficient manner.
- Operational and/or constructional interventions need to be optimised since often a broad
range of solutions are possible. In this context, the issue of proportionality of an
intervention needs to be addressed.
Consequently, the codes have to treat all these items to allow for a systematic and rational
engineering approach to existing structures.
3 OVERVIEW OF THE PROJECT
The Code SIA 269 Existing structures - Basis (presented in this paper) is the basic document
which is complemented by a series of codes which treat specific items:
- Code SIA 269/1 Existing structures Actions
- Code SIA 269/2 Existing structures Concrete structures
- Code SIA 269/3 Existing structures Steel structures
- Code SIA 269/4 Existing structures Steel-concrete composite structures
- Code SIA 269/5 Existing structures Timber structures
- Code SIA 269/6 Existing structures Masonry structures
- Code SIA 269/7 Existing structures Geotechnical aspects
- Code SIA 269/8 Existing structures Earthquake.

The Code SIA 269/1 contains updated models for actions and action effects, and SIA 269/2 to 6
give specific indications for updating material and structural parameters and models valid for
the various types of construction, in particular when materials and structural systems from the
past (which are no longer used) are involved. The Code SIA 269/7 covers geotechnical aspects
specific to existing structures, and SIA 269/8 refers to aspects of earthquake engineering of
existing structures. This set-up of codes is analogous to the European codes valid for the design
of new structures. Issues valid for both existing and new structures will remain in the codes for
new construction, i.e., the codes on existing structures are complementary to the codes for the
design of new structures.
The project organisation is headed by the Project Management Team (composed of the
authors of this paper) and for each code, a Working Group consisting of specialists from
administrations (owners), consulting engineering companies, construction companies and
technical universities and research institutions (ETH, EPFL and EMPA) is editing the code. The
project is accompanied by a Steering Committee (composed of the main funding partners) and
the SIA Standard Committee on Structures.
At the time of writing of this paper, all codes are finalised for public consultation and are to
be published in summer 2010. The following sections present the Code SIA 269.
4 SIA 269 CONTENT, TERMINOLOGY AND PRINCIPLES
Figure 1 gives the table of contents of the Code SIA 269 Existing structures - Basis for which
elements were taken over from the International Organization for Standardization [1] and from
several national SIA Recommendations dealing with various aspects of existing structures. After
defining the framework, the basics and activities when dealing with existing structures are
covered. The definitions of the main terms are as follows:


284
- Updating: Specification of the existing information by collection of detailed information on
the existing structure.
- Proportionality of intervention: Comparison between effort and benefit of interventions
with the objective of an efficient use of means.
- Examination (often called assessment): Condition survey, condition evaluation and
intervention recommendation, triggered by a special cause.
- Interventions: Operational or constructional measures to prevent or mitigate risk in order to
ensure the continued existence of a structure, including preservation of its material and
cultural values.

Figure 1: Table of content and set-up of the Code SIA 269.
The first principle of the code states that activities related to existing structures are carried out
while duly respecting individual and societys safety needs as well as economical,
environmental, cultural and societal compatibility, thus following the principles of sustainable
development of the built environment.
Maintenance of an existing structure over its remaining service life has to fulfil the following
objectives:
- satisfying the requirements regarding the utilisation and legal conditions
- guaranteeing the structural safety and serviceability
- preserving the material and cultural values of a structure while accounting for economy and
aesthetics.
5 SIA 269 - BASICS
5.1 Updating
Structural engineering in the domain of existing structures relies on an inherent methodology
since the structure exists already for some time and has its history of performance. It is thus
possible to obtain and gain more or less detailed information on a specific existing structure. In
this way, uncertainties in structural parameters are reduced through updating. (This is a
fundamental difference with respect to the methodology used for the design of new structures
where uncertainties are dealt with by relying on information gained from experience.)
Consequently, variables describing actions and action effects as well as material and
structural behaviour are updated based on information gained from the existing structures.
Updating takes into account the experience gained from the surveying and monitoring of a
structure, the results of condition surveys and the foreseen modifications during the remaining
service life.
The code provides provisions regarding updating of actions and action effects, characteristic
values of material and soil properties, structural models and geometric quantities as well as
resistance and deformation capacity of structural elements.

Buildings
285
5.2 Structural analysis and verifications
The objective of the structural analysis is to determine the behaviour of an existing structure
regarding relevant situations and given conditions during the remaining service life for which
the structure has to fulfil the requirements for structural safety and serviceability. The
influencing parameters for the structural analysis are obtained through the updating process.
The structural safety and serviceability verifications are performed using updated values, so-
called examination values. In general, deterministic verification will be conducted; probabilistic
verifications are in particular appropriate in cases where either very little or a lot of information
on the structure is available as well as in cases of large consequences of structural failure.
Serviceability is verified qualitatively using the results of the condition survey in case live
loads will not be increased.
The notion of degree of compliance n is introduced in the deterministic verification of the
structural safety:
updated d
updated d
E
R
n
,
,
=
(1)
where
updated d
R
,
and
updated d
E
,
are the examination values of resistance and action effect,
respectively. This formulation not only gives the information whether the structural safety is
fulfilled, i.e. 0 . 1 > n , it also indicates by how much the verification is fulfilled (or not). The
latter is necessary for the evaluation of results and in view of the planning of interventions.
The code also defines (smaller) load factors for permanent actions (compared to those used
for the design of new structures) as uncertainties in the determination of permanent actions are
reduced through the updating process. The examination values may also be determined directly
according to the semi-probabilistic approach in case sufficient statistical data is available.
The probabilistic verification of structural safety is conducted according to the following
format:
0
r r s

(2)
P
r r
0
s
(3)
where r is the failure rate of a structure or a structural element;
0
r is the target value of the
failure rate and
P
r
0
the acceptable failure rate to limit the individual risk for persons.
5.3 Risk-based safety
The risk-based safety approach is adopted from the Probabilistic Model Code of the Joint
Committee on Structural Safety [2]. The target value of the failure rate depends on the
consequences of a failure and the efficiency of interventions (Table 1).
Table 1. Target values of the failure rate for structural safety according to SIA 269.
Consequences of structural failure
Efficiency of intervention EF
M

low
< 2
moderate
2 < < 5
high
5 < < 10
small:
M
EF < 0,5 10
-3
/year 310
-4
/year 10
-4
/year
medium: 0,5
M
EF 2,0 10
-4
/year 10
-5
/year 510
-6
/year
large:
M
EF > 2,0 10
-5
/year 510
-6
/year 10
-6
/year

The consequences of structural failure are expressed as the ratio of direct costs
F
C at failure
to the costs
W
C necessary to restore the structure:
W
F
C
C
=
(4)


286
5.4 Proportionality of interventions
The proportionality of interventions, in particular those related to safety, has to be verified by
expressing the efficiency of intervention, i.e., confrontation of effort and benefit, considering
safety requirements, availability of the structure, magnitude of damage and the preservation of
material and cultural values.
The efficiency of intervention
M
EF is evaluated by a comparison of risk reduction
M
R A with
respect to safety costs
M
SC , as expressed by the following ratio:
M
M
M
SC
R
EF
A
=
(5)
A safety related intervention is proportional when 0 . 1 >
M
EF .
6 SIA 269 ACTIVITIES
6.1 Examination
The examination is the central activity in connection with existing structures. It is triggered by a
special cause like a change in the use of a structure (e.g. increase in live loads), new hazard
scenarios, new findings about the structural behaviour, doubts regarding the structural safety or
unexpected condition (after detection of important damage and deterioration). In addition, the
material and cultural value of the structure needs to be evaluated.
The examination is conducted following a stepwise procedure with increasing focus on details.
The general examination comprises the whole structure with the objective to identify aspects
that need to be examined in more detail. One or more detailed examinations follow with the
objective to focus on the identified aspects.
In the chapter on the condition survey, provisions are given regarding collecting information
from an existing structure with respect to hazard scenarios, actions and the conditions. The
objectives of the study of the structure are to identify potential conceptual and constructive
deteriorations and deficiencies and to collect the information for the necessary update of
structural analysis models. The material characteristics should ideally be determined by means
of non-destructive testing methods. Provisions are also given for in-situ load testing of existing
structures.
The condition evaluation is usually based on the quantitative information about the structural
safety as expressed for example by the degree of compliance, and it includes a condition
forecast. In case the quantitative verifications of the structural safety are not conclusive, a so-
called empirical analysis may be performed, i.e., sufficient structural safety can be presumed if
the structure is in a satisfactory condition, it shows no abnormal behaviour, the live load is not
increased and an assessment rates the risk as being acceptable.
The intervention recommendation comprises a wide range of potential interventions from
accepting the existing condition to implementing a heavy construction intervention. The
intervention recommendation concludes the examination; it is clearly a recommendation to the
owner (and not a design of an intervention).
6.2 Interventions
Maintenance interventions are based on the concept of intervention which includes long-term
considerations and which is obtained by optimisation of intervention variants (as derived from
the results of the examination).
The design of an intervention is the implementation of the concept of intervention including
operational and/or constructional interventions to be performed. Operational interventions may
comprise intensified surveying (e.g. monitoring) or restrictions in the use of the structure;
constructional interventions include rehabilitation or modification of the structure (i.e.,
adaptation or transformation to respond to the new requirements of its utilization).
The resulting intervention project needs to be justified by checking it against technical,
economical and operational criteria. In particular, the code prescribes that the efficiency of

Buildings
287
measures to restore and guarantee the durability has to be demonstrated. In addition, the
influence of the intervention on the aesthetics of a structure and on its cultural value needs to be
assessed. The objective of this procedure is to obtain optimised maintenance interventions.
It is important to note that the code implicitly requires a systematic surveying (monitoring) of
all types of structures (i.e., not just bridges but also large span roofs or buildings with complex
structures). Regular preventive maintenance is optional but obviously highly recommended.
7 IMPLEMENTATION OF CODES
After editing of the codes, the second and maybe even more difficult task is to implement the
content and methodology provided by the codes for existing structures in the engineering
practice.
After printing and releasing all SIA Codes on existing structures, introductory courses will be
organized, and background documentation will be established.
Exemplar cases will be elaborated in order to demonstrate the application of certain articles in
the codes. Such cases include, for example, the determination of examination values following
the deterministic or semi-probabilistic approach; the fatigue safety verification and the
estimation of remaining fatigue life; the probabilistic safety verification of a structure exposed
to accidental hazards; the risk-based assessment of the required safety level; the evaluation of
the proportionality of maintenance interventions; etc.
Implementation of these codes will be a challenge since most structural engineers have no or
only a partial education and training in structural engineering related to existing structures.
Overcoming this challenge will make the real success of the whole project.
8 CONCLUSIONS
From a socio-economic viewpoint, novel engineering methods are needed to deal with existing
structures in an efficient way responding to the principles of sustainability. There is thus a great
need to establish corresponding codes for existing structures which the structural engineer can
rely on.
The series of SIA Codes for existing structures is an original and pioneering initiative to deal
with this challenge.
Structural engineering in the domain of existing structures follows an inherent methodology
which is addressed in these SIA Codes.
The SIA Codes for existing structures provide the Swiss structural engineers with the
necessary tools to react professionally and by means of efficient and cost-effective solutions.
The codes will also present an important opportunity for a thorough education and training of
the structural engineering community in the domain of existing structures.
REFERENCES
[1] ISO/CD 13822, 1999, Basis for design of structures Assessment of existing structures, Code of the
International Standard Organisation.
[2] JCSS, 2001, Probabilistic Model Code for design and assessment of structures, Joint Committee on
Structural Safety, Zurich.
ACKNOWLEDMENTS
The paper originally was presented at the 2010 Joint IABSE-fib Conference in Dubrovnik and
was newly edited for COST 25 purposes by M. Mensinger, TU Mnchen, Germany. The
authors acknowledge the contribution of the following persons belonging (in addition to the
authors) to the editing team of the Code SIA 269: M. Faber, A. Herwig, A. Isler, J.-C. Putallaz,
W. Schuler, P. Rietmann, P. Ritz, R. Steiger, P. Stoffel and L. Trausch.


288
D.M. Frangopol
Lehigh University, Bethlehem, PA, USA
N.M. Okasha
Lehigh University, Bethlehem, PA, USA
Bridge Maintenance, Repair and Rehabilitation in a Life-Cycle
Context
1 INTRODUCTION
Maintenance interventions are actions scheduled to prevent or react to structural damage. The
practice of scheduling maintenance interventions has emerged in recent decades as the numbers
of deficient structural systems, including buildings and bridges, are in continuous increase.
Optimization is a process that scholars have been implementing for determining the most
effective and least expensive schedules for maintenance interventions.
Devising a schedule for the maintenance intervention for a structure requires knowledge of:
(1) the performance of the structure over its service life, and (2) the effects of the available
maintenance actions on the structural performance. To know the performance of the structure
over its service life, one must resort to models that predict the time-variations in the different
components that influence the structural performance, such as structural resistance deterioration
and load increase. Structural resistance may deteriorate due to many mechanical or
environmental factors. For instance, a steel component that is exposed to cyclic loading in a
corrosive environment is bound to progressively lose parts of its cross section due to corrosion,
in turn reducing its axial, bending, shear and/or torsional strength, while its fatigue life is
continuously consumed as the cyclic loads are repeated. Increases in traffic demand potentially
result in more frequent occurrences of the live load and increase of their intensity.
Predicting the variations in resistance and load over time is a process met with a good deal of
uncertainty. Accordingly, probabilistic performance measures have been used in the
maintenance optimization process. Among these measures, which have proven to be
computationally efficient in conjunction with maintenance optimization solution techniques,
and considerably flexible in incorporating the effects of maintenance are the lifetime functions.
In fact, the efficiency and convenience of using lifetime functions even in non-automated
maintenance optimization has already been demonstrated (Yang et al. 2006; Okasha and
Frangopol 2010b).
Another element in the optimum maintenance scheduling process that sheds additional
uncertainty is the variation of the monetary value of the maintenance actions over time. On one
hand, the cost of these actions may increase or decrease at the time they are scheduled for
implementation in the future, while on the other hand, the funds currently available and
allocated for the future maintenance interventions may grow with interest. Either way, the exact
determination of the cost of the future maintenance interventions imposes an additional
challenge in the maintenance optimization process. A discount rate has traditionally been used
to account for the growth of money, but the value of the discount rate is a decision that is not
straightforwardly made. Simplistically, and due to the lack of better information, the discounted
rate of the cost of future maintenance interventions is traditionally treated as deterministic.
The mathematical complexity of the optimum maintenance scheduling process requires the

Bridges
289
automation of solving the optimization problem. Genetic algorithms (GAs) are usually used to
obtain solutions to the multi-objective optimization problems (Deb et al. 2002).
This paper discusses a method for the optimization of structural maintenance. Models that
reflect the effects of essential and preventive maintenance on the life-cycle system performance
indicators are presented.
2 ESSENTIAL AND PREVENTIVE MAINTENANCE ACTIONS
Maintenance interventions are scheduled to either delay the occurrence of reaching a
performance indicator to its threshold or improve the state of the performance indicator if its
threshold is reached. Thus, maintenance types may be categorized into two general groups:
preventive maintenance (PM) and essential maintenance (EM) (Kong and Frangopol 2003). PM
may result in stopping the deterioration of the performance or slowing down the rate of
deterioration and it may be applied at regular or irregular time-intervals (Frangopol et al. 2002).
PMs are usually time-based, that is, they are applied at prespecified time instants over the life-
cycle of the structure. In contrast, EMs are performance-based in that they are applied when
some performance indicators reach pre-defined target values (Liu and Frangopol 2005). Figure
1 shows the difference in the timing and the impact of EM and PM on the performance of a
deteriorating structural component.
THRESHOLD P
E
R
F
O
R
M
A
N
C
E
TIME
EM
PM
WITH MAINTENANCE
WITHOUT MAINTENANCE

Figure 1. Effects of preventive and essential maintenance on the life-cycle performance.
3 LIFE-CYCLE PERFORMANCE
Frangopol and Okasha (2009) have shown that the unavailability is the most computationally
effective performance indicator for consideration in maintenance optimization. Unavailability
can be measured based on lifetime functions. A component is unavailable at time t if it is not
functioning at time t. The unavailability An(t) can be calculated by integration of the probability
density function (PDF) of the time to failure. The time to failure is a random variable and is
defined as the time elapsing from the time the component is put into operation until it fails for
the first time (Hoyland and Rausand 1994). The unavailability of a component without
maintenance can be calculated by
}
= < =
t
du u f t T P t An
0
) ( ) ( ) ( (1)
where T is the time to failure and f(t) is its PDF.


290
4 EFFECTS OF MAINTENANCE
Okasha and Frangopol (2010a) provided complete formulations for the effects of applying EM
and PM on the unavailability, separately and combined. For instance, the unavailability of a
component considering the application of PM can be calculated by
> < s
<
=
+
=
[
1 , )) ( 1 ( )) ( 1 ( 1
) (
) (
1
1
1
1
*
i Tpm t Tpm if Tpm Tpm An Tpm t An
Tpm t if t An
t An
i i
i
j
j j i
(2)
where An
*
(t) = the unavailability after maintenance, Tpm
1
, Tpm
2
, , Tpm
q
are the times at
which q PM actions are applied to a deteriorating structure and Tpm
0
=0, and Tpm
n+1
=.
5 COMBINED MAINTENANCE OPTIMIZATION
The proposed approach finds the optimum maintenance intervention schedules by solving a
nested optimization problem as shown in Figure 2. On the main-level of this nested optimization
formulation, the genetic algorithm is used to solve the problem with design variables related to
the times of PM applications. In fact, a number of continuous design variables are considered
where each represents the time of application of a PM application, in addition to an integer
design variable for each PM type that determines the number of applications that are actually
considered out of the ones provided. This ensures that the program has the ability to decide the
optimum number of applications for each type of PM. This approach is capable of providing
irregular time-interval maintenance strategies, as well as regular time-interval ones if they are
optimum. On the sub-level of the nested optimization formulation, the EM is optimized.
Performance thresholds are treated as design variables that are provided in the main-level and
used in the sub-level. Each time the design variables are provided, a sub-level optimization
takes place to find the optimum set of EM applications, given the presence of the PM
applications and satisfying the given thresholds. An EM action is applied each time a
performance threshold is reached. The type of EM applied is the one that provides the lowest
present cost per year increase of service life. The sub-level optimization is performed during the
objective functions evaluation of each main-level optimization design variables. Based on the
results of the sub-level optimization, the worst values of the performance thresholds throughout
the life of the structure and the LCC are returned to the global optimization as values of the
considered objective functions. The global optimization thus is able to find the best set of PM
applications to fulfill the thresholds, provided that EM applications may also be applied and
their set is determined by the sub-level optimization.
The structural performance indicator used (i.e., the unavailability) is introduced in the
optimization process as criterion. The value of the unavailability objective functions is the worst
value it reaches throughout the service life of the structure. The goal of the optimization is to
improve this (worst) value. In other words, one objective is to minimize the maximum (worst)
value reached for the unavailability throughout the service life An
max
. The second objective is to
minimize the LCC. All maintenance costs are calculated considering the discount rate as
t PV
C
C
) 1 ( v +
= (3)
where C
PV
= present value of maintenance cost, C = cost of maintenance action at time of
application, v = discount rate of money (2% in this study), and t = time of application of
maintenance.
Thirteen design variables are considered in this problem. These design variables are: a
continuous design variable for the unavailability threshold An
th
, five continuous design variables
for the time of application variables for the preventive maintenance of the deck Td
i
(i=1,2,..5),
an integer design variable for the number of applications of the preventive maintenance for the
deck N
d
(where N
d
=0,1,2,,5), five continuous design variables for the time of application
variables for the preventive maintenance of the girders Tg
j
(j=1,2,..5); and an integer design

Bridges
291
variable for the number of applications of the preventive maintenance for the girders N
g
(where
N
g
=0,1,2,,5).
Generate random population of
PM times T
i
and number of
applications Q for all PM types
Perform GA operations
Repeat for each GA chromosome
Repeat for each PM type
Ts
i
= [T
1
, T
2
, , T
Q
]
Perform EM optimization
Compute objective functions
for the obtained strategy
GA converged?
Obtain Pareto-optimal solution
no
yes

Figure 2. Flowchart for the combined optimization of maintenance process.
The threshold for the unavailability is desired to range between 10
-1
and 10
-3
. The lower bound
for this threshold is imposed in order to ensure that the optimum solutions obtained have
adequate availability. On the other hand, the upper bound is imposed to eliminate solutions that
are found by initial analysis to require too frequent essential maintenance applications and
require long computational time for solving the optimization problem.
It is desired not to apply maintenance anytime in the first or last two years of the service life
of a structure. Thus, the application time variables are only allowed to vary from the year 2 to
the year 73. In addition, a minimum time span of two years between any two consecutive
maintenance applications of the same type is required.
Accordingly, the formulation of the problem is stated as follows (Okasha and Frangopol 2010a):
- Find: An
th
,Td
i
, N
d
, Tg
j
, N
g

- To achieve the following two
objectives:
Minimize An
max
Minimize LCC

(4a)
(4b)
- Subject to the constraints:
3 1
10 10

s s
th
An
years 73 years 2 s s
i
Td
years 73 years 2 s s
j
Tg
2
1
>
i i
Td Td years
2
1
>
j j
Tg Tg years
N
d
, N
g
={0, 1, 2, , 5}
where i, j = 1,2,..5.

(4c)
(4d)
(4e)
(4f)
(4g)
(4h)


292
10
-3
10
-2
10
-1
0
2
4
6
8
10
12
14
16
18
C
B
A
D
L
I
F
E
-
C
Y
C
L
E

C
O
S
T
,

L
C
C

(
$
1
0
5
)
MAXIMUM UNAVAILABILITY, An
max
Figure 3. Bidimensional Pareto-optimal solution set.
6 APPLICATION
In this application, the proposed maintenance optimization approach is applied to the
superstructure of the Colorado Bridge E-17-AH. A detailed description of this bridge can be
found in Estes (1997). The bridge has three spans of equal length. The reinforced concrete slab
is supported by nine non-composite steel girders (Estes 1997).Weibull distributions are adopted
to model the unavailability of the components of this bridge. The parameters of these
distributions are based on the best fit of the data summarized in (Maunsell 1999). The shape and
scale parameters, k and , respectively, are k = 2.37 and = 0.0077 for the deck, and k = 2.86
and = 0.0106 for the girders. The system failure is assumed to occur by the failure of any three
adjacent girders or the deck. It was found that including the deck in the redundancy calculations
completely undermines the redundancy offered to the system by the girders. Therefore, the deck
was excluded from the system model in the redundancy calculations. However, it was part of
the system model for the availability calculations.
Four essential maintenance options are considered to maintain the system unavailability and
redundancy within their thresholds during at least the 100 years of its service life. These
essential maintenance actions and their associated costs are (Estes and Frangopol 1999):
Replace deck ($225,600); Replace exterior girders ($229,200); Replace exterior girders and
deck ($341,800); and Replace superstructure ($487,100). For preventive maintenance, silane
treatment is considered for maintaining the deck and re-painting is considered for maintaining
the girders. The cost of silane treatment for the entire deck is assumed as $50,000 and the cost
of girder re-painting for all girders is assumed as $100,000.
The resulting two dimensional Pareto-optimal set of the optimization problem is shown in
Figure 3. Clearly, the reduction in the unavailability is associated with a higher LCC. Four
representative solutions (A, B, C, D) selected from the Pareto-optimal set shown in Figure 3 are
compared in Table 1. Details of the maintenance strategies for the selected representative
Pareto-optimum solutions are given in Table 2. The history profiles for the unavailability of
these selected solutions are plot in Figures 4. Figure 5 illustrates the history profiles for the
cumulative present LCC associated with the four selected solutions.

Bridges
293
Table 1. Values of the integer design variables, unavailability thresholds and LCC for the selected
representative Pareto-optimum solutions.
Solution N
d
N
g
An
max

LCC
($10
5
)
A 1 2 1.00010
-1
1.45
B 4 3 1.11510
-2
4.39
C 5 2 3.20610
-3
8.06
D 4 1 1.00010
-3
15.64
Table 2. Details of the maintenance strategies for the selected representative Pareto-optimum solutions.
Solution A Solution B Solution C Solution D
Time
(years)
Activity
Time
(years)
Activity
Time
(years)
Activity
Time
(years)
Activity
39.10 ST 11.53 ST 6.56 ST 2.13 ST
46.69 GR 22.45 ST 11.29 ST 8.90 DR
68.20 DR 23.26 GR 20.60 DR 10.08 ST
73.00 GR 36.40 DR 25.61 GR 11.42 GR
48.16 ST 29.00 ST 17.10 DR
50.99 GR 37.60 DR 24.20 DR
64.10 DR 40.92 GR 29.87 ST
70.44 GR 48.90 DR 34.40 DR
73.00 ST 58.42 ST 40.50 SR
89.40 DR 63.80 SR 47.90 DR
73.00 ST 55.00 DR
81.00 DR 62.00 DR
92.20 DR 68.50 DR
73.10 SR
80.20 DR
87.30 DR
94.30 DR
ST: Silane Treatment; GR: Girder Repainting; DR: Deck Replacement; SR: Superstructure Replacement.

Evidently, to keep the performance above higher levels, more frequent maintenance actions
have to be applied, and thus a higher LCC has to be paid. It is clear that the selected optimum
solutions have irregular time-interval schedules. Hence, the proposed approach is capable of
finding optimum solutions even if they have irregular time-interval schedules.
The frequent applications of maintenance actions associated with solution D and its high
associated LCC may deem this solution impractical. It is necessary however to solve the
optimization problem to realize this finding. Accordingly, decision makers could make a sound
decision with respect to the selection of the maintenance strategy based on the provided set of
optimal solutions. The interactions, trade-offs and compromises among these solutions provide
valuable insight for making these decisions. Additional details about the approach presented and
its applications can be found in Okasha and Frangopol (2010a).


294
0 10 20 30 40 50 60 70 80 90 100
10
-8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
0 10 20 30 40 50 60 70 80 90 100
10
-8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
0 10 20 30 40 50 60 70 80 90 100
10
-8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
0 10 20 30 40 50 60 70 80 90 100
10
-9
10
-8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
WITHOUT REPAIR
WITH REPAIR
SOLUTION A
U
N
A
V
A
I
L
A
B
I
L
I
T
Y
,

A
n
(
t
)
TIME, t (YEARS)
U
N
A
V
A
I
L
A
B
I
L
I
T
Y
,

A
n
(
t
)
TIME, t (YEARS)
SOLUTION B
WITHOUT REPAIR
WITH REPAIR
SOLUTION C
U
N
A
V
A
I
L
A
B
I
L
I
T
Y
,

A
n
(
t
)
TIME, t (YEARS)
WITHOUT REPAIR
WITH REPAIR
WITHOUT REPAIR
WITH REPAIR
SOLUTION D
U
N
A
V
A
I
L
A
B
I
L
I
T
Y
,

A
n
(
t
)
TIME, t (YEARS)
(a) (b)
(c) (d)

Figure 4. Unavailability profiles for selected solutions.
L
I
F
E
-
C
Y
C
L
E

C
O
S
T
,

L
C
C

(
$
1
0
5
)
TIME, t (YEARS)
0 10 20 30 40 50 60 70 80 90 100
0
2
4
6
8
10
12
14
16
SOLUTION A
B
C
D

Figure 5. Cost profiles of for selected solutions.

Bridges
295
7 CONCLUSIONS
This paper presented an approach for an automated multi-objective optimization of maintenance
strategies considering the lifetime unavailability and the life-cycle cost as criteria. The proposed
approach is able to provide optimum maintenance strategies considering a combination of
multiple essential maintenance options and multiple preventive maintenance options applied at
regular or irregular time-intervals. The lifetime unavailability is calculated based on lifetime
functions. Genetic algorithms are used to solve the optimization problem. The optimization
approach is applied to a bridge located in Colorado, USA. It was shown that the optimization
approach was able to provide a set of optimum maintenance strategies among which trade-offs
exist and can be identified. The interactions, trade-offs and compromises among these solutions
provide valuable insight for making these decisions.
ACKNOWLEDGMENTS
The support from (a) the National Science Foundation through grants CMS-0638728 and CMS-
0639428, (b) the Commonwealth of Pennsylvania, Department of Community and Economic
Development, through the Pennsylvania Infrastructure Technology Alliance (PITA), (c) the
U.S. Federal Highway Administration Cooperative Agreement Award DTFH61-07-H-00040,
and (d) the U.S. Office of Naval Research Contract Number N00014-08-1-0188 is gratefully
acknowledged. The opinions and conclusions presented in this paper are those of the authors
and do not necessarily reflect the views of the sponsoring organizations.
REFERENCES
Deb, K., Pratap, A., Agrawal, S. & Meyarivan, T. 2002. A fast and elitist multiobjective genetic
algorithm: NSGA-II. Transaction on Evolutionary Computation 6(2): 18297.
Estes, A.C. 1997. A system reliability approach to the lifetime optimization of inspection and repair of
highway bridges. PhD thesis, Department of Civil, Environmental, and Architectural Engineering,
University of Colorado at Boulder.
Estes, A.C. & Frangopol, D.M. 1999. Repair optimization of highway bridges using system reliability
approach. Journal of Structural Engineering 125(7): 766-775.
Frangopol, D.M. & Okasha, N.M. 2009. Multi-criteria optimization of life-cycle maintenance programs
using advanced modeling and computational tools. Chapter 1 in Trends in Civil and Structural
Computing, B.H.V. Topping, L.F. Costa Neves, and C. Barros, eds., Saxe-Coburg Publications,
Stirlingshire, Scotland, 1-26.
Frangopol, D.M., Miyake, M., Kong, J.S. & Gharaibeh, E.S. 2002. Reliability- and Cost-Oriented
Optimal Bridge Maintenance Planning. Chapter 10 in Recent Advances in Optimal Structural Design,
ASCE , Reston, Virginia, 257-270.
Hoyland, A. & Rausand, M. 1994. System Reliability Theory: Models and Statistical Methods. Wiley-
Interscience publication, John Wiley & Sons, NY, 518p.
Kong, J.S. & Frangopol, D.M. 2003. Life-cycle reliability based maintenance cost optimization of
deteriorating structures with emphasis on bridges. Journal of Structural Engineering, ASCE, 129(6):
818828.
Liu, M. & Frangopol, D.M. 2005. Bridge annual maintenance prioritization under uncertainty by
multiobjective combinatorial optimization. Computer-Aided Civil and Infrastructure Engineering,
20(5): 343353.
Maunsell Ltd. 1999. Serviceable life of highway structures and their components. Final report,
Birmingham (UK): Highways Agency, London.
Okasha, N.M. & Frangopol, D.M. 2010a. Novel Approach for Multi-Criteria Optimization of Life-Cycle
Preventive and Essential Maintenance of Deteriorating Structures. Journal of Structural Engineering,
ASCE, 136(8): 1009-1022.
Okasha, N.M. & Frangopol, D.M. 2010b. Redundancy of Structural Systems with and without
Maintenance: An Approach Based on Lifetime Functions. Reliability Engineering & System Safety,
Elsevier 95(5): 520-533.
Yang, S-I, Frangopol, D.M. & Neves, L.C. 2006. Optimum maintenance strategy for deteriorating
structures based on lifetime functions. Engineering Structures, Elsevier, 28(2): 196-206.


296
J. Andersson
Division of Structural Engineering, Lule University of Technology, Lule, Sweden
A. Geler
Lehrstuhl fr Stahlbau und Leichtmetallbau, RWTH Aachen University, Aachen, Germany
J. Naumes
Verheyen - Ingenieure, Bad Kreuznach, Germany
O. Hechler
ArcelorMittal, Commercial Sections S.A., Esch-sur-Alzette, Luxembourg
Strengthening of civil structures
1 INTRODUCTION
The demand on retrofitting or reconstruction of aged and deteriorated civil concrete structures
has been significantly increased in the recent years. Many structures are, after refined
calculations and/or follow up by monitoring, found to need strengthening to withhold or
improve their performance in durability, load carrying capacity, aesthetics or serviceability.
Constant deterioration of infrastructure elements punctuates the imperative need for effective
rehabilitation techniques with low material and maintenance costs and short installation time
which leads to reduced interruptions on the structure activities (Sustainable Bridges). The cause
of world-wide deterioration of infrastructure elements depends on a variety of factors including
time dependent material degradation, widespread fatigue damage, initial use of poor materials,
flaws in design and in some cases the use of de-icing salts. Many countries must enhance the
requests of upgrading and retrofitting the existing structures beyond the original design limits
due to rapid population increase, higher demands on logistics etc.
The objective of the project consists of compiling a series of user-friendly codes dealing with
all aspects of existing structures. In particular, the following typical challenges facing the
structural engineer are addressed:
- Higher live loads (such as traffic loads on bridges or live loads in buildings) are anticipated
for which an existing structure has not been initially designed. The structural engineer has
to prove that these can be carried by the existing structure without asking for costly
strengthening interventions.
- In case the structural safety for higher live loads can be verified, the fatigue safety and the
remaining fatigue life (of fatigue vulnerable structures such as bridges) and the
serviceability become predominant issues requiring advanced analysis methods.
- Risk and safety issues of structures subjected to accidental actions due to natural or man-
made hazards (e.g., earthquakes or impacts) need to be addressed. These issues have not
always been considered when the existing structure was designed.
- Durability of structures showing major damage and deterioration needs to be re-established
and improved in an efficient manner.
- Operational and/or constructional interventions need to be optimised since often a broad
range of solutions are possible. In this context, the issue of proportionality of an
intervention needs to be addressed.

Bridges
297

Consequently, the codes have to treat all these items to allow for a systematic and rational
engineering approach to existing structures. Actions, resulting from evaluation of the existing
structure, need to be outlined for further planning of strengthening and design of interventions
need to be specified.
When a structure is strengthened, this is usually done in the ultimate limit state (ULS).
However, many of the strengthening methods will also be applicable when measures are needed
in the serviceability limit state (SLS), for example decreased crack sizes for concrete structures
or increased stiffness for structural components. Repair and strengthening methods need
primary to be appropriate for reinstalling the functionality of the member. In addition, they also
need to be cost effective, environmental friendly and keep the traffic disturbance to a minimum.
In the following, various strengthening methods are presented and their suitability is approved
by the application examples given.

2 STRENGTHENING TECHNIQUES
2.1 Concrete bridges
Not all bridges need to be strengthened. More refined calculation models or following up
through monitoring may show that strengthening is obsolete. Nevertheless, if it is found that
strengthening is necessary then it would be advantageous if the owner, designer and contractor
are given guidance.
In general when speaking about repair and strengthening of structures the following
definitions are normally used:
Maintenance: To keep a structure performance at its original level;
Repair: To upgrade a structure performance to its original level;
Upgrading: To increase the performance of a structure.

Performance is related to durability, load carrying capacity, aesthetics and the serviceability
of a structure. Relatively many concrete structures have reserves in compression capacity while
the amount of existing steel reinforcement limits the load carrying capacity. In cases when
compressive strength limits the capacity, it is not possible to give general suggestions for
improvement. In this section possible strengthening methods will be given for beam-girder,
trough, box-girder and arch bridges, see Figures 1 to 5.
The methods presented in this chapter have been identified and analyzed in the frame of the
project (Sustainable Bridges) with regard to the applicable strengthening methods, cost and
experience with existing railways. It is important to stress that costs for strengthening measures
zones are very site specific, and have different relative costs in different countries.

Figure 1. Principle sketch of strengthening zones in
a beam-girder bridge (Sustainable Bridges).
a through bridge (Sustainable Bridges).


298

a box-girder bridge (Sustainable Bridges).
an arch bridge (Sustainable Bridges).

a column (Sustainable Bridges).

For the above shown concrete bridges and concrete structural components letters have been
added to the figures showing possible deficiencies. These letters denote different types of codes
related to typical problems that can result in strengthening measures:

Code: Problem:
A Deficient flexural bearing capacity. Cracks starting at the bottom or top of the
beam, propagating towards neutral layer, may be an indication that the member
needs strengthening in flexure.
Strengthening may be done with several methods: External CFRP plates or
sheets, MBC (Mineral Based Composite), NSMR (Near Surface Mounted
Reinforcement), longitudinal external prestressing or increased reinforced cross-
section.

B Deficient flexural bearing capacity in regions which cannot be directly accessed.
Strengthening with external prestressing is recommended. Internal steel/CFRP
stays may be an alternative.

Bridges
299
C Deficient shear bearing capacity of beam-like structural elements. The problem
may be addressed by calculations or inclined cracks in middle of the beams
height, close to support or where amount of shear reinforcement changes.
Strengthening may be done with several methods. External CFRP plates or
sheets, MBC, NSMR, longitudinal or transversal external prestressing or fibre
reinforced shotcrete. There are also other methods such as new stirrups, internal
steel/CFRP stays and stiching.

D Deficient shear bearing capacity of slab-like structural elements. Problems with
low shear bearing capacity of slabs are likely near supports, i.e. at connections
to other structural members such as beams or columns.
Strengthening with internal or external prestressing is recommended. Internal
steel/CFRP stays or an increased load transfer area may also be considered.

E Deficient compression bearing capacity in a column. In addition to high
compressive forces, columns are also likely to be subjected to bending and shear
forces. Strengthening methods for bending and shear are presented for code A
D.
Strengthening with external CFRP, steel jacket or increased reinforced cross-
section is recommended.

The following strengthening techniques are mentioned in the project:
Adhesively Bonded CFRP Plates to Concrete Structures.
Adhesively Bonded CFRP Sheets to Concrete Structures.
Strengthening of Concrete Structures with Mineral Based Composites.
Near Surface Mounted Reinforcement (NSMR) to Concrete Structures.
External Prestressing of Concrete Structures.

For detailed method descriptions, case studies, advantages/disadvantages, standard/ guidelines
etc. it is referenced to (Sustainable Bridges).
2.2 Metallic bridges
2.2.1 General
Metallic bridges can be made of cast iron, wrought iron, steel, stainless steel or aluminum
(Sustainable Bridges). Further, several different designs such as girder, truss and box girder
bridges are on the market. In addition to this, metallic bridges can be found in almost any design
with different material qualities for different structural members within the same structure.
The most common metallic bridges are made of different compilation of welded, riveted or
bolted beams. In Figure 6, a typical riveted beam and in Figure 7, a typical welded beam profile
is shown. Metallic box girders, shown in Figure 8, are particularly an alternative for large spans.
All three variants commonly have a deck of concrete and, depending on the beam-to-plate
connection, are in general composite bridges. For particularly lightweight structures a steel deck
plate, so called orthotropic bridge decks are an alternative.
The superstructure of a truss bridge commonly consists of two vertical trusses on each side of
the deck, see Figure 9. Rail tracks can be located above or below the truss beams. These trusses
are the primary structural elements of the bridge and they consist of many members, such as top
and bottom elements, diagonal and vertical struts, and end struts. These members form the truss
and are mainly subjected to tensile or compressive forces. Horizontal stability is given by a
horizontal truss at the top or bottom of the bridge. The horizontal truss gives bearing capacity
for horizontal loads from wind and horizontal accelerations of the train. All truss members are
mutually connected by bolts or rivets. The truss walls have the same function in a truss bridge
as a beam in a beam bridge. The top and bottom members of a truss are similar to the top and
bottom flange of a beam, and the end, vertical and diagonal struts are similar to the web of the
beam.


300

Figure 6. Riveted metallic beam element (Sustainable
Bridges).
Figure 7. Welded metallic beam element (Sustainable
Bridges).

Figure 8. Metallic box girder bridge [(Sustainable
Bridges).
Figure 9. Metallic truss bridge (Sustainable
Bridges).
The easiest and most common way to strengthen metallic structures is adding sectional area by
welding additional steel plates. Although, recently, alternative methods have been developed.
These methods have been reviewed and assessed for typical problems (identified by the code
letters F to L, added to the above Figures) in the frame of the project (Sustainable Bridges) with
regards to their applicability, costs and experience for existing railway bridges. It is important to
stress that costs for strengthening measures zones are very site specific, and have different
relative costs in different countries. In the following these problems and related strengthening
methods are presented.

Code: Problem:
F Deficient global bearing capacity or stiffness. Deficient tension capacity in
structural elements and joints.
Strengthening with longitudinal external prestressing is recommended.

G Deficient compression bearing capacity of beam-like elements. Relatively many
structures are slender and have reserve in compression capacity hence global
buckling limits the capacity.
For strengthening, application of an external CFRP plate or external CFRP
Sheet is recommended. In addition, an increase of the cross section or change of
the structure, i.e. buckling length, should be thought of.

Bridges
301
H Deficient bearing capacity of joints, including rivets. Cross-sections are often
changed in joints causing stress concentration, which may lead to fatigue
problems.
External prestressing with longitudinal tendons and external prestressing with a
prestressed CFRP plate are adequate strengthening techniques. Although the
replacement of components, rivets, bolts, plates are common refurbishment
measures.

J Deficient shear bearing or buckling capacity of webs or flanges.
Application of longitudinal stiffeners, transverse stiffeners and external CFRP
are strengthening options.

K In conjunction with welded joints, cross sections are also typically changed
causing stress concentration which may lead to fatigue problems with possible
cracks.
Standard measures are welding the cracks and/or drilling stop holes.
Further external prestressing and external CFRP are possible strengthening
techniques that are rather adequate.

L Deficient flexural load bearing capacity often addressed by calculations.
Insufficient stiffness can also be solved by strengthening.
Strengthening with external prestressing and external CFRP is recommended.

Strengthening with CFRP laminates as an upcoming (but yet rather unknown) technique is
described in the following in more detail. For further detailed method descriptions, case studies,
advantages/disadvantages, standards/guidelines etc. it is referenced to (Sustainable Bridges).

2.2.2 Strengthening by CFRP laminates for metallic bridges
Bonded CFRP laminates become more and more important with regard to the sustainability of
old buildings and bridges and thus, also for steel structures. Therefore a lot of laboratory and
field tests have been performed in the last years to analyse and optimise this modern
reinforcement system.
Beside the research of the mechanical properties of such advanced strengthening systems,
focus should be set on its durability. Concerning the long-term performance of FRP repair- and
strengthening systems, environmental influences play a very important role. Extreme
temperatures, humidity, aggressive exposure, UV-radiation and their combinations can
extremely shorten the lifetime of such components (Feldmann, Naumes, 2007).
Bonding has many advantages compared to conventional jointing methods such as bolting or
welding, e.g.:
- the possibility to join different kinds of materials,
- no heat introduction into the base material,
- no sectional weakening due to drill holes.

However, bonding is a very labour intensive and sensitive technique. The way of
implementation (in particular surface preparation) is a decisive factor for the quality and
durability of the bonding. Like welding, bonding should only be done by special skilled
workers. Bad surface preparation and bonding with contaminated adhesive can lead to
premature debonding of the CFRP sheets and thus adequate surface preparation of the substrate
should be done with great care.
The main working steps of the installation of a CFRP repair and strengthening system can be
outlined in Table 1.


302
Table 1. Main working steps for the installation of a CFRP strip (Feldmann, Naumes, 2007)
Step Description Characterisation
1 Cleaning of the
metal surface
- Mechanical cleaning to remove any dust, paint or un-sound material on the
bonding surface according to ISO 8504.
- Recommended methods are sandblasting (at least Sa 2- according to ISO
8503) or SACO treatment (for further information, see (Sustainable Bridges
2006)).
- Mechanical surface preparation is accomplished by a chemical cleaning of
the bonding surface to remove all dirt and debris using acetone or other
appropriate cleaning agent.
2 Application of
primer
- A thin layer of polar primer has to be applied to optimize adherence between
steel and adhesive and to avoid any corrosion on the steel surface until the
adhesive will be applied.
- The primer is usually based on the applied adhesive and forms part of the
bonding system.
- Recommendation of the adhesive producer for the choice of primer should
be considered.
3 Cleaning of
CFRP
- CFRP surface may be cleaned with acetone or adequates before bonding
- Some producers protect the laminate surface with a special protection layer,
which has to be removed before bonding (in such cases cleaning via acetone
may not be necessary).
4 Preparation of
the adhesive
- During the whole bonding procedure absolute cleanliness is required to
avoid any contamination of the adhesive.
- In case of a two-component adhesive system, new stirring rods and cups
should be used for each mixture.
- Special attention has to be paid on the correct component ratio (for two-
component adhesives) and sufficient stirring to assure the optimal bonding
performance.
5 Application of
the adhesive
- After curing of the primer and after step 3 and 4, the adhesive can be
applied.
- To reach a high homogeneity of the adhesive layer without any air locks, the
adhesive should be applied single sided (preferentially on the CFRP side)
and in a triangular shape (triangular cross-section).
6 Assembling - CFRP or laminate has to be placed during the potlife time of the adhesive.
- Assembling has to be carried out with great care to avoid any contamination
or air-locks.
- The CFRP should be attached carefully in one working step, which means
that an adequate number of workers and fixing equipment has to be
available.
- Any detaching and/or replacing may cause grave air inclusions or
contamination of the adhesive.

General recommendations about an appropriate bonding procedure for CFRP-metal joints
have been given. These are based on test on surface preparation methods (test series of 71
samples) and information available in (Feldmann, Naumes 2007). Overall the following
conclusions can be summarised:
1. A good surface preparation and an accurate bonding procedure are essential for the
durability of a bonded joint.
2. Like welding, bonding should only be done by special skilled workers.
3. To reach a better adherence between steel surface and adhesive the application of new
surface preparation methods, which combine cleaning and coating of the steel surface in one
working step, should be considered.
4. In general an additional coating of the whole reinforcement via durable protection paint is
highly recommended.

Bridges
303
5. In case of absence or destruction of any protective coating, bond-line corrosion leads to an
increasing deterioration of the bonding.
6. A part of the bonding area which is affected by bond-line corrosion is completely
inoperative, which means that a direct correlation between ultimate strength and percentage
of corroded bonding area can be drawn.
7. The bond-line corrosion progress can be decelerated to nearly a third by the use of advanced
surface preparation techniques such as the mentioned SACO method compared to
conventional sandblasting.

It has to be mentioned that these conclusions result from ageing tests which can only point
out tendencies but give no precise correlation between real and accelerated ageing. In particular
if the real environmental conditions are not known exactly.
2.3 Subsoil
Bridges abutments are normally supported on rigid foundations; however the railway
embankment (ballast/sub-ballast/backfill) in the transition zone is typically constructed above
the native subsoil and is normally not supported on foundations. As a result, the transition zone
often experience problems related to stability and settlements or deformations (Sustainable
Bridges). These problems are exacerbated as train traffic becomes heavier and faster.
The subsoil conditions govern to a large extent the behaviour of the transition zone. In order
to fulfil safety and settlement requirements for e.g. future train traffic with higher axle loads and
higher speeds, strengthening of the subsoil might be required. Most currently available methods
for strengthening the subsoil adjacent to existing railway bridges require railway traffic to be
interrupted. These methods usually require that the railway tracks and sleepers are removed, and
the ballast, sub-ballast and part of the embankment fill to be excavated in order to perform the
strengthening works. Such methods are complicated, time consuming and expensive. Ideally,
strengthening works should be performed with minimal impact on the railway traffic and
without, or with only marginal, reduction of train speed and axle loads. A number of methods
have been identified in this project based on the:
strengthening effects possible to achieve,
influence on the existing railway and train traffic,
applicability for European soil and railway conditions, and
availability in Europe.

The methods presented in this chapter have been identified and analyzed in the frame of the
project (Sustainable Bridges) with regards to the applicable soils, anticipated increase in
stability and/or reduction of settlements, cost and experience with existing railways. Again it is
important to emphasize that costs for strengthening measures zones are very site specific, and
have different relative costs in different countries.
Information on subsoil strengthening methods and descriptions of several subsoil
strengthening methods, including installation procedures and advantages/disadvantages are
further given as rough guidance in evaluating potential subsoil strengthening methods in order
to identify the most suitable subsoil strengthening method for a given site.
Location of subsoil and example of foundation are given in Figure 10.

Figure 10. Subsoil and foundation for an arbitrary bridge (Sustainable Bridges).


304

The subsoil conditions in the transition zones at the bridge abutments govern the applicable
strengthening method. Low stability of embankment in transition zones might be a problem, see
Figure 11.

Figure 11. Principle for the analysis of stability of embankment in transition zones (Sustainable Bridges).
Depending on the subsoil conditions, the foundation of bridge abutments and piers can be a
shallow foundation or a deep foundation. At subsoil with high bearing capacity shallow
foundation is normally used. At subsoil with poor bearing capacity deep foundations are used.
The principle zones of improvement at shallow and deep foundation of the bridge abutments are
shown in Figure 12 and Figure 13 respectively.

Figure 12. Principle sketch of zone of improvement at
shallow foundation of bridge abutment (Sustainable
Bridges).
Figure 13. Principle sketch of zone of improvement at
deep foundation of bridge abutment (Sustainable
Bridges).
To avoid the problem of settlements in subsoil layer or low stability of railway embankment
with respect to the subsoil layer, the following methods are available:
Deep Mixing,
Jet Grouting,
Sheet Pile Walls / Stabilising Berms,
Compaction Grouting,
Embankment Piling.
Possible potential strengthening methods for foundation, where the problem of a low bearing
capacity of the foundation of the bridge has been identified: Jet Grouting, Compaction Grouting,
Shaft Grouting and Base Grouting.
For detailed method descriptions, the applicable soils types, advantages/disadvantages,
standard/ guidelines etc. it is referenced to (Sustainable Bridges).

Bridges
305
3 EXAMPLE: INNOVATIVE STRENGTHENING OF ORTHOTROPIC BRIDGE DECKS
3.1 Orthotropic steel bridge decks
Steel bridges with steel decks (the so called orthotropic decks) play an important role within the
road network, in particular for bridges with larger spans. The design of orthotropic bridge decks
is defined in a standard solution since the 1970s, see (Sedlacek et al. 1992). Detailing,
tolerances and partly also production processes are specified in chapter IV of EN 1993
Eurocode 3 Part 2 - Design of steel structures - Design of steel bridges (EN1993 Part 2 2001).
In addition to those design rules for the steel deck, it should be noted that in most countries a
specified asphalt layer of 75-80mm thickness is required, thus forming composite action with
the steel deck plate.
However, the development of
orthotropic plates on their way from
the first application to the present
standard solution has produced some
variants showing particularly
sensitiveness to fatigue, some of
those bridges are actually already
suffering from damages. Moreover a
recent study (Sedlacek et al. 2010b)
even showed, that in spite of
following the proven details in the
above mentioned standards, there is a
potential risk in the future to be
confronted with fatigue damages,
mainly due to the steadily increasing
volume of traffic and magnitudes of
vehicle loads. Therefore the
maintenance and refurbishment of
bridges and their decks are becoming
a major issue.
Figure 14. Conventional design for steel bridges with
orthotropic deck, connection details
3.2 Potential risk of damages for orthotropic steel decks
A
300 300 300

>

1
2
1a
1c 1b
Quertrgersteg
pot
Ri
potentielle
Ristellen

A Detail
1a
1b 1c
ntielle
tellen

The main potential risk of damage has
been identified at the deckplate and its
connections to the longitudinal stiffeners
(Category 1, see Figure 15). This
connection is essentially dependant on the
detail of the welded connections and the
quality of the execution of those welds. In
particular the damages of Category 1b
and 1c occur more frequent the thinner
the asphalt layers are. This clearly
indicates the influence of the composite
action between the deck plate and the
asphalt layer. It was also found that most
damages are in the heavy traffic lane, thus
indicating that this Category 1 will be
particularly sensitive to the increasing
loads in the future, meaning at the same
time, that methods to strengthen the
structure to prevent those Category 1
damages need to be found.
Figure 15. Risk for potential damages connections
stiffener-deckplate, Category 1 (Sedlacek et al.
2010b)


306
3.3 Strengthening techniques to prevent potential damages
If damages have already occurred then repair welding methods need to be specified and the
quality of such welds needs to be ensured. For this purpose guideline is existing (e.g. DVS
M1709 2008). However, to prevent that damage will occur at the same spots again and in order
to reduce the risk of damage where they have not yet been detected, precautionary measures
need to be ensured. In particular for orthotropic plates, it is necessary to reduce the deck plate
curvature under local wheel loads in order to reduce the bending stresses in the weld between
the stiffeners and the deck plate, to ensure stress range levels below the endurance limit.
In order to achieve such reduction of bending stresses, several methods have been taken into
consideration. On the one hand there are research activities going on to improve the composite
action by measures to raise the stiffness of the asphalt layer, like:
Specially designed fiber-reinforced asphalt;
Polymer modified asphalt bitumen;
Asphalt with voids filled with epoxy resin.

On the other hand, techniques have been developed to strengthen the deck plate directly by
some supplementary component. Whereas those measures to raise the stiffness of the asphalt
layer are showing their potential in lab tests, but not to a degree enough yet, some of these other
methods to strengthen the deck plate directly have already been proven their effectiveness not
only in lab tests but also already by prototype applications on road bridges. These methods are:
- Additional reinforcement plates bonded to the deck plate;
- Additional composite layer of high-strength concrete B110 (with steel fibres);
- Additional steel-polymer-sandwich (Sandwich Plate System SPS).

In particular the strengthening method with SPS has been indentified to be promising.
3.4 Strengthening by sandwich overlay using SPS
The Sandwich Plate System (SPS) is a sandwich composed of two steel plates with a core of
solid polymer (polyurethane), developed by i.e. (Intelligent Engineering Holdings Limited),
UK, together with Elastogran GmbH, Germany. Sandwich-action is generated through the bond
between the polymer core and the steel plates (Feldmann et al., 2007), Figure 16.
3-LAYER PLATE STRUCTURE

- face sheet (skin) - here: steel
- interface layer (core)- here: elastomer

Figure 16. Sandwich Plate System (SPS) (Kennedy, 2007)
SPS plates are suitable for strengthening plates as they:
- Have a high bending resistance and bending stiffness in plate-direction (no stiffeners usually
needed).
- Are light because of the relatively low density of the core material ( = 1,1 g/cm3).
- Provide minimum steel surfaces exposed to corrosion.
- Have good fatigue properties due to the lack of welds.
- Have good damping properties (for noise emission) and insulation properties (for
temperature or fire resistance).

Bridges
307
For the use as overlay-technique to strengthen existing steel decks, SPS requires the
preparation and filling of sandwich cavities on site following a specified procedure:
- Sand-grinding of the existing bare steel plate with class SA 2 (surface roughness of 60
m).
- Mounting of perimeter bars on the existing plate (welded or bonded).
- Fixing of new steel deckplate (ground to SA 2 ) onto those perimeter bars to create the
cavities; between the perimeter bars elastomer-spacers provide the constant height inside,
while on the outer side a temporary down-holder-structure keeps the top plate in place to
avoid deformations during filling and polymerization.
- Filling of the cavities with an elastomeric mixture of Polyol and Isocyanat at a filling rate of
2.5 l/sec.

Polymerization to serviceability strength lasts less than one hour and full strength is achieved
after about 24 hours.
3.5 Pilot project for strengthening by sandwich overlay using SPS
An initial pilot application for strengthening an orthotropic road bridge deck of a permanent
bridge in Germany with SPS-Overlay technique was on the Schnwasserpark-bridge in the
highway A57 close to Krefeld (Minten et al., 2007). That building K3 Schnwasserpark is a
steel structure erected in the year 1972 for bridge loading class 60 (DIN 1072) with a span of
65 m to cross a railway track. The bridge consists of two separate superstructures for each
traffic direction. The deck is an orthotropic 14 mm steel deck with trapezoid troughs. Views of
the bridge and the cross-section of both superstructures are given in Figure 17.
This in-field application showed the applicability of the SPS-Overlay technique under practical
conditions and time restrictions for construction sites in a highway.
However, the main purpose of the application of the SPS-Overlay technique was to improve
the load bearing capacity and the durability of the bridge deck. This beneficial effect had to be
proven by calculations and tests. To this end calculations were carried out which showed that
the stress ranges next to the connection-welds between deckplate and stiffener were reduced due
to the SPS-Overlay from down to k , with k being approximately 0.5, implicating that
the life-time was enlarged by approximately (1/k)
5
32.

berbau II
2%
berbau I

Figure 17. Schnwasserpark-bridge along A57, near Krefeld, Germany: views and cross-section
Superstructure I Superstructure II


308
For the verification of the reliability of the strengthening, measurements have been
performed. For further information it is referenced to (Feldmann 2007, Minten 2007). Taking
the investigations in the run-up to these prototype applications on material characteristics
(Lloyds Register, 2006), (Feldmann & Geler, 2006a, 2006b) together with the results of the
research work on those applications as described herewith it has been proven that the SPS-
Overlay method is suitable for strengthening orthotropic bridge decks to make them fit and
lasting for the future.
4 SUMMARY AND CONCLUSION
When there is a need to strengthen a civil structure, due to increased loading, insufficient
bearing capacity, aesthetics or serviceability, guidance is needed in order to find the most
suitable solution for strengthening the structure. In the project Sustainable Bridges, among other
things, different strengthening methods were analysed which resulted in a Guideline for
strengthening of railway bridges. Most of the methods and applications from this guideline can
of course also be used on road bridges, since the problems are similar.
For concrete bridges, strengthening with CFRP, Mineral Based Composites, external
prestressing and NSMR are suggested, depending on which part of the concrete structure that is
in need of strengthening.
For metallic bridges, the easiest and most common way to strengthen metallic structures is
adding sectional area by welding additional steel plates. Although, recently, alternative methods
have been developed, such as external prestressing with tendons or CFRP plates or sheets.
Bonded CFRP laminates become more and more important in the sustainability of old metallic
buildings and bridges. Bonding is the most sensitive part of these strengthening systems, special
focus should be set on its durability. A new method for surface pretreatment, the SACO method,
has been compared to conventional methods. Conclusions from the tests are that the surface
preparation and an accurate bonding is essential for the durability of a bonded joint. The
preparation also has a significant influence on the bond-line corrosion progress, and should only
be done by special skilled workers.
The transition zone between the embankment and the bridge are problem areas, since there is
a difference between the stability and settlements in the embankment compared to the
abutments. The subsoil conditions govern the behaviour, normally deep foundations are used for
subsoil with poor bearing capacity and shallow foundations are used when the subsoil have high
bearing capacity. Possible strengthening methods if the embankment stability is a problem are
deep mixing, sheet pile walls, jet/compaction grouting and piling. If the foundation of the bridge
has a low bearing capacity, different grouting methods should be used.
Orthotropic decks play an important role, especially for steel bridges with larger spans. Some
of the first versions of these decks already suffer from fatigue damages. Especially since the
traffic volume and loads are increasing, the maintenance and refurbishment may become a
major issue. One method to enhance the durability of the deck-plate and of its welded
connection is to reinforce the deckplate by a steel-elastomer-sandwich (Sandwich Plate System
SPS).
There is already a lot of knowledge on where and how to strengthen civil structures, and there
is also guidance available for owners, designers and contractors. Also, there is a lot of on-going
and future research in this area, so the knowledge is always in motion. With this document and
its references, the use of strengthening methods will hopefully increase, and the need of
replacement for many bridges may in many cases be postponed. One small step on the long road
to find and create a more sustainable society.
REFERENCES
DIN 1072, Straen- und Wegbrcken; Lastannahmen. Beuth Verlag, edition:1985-12.
DIN-Fachbericht 103, Stahlbrcken. Beuth Verlag, edition:2003.
Leitfaden zum DIN-Fachbericht 103, Stahlbrcken - edition March 2003. Beuth Verlag, edition:2004.
DVS-Merkblatt M1709, Instandsetzung und Verstrkung orthotroper Fahrbahnplatten. Edition 05/2008.

Bridges
309
EN 1993 Eurocode 3 Part 2 - Design of steel structures - Design of steel bridges. Edition: DIN V
ENV 1993 2, Feb. 2001.
Feldmann, M. & Geler, A. 2006a: Durchfhrung eines Versuchsprogramms zur Eignungsprfung der
SPS-Technologie hinsichtlich Wrmeeinwirkung an der D-Brckenfahrbahnplatte. Lehrstuhl fr
Stahlbau und Leichtmetallbau, RWTH Aachen, November 2006.
Feldmann, M. & Geler, A. 2006b: Durchfhrung eines Versuchsprogramms zur Eignungsprfung der
SPS-Technologie mit geklebter Leiste an Kleinteilversuchen Klebstoffe. Lehrstuhl fr Stahlbau und
Leichtmetallbau, RWTH Aachen, November 2006.
Feldmann, M.; Sedlacek, G.; Geler, A. 2007. A System of Steel-Elastomer Sandwich Plates for
Strengthening Orthotropic Bridge Decks. Mechanics of Composite Materials, journal, Vol. 43, No. 2,
Institute of Polymer Mechanics, University of Latvia 2007, Seite 271-282.
Geler, A.: Wirksamkeit der Sandwich-Technologie bei der Verstrkung orthotroper Fahrbahnplatten.
Dissertation RWTH Aachen (in preparation).
Intelligent Engineering Limited (IEL) 2004. SPS Elastomer Type Approval - Submission Document,
Prepared for Classification Societies and Regulatory Authorities, Submitted to ABS, May 2004
(unpublished).
Kennedy, S. 2007. Das Sandwich-Platten-System (SPS). Stahlbau 76, H. 7, P 455-464.
Lloyds Register 2006. Provisional Rules for the Application of Sandwich Panel Construction to Ship
Structure. London, April 2006.
Minten, J., Sedlacek, G., Paschen, M., Feldmann, M. & Geler, A. 2007. SPS ein neues Verfahren zur
Instandsetzung und Ertchtigung von sthlernen orthotropen Fahrbahnplatten. Stahlbau 76, H. 7, P
438-454.
Sedlacek, G.; Dowling, P.J.; Harding, J.E.; Bjorhovde, R. 1992. Orthotropic Plate Bridge Decks,
Constructional Steel Design. An International Guide, Elsevier Applied Science, London, 1992, Seite
227.
Sedlacek, G., Feldmann, M., Geler, A., Mller, S., Ungermann, D. & Kalameya, J. 2010a: Herstellung
von Stahlfahrbahnen in Sandwichbauweise mit verringertem Schweiaufwand. Bericht FOSTA,
Projekt P 628
Sedlacek, G., Paschen, M., Feldmann, M., Geler, A., Steinauer, B. & Scharnigg, K. 2010b.
Untersuchungen zur nachhaltigen Instandsetzung von orthotropen Fahrbahnplatten von Stahlbrcken
unter Bercksichtigung des Belagsystems. Bericht FE 15.405/2004/CRB der Bundesanstalt fr
Straenwesen, Bergisch Gladbach.
Sustainable Bridges, Assessment for Future Traffic Demands and Longer Lives. Edited by Bien, J.,
Elfgren L. & Olofsson, J., 2007.
Tljsten, B, 2006. FRP Strengthening of Existing Concrete Structures Design Guideline. Lule
University of Technology, Sweden, Fourth edition.
Feldmann, M., Naumes, J. 2007: Durability increase of bonded CFRP-steel joints via appropriate surface
preparation. International Conference "Sustainable Bridges Assessment for Future Traffic Demands
and Longer Lives", Wrocaw, Poland.
Feldmann, M., Naumes, J., Kradjian, A.: Strengthening and Repair of Steel Bridges using FRP
Durability under extreme Environmental Infuences, Lehrstuhl fr Stahlbau und Leichtmetallbau,
RWTH Aachen, Sustainable Bridges Report - D.6.2.4, 2006


310
O. Hechler
Conclusion
This chapter addressed maintenance, repair and rehabilitation techniques and planning for
structures.
As general introduction to this thematic management systems for maintenance, repair and
rehabilitation are presented. As general architecture of the recommended management and
planning system a concept of a holistic design platform is here recommended for the
management and planning of MR&R activities in the future. In particular, the platform should
allow connection of design and analysis tools to a central database in a plug-and-play fashion.
The structure of data is based on BIM and specially designed for life-cycle design and
management. The design and management tools implanted on a common platform and
connected to a central database will serve as an easy-to-use package for practical designers and
managers to work with. The platform with appropriate applications would improve the
information management and the quality of design. The system should be preferably predictive,
life-cycle based and optimizing with multiple attributes. The attributes for optimization would
contain safety, performance, costs, environmental impacts and cultural aspects.
In the second part of the chapter techniques of maintenance, repair and rehabilitation for
buildings have been reviewed. The link between general criteria, that govern structural
consolidation, with levels of intervention have been outlined and pinpointing on the most
appropriate level of intervention has been done. Further the levels of intervention - safeguard,
repair, reinforcing and restructuring - including suitable techniques are addressed and discussed
in detail.
For safeguard it is concluded, that one must not forget about works done as a precaution,
capable to prevent landslides or any other phenomena, not entirely controlled, tend to cause
degradation of the existing structure.
For repair, from the point of view of safety, it is concludd that repairing offers a simple
restoration of structural performance without introducing supplemental elements or
strengthening of the building. Recently, the use of fibre reinforced polymers (FRP) has greatly
increased in the repair of existing buildings. The potential of this material in the future is
pointed out. However, the development of FRP is mainly justified by the absence of the
corrosion phenomena that affect both steel and reinforcement in the case of reinforced concrete
structures. The disadvantages of this material are its fairly high price, lack of ductility and
insufficient research of its behaviour in time. In the matter of ductility, those who promote FRP
support the existence of a global ductility, in the absence of the material ductility.
Reinforcing means, that the consolidation does not produce any significant change in the
structural scheme but brings new structural elements that need to be integrated with existing
elements without changing or significantly altering the mass and rigidity distribution of the
structure. Unlike a simple repairing intervention, the consolidation can be made at different
performance levels that depend on the level of safety that is desired to be achieved.
The accepted technologies for reinforcing can be schematically represented in a matrix. From
this matrix, it could be observed that not all the materials are suitable to be used for the
consolidation of structural types defined by the material used. The use of materials different
from the base material can create the so-called mixed technologies based on the composite
action of different materials, which form new mixed materials. Thus it is noticed the

Conclusion
311
impossibility to combine all the materials, but that the composite action of two different
materials can result in a behaviour superior to individual materials.
Further not all the possibilities presented before fulfil one of fundamental demands in modern
consolidation philosophy, i.e. the reversibility. This requirement becomes very important in the
case of monumental buildings. If the designer decides to use irreversible techniques, aspects
have to be counterbalanced.
The highest general level of consolidation is the restructuring, which is the partial or total
modification of volume distribution and implies a radical change of the initial static scheme.
This can be done when the destination change is needed, the creation of new functional spaces,
new volumes, or when the imposed capacity level is impossible to be attained without a
substantial resistance skeleton change. Because of the wide range of steel types that can be
found on the market, hot-rolled or cold-rolled profiles, steel sheeting, tube sections, I, H
sections etc., and of the wide range of mechanical characteristics that it can possess, steel can be
used with a high degree of flexibility and it is able to solve almost any consolidation issue. The
possibilities are thus numerous and can be extended to a wide range of operations from a simple
intervention of consolidating a single structural element to restoring the whole structural
assembly taking into account the anti-seismic behaviour of the building.
Overall, steel is an optimum material for all types of consolidation works of different
elements and assemblies, especially those in seismic areas.
Finally application examples for various types of structures and reasons are presented to
provide guidance in the choice of the appropriate measures.
The constribution to techniques of maintenance, repair and rehabilitation.for buildings
presents the Swiss codes for existing structures with special focus on its principles and
challenges. From a socio-economic viewpoint, this kind of code is needed to deal with existing
structures in an efficient way responding to the principles of sustainability. Thus the series of
SIA Codes for existing structures is an original and pioneering initiative to deal with this
challenge. Structural engineering in the domain of existing structures follows an inherent
methodology which is addressed in these SIA Codes. The SIA Codes for existing structures
provide the Swiss structural engineers with the necessary tools to react professionally and by
means of efficient and cost-effective solutions. The codes will also present an important
opportunity for a thorough education and training of the structural engineering community in
the domain of existing structures.
In the third part of the chapter techniques of maintenance, repair and rehabilitation for
bridges in the life-cycle context has been discussed. This part presented an approach for an
automated multi-objective optimization of maintenance strategies considering the lifetime
unavailability and the life-cycle cost as criteria. The proposed approach is able to provide
optimum maintenance strategies considering a combination of multiple essential maintenance
options and multiple preventive maintenance options applied at regular or irregular time-
intervals. The lifetime unavailability is calculated based on lifetime functions. Genetic
algorithms are used to solve the optimization problem. The optimization approach is applied to
a bridge located in Colorado, USA. It was shown that the optimization approach was able to
provide a set of optimum maintenance strategies among which trade-offs exist and can be
identified. The interactions, trade-offs and compromises among these solutions provide valuable
insight for making these decisions.
The last contribution provides general guidance when there is a need to strengthen a civil
structure, due to increased loading, insufficient bearing capacity, aesthetics or serviceability in
order to find the most suitable solution for strengthening the structure. There is already a lot of
knowledge on where and how to strengthen civil structures, and there is also guidance available
for owners, designers and contractors. Also, there is a lot of on-going and future research in this
area, so the knowledge is always in motion.
For concrete bridges, strengthening with CFRP, Mineral Based Composites, external
prestressing and NSMR are suggested, depending on which part of the concrete structure that is
in need of strengthening.
Further special focus is put on strengthening by bonded CFRP laminates. For metallic
bridges, although, the easiest and most common way to strengthen metallic structures is adding
sectional area by welding additional steel plates and alternative methods maybe external
prestressing with tendons or CFRP plates or sheets. Hereby bonding is the most sensitive part of


312
these strengthening systems, special focus should be set on its durability. A new method for
surface pretreatment, the SACO method, has been compared to conventional methods.
Conclusions from the tests are that the surface preparation and an accurate bonding is essential
for the durability of a bonded joint. The preparation also has a significant influence on the bond-
line corrosion progress, and should only be done by special skilled workers.
The transition zone between the embankment and the bridge have further been identified as
problem areas, since there is a difference between the stability and settlements in the
embankment compared to the abutments. Possible strengthening methods if the embankment
stability is a problem are deep mixing, sheet pile walls, jet/compaction grouting and piling. If
the foundation of the bridge has a low bearing capacity, different grouting methods should be
used.
Another particular field of investigation, with regard to strengthening efforts on steel bridges,
has been the fatigue issue of orthotropic steel decks. Orthotropic decks play an important role,
especially for steel bridges with larger spans. Some of the first versions of these decks already
suffer from fatigue damages. Especially since the traffic volume and loads are increasing, the
maintenance and refurbishment may become a major issue. One method to enhance the
durability of the deck-plate and of its welded connection is to reinforce the deckplate by a steel-
elastomer-sandwich (Sandwich Plate System SPS).

Chapter 5. Demolition and deconstruction

Chapter 5. Demolition and deconstruction:
Introduction

315
F. Portioli
O. Hechler
Introduction
The total production of construction and demolition waste in the European Union is about 450
million tonnes per year and represents the most part of waste stream. Excluding earth and
excavated road material, the amount of generated construction and demolition waste is
estimated to be roughly 180 million tonnes per year, which represents 480 kg per person.
Roughly 75% of waste is landfilled, despite its major recycling potential. This means that more
than 3 million tonnes per year have to be occupying existing landfills.
Even if the technical and economic feasibility of recycling has been proven, very few
Member States have specific management legislation. However, those countries which have
introduced measures to improve its management have achieved high levels of recycling, thus
demonstrating that construction and demolition waste has a very high recycling potential. The
south European countries recycle very little of their construction and demolition waste. Their
natural resources are of a sufficient quality and quantity to meet the demand for building
materials at a moderate cost. The market for recycled materials is not highly developed in those
countries.
To reduce the impact of buildings on the environment, in recent years different strategies
have been developed that are mainly based on prevention policies and improved construction
and demolition waste management.
In this chapter, different contributions to the main issues related to the sustainable waste
management and other relevant questions such as health and safety are presented with reference
to demolition techniques and design strategies for deconstruction of structures. In particular, the
key principles of different design strategies for re-use and recycling of buildings are presented,
including design for adaptability, for dismantling and design for deconstruction. Effective
actions to prevent the generation of waste are discussed. Moreover, an overview of the
demolition methods is provided in order to reduce the amount of waste and to make it easier to
re-use, including selective demolition, sorting and crushing operations. Different demolition
methods are presented for the main structural typologies and materials. Also the factors relevant
to the choice of the demolition technique and its influence on the eventual recovery of materials
are investigated.
This chapter is organized in three parts. In the introduction, general questions on the selection
of optimized strategies for deconstruction and demolition are dealt with, as well as case studies
on the assessment of the structural materials at high stain-rates in tension for the planning of
explosive demolition and main principles governing design for deconstruction and demolition
and recycling of demolition rubble after deconstruction. The next two parts are concerned with
demolition and deconstruction of buildings and bridges.
In particular, the first paper gives an overview of selection methods for sustainable strategies
in the refurbishment, deconstruction and demolition of existing structures. Following a brief
introduction of the most popular methods currently available in the scientific literature,
significant applications and case studies are discussed in order to provide information on types
of alternatives and criteria that can be considered to select the best solution, with particular
reference to sustainable interventions.

316
In the second paper, problems related to the assessment of the structures dynamically loaded
at high strain rate, such as those subjected to impacts and explosions, are discussed. Also in the
case of the deconstruction, the use of particular demolition techniques underlines the necessity
to know the behavior of the structural materials subjected to high strain rate. The optimization
of the demolition tools requires the knowledge of the structural materials under high strain rate
to respond the requirement dictated to the sustainability. This paper presents the techniques used
to determine the dynamic mechanical characteristics of structural materials. Finally, as example
some results describing the dynamic behavior of structural materials in tension are reported.
The next paper focuses on Design for Deconstruction at the end of life as an integral
approach. It gives an overview of advantages and challenges and provides recommendations. It
also gives examples of successful applications with an overview of possible environmental
savings. Different materials, structural parts, elements and whole buildings are presented. In
that, the Design for Deconstruction concept has been examined in the context of life cycle
approaches to design and construction and the methodology and approach adopted in Design for
Deconstruction has been assessed with reference to the key parameters.
The introduction section is closed with a paper on waste management in reference to
demolition techniques. Demolition methods have to be taken into account in order to reduce the
amount of waste and to make it easier to re-use, including selective demolition, sorting and
crushing operations. In particular different demolition methods are presented for the main
structural typologies and materials. Also the factors relevant to the choice of the demolition
technique and its influence on the eventual recovery of materials are investigated. Further an
overview on the most common fractions of demolition rubble as well as it answers the question,
when on-site recycling of mineral debris is ecologically and economically reasonable and when
off-site recycling or to deposit the demolition rubble is inevitable. Recycling is not only a
question of the total costs of a deconstruction, but also depends on the fractions of the materials,
which do not all permit re-cycling.
The building section includes a paper which is concerned with masses and housing types
involved in the deconstruction process. In particular, the paper uses 25 different data sources of
deconstructed buildings to give an overview of the total masses. The results show how waste
production for de-construction works is influenced by different housing types as well as the year
the building was build. In the end, the paper is discussing the question, if there are possibilities
to avoid an in-creasing rate of land filling in the future.
For bridges, an overview of different demolition and deconstruction techniques is presented
with particular reference to few sustainability issues, such as health and safety and waste
management. The criteria and attributes which might affect the selection of the most suitable
solution strategy for end-of-life management are briefly illustrated. Finally, few case studies of
bridge demolitions and deconstructions are presented with particular emphasis on structural and
technological issues.

General

317
F. Portioli, L. Cascini, R. Landolfo
V. Ungureanu
Politehnica University of Timisoara, Romania
Sustainable refurbishment, deconstruction and demolition:
Applications of MCDM methods
1 INTRODUCTION
A number of techniques and design principles are available in the literature on refurbishment,
deconstruction and demolition of existing structures.
The practices involved in these types of interventions strongly influence sustainability of
buildings, as for example possibility of reusing and recycling of materials in demolition and
deconstruction.
The selection of the type of intervention to be undertaken is dependant on several factors, or
criteria, including the assessment of structural performance, the location of the building, the
type of structure and materials involved and, of course, on the time and money available. In case
of demolition other factors should be taken into account as the space available on site for
segregation and storage, health and safety of operatives undertaking demolition work and
permitted levels of nuisance (Portioli & Ungureanu 2008).
The different methods available for structural retrofitting and for tackling a demolition
present various advantages related to the factors above. The main issue is to identify and to
evaluate the factors relevant to the choice of the intervention, with reference in particular to the
sustainability of considered solution.
On the basis of these factors, one of the main questions for designers is related to the
selection of the optimal solution.
In general, the first step in defining intervention strategy should be to assess whether the
building has to be demolished or retrofitted. Each solution, or alternative, should be evaluated
on the basis of several criteria, as installation, maintenance and demolition cost and damage
risk.
Since different criteria and alternatives are involved, a multi-criteria decision-making
(MCDM) approach is required to select the type of the intervention to be undertaken on existing
buildings.
To show the capability of application of multi-criteria decision making methods to the
refurbishment, demolition and deconstruction of existing buildings, a state-of-the-art report is
presented in this paper, focusing on most common used methods and considered criteria and
alternatives.
2 MULTI-CRITERIA DECISION MAKING METHODS
Multi-Criteria Decision Making (MCDM) methods are techniques aimed at identifying and
choosing alternatives starting from several criteria. The existing literature offer many different
approaches to the optimization problem, that is the problem of selecting the best among

318
different alternatives (or actions) given a set of selection criteria (or constraints). According to
Zimmermann (1996), MCDM is divided into multi-objective decision making (MODM) when
the decision space is continuous, and multi-attribute decision making (MADM) when the
decision space is discrete.
The latter, following the classification by Chen and Hwang (1991), includes: weighted sum
model, usually adopted for single dimensional problems (Fishburn 1967); weighted product
model (Bridgman, 1922; Miller &Starr, 1969), whose structure allows to deal with multi-
dimensional problems; the analytic hierarchy process (Saaty 1980, 1994, 1999), based on
pairwise comparisons, that makes the cost/benefit ratio among qualitative and quantitative
criteria possible; the ELECTRE method (Benayoun & Sussman 1966, Roy 1973), mainly used
for problems characterized by few criteria versus several alternative; the TOPSIS method
(Hwang & Yoon, 1981) that selects alternative which have the shortest distance from the ideal
solution and the farthest distance from the negative-ideal solution in some geometrical sense.
Beside these techniques, the VIKOR method (Triantaphyllou 2000) determines a compromise
ranking list and a set of compromise solutions. So far, one of the crucial question in MCDM is
the selection of the best methods for a given problem.
In this section a brief description of multi-criteria decision making methods (MCDM) which are
most commonly used in retrofitting, deconstruction and demolition of buildings is presented.
A multi-criteria decision problem consists of determining the optimal alternative A
*
among a
set of solutions A
i
which are evaluated with respect to a set of criteria C
j
.
Criteria can be expressed with different units of measure and may be qualitative or
quantitative. As a consequence, the application of MCDM methods involve the normalization of
variables, the quantification of qualitative data and determination of criteria weights. Criteria
can be classified as benefit-type or cost-type, according to the interest of decision maker in
maximizing or minimizing their evaluation.
In general, MCDM problems can be formulated in a matrix format. Let a
ij
be the performance
of each alternative A
i
with respect to different criteria C
j
whose weights are w
j
. Weights are
generally expressed in relative terms, that is
j
j
w
=1. The typical decision matrix is then:

| |
1
2
m x n
m
... ... ... A
A
A
A
1 2 3 4 n
11 12 1n
21

m1 mn
C C C C C
a a a
a

a a
(
(
(
=
(
(

(1)

Any decision-making technique can be divided in three steps: l) Determination of the relevant
criteria and alternatives; 2) Attachment of numerical measures to the relative importance of the
criteria and to the impacts of the alternatives on these criteria; 3) Processing of numerical values
to determine a ranking of each alternative.
In the following, a brief description of most common MCDM methods and most used
approaches in demolition, deconstruction and retrofitting of existing structures is reported. In
particular, details on weighted sum model, weighted product model, analytic hierarchy process
and TOPSIS method are provided. Concerning other approaches as ELECTRE, MAUT, VIKOR
and PROMETHEE methods the reader is referred to existing literature.

The WSM and WPM methods
In the weighted sum model (WSM) the best alternative is selected according to the following
espression:

n
j 1
max for
*
WSM-score ij j
i
A a w i=1,...., m.
=
=

(2)

General

319
In general, this method is suitable for single-dimensional problems while in multi-
dimensional cases difficulties occur. Moreover, involving the maximization or minimization of
different performances, the WSM method is applicable to problems with criteria all of the same
type, that is benefit or cost-type criteria, while it is clear that in some cases their value can
conflict each other.
The weighted product model (WPM) is similar to the WSM. Each alternative is compared
with the others according to the following product:
n
wj
k L Kj Lj
j 1
R(A / A ) (a / a )
=
=
[
(3)
In the maximization case, the alternative A
k
is to be preferred with respect to A
L
if the term R
is greater than or equal to one. The best alternative is the one that maximizes the value of R.
The calculation of the ratios of different performances allows the elimination of units of
measure and for this reason the method id called dimensionless analysis. However, limitations
concerning the types of criteria are similar to WSM method.

The AHP method
The analytic hierarchy process (AHP) was developed by Saaty. The decision matrix is
constructed by using the relative importance of the alternatives with respect to considered
criterion.
Each entry a
ij
in the m x n matrix [A] represents the relative value of alternative Ai in terms
of the criterion Cj. In general, it is assumed that
n
i 1
1
ij
a
=
=
. As a consequence, in the AHP

method the columns of decision matrix are normalized to add up one.
The relationship used to evaluate the best alternative is similar to WSM but it is based on
relative values instead of actual ones and as a consequence in this case the method may be
applied to multi-dimensional problems.
n
j 1
max for
*
WSM-score ij j
i
A a w i=1,...., m.
=
=

(4)
In the AHP method the columns of decision matrix are defined on the basis of a series of
pairwise comparisons of alternatives with respect to a selected criterion. The pairwise
comparison approach allows to quantify qualitative data and to decompose a complex MCDM
problem into a system of hierarchies. The same comparisons can be carried out on considered
criteria as well, to define relative weights. The value of pairwise comparison is expressed by
means of a linguistic judgment, as A is more important than B or A is of the same
importance as B and so on. The quantification of linguistic choices is determined on the basis
of proper scales, as the linear one proposed by Saaty (Table 1).
The results of pairwise comparisons are used to define judgment matrices, also called
Pairwise matrices or P matrices.
Each entry p
ij
of these matrices corresponds to the number that estimates according to
selected scale the relative performance (or weights of importance) of entity E
i
when it is
compared with E
j
. Entities can be represented by alternatives A
i
or criteria C
i
that have to be
compared in terms of selected criterion or decision goal.

| |
n x n
... ... ...
P
1 2 3 4 n
1 11 12 1n
2 21

n m1 mn
E E E E E
E p p p
E p

E p p
(
(
(
=
(
(

(5)

The matrix of pairwise comparisons is reciprocal, being p
ij
=1/p
ij
and p
ii
=1. In case it is
possible to have exact value of p
ij
= w
i
/w
j
, with w
i
actual weigth of importance of entity E
i
, the
matrix [P] is consistent, that is p
ij
=p
ik
p
kj
. It can be shown that for real values of p
ij
the rank of

320
Table 1. Fundamental Scale of Relative Importances (Saaty 2008).
Intensity of
Importance
Definition Explanation
1 Equal Importance Two activities contribute equally to the
objective
2 Weak or slight
3 Moderate importance Experience and judgment slightly favor
one activity over another
4 Moderate plus
5 Strong importance Experience and judgment strongly favor
one activity over another
6 Strong plus
7 Very strong or demonstrated importance An activity is favored very strongly over
another; its dominance demonstrated in
practice
8 Very,very strong
9 Extreme importance The evidence favoring one activity over
another is of the highest possible order
of affirmation
Reciprocals of
above
If activity i has one of the above nonzero
numbers assigned to it when compared
with activity j, then j has the reciprocal
value when compared with i
A logical assumption

[P] is 1 and that =n is the principal eigenvalue. When the values of p
ij
are close to actual ones
but different, which is the most common case in applications, the [P] matrix is not consistent,
the rank is greater than 1 and the maximum eigenvalue
max
is greater than or equal to n.
It can be demonstrated that the normalized principal eigenvector w of P matrix that
corresponds to the maximum eigenvalue
max
represents the vector (a
i1
, a
i2
,., a
in
) with relative
performance, or weights, of compared entities, that is the i-th colum of matrix [A] or the vector
of relative weights of criteria.
In particular, it results:

| |
max
P w w = with
max
n > (6)

,..., )
i1 i2 in
w=(a , a a (7)

A simplified calculation of eigenvector w can be performed by multiplying the entries in each
row of matrix P together and taking the n-th root. In order to have values that add up to 1, a
normalization by the sum of obtained values has to be carried out. A simplified procedure to
obtain
max
is to add the columns of matrix P and the multiplying the resulting vector with w.
To estimate the consistency of pairwise comparison matrix, Saaty proposed the Consistency
Ratio CR, obtained dividing the consistency index by the Random Consistency Index RCI given
in table 2.
CI is defined as follows:

max
CI ( n) /(n 1) = (8)

RCI is an average random consistency index derived from randomly generated reciprocal
matrices.
According to Saaty, the CR should be less than 10% in order to have consistency in pairwise
judgments.

General

321

Table 2. RCI values of sets of different order n.
n 1 2 3 4 5 6 7 8 9
RCI 0 0 0.58 0.9 1.12 1.24 1.32 1.41 1.45

The TOPSIS method
TOPSIS method is based on the assumption that the best alternative should have the shortest
distance from an ideal solution and the farthest distance from the negative-ideal solution. It is
assumed that each criterion has a tendency of monotonically increasing or decreasing utility.
The preference order of the alternatives is obtained on the basis of these relative distances. The
method is organized in different steps as follows. In the first step the normalized decision matrix
R is constructed.
| |
m x n
... ... ... ...
R
11 1n
21
ij
ij

m
2
kj
k=1
m1 mn

r r
r
a
r
a
r r
(
(
(
(
=
( =
(
(
(
(

(9)

In the following step, the weighted normalized decision matrix [V] is generated, by
multiplying the columns of matrix R with weights of criteria (w
1
, w
2
, , w
n
) estimated by the
decision maker.

| |
m x n
... ... ... ...
V
1 11 n 1n
1 21

ij i ij
1 m1 n mn

w r w r
w r

v w r
w r w r
(
(
(
=
( =
(

(10)

In the third step, the ideal and negative ideal solutions are determined, denoted as A
*
and A
-
,
respectively:

{ }
{ }
i i
(max ), (min ), ,....,
* b c
ij ij 1* n*
A v j J v j J i=1,..., m v v = e e = (11)

{ }
{ }
i i
(min ), (max ), ,....,
- b c
ij ij 1- n-
A v j J v j J i=1,..., m v v = e e = (12)

where indexes J
b
and J
c
are associated to benefit and cost criteria.

To measure the separation S
i*
and S
i-
of different alternatives from ideal solutions, the n-
dimensional Euclidean distance is determined according to the following expressions:

2
( )
n
i* ij j*
j=1
S v v =
for i=1,,m (13)

322
2
( )
n
i- ij j-
j=1
S v v =
(14)
The best alternative is the one with the shortest distance from ideal solution, that is the one
which maximizes the relative closeness index C
i*
, calculated as follows:

i-
i*
i* i-
S
C
S S
=
+
(15)

with C
i*
=1 if A
i
=A
*
.
3 APPLICATIONS TO STRUCTURAL RETROFITTING
An application of MCDM to seismic retrofitting of existing buildings is presented in (Caterino
et al. 2009) and in (Bostenaru 2004).
In the first case, a comparison of different decision making methods is carried out as well. A
case study developed in a previous work by the same authors is used to apply and compare each
retrofitting and decision making method. It refers to an existing underdesigned structure and
allows assessing the specific features of each MCDM procedure. The selected case study is the
SPEAR building, a three-story reinforced concrete (RC) building which is a replica of an
existing non-seismic building, full-scale tested at the ELSA Laboratory. The structure was
selected in order to be representative of preseismic code constructions in southern Europe. Five
alternative retrofit interventions are designed and analyzed by MCDM methods, considering
eight different criteria of evaluation. The first alternative consists of confinement by glass fiber
reinforced plastic (GFRP) of columns and joints and results in an increase of the building
displacement capacity. The second retrofitting solution is based on addition of steel braces.
Other considered seismic strengthening techniques are the concrete jacketing of selected
columns, base isolation of the structure and use of viscous dampers at the first story of the
building to produce attenuation of the seismic demand through a drastic increasing of the
dissipation capacity of the structural system.

Figure 1. Spear Building at ELSA Laboratory. Figure 2. Retrofitting with FRP.

The selected criteria include both qualitative and quantitative measures. In particular, four
socio-economic criteria as well as four technical criteria are defined as summarised in Table 3.

General

323
Table 3. Selected criteria by Caterino et al. (2009).
Socio-economic criteria
Description Unit Typology
Installation cost cost
Maintenance cost cost
Duration of work days cost
Functional compatibility qualitative benefit
Technical Criteria
Description Unit Typology
Required skill of labour qualitative cost
Foundation retrofit strengthened required [ %] cost
Risk of severe damage probability of attainment [ %] cost
Risk of low damage probability of attainment [ %] cost

In order to perform the comparison among the selected alternative, the authors assume that
the decision maker is the owner of the building, thus the matrix of pairwise comparison is
defined accordingly.
The MCDM methods that were considered for the evaluation of alternatives include among
the others WPM, ELECTRE, VIKOR and TOPSIS.
According to the authors, the last two methods result to be, among those investigated, the
most suitable to the focused decision task, because of their capability to manage each kind of
judgment criteria and variables, the clarity of their results, and the reduced difficulty to deal
with parameters and choices they involve.
Another example of MCDM application to retrofitting of existing building is provided in
(Bostenaru 2004). In this case, the type alternatives of retrofit measures included addition of
structural walls, steel braces and side walls respectively, as well as steel mantling and finally the
status quo.
As for criteria, also in this case quantitative and qualitative parameters were considered.
Among quantitative criteria, the strengthening costs using the respective retrofit alternative for
the model building, the equivalent damage elements (total reparation costs for the whole
building divided by the reparation costs for the highest damage in an element) were considered.
The qualitative criteria included influence on building appearance based on the degree influence
the retrofit alternative has on flexibility in facade conformation. Finally, a criterion related to
the extent to which activities inside the building are disturbed during the measure is considered.
In this case an AHP method is used for decision process.

Table 4. Example of a decision matrix for different retrofitting strategies (Bostenaru 2004).
Retrofitting alternatives

Structural
walls
Steel bracing Side walls Steel
mantling
Status quo
Criteria
Costs 103622 87624 102960 55152 0
Equivalent
Elements
345950/
2000
353700/
2000
411170/
2000
273885/
2000
376411/
2000
Architecture E C D B A
Project
Management
V II IV III I

Another example of application of Topsis method for the selection of the optimal retrofit
solutions of an historical church, is being developed by the Authors at the University of Naples
in the Framework of the Research Project FARO (FARO 2009). The building under

324
investigation was heavily damaged by a near-fault earthquake (moment magnitude Mw=6.3),
and following the seismic event, very severe vertical crack in the connection between the facade
and orthogonal walls fatally impaired the structural behaviour of building. An incipient out of
plane failure of the facade was likely to occur. In order repair the building, four structural
interventions have been proposed, which are: (1) steel ties, (2) fiber reinforced concrete
wrapping, (3) SRG wrapping and (4) base isoltation system.

Figure 3. The Visitazione Church in poggio
Picenze (Italy). Post-earthquake damage.
Figure 4. Interior view.

The comparison among the selected rehabilitation intervention has been carried out
considering the four design alternatives listed before versus eight criteria which include: (i) cost
criteria (installation cost, maintenance cost, duration of work); (ii) conservation of historical
heritage criteria (reversibility, compatibility, minimum intervention criteria); (iii) technical
criteria (vulnerability index of repaired building; feasibility).
4 DEMOLITION AND DECONSTRUCTION
As for demolition and deconstruction, alternatives should include issues related to sustainable
waste management and selection of proper demolition techniques.
In (Roussat et al. 2009) an interesting application is presented to waste management. In this
case, nine alternatives are considered, corresponding to different demolition strategies, sorting
and recycling of materials. These solutions range from demolition without sorting of the
different materials to deconstruction and specific treatments of hazardous wastes. A synthesis of
waste treatments considered for each alternative is presented in Table 5.
The criteria taken into consideration for this study are summarized in Fig. 6, where the three
spheres of sustainable development are represented, which include its social, environmental, and
economic aspects. Each decision criteria used in this study takes place in these spheres and
interspheres of sustainable development.
The selected method for decision making is ELECTREIII and a sensitivity analysis is
presented in the paper as well.

General

325
Table 5. Types of alternative waste managements A
i
and treatments (Roussat et al. 2009).

Waste treatments
A
1
A
2
A
3
A
4
A
5
A
6
A
7
A
8
A
9

Demolition x x x
Deconstruction x x x x x x
Sorting platform x x
Illegal dump x
Inert waste in landfill x x
Recycling of metals x x x x x x x x

In (Anumba et al. 2008), the AHP method is applied to the selection of the most appropriate
demolition technique. According to Abdullah and Anumba (2002a), there are six main criteria,
which affect the choice of demolition techniques. The main criteria are structural characteristics,
site conditions, demolition cost, past experience, time, and potential for reuse and recycling. As
far as demolition techniques, the following alternatives are considered: progressive demolition,
deliberate collapse mechanisms and deconstruction.

Figure 5. Criteria taken into account for a
sustainable demolition waste management (Roussat
et al. 2009.)
Figure 6. Example of progressive demolition as
alternative solution techniques in end of life
management (http://www.wreckingcorp.com).
5 CONCLUSIVE REMARKS
In this paper a brief outline of decision making methods and issues related to demolition,
deconstruction and retrofitting of existing buildings was provided.
In particular, the applicability of most common methods was discussed, focusing on types of
alternatives and criteria that usually are taken into account in the literature for considered
problems.
On the basis of presented applications, it can be concluded that different approaches are used
to solve partial problems mainly concerning retrofitting or demolition and deconstruction.
Applications including all possible alternatives available to decision makers could represent
further development of current studies.
REFERENCES
Anumba C. J., Abdullah A. and Ruikar K. 2008. An Integrated System for Demolition Techniques
Selection. Architectural engineering and design management (2008). Vol. 4. Pages 130148.
Bostenaru, D. 2004. Multi-criteria decision model for retrofitting existing buildings. Natural Hazards
and Earth System Sciences (2004) 4: 485499.

326
Benayoun, R.; Roy, B.; Sussman, N., Manual de Reference du Programme Electre. Note de Synthese et
formaton, no.25, Direction Scientifique SEMA, Paris, France, 1966.
Bridgman, P.W. 1922. Dimensional analysis. Yale University Press. New Haven, CT, USA.
Caterino, N. Iervolino, I. Manfredi G. & Cosenza E. 2009. Comparative Analysis of Multi-Criteria
Decision-Making Methods for Seismic Structural Retrofitting. Computer-Aided Civil and
Infrastructure Engineering 24 (2009) 432445.
Chen, S.J., Hwang, C.L. 1991. Fuzzy Multiple Attribute Decision Making: Methods and Applications,
Lecture Notes in Economics and Mathematical Systems, No. 375, Springer-Verlag, Berlin, Germany.
FARO 2009. Local Research Project on Innovation and Sustainability of Requalification Interventions of
Existing Buildings: Best Practice for Retrofitting and Maintenance, University of Naples Federico
II-School of Sciences and Technologies.
Fishburn, P.C. 1967. Additive Utilities with incomplete product set: applications to priority and
assignments. Operation research society of America. Baltimore, MD, USA.
Hwang, C.L.; Yoon, K. 1981. Multiple Attribute Decision Making, Lecture Notes in Economics and
Mathematical Systems 186, Springer-Verlag: Berlin.
Opricovic, S. 1998. Multicriteria Optimization of Civil Engineering Systems, Faculty of Civil
Engineering, Belgrado, Serbia.
Marler Rt, Arora, Js. 2004. Survey of multi-objective optimization methods for engineering, Structural
and Multidisciplinary Optimization, 26(6): 369-395.
Miller, D.W. and Starr, M.K. 1969. Executive decisions and operations research, Prentice Hall, Inc.,
Englewood Cliffs, NJ, USA.
Portioli, F. & Ungureanu, V. 2008. Demolition and deconstruction of building structures. In Proc. of
COST C25 Seminar Sustainability of Constructions, Dresden 6,7 October 2008. (pp. 4.124-4.131).
ISBN: 978-3-86780-094-5. Ed. L. Bragana et al. COST Bruxelles.
Roy, B. 1973. How outranking relation helps multiple criteria decision making, Multiple Criteria
Decision Making, Actes du Sminaire Thorie de la Dcision, Beaulieu-Sainte-Assise, Francia,
1973.
Roussat, N., Dujet, C., Mhu, J. 2009. Choosing a sustainable demolition waste management strategy
using multicriteria decision analysis. Waste Management 29 (2009) 1220.
Saaty, T.L. 1980. The Analytic Hierarchy Process, McGraw-Hill, New York, NY, USA, 1980.
Saaty, T.L. 1999. Decision making for leaders : the Analytic Hierarchy Process for decision in a complex
world, RWS Publications, Pittsburgh, PA., USA.
Saaty, T.L. 1994. Fundamentals of Decision Making and Priority Theory with the AHP, RWS
Publications, Pittsburgh, PA, USA.
Saaty, T. L. 2008. Relative Measurement and its Generalization in Decision Making: Why Pairwise
Comparisons are Central in Mathematics for the Measurement of Intangible Factors - The Analytic
Hierarchy/Network Process. Review of the Royal Spanish Academy of Sciences, Series A,
Mathematics. Vol. 102 (2): 251318.
Triantaphillou, E. 2000. Multi-criteria decision making methods: A comparative study. Dordrecht, itd:
Kluwer Academic Publishers, Dordrecht, The Netherlands.
Zimmermann, H.J. 1996. Fuzzy Set Theory and Its Applications, Kluwer Academic Publishers, 3rd
Revised Edition, Boston, MA, USA.

General

327
E. Cadoni
University of Applied Sciences of Southern Switzerland
Assessment of the structural materials at high stain-rates in
tension
1 INTRODUCTION
The optimization of the tools used in the demolition, whether these are explosives or dynamic
devices (i.e. wrecking ball), is a fundamental requisite to make sustainable also the last phase of
life of the structure. In fact, one of the goal of the deconstruction is to minimize the amount of
materials going to landfills. The knowledge of the dynamic characteristics could improve the
capacity to optimize the debris and to facilitate the segregation of different materials for reuse or
recycling.
Based on these knowledge the deconstruction can be conducted following proper planning, as
a consequence both the environmental and economical costs can be reduced. Moreover,
considering the rising costs of landfills, this approach could result the deconstruction more
economical than traditional demolition, and to contribute to a healthier environment.
Demolitions in urban areas must meet very strict boundary conditions. Constraints can be
different in nature and is increasingly demanded that the demolition should be carried out with
limitation of dust, noise, vibration. This goal can be achieved through proper design and
subsequent virtual simulation using computer codes.
The accuracy of these simulations depends on the precision of the constitutive law in
reproducing the real behavior of the structural materials. In this paper, the experimental
techniques used to determine the mechanical characteristics of the structural materials at high
strain rate in tension as well as some representative results obtained in the recent period are
described briefly.
Because of the restricted space at disposal, this paper cannot be a state of the art review but it
would be only a contribution to illustrate the behavior of the structural materials behavior at
high strain rate and the description of the experimental techniques used to determine them.
2 EXPERIMENTAL TECHNIQUES AT HIGH STRAIN RATE
The design of structures resistant to dynamic loading, requires the use of modern analytical
methods, such as finite element codes, in which the constitutive equations developed are
implemented and calibrated on the dynamic mechanical properties of the materials. The
mechanical properties should be measured with precision tests in a wide range of strain rate in
the most common deformation modes like tension, compression, shear and under combined
loading. Testing of structural materials is then necessary for a basic understanding and safe
design of structures. Moreover the methodologies adopted for testing materials can be
opportunely used for testing of structures.
The dynamic process involved in the demolition of the structures requires the explanation of
the dynamic response of the structural materials. The demolition can be produced in the
structure by means of dynamic loading such as explosives, wrecking ball etc. The experimental
techniques may be the same for material and structural testing.

328
The experimental techniques used to determine the mechanical characteristics of structural
material at high strain rate can be resumed in the three following categories:
- Techniques based on the dropping weight system
- Techniques based on the Split Hopkinson Pressure Bar
- Techniques based on the horizontal sledge or gas gun

In this contribution, the attention will be focused on the second category because it has been
demonstrated that these techniques have shown distinct advantages and results.
Hopkinson bar testing methods with high strain rate applied to structural materials such as
concrete, reinforcing and prestressing steels for investigating the tensile strength are treated
here.
2.1 The Split Hopkinson Tension Bar
The Split Hopkinson Tension Bar (SHTB), developed by the JRC in seventies for direct tensile
testing, has been used to determine the results in the next chapters. The SHTB used are installed
in the DynaMat laboratory of the University of Applied Sciences of Southern Switzerland. The
SHTB consists of two cylindrical high strength steel bars having a diameter of 10 mm, with a
length of 9 and 6 m for input and output bar respectively and the specimen is screwed to the two
bars as shown in Fig. 1.

Figure 1. Split Hopkinson Tension Bar scheme

The test with the SHTB is performed as follows: 1) first, a hydraulic actuator is pulling part
of the input bar (6m) as pretension bar with a diameter of 10 mm; the pretension stored in this
bar is resisted by the blocking device; 2) second operation is the rupture of the brittle bolt in the
blocking device which gives rise to a tensile mechanical pulse of 2.4 ms duration with linear
loading rate during the rise time (30s), propagating along the input and output bars bringing to
fracture the specimen. The input and output bars are instrumented with strain gauges which
measure the incident, reflected and transmitted pulses acting on the cross section of the
specimen. The part of the input bar is used as pretension bar. On the basis of the incident (
I
),
reflected (
R
) and transmitted (
T
) records, the basic constitutive equation of the input and output
elastic bar material and one-dimensional wave propagation theory, it is possible to calculate the
stress, strain and strain-rate curves by the following equations (Lindholm, 1971):
( ) ( ) t
A
A
E t
T E
c o
0
0
= (1) ( ) ( ) dt t
L
C
t
t
R E
}
=
0
0
2
c c (2) ( ) ( ) t
L
C
t
R
c c
0
2
= (3)
Where, E
0
is the elastic modulus of the bars; A
0
is the cross-sectional area of the bars; A is the
cross-sectional area of the specimen; L is the gauge length of the specimen and C
0
is the
velocity of sound in the bar.
3 DYNAMIC BEHAVIOR OF DUCTILE STRUCTURAL MATERIALS
The behavior of ductile materials as steels can be described by their stress versus strain curves.
These curves indicate an enhanced dynamic strength at all values of strain until failure of the
materials. As experienced with many other metallic materials, dynamic mechanical behavior of
reinforcing steel presents significant differences if compared with mechanical behavior
exhibited under static load conditions. The explanation of these phenomena can be found in the

General

329
description of internal mechanisms as thermally activated movement of dislocations. Available
scientific data outline that, as strain rate increases, the following changes in mechanical
properties of reinforcing steel can be noticed. The yield strength and the subsequent flow stress
are seen to increase with the strain rate. On the contrary no changes are experienced in terms of
Young modulus. In the following paragraphs results of some steel used in constructions are
briefly described.
3.1 Reinforcing steel
In literature, unfortunately only limited information are available on the behavior of reinforcing
bars under high dynamic loading, in particular for strain rate characterizing the blast effects (up
to 1000 s
-1
). Reviews of some of the published data were presented in Malvar (1998) and
Mainstone (1975) in which an enhancement of the dynamic strength with increasing of strain
rate was observed. The reinforcing steel with higher yield strength usually show less
enhancement of strength under dynamic conditions than those with lower static strength. The
effect of the strain rate on the materials properties is generally described by the ratio between
the dynamic and static values and it is called Dynamic Increase Factor (DIF). At intermediate
strain rate Nadai (1950) reports a DIF of yield stress of 1.65 for mild steel (
y
=192MPa).
Feldman et al. (1962) recorded a value of about 1.40 (
y
=276-338MPa). Flathau (1971) reports
for steel with a static yield stress of 482 MPa an average DIF value of 1.10-1.15. The
comparison of the dynamic and static stress vs. strain curves of different grade of reinforcing
steels (Mainstone, 1975) is shown in Fig. 2. It can be observed that the increase of the grade in
strength imply the reduction of the uniform strain and the fracture strain.

0
500
1000
1500
0 0.1 0.2 0.3 0.4 0.5
reinforcing steel
200-300 0.001 1/s
200-300 106 1/s
300-400 0.001 1/s
300-400 2600 1/s
over 400 0.001 1/s
over 400 2300 1/s
e
n
g
i
n
e
e
r
i
n
g

s
t
e
s
s

[
M
P
a
]
engineering strain

Figure 2. Comparison between static and dynamic stress vs. strain curves with different grades of steel.

Recently, Cadoni et al. (2007) and Asprone et al. (2009) have been presented the data at high
strain rate of the rod bar reinforcing steel from an existing bridge built in the sixty. They
conclude that as strain-rate increases, yield stress increases more the ultimate stress; in fact yield
stress assumes a maximum DIF value of 1.62 for a strain-rate of 629 s
-1
, while ultimate stress
reaches a DIF of 1.17, with the same strain-rate level. In Fig. 3, the dynamic and static curves
are shown to compare the variation in mechanical properties.

330
0
200
400
600
800
1000
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
629 1/s
184 1/s
0.001 1/s
e
n
g
i
n
e
e
r
i
n
g

s
t
r
e
s
s

[
M
P
a
]
engineering strain

Figure 3. Stress vs. strain curves for rod bar reinforcing steel.

New results on cold-rolled steel B500A steel (EN10080, 2005) are here reported. In Fig. 4, the
stress versus strain curves for different strain rates are depicted. This steel shows an
enhancement of all mechanical parameters when the strain rate increases. The DIF for ultimate
tensile strength at 500s
-1
is 19% higher than the static one. This increase have to be considered
when the dynamic actions like explosion take place.

0
200
400
600
800
1000
0 0.1 0.2 0.3 0.4 0.5
0.001 1/s
5 1/s
250 1/s
500 1/s
e
n
g
i
n
e
e
r
i
n
g

s
t
r
e
s
s

[
M
P
a
]
engineering strain
B500A

Figure 4. Stress vs. strain curves for B500A steel at different strain rates.

3.2 Stainless Steel
The use of stainless steels in constructions is more and more increased in the last years. The
design philosophy moves towards life cycle costing. The increasingly stringent environmental
and economic constraints have encouraged engineers to provide a structure which may be
slightly more expensive at the outset, but over the design life a structure`s final cost will be
significantly reduced. In Fig. 5, the stress versus strain curves at different strain rates are shown.

General

331
This steel enhances its resisting properties increasing the yield and ultimate tensile stress but the
reduction of the uniform strain is also observed.

0
200
400
600
800
1000
0 0.1 0.2 0.3 0.4 0.5 0.6
1000 1/s
500 1/s
250 1/s
30 1/s
0.001 1/s
e
n
g
i
n
e
e
r
i
n
g

s
t
r
e
s
s

[
M
P
a
]
engineering strain
AISI 304

Figure 5. Stress vs. strain curves for AISI304 stainless reinforcing bar steel at different strain rates.
3.3 Structural steel
The structural steel is normally used metallic structures, as an example in Fig. 6, the stress
versus strain curves of the Fe430 steel are shown. The trends confirm the tendency shown for
the previous described steels.

0
200
400
600
800
1000
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Fe430
0.00045 1/s
0.045 1/s
0.5 1/s
5 1/s
550 1/s
1400 1/s
e
n
g
i
n
e
e
r
i
n
g

s
t
r
e
s
s

[
M
P
a
]
engineering strain

Figure 6. Stress vs. strain curves for Fe430 steel at different strain rates
3.4 Pre-stressed steel strands
The pre-stressed steel strands are widely used in civil engineering field as tensioning
components in order to obtain and ensure a high mechanical strength and a long service life to
the structures. Used examples include numerous civil structures, such as cable-stayed and
suspension bridges, tendon-based in offshore platforms and pre-stressed strands in concrete

332
structures (bridges, dams, nuclear power plants, etc.). This steel has been studied and the
mechanical properties of the core wire is analyzed as illustrated in Fig. 7 where the two curves
(dynamic and static) are shown.
The enhancement of the mechanical properties is quite limited compared the reinforcing
steels. This high strength steel shows in dynamic an increase of the 2.8% of the ultimate tensile
strength and of the 3.6% of the yield stress.

0
500
1000
1500
2000
2500
0 0.05 0.1 0.15 0.2 0.25
0,0001 1/s
500 1/s RT
e
n
g
i
n
e
e
r
i
n
g

s
t
r
e
s
s

[
M
P
a
]
engineering strain

Figure 7. Stress vs. strain curves for pre-stressed strands steel
4 DYNAMIC BEHAVIOR OF BRITTLE STRUCTURAL MATERIALS
Brittle materials used in constructions, as rocks and concrete, have been studied by several
researchers in the last three decades. In particular, the concrete has been studied in both tension
and compression under dynamic loading, some reviews can be found in Malvar & Ross (1998)
and Bishoff & Perry (1991). The experimental techniques used for such studies were principally
Hopkinson bar with traditional set-up or with important modifications. The results, reported
here, have been taken from the reference: SHTB (Cadoni et al. 2000, 2001, 2009).
4.1 Concrete
In the reviews published so far an increase of the concrete mechanical characteristics, such as
strength, ultimate strain, elastic modulus and fracture energy, is observed as the strain-rate
increases. Often the data have been obtained by different experimental techniques and the
results are not easily comparable and even sometimes contradictory. The rate sensitivity in
concrete is normally expressed as the ratio of the dynamic to the static value of a particular
mechanical property. Special attention has been given to the concrete strength (DIF) and can be
useful in numerical code formulations. Such descriptions of the concrete behavior under impact
can be found in (CEB 1988). It should also be mentioned that the current constitutive laws for
concrete tend to encompass more information about the material characteristics, for example,
the ultimate strain or fracture energy, and this imposes further demands on testing.
The tensile strength of concrete can be determined by means of several tests, such as direct
tension, splitting, bending or double-punching. The direct tension test, which is naturally
considered as reference, is the more complex to perform both under static as well as under
dynamic conditions. The other tests produced, in general, differing in experimental results, can
be referred to the direct tension by a correction factor. In dynamic testing these differences
could be more marked because of the wave propagation and inertia effects. No comparative
studies have until now been carried out in this direction.

General

333
The high strain-rates testing apparatuses which are able to measure the wave propagation,
should be used for the mechanical characterization of concrete. It is nowadays recognized that
the widely used Split Hopkinson Pressure Bar (SHPB) is the more suitable technique for the
determination of the mechanical properties of structural materials under high loading rates.
While standard SHPB with a diameter in the range of 10-20 mm are sufficient for dynamic
testing of fine-grained materials, like steel, larger bars are needed to load concrete specimens of
meaningful size. In fact, for concrete, the specimen size should be larger than four times the
maximum aggregate size (as a rule-of-thumb), so that a representative volume of material is
tested. The material and the dimension of the bar have also to satisfy the appropriate conditions
in order for the one-dimensional theory of wave propagation to be applicable.
Direct tensile set-up testing was developed in the last decades, for example at the Delft
University in the Netherlands and at the Joint Research Centre (JRC) in Ispra, Italy. In the
Dutch version, a tensile stress wave is generated by striking the end of the input bar
(compression pulse converted to tension). In the JRC modified Hopkinson bar (Albertini et al.
1999), the pulse is generated by the sudden release of the energy stored in a high strength steel
bar and not by an impacting projectile (direct tension wave). By exploiting this principle at the
JRC, special testing equipment has been developed allowing the dynamic characterization of
real size concrete specimens (cube of 200 mm side) (Albertini et al. 1999, Cadoni et al. 2000)
by means of a very large Hopkinson bar set-up (200 m long).
The data, here analysed, have been obtained during a broad experimental campaign on
several concretes carried out over the last decades.

Figure 8. Set-up for Split Hopkinson Tension Bar for concrete.
The stress versus strain curves are obtained using equations (1) and (2). The comparison
between three curves measured at different strain rates is shown in Fig. 9. It can be observed
easily that all mechanical parameters are enhanced as the strain rate increases. This clearly
implies that the concrete is very rate sensitive and the used apparatus is capable of capturing
these phenomena in all their aspects (strength, strain, fracture energy and so on). The dynamic
elastic modulus has been calculated in the stress versus strain curve between the points at one
third and two thirds of the maximum tensile strength. The load versus displacement curve has
been used to obtain the fracture energy, which is considered as the integral of this curve after
the subtraction of the elastic contribution.

334
0
2
4
6
8
10
12
0 0.0005 0.001 0.0015 0.002 0.0025
d
max
= 10 mm
10-6 s-1
1 s-1
10 s-1
s
t
r
e
s
s

[
M
P
a
]
strain

Figure 9. Stress vs. strain curves for different strain rates for concrete

The results produced in the recent years are now compared with some formulations proposed
in the literature. The data used in this comparison have been chosen with the criterion that they
must have been obtained with similar concrete in terms of compressive strength (between 35
and 40 MPa) and specimen size. Unfortunately, only few experimental data have been found for
direct tensile tests, as the majority has been obtained by means of spalling tests.
4.1.1 Tensile strength
As starting point for assessing the influence of strain rate on the mechanical properties of
concrete, the dynamic tensile strength has been compared to the static tensile strength and the
relevant DIF values have been calculated. Further, in order to verify that the current
experimental data are consistent with other results available in the literature, a comparison with
existing theories has been performed. As reference, the (CEB 1988) and (Malvar & Crawford
1998) expressions have been considered. Both of these formulations try to predict dynamic
tensile strength, expressing the DIF as a function of the applied strain rate. The CEB formula
reads:
1
016 . 1
0 s , t
d , t
s 30 for
f
f
DIF

o
s c
|
|
.
|
\
|
c
c
= =
(4)
1
3
1
0 s , t
d , t
s 30 for
f
f
DIF

> c
|
|
.
|
\
|
c
c
q = =
(5)
Where, f
t,d
is the dynamic tensile strength; f
t,s
is the static tensile strength; c is the strain-rate;
0
c is a constant equal to 3010
-6
s
-1
and has the meaning of the static strain rate;
and log 7.11 2.33 q o = ,
where
0
cs
f
f
6 10
1
+
= o
,
f
cs
is the static compression strength; f
0
is a constant equal to 10 MPa.

This formulation gives the DIF as a bilinear function of the strain rate in the logarithmic
scale, and presents a slope variation at 30 s
-1
.
The Malvar formulation, instead, while expressing the DIF in a similar way, fixes the
reference static strain rate at 10
-6
s
-1
and moves the slope variation to 1 s
-1
:

General

335
1
M 0 s , t
d , t
s 1 for
f
f
DIF
M
o
s c
|
|
.
|
\
|
c
c
= =
(6)
1
3
1
M 0 s , t
d , t
s 1 for
f
f
DIF

> c
|
|
.
|
\
|
c
c
| = =
(7)

where, f
t,d
, f
t
,
s
and c are as already explained above;
M 0
c is a constant equal to 10
-6
s
-1
;
and 2 6 log =
M
o |
where
0
cs
M
f
f
8 1
1
+
= o , f
cs
, f
0
are as already defined above.

Next, plotting these relationships layered on the selected experimental results (see Fig. 10a),
allows a direct evaluation of the consistency of the presently derived data. It is evident that, as
Malvars formulation suggests, for the current data the slope variation is effectively perceptible
at 1 s
-1
, even if experimental data are not available for strain rates higher than 10 s
-1
. In general,
it seems that the CEB expression often underestimates or just lightly overestimates the
experimental results while, Malvars expression often foresees a higher DIF than the test value.
For this reason the CEB expression may be preferable for design calculations, where more
conservative values should be considered.

0
2
4
6
8
10
12
14
16
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
10
1
10
2
10
3
JRC, 1997, Dmax=5mm
JRC, 1997, Dmax=10mm
Weerhijm&Van Doormaal, 2007, Dmax=8mm
Brara&Klepaczko, 2006, Dmax=2mm, wet
Brara&Klepaczko, 2006, Dmax=2mm, dry
Schuler et al., 2006, Dmax=8mm
Toutlemonde, 1994, Dmax=10mm
CEB, 1988
Malvar, 1998
Krmeling et al., 1980, mix 17, Dmax=8mm
Krmeling et al.,1980, mix 15, Dmax=8mm
Krmeling et al., 1980, mix 14, Dmax=16mm
d
y
n
a
m
i
c

s
t
r
e
n
g
t
h

/

s
t
a
t
i
c

s
t
r
e
n
g
t
h
strain-rate [s
-1
]
a)
0
2
4
6
8
10
12
14
10
-5
10
-3
10
-1
10
1
10
3
JRC, 1997, Dmax=5mm
JRC, 1997, Dmax=10mm
Weerhijm&Van Doormaal, 2007, Dmax=8mm
CEB, 1988
d
y
n
a
m
i
c

s
t
r
e
n
g
t
h

/

s
t
a
t
i
c

s
t
r
e
n
g
t
h
loading-rate [GPa/s]
b)
Figure 10. Dynamic Increase facto in function of: a) strain curves and b) loading rate

The results of the tensile tests have also been expressed in terms of the applied stress rate, as
proposed in (CEB 1988):
s / MPa 10 for
f
f
DIF
6
0 s , t
d , t
s o
|
|
.
|
\
|
o
o
= =
o
(8)
s / MPa 10 for
f
f
DIF
6
3
1
0 s , t
d , t
> o
|
|
.
|
\
|
o
o
q = =
(9)
where: f
t,d
, f
t,s
, and log have already been defined above and o is the stress rate and
0
o is a
constant equal to 1 MPa/s and has the meaning of the static stress rate;
Fig. 10b shows both the current experimental results and the literature data with the CEB
formulation. The conservative approach of CEB is again clearly seen.

336
It is worth commenting that the formulations up to now proposed, including those above,
consider only the concrete grade and they do not take into account other important variables.
There is, however, strong evidence that the enhancement of the mechanical properties depends,
for example, on the maximum aggregate size, as demonstrated in (Cadoni et al. 2001b), and that
the free water content greatly influences the strain-rate behaviour (Cadoni et al. 2001a, Rossi
1991). To properly analyse the DIF a new formulation should be developed, and this in turn
would require more studies for describing the influence of the single parameters, such as the
cement content, cement type, aggregate type, water/cement ratio, etc.
4.1.2 Fracture Energy
The fracture energy is an important parameter, which is crucial to know because it governs
crack propagation in concrete and consequently crucial aspects in demolition. The CEB
proposes a relation between the total dynamic fracture energy and the rate of deformation, in
particular involving the crack opening displacement. For the calculations, it is recalled that for a
specimen subjected to uniaxial loading the deformation consists of an elastic part and the
opening of a crack. Since the elastic deformation is by definition reversible, the fracture energy
is defined as the area under the stress-crack opening curve multiplied by the cross-sectional area
of the specimen. The CEB relation is expressed as:

w
w
G
G
045 . 0
0 s , tot
d , tot
|
|
.
|
\
|
=
(10)

with w the crack opening velocity and
0
w = 10
-3
mm/s. G
tot,d
and G
tot,s
denote the dynamic and
static total fracture energy, respectively.
In Fig. 11a, the comparison between the current experimental data and this relation is shown.
The data obtained in this campaign are also compared with the available literature data and
presented in Figure 11b. Clearly, they follow well the general trend. It is worth observing how
the fracture energy increases with the strain-rate; as noted by Brara & Klepachko (2007),
fracture energies can be more than 10 times the corresponding static values for high strain-rates.
An explanation of this behavior can be reached through the theory of multi-activation of cracks
(Cadoni et al. 2000) and by the nature of the loading. Due to the wave propagation, a crack does
not choose the path with the minimum energy (through the cement matrix around the
aggregates), but it is forced to pass through the aggregates, thus incrementing the dynamic
characteristics of the specimen in tension.

1
2
3
4
5
6
7
8
10
-3
10
-2
10
-1
10
0
10
1
10
2
10
3
10
4
CEB 187, 1988
Dmax= 5mm
Dmax= 10mm
G
f
,
d
/

G
f
,
s
w/w
0
. .
a)
0
200
400
600
800
1000
1200
1400
1600
10
-7
10
-5
10
-3
10
-1
10
1
10
3
JRC, Dmax=5mm
JRC, Dmax=10mm
Weerheijm et al., 2007
f
r
a
c
t
u
r
e

e
n
e
r
g
y

[
J
/
m
]
strain-rate [s
-1
]
b)
Figure 11. Fracture energy: a) DIF and b) evolution in function of strain rate

General

337
1
1.5
2
2.5
3
3.5
4
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
10
1
10
2
Dmax=10mm
Dmax=5mm
CEB, 1988
d
y
n
a
m
i
c

u
l
t
i
m
a
t
e

s
t
r
a
i
n

/

s
t
a
t
i
c

u
l
t
i
m
a
t
e

s
t
r
a
i
n
strain-rate [s
-1
]
a)
1
1.5
2
2.5
10
-6
10
-5
0.0001 0.001 0.01 0.1 1 10 100
CEB 187, 1988
Dmax= 5mm
Dmax= 10mm
Dargel, 1984
d
y
n
a
m
i
c

e
l
a
s
t
i
c

m
o
d
u
l
u
s

/

s
t
a
t
i
c

e
l
a
s
t
i
c

m
o
d
u
l
u
s
strain-rate [s
-1
]
b)
Figure 12. DIF of a) Ultimate strain and b) elastic modulus in function of strain rate
4.1.3 Ultimate strain
In the current dynamic direct tensile tests, carried out with the SHTB, the strain is obtained
through equation (2). The ultimate strain is defined as the strain at the maximum stress.
Obviously, the strain calculation is valid only up to the instant where macro-cracks initiate to
open, which occurs after the maximum stress has been reached. Beyond this point the
measurements of the SHTB record the crack opening displacement and the remaining parts of
the concrete unloading. Fig. 12a shows the experimental results, which are compared with the
CEB relationship between the ultimate strain and the stress or strain-rate, described by:

020 . 0
0 s , u
d , u
|
|
.
|
\
|
c
c
=
c
c
(11)
with
0
c =310
-6
s
-1

In the above relations c
u,d
and c
u,s
represent the ultimate dynamic and static strains,
respectively. These relations have been derived on the basis of few experimental data. As well
evidenced in Fig. 12a, they fail to describe the increase of the ultimate strain after the strain-rate
of 10 s
-1
. A new formulation would be needed, for example a bilinear relation should be
proposed. Of course, to do this, more experimental data is required.
4.1.4 Youngs modulus:
The influence of the strain-rate on the Youngs modulus for tension is described by CEB with
the following formulation:

E
E
016 . 0
0 s
d
|
|
.
|
\
|
c
c
=
(12)
with
0
c =310
-6
s
-1
. E
d
and E
s
denote the dynamic and static Young modulus, respectively.

These relationships are valid for all strain-rates and all concrete grades. Comparing, in
Fig.12b, the present data against this expression and with the literature data, it can be concluded
that eq.(12) fits well the experimental results up to 1 s
-1
. For higher strain-rates, also for this
property, a bilinear expression should probably be proposed, as for the DIF. More experimental
studies should be performed in order to validate the formulation and also physical explanation
of this increase is necessary.

338
5 CONCLUSION
The assessment of the structural materials at high strain rate in tension is the objective of this
paper. The experimental results as well as the experimental techniques here presented have the
aim to underline the importance of the material characterization to perform optimized design of
the deconstruction phase. The knowledge of the dynamic material behavior is strictly necessary
and represents a very important point for a sustainable approach to the demolition phase of a
structure so that costs and impacts can strongly reduced.
Material characterization at high strain rate of ductile and brittle materials has been presented
recalling some formulation proposed for the dynamic materials characteristics. The mechanical
properties of the structural materials presented have shown a strain rate dependence consisting
of a non linear increase of strength, ductility and energy absorption capability with increasing
strain-rate.
REFERENCES
Albertini C., Cadoni E., Labibes K. 1999. Study of the mechanical properties of plain concrete under
dynamic loading, Experimental Mechanics, 39 (2): 137-141
Asprone D., Cadoni E., Prota A. 2009. Tensile dynamic behavior of existing reinforcing steel under high
strain rates, ACI Structural Journal, vol. 106, n. 4, 523-529
Bischoff, P.H., Perry, S.H. 1991. Compressive Behavior of Concrete at High Strain Rates, Materials and
Structures, Vol. 24, pp. 425-450.
Brara A., Klepaczko, J. R. 2007. Fracture energy of concrete at high loading rates in tension,
International Journal of Impact Engineering 34: 424435
Cadoni E., Labibes K., Berra M., Giangrasso M., Albertini C. 2000. High strain-rate tensile concrete
behaviour, Magazine of Concrete Research, 52, No.5, Oct., 365-370.
Cadoni E., Labibes K., Albertini C., Berra M., Giangrasso M. 2001a. Strain-rate effect on concrete with
different level of the relative humidity, Materials and Structures, 34(235), 21-26.
Cadoni E., Labibes K., Berra M., Giangrasso M., Albertini C. 2001b. Influence of the aggregate size on
the strain-rate tensile behaviour of concrete, ACI Materials J., 98(3), 220-223.
Cadoni E., Asprone D. and Prota A. 2007. High strain-rate testing analysis of concrete and steel for the
assessment of the Tenza Bridge under blast loading. in New trends in fractures mechanics of
concrete, eds. Carpinteri-Gambarova-Ferro-Plizzari, Balkema, pp 627-635.
Comit Euro-International du Bton. 1988. Concrete structures under impact and impulsive loading,
CEB Bulletin 187, Lausanne, Switzerland.
European Standards. 2005. EN 10080:2005 Steel for the reinforcement of concrete.
Feldman, A., Keenan, W.A., Seiss, C.P. 1962. Investigation of resistance and behavior of reinforced
concrete members subjected to dynamic loading, part III, civil Engineering studies, Structural
research Series, No. 243, University of Illinois, Urbana, Feb.
Flathau W.J. 1971. Dynamic tests of large reinforcing bar splices. Technical report N-71-2, U.S. Army
engineer division, Huntsville, Ala.
Lindholm U.S. 1971. High Strain Rate Tests, in Measurement of Mechanical Properties, vol. 5,
Techniques of Metals Research, Bunshah ed., Interscience, Wiley, 199-271.
Mainstone, R.J. 1975. Properties of materials at high rates of straining or loading, State-of-the-Art- report
on Impact loading of structures, Materials & Structures, Vol. 8, N. 44, 102-116.
Malvar, L.J. 1998. Review of static and dynamic properties of steel reinforcing bars, ACI Materials
Journal, vol. 95, pp. 609-616.
Malvar L.J., Crawford J.E. 1998. Dynamic Increase Factors for Concrete, Proc. 28th DDESB Seminar,
Orlando.
Malvar L.J., Ross C.A. 1998. Review of strain rate effects for concrete in tension, ACI Materials Journal,
95(6), 735-739.
Nadai, A. 1950. Theory of flow and fracture of solids, Engineering Societies Nomographs, Vol. I, 2
nd
Ed.
McGraw-Hill Book Co., New York.
Rossi P. 1991. Influence of cracking in the presence of free water on mechanical behaviour of concrete,
Magazine of Concrete Research, 43, N.154, Mar, pp.53-57.

General

339
O. Hechler
O. Popovic Larsen, S. Nielsen
The Royal Danish Academy of Fine Arts, School of Architecture, Institute of Technology, Copenhagen,
Denmark
Design for Deconstruction
1 INTRODUCTION
1.1 Life cycle assessment and relevance of design for deconstruction
Cradle-to-grave is the full Life Cycle Assessment from manufacture ('cradle') to use phase and
disposal phase ('grave'), see Figure 1. All inputs and outputs are considered for all phases of the
life cycle. For buildings, 50 years of design life are generally imposed - so far, it has been
assumed, that ca. 80% of the energy input and emission arise in this use phase (service life).

Extraction
Construction
Installation
End of life
Use
phase
Material
production
Product
fabrication
Packaging
Transport
E
n
e
r
g
y
M
a
t
e
r
i
a
l
E
n
e
r
g
y
M
a
t
e
r
i
a
l
E
n
e
r
g
y
M
a
t
e
r
i
a
l
E
n
e
r
g
y
E
n
e
r
g
y
E
m
i
s
s
i
o
n
E
m
i
s
s
i
o
n
E
m
i
s
s
i
o
n
E
m
i
s
s
i
o
n
E
m
i
s
s
i
o
n
E
m
i
s
s
i
o
n
S
o
i
l
A
i
r
W
a
t
e
r

Reuse
Recycling
E
n
e
r
g
y
Extraction
Construction
Installation
End of life
Use
phase
Material
production
Product
fabrication
Packaging
Transport
E
n
e
r
g
y
E
n
e
r
g
y
M
a
t
e
r
i
a
l
E
n
e
r
g
y
E
n
e
r
g
y
M
a
t
e
r
i
a
l
E
n
e
r
g
y
E
n
e
r
g
y
M
a
t
e
r
i
a
l
E
n
e
r
g
y
E
n
e
r
g
y
E
n
e
r
g
y
E
m
i
s
s
i
o
n
E
m
i
s
s
i
o
n
E
m
i
s
s
i
o
n
E
m
i
s
s
i
o
n
E
m
i
s
s
i
o
n
E
m
i
s
s
i
o
n
S
o
i
l
A
i
r
W
a
t
e
r

Reuse
Recycling
E
n
e
r
g
y

Figure 1. Diagram showing cradle-to-grave approach
Since the first oil-crisis in 1973 most countries have implemented progressive building codes
for insulation standards, resulting in a massive saving of energy for operation heating, cooling
and domestic services. Technological development in HVAC-systems (Heating, Ventilation and
Air-Conditioning systems) and sealing methods has enabled a decrease in buildings' operational
energy consumption to very low levels up to the extend known as passive-house projects.

340
These changes have induced the relevance of widening the scope of energy saving from a
former exclusive focus on operational energy in the use phase to the inclusion of process energy
i.e. energy for mining, processing, transportation, assembly and building site operations.
Through the processes, the assembled building and its materials come to represent an
accumulated energy capital (embodied energy or grey energy) which should be administered
appropriately. Therefore, resources consumed in relation to buildings must be viewed in a
lifecycle perception which implies a perspective beyond operation and amortization the
management hierarchy proposed is: Reduce, Reuse, and Recycle (the 3 Rs). Three significant
challenges evoke from this perspective:
1. Sophisticated integral building management combined with high quality cladding systems as
well as optimally tailored thermal mass and minimal weight for reducing the embodied
energy should be designed; for this purpose advanced, cost-effective solutions combining
structural steel with concrete (ca. 10-14cm concrete slab are sufficient for passive thermal
activation (www.corusconstruction.com)) have already been developed. Hereby the use of
high strength materials (e.g. S460 steel and higher strength concrete) should be focused on.
2. Disassembly methodologies must be developed and employed in contemporary building
practice in order to enable future reuse of building parts with the lowest possible
consumption of energy for transformation.
3. The understanding of buildings structures as a capital amount of embodied energy certainly
includes the existing building stock, and hereby the relevance of treating construction waste
in an upgrading process similar to the processing of virgin mining resources.
2 END OF LIFE SCENARIOS AND IMPACTS
2.1 Life cycle assessment and relevance of design for deconstruction
The lifetime of structures is conditioned by market economy, socio-political constraints,
popularity and change of architectural needs. At end of their life, buildings should be
deconstructed. The end-of-life scenario can be broken down into three separate sub-phases:
1) Environmental impact of deconstruction activities (dust, noise, etc),
2) Re-use and recycling rates of materials,
3) Environmental impact of waste processing activity (e.g. scrap processing).

Recycling is hereby defined as the end-of-life recovery and reprocessing of a product (e.g. by
re-melting of steel construction products to form new steel products) and Reuse is defined as the
end-of-life recovery and reuse (e.g. of steel construction products as a product filling the same
function with or without some reprocessing). The recycling rate (within a defined system) is
defined as The tonnage of a product recycled / Tonnage of the product arising on demolition
sites. The reuse rate is similar defined by replacing the word recycled with reuse.
In the following each end-of-life scenario is shortly reviewed.
2.2 Environmental impact of deconstruction activities
Disassembly technique is a new discipline within the field of architectural/engineering practice
and theory. It should be pointed out that it is absolutely not part of the current daily routine or
responsibility among professional architects to plan for and to explain how buildings are
demolished, or how materials can be reused and respectively recycled. However, in a resource
perspective, this is a crucial ability to cultivate the implementation of documentation for
possible disassembly similar to the documentation of operational energy use in building
regulations, which today is required in most countries
1
.

1
For demolition concepts and detailed deconstruction activities information refer to [Kamrath &
Hechler 2010].

General

341
2.3 Re-use and recycling rates of materials
The large quantities of Construction and Demolition (C&D) waste, produced from Construction
related activity, are resulting in a significant burden on society. Thus architects and engineers
need to already in design stage to specify how they can stimulate higher reuse and recycling
rates of materials at end of life. The so-called design for deconstruction is thus defined as
design for deconstruction of structures and complex building parts into basic materials so that
they enable resubmission of the materials in the resource loop at the end of life of a building or
structure. End-of-life scenarios are hereby:
Building reuse or relocation,
Component reuse or relocation in a new building,
Material reuse in the manufacture of new component,
Material recycling into new materials.

The ultimate target is to aspire towards a zero-waste economy, leading to buildings which do
not consume non-renewable sources of energy. In the case of building materials, this is seen as
possibly using over and over again different building components, once the life-time of a
structure comes to an end (CIRIA).
Therefore, parallel to a classification system and quality control, an economic loop has to be
established, in which resources taken from a previous life cycle are used. This loop will need a
high rate of standardization, planning, logistics and technical processing. However this
approach leads to effective mapping out of future risk and cost, by ensuring that building
components and products used in the structure can be effectively maintained and replaced. In
principle, design and detailing for deconstruction has also economic dimensions, in that a
property is seen as more attractive as an investment opportunity (SEDA).
2.4 Environmental impact of waste processing activity (e.g. scrap processing)
The recycling of C&D waste only partially addresses the problem, because it can lead to
considerable consumption of resources in re-processing and transportation. Although, in
addition to reduced resource extraction and waste creation, environmental impacts can be
significantly reduced due to recycling (e.g. 50% less CO
2
emission for steel due to recycling
estimated via the closed loop analysis (EPD Steel 2010), the energy input is e.g. reduced on
25% (SRI)). The waste processing activity is however not addressed in this paper. For further
information reference is given to e.g. ((Recycling concrete 2008), (SRI), (WSA), (Hettinger et al
2011)).
2.5 Benefits and barriers for deconstruction and reuse/recycling
Many benefits are associated with deconstruction and reuse/recycling. However only a fraction
of construction elements can be reclaimed and reused for their original purpose as barriers
against deconstruction are still present. Table 1 gives an overview of the benefits from and
barriers to deconstruction (based on (CBPR 2003)). To increase the application of
deconstruction the opportunities need to be outlined in all levels of decision making and the
barriers need to be reduced or overcome.

342
Table 1. Benefits and barriers to deconstruction
Benefits of deconstruction Barriers to deconstruction

Environmental:
Reduced primary resource use;
Reduced waste to landfill;
Opportunities for recycling;
Reduction of site impacts caused
by demolition (compaction, dust
etc.).

Economic:
Profits due to on-sale of salvaged
goods and reduced landfill costs;
Small business development to
handle salvaged material for reuse
(NAHB 2000);
Promotion and increased sales of
green products to be accounted
for in e.g. LCA.

Social:
Creation of jobs in deconstruction
(opportunity for unemployed and
unskilled workers (NAHB 2000));
Deconstruction trains workers for
the construction industry;
Cultural values preservation and
reflection of trend to sustainable
living in the population;
Aesthetic qualities of reused
former local materials may be used
for architectural identification or
the aged look may be celebrated
(Guy & Timothy 2003);
Deconstruction could provide low
cost materials to low income
communities;
Increased networking stimulated as
deconstruction opens the potential
to make stronger communities
through greater communication.

Health:
Off-gassing characteristics of old
materials are better than that of
new materials;
Deconstruction means less new
waste facilities.

Perception and education:
Designers/public/builder attitude: new is better or
new is easier;
Lack of resources for education on deconstruction;
Lack of research into deconstruction;
Lack of information and tools to implement
deconstruction.

Design for Deconstruction:
Design for deconstruction in new buildings is hardly
considered (failure of codes to address the reuse of
building materials (Guy & Williams 2003));
In most construction segments existing buildings are
not designed to be deconstructed;
Lack of education on design for deconstruction;
Lack of understanding benefits and opportunities
associated with deconstruction;
Lack of understanding and use of LCA tools or
concepts.

Market development:
High cost of transport and storage of recycled
components and materials;
Uses for some salvaged materials are undeveloped;
Guaranteed quality/quantities of reused materials are
difficult.

Economics:
Low cost of some new raw materials;
Low tipping rates (including landfill) in some
countries;
Deconstruction needs a more skilled workforce than
demolition (Kamrath & Hechler 2010);
Benefits of deconstruction are long term and collective
but at first costs focus is dominant;
Market pressure the current climate of as fast as
possible;
Highly speculative nature of many buildings, whereby
there is no long-term ownership and adaption,
renovation and demolition costs are not borne by the
original owner.

C&D Industry:
Hardly regulated industry;
Lack of communication and networking in the C&D
industry and with waste minimisation organizations;
Demolition is usually a low profit margin industry.

General

343
Legislation:
Contributes to meeting Local
Authorities and Central
Government obligations for waste
targets, zero waste, Kyoto targets
and energy efficiency targets.
Legislation:
Confusion may be present on what Government
legislation is, relating to environmental responsibility;
C&D waste minimisation may not be a priority as the
agenda is relatively new, vaguely exposed and thus
unknown;
Lack of politically provided incentives to stimulate the
overcoming of constraints; economical, technological,
cultural and logistical.

Technical Issues:
Liability in certification and avocation of reused
components or materials not clear (lack of grading
system for reused structural elements);
Lack of documentation on existing buildings to plan
for deconstruction;
Some new materials are subsidized, creating unfair
competition with reused materials;
Increase in use of non-reversible technology, systems,
chemical bonds and plastic sealants etc.
Seismic areas may make design for disassembly more
difficult;
New construction systems make recovery more
difficult and less financially rewarding.
3 DESIGN FOR DECONSTRUCTION
3.1 General
Literature gives numerous references with recommendations for the dimensioning of building
elements with implications for their practical performance in terms of conversion and reuse, e.g.
(Nordby 2009). However many of these are based on case studies and do not provide general
guidance. Consequently a set of recommendations has been developed based on literature and
own experience. The recommendations are differentiated in general principles, principles in the
choice of materials, structural design principles, principles for the connections and for the
foundations. Further a short example on a good design for deconstruction is presented.
It is apparent from the following list of design for disassembly principles that there will be
many occasions when there will be a conflict between some of them. The principles in
themselves offer guidance on how to design for future disassembly, but as already noted, there
are broader themes that must be engaged with in order to answer the more challenging questions
of what, where, when, and indeed if, to disassemble (Crowther). Thus, the guidance offered here
must be taken as a starting point for the development of individual strategies for individual
buildings and reflect in general good design practice.
3.2 General principles for design for deconstruction
The following general principles should be followed to maximize reuse of elements or recycling
of materials:
The building needs to be documented in detail with as-built drawings, photographs of hidden
components, connections and material descriptions. A list of building elements with their
design life and potential for reuse should be added (Webster&Costello 2005). Labeling of
the elements by a bar code, stamp or RFID-microchip speeds up the identification at element
level and is optionally to be considered (information stored e.g.: material type, characteristic
properties, origin, place, time, dimensions, toxic risks, etc.).
The deconstruction concept including instructions for disassembly should be added in the
building documentation. Selectively dismantling and removing of elements before the

344
structure is demolished or disassembling of the entire structure in the reverse order to which
it was constructed are optional concepts (Hagen 2007). The deconstruction steps and aids
should be ordinary (e.g. crane usage may be impractically for certain steps). Further points of
attachment for lifting equipment as well as temporary supports should be envisioned.
Layering in design is recommended. Layering is a prominent strategy for handling the
different lifecycles of building elements ((Brand 1994), (Durmisevic 2006)). It accounts for
differences in expected lifetime for respectably structure (30 300 years life expectation;
typically 60 years for buildings; 50 years is the design life for buildings, 100 years for
bridges in the Eurocode), skin (cladding, roofing: ca. 20 years due to wholesale maintenance,
changing technology and fashion), services (electrical, hydraulic and data, etc: ca. 7-15
years), space plan (change each 3 years in commercial buildings, every 30 years in domestic
buildings) and stuff (furniture and non-attached space elements: may change daily to
monthly), see Figure 2. Site is the ground on which the building sits and is external. An
overview on life expectations from various sources in literature is given in (Crowther).

The idea of shearing layers in a building
is to technically separate the layers.
Hereby it becomes possible to make
changes with a minimal resources use and
cost.. By making conversions easier and
cheaper, a total economic gain can be
obtained which in turn decreases the
probability of demolishing. Thus, also a
recycling hierarchy should be accounted
for in layering in such a way that
independent exchange of building parts
according to lifecycles is enabled.
Figure 2. A building as Layers (Brand 1994).
The use of small and light-weight building modules sometimes offers a similar flexibility as
the layering principle for adaptation to varying lifecycles (Nordby 2009).
Provide realistic tolerances to allow for handling during disassembly/reassembly (greater
tolerances may be required for repeated assembly/disassembly process (Crowther)).
Massive construction should be avoided in addition to problems in due to heavy weights
and large volume it is against the European Waste Regulation (2008/98/EC).
Plaster or screed should not be foreseen.
On the level of service integrations, concealed wires should be avoided; services in channels
(e.g. integrated in facade or slabs) are recommended (an example of the structural integration
of installation services is given in Figure 3). The function of the installation devices may be
remote controlled.

Figure 3. Examples of integration of structure and services.

General

345
3.3 Principles for the choice of materials
With the right choice of materials the reuse/recycling rate can be stimulated significantly. In the
following principles for material selection are given:
Use recycled and recyclable materials to allow for all levels of the recycling hierarchy;
increased use of recycled materials will encourage industry and government to develop
technologies for recycling and to create larger support networks and future markets
(Crowther).
In general, a limitation of the number of materials/components leads to higher reuse/recycle
rates this will simplify the process of sorting during disassembly, reduce transport to
different recycling locations and result in greater quantities of each material (Crowther).
Durable materials and components with good tolerances are to be preferred.
Avoid paintings, fire protection spray or intumescent paints to get back clean material (e.g.
sorted steel); e.g. corrosion of exterior steel should be rendered unnecessary
constructively; fire engineering is an option for hot design. Note, some protective finishes
such as galvanization may still on balance be desirable despite disassembly problems
(Crowther).
With regard to structural masonry, design for reuse involves the using of mortar weaker than
the masonry in order to facilitate separation of bricks and of prefabricated brickwork panels.
For the recycling of brickwork as fill or aggregate into low-grade concrete, it is important
that no chemicals are contained that could contaminate the water in the concrete mix.
The use of non-recyclable materials like thermal insulation (glass wool, polystyrene), hard
plaster, light-weight materials with low recycling possibilities should be avoided.
Avoid toxic and hazardous materials - this will reduce the potential for contaminating
materials that are being sorted for recycling, and will reduce the potential for health risks that
might otherwise discourage disassembly (Crowther).
3.4 Principles for the structural design
In general, the structural system of a building comprises over 50% of the building weight and,
traditionally, has the lowest potential for reuse (Webster & Costello 2005). Thus by respecting
the following principles in the structural design will increase the potential for reuse/recycling:
Use a standard column grid and inter-storey height.
Use as wide of structural grid as possible to maximize the non-structural wall elements
(Seattle 2006). Further, for steel structures, the design of long span steel beams is
recommended to allow for flexibility of reuse, e.g. by trimming to a new length.
Use prefabricated subassemblies and a system of mass production - to reduce site work and
allow greater control over component quality and conformity (Crowther).
Use an open building system where parts of the building are more freely interchangeable and
less unique to one application this will allow alterations in the building layout through
relocation of component without significant modification (Crowther).
Minimize the number of different types of components for simplification of sorting,
reduction of different disassembly procedures to be undertaken and for making recycling and
reuse more attractive due to greater numbers of fewer components (Crowther).
Simply supported beams should be used in concrete structures to make disassembly possible
and avoid damaging during deconstruction; for steel structures this is not to be considered as
connections are mainly foreseen in the frame corners and therefore the full potential can be
capitalized.
Roof structures are generally suitable for reuse after deconstruction if made of truss or
framework systems. After dismantling, they may be reassembled in their original
configuration or in alternative layout.
Design modular structures (it is possible to built ca. 80% of high-rise buildings industrialized
with a high degree of prefabrication with savings of 10% costs and 20% time during
construction (FOSTA)) thus, the use of dimensionally and functionally compatible
components and materials is supported.
With regard to recycling after deconstruction, the use of structural systems that are easier to
deconstruct and to demolish is recommended (elementary and /or modular construction).

346
Sequential/Partial deconstruction process should be taken into account during design of the
structure (partial reuse of building / modular design).
Overall, the smaller and less complicated the elements are, the better their potential for reuse
(Nordby 2009). Further handling becomes easier during assembly, disassembly, transport,
reprocessing, and re-assembly.
3.5 Principles for connection design for deconstruction
A primary way to ensure easy deconstruction and increase the recycling/reuse rate is seen in
appropriate design of connections. Further, connections are being responsible for the lions
share of design, fabrication and erection issues (for information: from the total costs for a
structural steel work [without transport] the approximate share of costs is as follows: 48% are
materials costs, 20% fabrication, 17% erection, 10% design and 5% painting) and thus, the
necessity and pressure for an appropriate connection design becomes obvious. The following
should be considered for connection design for deconstruction:
Connection of element should facilitate dismantling and mutual independency between the
elements should be considered.
Use of standard and reusable fixing should be considered (e.g. bolted connections instead of
welded joints for steel structures, frictional fixations [best, as no material penetration] or
expansion sealant).
Connections as cast joints, glued fixation and elastic sealant should be avoided.
Steel construction provides the largest variety and flexibility of dismountable connection
systems and is therefore most appropriate for the reuse of elements.
Timber structures may easily be disassembled (large down-cycling potential of timber
through many generations is present) if mechanical connections instead of glued joints are
used. With regard to recycling, screws are to be preferred to nails, which may remain
embedded in the timber during separation.
For concrete structures steel end fixing and bolted connections are to be preferred to grouted
connections to facilitate reuse. Elements with damaged connections can hardly be repaired
for reuse. Prototypes of such kind of dismountable steel connections for prefabricated
concrete elements are available in literature, e.g. see (Vambersky 1994).
Non-varying connection systems (e.g. only one type of bolts) should be used.
Easy and permanent access to connection should be guaranteed.
Further, the stronger and continues the composite action, the more difficult it is to separate;
this applies for e.g. sandwich panels with PU core, fiber reinforced concrete, etc. Further
recyclable material may be contaminated by a small amount of a foreign material which can
not be completely separated.
An example of an innovative steel frame connection is the Quicon System, used for the first
time in 2002 to construct a mezzanine floor within a warehouse unit in Dartford (SCSSC 2003).
With this connection typology it is possible to reduce the construction programme by 50%
compared to traditional connections. The Quicon system includes a slotted T-piece, which is
welded to the main column or beam in the factory. On site, the T-piece is attached to the
supporting beam or column via a series of short studs attached to the beam. Figures 4 show the
beam-to-column connection.
An overall structural system has been developed by the ArcelorMittal Group including the
Quick erect system (AOB) (see Figure 5). By means of cut-outs in the end plates, the beams are
fitted directly onto threaded rods passing through the columns. This connection has been
developed for the Slim Floor system being a fast, innovative and economical solution which
marries prefabricated concrete slabs with built-in steel beams. An important feature of the
design is a special kind of girder with a lower flange wider than the upper flange. This
arrangement enables to fit the floor slabs directly onto the lower flange plate of the beam, so
that the two constituents make up the floor. Created to eliminate beam downstands at the level
of the floor slabs, this reliable and economic prefabricated component unit gives the architect
new scope for imagination and guaranteed economy for working spans of up to 8 meters, in
composite action even up to 14m (Hechler et al 2010). The reduced height of the slabs and the
advantage of a considerable degree of fire resistance without any protection provide a maximum

General

347
freedom in creativity and the structural elements are easy to separate for deconstruction and
recycling.

Figure 4. The Quicon system (SCSSC 2003). Figure 5. Quick Erect System for slim-floor beams.
A number of pre-cast concrete building systems designed for disassembly are also available,
e.g. MXB-5, Bestcon-30, CD-20, Moducon-2000, and SMT systems (Portioli & Ungureanu
2008). These systems have mainly been developed in the Netherlands. In principle all systems
aim on using a limited number of types of pre-cast concrete components to create a large
enough stock of compatible components that may attract interest for reuse in the future.
For further examples and investigations it is referenced to e.g. ((Lindapter), (Sassi 2008),
(Durmisevic&Brouwer 2002)).
3.6 Principles for foundation design for deconstruction
The foundations are the basis of any construction. For their design for deconstruction the
following should be considered:
Pre-cast concrete pad foundations or steel screw piles should be preferred for easy removal
and reusing.
Steel and pre-cast concrete driven piles can be used for deconstruction.
Sheet piles are commonly used as temporary retaining structures a stock market for reused
element already exists.
Generally, in-situ reinforced concrete piles are difficult to remove and separate into base
materials and thus should be avoided even for recycling.
Design foundations to allow for potential vertical expansion in lieu of demolition.
3.7 Example in design for deconstruction Parking Marignane
One recent example of the economic and sustainable design for reuse is given by the Parking
Marignane in Marseilles, France. The overall parking concepts consists of 2 parkings with
5000m
2
each, providing parking space for 1000 places each on a ground floor and upper deck
(overview see Figure 6, A). Main aim during design has already been to be deconstructed and
reused at least one of the parkings for the airport of Marseilles in the future.
Consequently a steel structure was been designed to enable easy and non-destructive
dismounting. The columns are circular hollow sections and the beams IPE 300 (see Figure 6, B).
All steel members were galvanized for corrosion protection. The galvanization was chosen to
provide robustness of the protection during use, dismounting, transport and 2
nd
life erection. As
slab COFRADAL 200 elements were chosen to design a light slab system with a high degree of
prefabrication, as easy and fast to be placed (see Figure 6, C).
Additionally focus has been put on the connections used. All connections are bolted and
therefore detachable (see Figure 6, D and E). Further, the connection column beam was
designed with beams on top of column to assure easy dismounting. Further the connection slab
beam is detachable (see Figure 6, E); the slab elements were attached by clamps. Finally, also
the installations are fixed by detachable connections (www.duplipark.com).

348
A

B
C

D
E

F
Figure 6. Design and construction of the Parking Marignane, Marseilles, France.
Further case studies with various construction types are given in references (NAHB 2000),
(Guy & Timothy 2003), (Seattle 2006).
4 REUSE OF ELEMENTS AND MATERIALS
4.1 General

Recovery and reuse of elements and/or materials
2
nd
life is normally the best environmentally
sustainable option, especially on a local level. The
only energy input required is the costs for
transportation to the new site (Webser 2005). In this
context a new hierarchy for managing the reuse of
building materials is created: Repeat, Rethink,
and Renew", see Figure 7 (Guy & Timothy 2003).
Figure 7. Design for Reuse Hierarchy.
Repeat refers to the direct reuse of a material in new construction exactly as it was used
previously; a wall stud becomes a new wall stud, flooring becomes new flooring. It involves the

General

349
minimal additional energy expenditure and is consequently the most resource efficient of the
reuse methods.
Rethink involves the reuse of a material, without significant modification, in a manner that is
different from its intended reuse, and consistent with its inherent properties (utilization based
upon its performance and aesthetic properties). This method allows for more flexibility in the
design for reuse, including where damage to the material makes it only useable in a different
way than its original use.
Renew describes the combination of new materials and salvaged materials, or significant
processing, in order to bring about successful reuse. This is particularly applicable when the
used material will be covered or have additional treatments.
In the following reuse is introduced with differentiation in reuse of materials, elements and
constructions.
4.2 Reuse of materials
The largest benefit in the strategy for resource saving is localized on the level of materials
today; besides the ensuring of a long lifetime of the building, focusing on the reusability of
materials is important - partly by integration in a demountable composition of building elements
with the least possible deformation in processing, partly by choosing materials which can be
reused without new processing (second life), for instance bricks and lime mortar.
In particular ancient materials, not produced anymore, do have a large potential for reuse.
One example is the reuse of large cobblestone, expensive in production and not common.
Especially historic centers are re-developed today in their former design creating a demand for
cobblestone. One important source for cobblestone is thus reused elements from
deconstructions, see Figure 8.

Figure 8. Cobblestone pavement (left top: old pavement cobblestones for recovering; left center:
different cobblestones for landscaping; left bottom: old cobblestone reused in street marking; right:
Reuse of recovered cobblestone for refurbished church place Heiligkreuz, Trier, Germany, 2010.

350
4.3 Reuse of elements
The reuse of elements is already well established for certain structural products. Especially steel
products provide the opportunity to be reused as steel construction is characterized by discrete
member design with dismountable connections. Composite sandwich cladding panels are e.g.
reused to 37% (53% recycling rate, 10% landfill), road barriers have a reuse rate of 34% (65%
recycling rate, 1% landfill). Structural sections are by now reused to 11% (87% recycling rate,
2% landfill). An overview on the reuse / recycling rates for steel products is to be found in
(Sansom & Meijer 2002).
On the basis of the reuse experience steel products the need of a stock of standardized
products (standard length, selection of profiles, standard grades, etc.) with certified properties
for structural products becomes obvious to create 2
nd
life material market. Even the creation of a
rental business has already been started for e.g. sheet piles for excavation pits (rent a pile).
These sheet piles can be reused about 10 times with minor repair works.
An example on the reuse of sheetpiles is the Vienna tube extension, Line 1, constructed from
2001 2003, opened in 2006 (RPS 2003). The construction works were carried out in segments.
As start for each segment, the sheetpile cofferdams were driven to the design depth using a
vibro-hammer mounted on a leader (see Figure 9, A) (in parts with assistance of jetting to ease
installation and later by guide holes for difficult driving conditions). Once the sheetpile
cofferdam had been driven, struts were installed (see Figure 9, B) and the soil within the
cofferdam was excavated (see Figure 9, C). At final depth, the trench was sealed by a concrete
plate, the ground water was pumped out (see Figure 9, D)and the tunnel box section for the tube
was constructed with water proof concrete in dry conditions (see Figure 10, E). Finally the struts
were removed, the trench backfilled and the sheetpiles were removed for reuse in the next
segment (see Figure 10, F).

A

B
C

D

Figure 9. Working steps A-D for the Vienna tube extension, Austria, 2003.

General

351
E

F

Figure 10. Working steps E and F for the Vienna tube extension, Austria, 2003.
4.4 Reuse of structures
The relocation of complete structures is already established in selected sectors of the
construction markets, e.g. for temporary structures (e.g. rental service of temporary bridges),
portable buildings, and industrial steel halls. Well known are the examples of the complete
reconstruction of industrial facilities reused in countries with high demands and market growth
(this must however not be a sustainable benefit from reuse).
Excellent examples for the reuse of structures are currently to be found from the Expo 58.
Overall 9 structures have been reconstructed after the exhibition (Infosteel). The footbridge to
the German pavilion has been reused even twice, until its present location in Duisburg,
Germany, see Figure 11.

Figure 11. Footbridge to the German pavilion (left: EXPO58; right: reused in Duisburg, Germany, 2008.

The reuse of buildings requires an upgrade to building standards meeting the demands at
present. The buildings reused from the EXPO58 have therefore been refurbished, not only in
terms of service installation but also faade design and performance. It is evident that, in most
of the examples, the architecture was still attractive to buildings owners of today, see Figure 12.

352

Figure 12. Restaurant of Cz pavilion (left: EXPO 58, right: reused as Office building in Praha, Cz, 2008.

The Christ Building for the EXPO2000 in Hannover was designed in the tradition the
EXPO58, see Figure 13. From the beginning the reuse of this building was considered. Today
the former exhibition pavilion serves as a church part of the extension of the cloister in
Volkenroda, D and partly as a laboratory at RWTH Aachen, Germany.

Figure 13. Christ pavilion (left: EXPO2000, right: reused in monastery in Volkenroda, Germany.
5 DISCUSSION AND CONCLUSIONS
Large quantities of C&D waste are resulting in a significant burden on the Industry in general.
Therefore new approaches in design and construction need to be found to increase recycling and
reuse rates design for deconstruction may contribute to an integral approach in construction.
At present, however, there are two obstacles to broader application of DfD: Buildings
deconstructed today are at their end of lifetime. Unfortunately, at the time when those buildings
were created, end-of-lifetime scenarios were not discussed yet. Thus, many materials that were
used have been classified as carcinogenic including asbestos, some older kinds of glass wool or
tar cork. These old materials decrease the rate of possible recycling and create the need for
modern disposal sites.
At present energy efficiency is one leading aspect of modern buildings. To decrease
construction, pre-cast segments are preferred. Sometimes heat insulation is surrounded by
concrete on both sides of the pre-cast segments. Even parts of modern frontage elements are
made of heat insulation with attached iron sheeting. It is clear that at the end of lifetime it will
be difficult to detach and separate the materials from each other. How can we efficiently get the
attached iron back from insulated frontage elements? Besides the contamination problem of

General

353
earlier products, the modern sandwich materials are even more difficult to separate and avoid
higher recycling rates in the future. At last, the rate of synthetic materials as polystyrene rises.
The need for energy efficiency contradicts the zero-waste deconstruction.
Overall, the outline is appearing of a new rationale for the handling of resource-consumption
related to building, according to which the involvement of process energy will have far-reaching
consequences for architectural/engineering practice and priorities. Architectural and structural
design based upon lifecycle considerations demands far more holistic strategies than needed for
saving operational energy only. These strategies pose a new framework for the
architects/engineers freedom to design which is in many ways more restrictive but is
simultaneously containing a potential for architectural renewal, and possibly a change in the
cultural practice of coding and de-coding building identity.
With the experience, that changes are mainly stimulated by changes in the legislation, politics
are urged to define these future ways to cope with the present scenario. A requirement for
design for deconstruction may however result in an immediate confrontation with the concrete
regiment which is dominating the building industry, building code and building practice in most
countries.
If we look at concrete construction today, concrete building elements tend to be heavily
integrated by being assembled in cast joints with reinforcement locks. Thus, conversions are
hampered, flexibility is weakened, and reuse of building elements is made impossible. Concrete
building demands a high process energy use.
However, crushed concrete is widely used as fill and to a minor extend (6-8% (Recycling
Concrete 2008)) as secondary aggregate in low-strength concrete. Although, a maximum of
20% recycled material can be included in new casting processes and thus a large quantity of
waste prevention is still not idle.
Steel Construction on the other hand has high recycling and reuse rates (e.g. recycling rate of
87% and a reuse rate of 11% for heavy sections (LCA for steel construction 2002)). As such
steel construction is considered to be the most potential way to design for deconstruction. These
shares are mainly stimulated by an economic market for steel scrap and the structural member
character of steel construction itself. Though the DfD movement recognizes that steel is well-
suited to being salvaged such that a larger proportion of the steel can be reused from one
building to the next (Webster 2007) and seeks strategies for appropriately designing steel
structures to ensure the steel elements can be reused, not recycled. Also, in order to minimize
carbon-footprint, the combination of wooden building parts assembled with steel connections
has a future potential.
To conclude it has to be stated that, in order to facilitate best deconstruction practice in the
future, we must practice design for disassembly now.
6 ACKNOWLEDGEMENTS
The authors wish to express their deep gratitude to the partners of COST C25 Sustainability in
Construction Integrated Approach to life-time Structural Engineering, especially to Ruben
Paul Borg (University of Malta), Francesco Portioli (University of Naples, Italy), Viorel
Ungureanu (University of Timisoara, Rumania) and Paul Kamrath (Paul Kamrath
Abbruchtechnik und Entsorgung GmbH, Dortmund, Germany) for their inspiring cooperation.
REFERENCES
A guide to deconstruction An overview of deconstruction with a focus on community development
opportunities complete with deconstruction project profiles and case studies. 2000. NAHB Research
Center, Inc., Upper Marlboro, MD.
Addis, W. & Schouten, J. 2004. Design for deconstruction. Principles of design to facilitate reuse and
recycling. London: CIRIA.
AOB 2009. ArcelorMittal, Long Carbon Europe, Sections and merchant bars. ArcelorMittal Office
Building in Esch/Alzette. Version 1-2009.
ArcelorMittal, Long Carbon Europe, Commercial RPS. 2003. Vienna Tube Extension.

354
Brand, S. 1994. How Buildings Learn and what happens after they are built. Penguin.
Crowther, P. Design of Buildings and Components for Deconstruction. Deconstruction: Techniques,
Economics, and Safety. Charles J. Kibert, Abdol Chini, and Charles Hendriks.
Directive 2008/98/EC of the European Parliament and of the Council of 19 November 2008 on waste and
repealing certain Directives
Durmisevic, E. 2006. Transformable Building Structures, Doctoral thesis TUD.
Durmisevic, E. & Brouwer, J. 2002. Design aspects of decomposable building structures. Proceedings of
the deconstruction and materials reuse conference, Karlsruhe.
Environmental product declaration - Structural Steel: Sections and Plates. Declaration number EPD-Bfs-
2010061-EN. Institute for construction and environment. www.bau-umwelt.com. 2010
FOSTA-Forschungsvorhaben Komponentensystem in Stahlbauweise (Nr. 753/2007)
Guy, B. & Ciarimboli, N. 2006. DfD Design for Disassembly in the built environment: a guide to
closed-loop design and building. City of Seattle, WA, and Resource Venture, Inc. by the Hamer
Center for Community Design. The Pennsylvania State University.
Guy, B. & Timothy, W. 2003. Design for deconstruction and reuse. Final Report. Powell Center for
Construction and Environment, Florida, USA.
Hagen, K. 2007. Deconstruction as an Alternative to Demolition Helping the Environment, Creating
Jobs, and Saving Resouces. www.associatedcontent.com.
Hechler, O. & Braun, M. 2010. New structural concepts in high-rise buildings. High-rise towers and tall
buildings. Design and construction of safe and sustainable high-rise structures. International
Conference, Munich, 2010.
Hettinger, AL., Thomas, JS., Hechler, O., Conan, Y. 2011. Life Cycle Assessment of Composite Bridges.
Final Conference. COST C25: Sustainability in Construction Integrated Approach to life-time
Structural Engineering, Innsbruck. 2011.
Kamrath, P & Hechler O. 2010. Demolition and recycling of demolition rubble after deconstruction.
Final Report WG3. COST C25: Sustainability in Construction Integrated Approach to life-time
Structural Engineering, 2010.
Lindapter. Steelwork Fixing. http://lindapter.com/PDFs/steelworksection.pdf
Morgan, C. & Stevensen, F. 2005. Design and Detailing for Deconstruction - SEDA Design Guide for
Scotland. Scottish Ecological Design Association.
Nordby, A.S. 2009. Salvageability of building materials. Doctoral thesis NTNU
Portioli, F. & Ungureanu, V. 2008. Demolition and deconstruction of building structures. COST C25
First Seminar: Sustainability in Construction Integrated Approach to life-time Structural
Engineering, Dresden.
Sansom, M. & Meijer, J. 2002. Life-cycle assessment (LCA) for steel construction. European
Commission, Directorate-General for Research. ECSC Final Report 20570. ISBN: 92-894-4637-4.
SCSSC 2003. Sustainable steel construction Building a Better Future. First annual report on progress.
UK.
Sassi, P. 2008. Defining closed-loop material cycle construction. Building Research & Information, v.36,
n.5.
SRI - The Steel Recycling Institute. http://www.recycle-steel.org . 2010-08-11.
Storey, J., Gjerde, M., Charleson, A., Pedersen, M. 2003. The state of deconstruction in New Zealand
2003 Synopsis. Centre for Building Performance research, Victoria University, New Zealand.
Vambersky, J.N.J.A. 1994. Precast concrete in buildings today and in the future. The Structural
Engineer. Vol. 72 No 15: 237-242.
Webster, MD. 2007. Structural Design for adaptability and deconstruction: A strategy for closing the
materials loop and increasing building value. Proceedings of the 2007 ACSE Structures Conference,
Long Beach, USA.
Webster, MD. & Costello, DT. 2005. Designing structural systems for deconstruction: How to extend a
new buildings useful life and prevent it from going to waste when the end finally comes. Proceedings
of the 2005 Greenbuild Conference, Atlanta, USA.
World Business Council for Sustainable Development. 2008. The Cement Sustainability Initiative -
Recycling concrete.
WSA - WorldSteel Association. http://www.worldsteel.org. 2010
www.corusconstruction.com/en/sustainability/thermalmass/
www.duplipark.com. 2010
www.infosteel.be

General

355
P. Kamrath
Paul Kamrath Abbruchtechnik und Entsorgung GmbH, Dortmund, Germany
O. Hechler
ArcelorMittal, Commercial Sections S.A., Esch-sur-Alzette, Luxembourg
Demolition and recycling of demolition rubble after
deconstruction
1 INTRODUCTION
1.1 Recycling of construction and demolition waste
The total production of construction and demolition waste (C&DW) in the European Union is
about 450 million tons per year and represents the most part of waste stream (WMU, 2000). If
earth and excavated road material are excluded, the amount of generated construction and
demolition waste is estimated to be roughly 180 million tons per year, which represents 480 kg
per person. Roughly 75% of waste is landfilled, despite its major recycling potential. This
means that more than 3 million tons per year have to be occupying existing landfills.
In Table 1 the construction and demolition recycling rates (apart from earth and excavation
waste) in the 15 Member States are shown (SYMONDS, 1999). The wide variety in the
management of that waste within the European Union is to be linked with specific waste
management legislation. Those countries (in particular the Netherlands, Denmark and Belgium)
with measures to improve its management have achieved high levels of recycling, thus
demonstrate the recycling ability and potential of construction and demolition waste.
Table 1. Produced and recycled C&DW in EU (SYMONDS, 1999)
________________________________________________________________________________________ _______
Member State Core C&DW Arising % Re-Used or Recycled % Incinerated or Landfilled
_________________________________________________________________________________________ _______
Germany 59 17 83
UK 30 45 55
France 24 15 85
Italy 20 9 91
Spain 13 <5 >95
Netherlands 11 90 10
Belgium 7 87 13
Austria 5 41 59
Portugal 3 <5 >95
Denmark 3 81 19
Greece 2 <5 >95
Sweden 2 21 79
Finland 1 45 55
Ireland 1 <5 >95
Luxembourg 0 n/a n/a
EU-15 180 28 72

356
The south European countries (Italy, Spain, Portugal, Greece) recycle little of their
construction and demolition waste. Their natural resources are of a sufficient quality and
quantity to meet the demand for building materials at a moderate cost. The market for recycled
materials is not highly developed in those countries.
To reduce the impact of buildings on the environment, in recent years different strategies
have been developed that are mainly based on prevention policies and improved construction
and demolition waste management. In Germany for example, the recycling rate has increased
for the year 2005 to 68,5% according to statistical information. KWTBau (2005) determined,
that from the 52.1 Mt corresponding to the total amount of building demolition waste, 35.7 Mt
were recycled. 7.0 Mt (13.4%) were used in mining installations and 3.8 Mt (7.3%) were
directly used by public authorities. 1.6 Mt (3.1%) were used for landfill construction and only
4.0Mt were landfilled (Figure 1) and thus not reused or recycled.

Figure 1. Composition of C&D waste in Germany. Building demolition waste has an amount of 52,1 Mt,
general excavated material is roughly 3 times as much with 140,9 Mt.
1.2 Construction and demolition waste management strategies
A description of the life cycle of materials is provided by the Delft Ladder (Figure 2, left). In
this case, 10 separate processes are identified for the life cycle of materials, including both
material cycle and building stages. On the basis of the Delft Ladder, different actions can be
undertaken to use materials and elements at the highest possible level during their life span,
reducing waste production and disposal to landfill (D&D, 2008). To aid thinking about
management of waste, a waste hierarchy has been developed, where the different options are
ranked from prevention to disposal (Figure 2, right). The waste hierarchy is a reduced and
simplified version of the Delft Ladder.
Generally, different design strategies can be adopted for the sustainable management of
construction and demolition waste according to building stages.
The first stage is the prevention of waste by design. Waste can be avoided by designing a
building system that allows dismantling and reassembly or designing a building component that
allows the materials to be recycled. The waste reduction corresponding to the two first levels of
waste hierarchy, i.e. prevention and minimisation, has two components, corresponding to the
reduction of the amount and of the hazard of waste produced.
The other actions, according to waste hierarchy, are the reuse of elements and recycling of
materials. Elements removed from a building should be improved by maintenance and

General

357
refurbishment and reused for their original purpose in a new situation. Waste materials from
production processes can be collected and improved to make them suitable for returning to the
production process. In particular, C&DW can be recycled, down cycled or up cycled
(Dorsthorst & Kowalczyk, 2002). When the material is used for the same function again, it is
called recycling (steel scrap used for the production of steel). When the material is used for
another function it is called down cycling (mixed granulates used as a road base material) and
when the recycled material is used for a better function than the original material it is called up
cycling (fly ash used in cement or concrete).
On the basis of different levels of reuse, two main design strategies can be identified for
building structures: Design for Adaptability (DfA) and Design for Deconstruction (DfD). DfA is
suitable for constructions having a longer lifetime than its function. So buildings must be easy
to adapt to other functions, in order to create a longer lifetime. DfD is aimed to reusing and
recycling of elements and materials respectively.

Figure 2. The Delft Ladder (Kristinsson et al., 2001), left. The waste hierarchy (Sligo Counting Council,
2004), right
2 DECONSTRUCTION OF BUILDINGS AN OVERVIEW
2.1 Deconstruction in the 2nd half of the 20th century
The deconstruction of buildings became part of the construction industry in the 2
nd
half of the
20
th
century, when buildings and industrial sites of the time of the industrial revolution reached
their end of lifetime. Simultaneously, the needed machinery such as excavators and trailer
trucks became available for medium-sized companies as well as small sized deconstruction
companies. The scope of duty was mainly the complete demolition of a building. Up to the
1970s the complete demolition rubble was deposited on dumpsites. Thus, the waste was not
reused or recycled. Even contaminations were not strictly handled as hazardous material.
Deconstruction was not so much a matter of knowledge but of logistics only.

358
2.2 Modern deconstruction concept with four phases
In our days, the deconstruction is a much more sophisticated task. In general, a typical concept
differentiates four main tasks, which are (Figure 3):

1. Decontamination of the building
2. Core removal of all in-house dry constructions
3. The deconstruction by machines
4. Disposal and / or recycling

While earlier deconstruction concepts do not consider any possibilities of reuse or recycling of
the debris, the modern concept with four phases separates the debris at any stage of the
deconstruction. Nevertheless, the recycling-rate depends on the used materials of the
construction (Rbenack et al., 2007).
During the decontamination phase (Figure 4, left) all pollutants will be removed from the
building. The typical pollutants are cement asbestos, tar-bitumen roof sheeting and
polychlorinated biphenyles fillings in jointings. All materials are not allowed to use anymore.
Besides, man-made fibres and wood painted with old preserver colors need to be separated
during the decontamination phase, too. In general, all pollutants have to be deposited on special
dumpsites. Because of the high thermal energy rate of roof sheeting and wood, waste
incineration as the lowest level of reuse is preferred instead of deposition.

Figure 3: Phases of deconstruction: Simple deconstruction model in two phases without any recycling or
contamination management (left) and modern concept with four phases (right). The modern concept
concerns sorting of waste into recycling fractions, contaminated waste and disposal of mixed waste.

General

359

Figure 4: Decontamination phase: Removal of cement asbestos (left) and core removal (right). At the end
of phase II (core removal) the building is prepared for deconstruction with machines.

Figure 5: States of concrete during recycling progress: Concrete as given after deconstruction by
machines (left), separated core wires (middle) and gravel of concrete (right)
High recycling rates are possible only, if separation starts already before the deconstruction
phase with machines as e.g. excavators (Wei, 2002). Especially wooden materials and plastics
tend to split in small pieces, if not separated. Thus, a poor quality of the recycled gravel of
stones by extraneous material should be avoided. In the core removal phase all non-mineral dry
constructions will be removed out of the building and separated. Typical materials are wood
(doors, frames, etc.), metals (water pipes, electrical lines) but also mixed waste (carpets,
plastics, etc.).
Deconstruction by machines is the phase, when the structure of the building (i.e. concrete,
steel constructions, walls) are teared down by excavators, cranes and other machines. In this
phase, depending on the structure, mostly mineral debris or steel arises. While recycling or
reuse of a steel construction is generally preferred, the recycling rate of concrete and stonework
depends on the quality of the material. Pumice stone is not applicable for recycling, because the
strength of the material does not suite the requirements for gravel. Concrete needs to be crushed
to separate core wires, which else could damage recycling machines (Figure 5).

360
3 DEMOLITION OF BUIDLINGS
3.1 Demolition techniques
The practices involved in the demolition of buildings strongly influence the reuse and recycling
of components and materials (D&D, 2008). The demolition industry has undergone major
transformation within the last 20 years (Hurley & Hobbs, 2003). Traditionally it has been a
labour intensive, low skill, low technology and poorly regulated activity, dealing mainly with
the disassembly and demolition of simply constructed buildings (e.g. from masonary). Common
methods have been by hand (with portable tools) and pulling (with e.g. ropes).
More recently, mechanized processes with specialized equipments replacing manual labour
have been invented. This trend results from the increased complexity in building design, the
financial pressures from clients, health and safety issues, regulatory and legal requirements and
advances in plant design. Traditionally, much of the demolition contractors income was from
the sale of salvaged and recycled materials. Today income is mostly generated from the contract
fee - demolishing as quickly and as safely as possible. Nevertheless, substantial amounts of
materials and components are recovered or reclaimed but, for down-cycling, is not used to its
fullest potential.
Todays demolition process relies on one of eight basic methods: pulling, impact, percussion,
abrasion, heating (or freezing), expanding, exploding or bending. Most of all deconstruction
work is done by excavators (82%, table 2). Wrecking balls are effective against masonry, but
are less easily controlled and often less efficient than other methods. Newer methods may use
rotational hydraulic shears and silenced rock-breakers attached to excavators to cut or break
through wood, steel, and concrete. The use of shears is especially common, because controlling
the deconstruction process is important to avoid unstable states or unwanted demolition of
neighboring houses. Two different kinds of sheers are used depending on the main material:
Scrap metal shears are able to cut steel or wood highly controlled and cracking shears to crack
concrete. The needed oil pressure varies up to 320bar. If cracking shears can not be used
which is possible especially for parts of the foundation, hydraulic hammers are used. The two
disadvantages are the sound level and the implicit vibrations. General waste is separated with
sorting grapples (Figure 6). However innovative techniques like expansion/bursting, either static
(buster with wedges or chemical expansion agent) or dynamic (explosion, high pressure
water/gas and CARDOX), or abrasive methods (hammer drills and diamond boring machines)
are available. They are designed to reduce noise, dust and increase the safety of the demolition
site. The size of an excavator used for deconstruction depends mainly on the height of the
building. If the height over ground is more than 15m, high-reach or longfront excavators are
used. With those types of machines, buildings up to 50m can be deconstructed by heavy
machine use.

Table 2. Main techniques of building demolition and deconstruction
in Germany (Lippok and Korth, 2004)
Main techniques (Usage quota in Germany, 2002)
Deconstruction by excavators with shears etc. 82%
Exploding 4%
Deconstruction with wrecking balls 3%
Other machines 3%
Percussion, abrasion, heating etc. 3%
Robots 0.3%
Other 1.7%

General

361

Figure 6: Use of a cracking shear on a longfront excavator. In opposite to cutting shears for steel the
concrete is cracked by high pressures. The structure of the concrete will be destroyed. With this
technique, the cracking shear cuts through the concrete.
The selection of demolition method is dependant on several factors concerning the physical
aspects of the building to be demolished, with safety and economic issues, i.e. the location of
the building, the type of structure and materials involved, the space available on site for
segregation and storage, health and safety of operatives undertaking demolition work, permitted
levels of nuisance and, finally, on the time and money available. The development of new tools,
techniques and working practices is hereby an essential pre-requisite to deconstruction and reuse
of structural elements.

362
3.2 Safety measures during the phases of deconstruction
The demolition respectively deconstruction process has to be well planned to minimize the risks
of:

- causing damage to persons and properties of the public,
- endangering the health and safety of site personnel and
- damaging the neighborhood environment (HKBD, 2004).

To avoid most risks and guarantee a high safety level, the deconstruction process has to be
planned quite as exactly as any building process. Especially the work during the 3
rd
phase (the
deconstruction by machines) needs a structural analysis of the building and a concept how to
perform the deconstruction. The most important planning process is to define stable states of the
partly destroyed object and to ensure, that no instable states occur at the end of a day.
Damage to persons and properties of the public needs to be strictly avoided. Consequently
suitable hoarding may be required to be erected to enclose the site during demolition. Also
safety measures for site access need to be considered. A person carrying on demolition
operations should make sure that the construction which is partly demolished (and the site
thereof) is so far as is reasonably practicable, properly secured or closed against entry at all
times.
The health and safety of site personnel should be assured during the whole deconstruction
process. Next to job safety trainings of deconstruction workers the communication on the
demolition planning to the workers is essential. Appropriate insurances need to be provided.
Apart from the training the equipment, protection and site installation need to be appropriate for
the demolition works. Current trends for demolition techniques consider safety on demolition
sites as major factor, especially for steel and concrete structures. The track loader and crawler
crane and drop ball have been replaced by excavators with long reach booms to distance the
operator from the building being demolished, consequently taking personnel out of potentially
dangerous situations. Specialist tools and attachments to excavators and mini-excavators are
additionally providing operation safety.
The risk of damaging the neighborhood environment needs to be minimized. Thus the
concept of 4 main phases have to be planned to avoid:

1. Air pollution: Concrete breaking, handling of debris and hauling process are main
sources of dust from demolition. Dust mitigation measures complying with the Air
Pollution Control (Construction Dust) Regulations of the specific country shall be
adopted to minimize dust emissions. Burning of waste shall not be allowed. Diesel
fumes generated by mechanical plant or equipment shall be subject to the control of the
Air Pollution Control (Smoke) Regulations.
2. Noise: Noise pollution arising from the demolition works including, but not limited to,
the use of specified powered mechanical equipment (SPME), powered mechanical
equipment (PME), such as pneumatic breakers, excavators and generators, etc.,
scaffolding, erection of temporary works, loading and transportation of debris, etc.
affects the workers, and the sensitive receivers in the vicinity of the demolition site.
Silent type PME shall be used to reduce noise impact as much as practicable.
Demolition activity shall not be performed within the restricted hours as established.
3. Hazardous Materials pollution: Of course measures have to be taken to avoid the escape
of hazardous materials in the environment as dust or suspensions. Thus,
decontamination should be preliminary to the main deconstruction phase.

Further environmental risks are the discharge of wastewater from demolition sites, which
require a valid discharge license, and possible traffic disturbances: Deconstruction needs to take
place without essential interference to the traffic. Especially for the deconstruction of bridges or
buildings close to main streets the increase in accident potential has to be addressed, which
however is controlled due to adequate traffic control measures (see Figure 7).

General

363
4 POSSIBILITIES AND LIMITATIONS OF RECYCLING
4.1 Types of construction and demolition waste
Wastes are defined as residue, substances or materials, generated by the production,
transformation or usage which have been or are planed to be abandoned. Final waste is a waste,
which is not treated anymore e.g. in France only final waste can be landfilled (FRANCE,
2009).
The aspect concerning the construction and demolition waste (C&D waste) includes concrete,
stones and dirt generated during excavation (sometimes collectively referred to as fill material
or rubble), as well as asphalt, wood (treated, painted and clean), metal (ferrous and non-ferrous),
and miscellaneous materials (DDC, 2003).
The deconstruction materials are generally considered as waste. During construction,
renovation and demolition activities, you may produce one or more of the following types of
residuals, which are Clean fill, Recovered materials, Regulated construction and
demolition waste, Hazardous materials and hazardous wastes (e.g. asbestos-containing
materials).
Clean fill e.g. is uncontaminated soil, rock, sand, gravel, concrete, asphaltic concrete, cinder
blocks, brick, minimal amounts of wood and metal and inert (non-reactive) solids. When
specified to be uncontaminated by e.g. metal-based paints including lead and other heavy metals
these materials can directly be used for fill, reclamation or other beneficial use.
Recovered Materials are those removed for reuse and those removed to be recycled into new
products. Potentially recyclable construction and demolition wastes may include scrap metals,
asphalt shingles, sheet rock, lumber, glass and electrical wire.
Regulated construction and demolition wastes are those not classified as clean fill and not
being reused or recycled. Regulated non-hazardous construction and demolition wastes must be
disposed of at a permitted landfill or transfer station and are regulated by law.

Figure 7: Deconstruction close to a main street. For safety reasons, the street is mostly closed for traffic.
In consequence, the whole traffic around the project side is temporally redirected.

364
4.2 Regulations by law
In general, the government of any European country tries, to keep the amount of waste little. In
Germany this aim is regulated in the Kreislaufwirtschaftsgesetz (German law that generally
claims the ReduceReuseRecycle approach) and the TA Siedlungsabfall (technical aspects
of the preliminary law, see (Knpple, 2007) for example). Other European countries may have
similar regularities. The guideline of this laws are, to a) avoid waste, b) minimize waste, c)
reuse the waste and finally, if there is no other possibility d) deposit the waste. Thus, there are
strong regularities. Nevertheless, some of the waste can not be recycled but only deposited. The
reasons are, that there is no need for the materials anymore (e.g. pumice stone), contaminations
(asbestos) or missing separation techniques (e.g. roof sheeting with attached fibre). But even if
reuse or recycling is not possible, some materials can be used as energy sources.
4.3 ReduceReuseRecycle
Opportunities for reducing construction and demolition waste focus on three approaches,
typically expressed as ReduceReuseRecycle (DDC, 2003).
Reducing waste, the first approach, yields the greatest environmental benefits. Using less
material costs less, reduces pollution from its manufacture and transportation, saves energy and
water, and keeps material out of landfills. Waste reduction should be the top priority in waste
management plans. Therefore economic and sophisticated bridge design concepts are required,
e.g. presented in (AMCS, 2009).
Reusing, the second approach, extends the life of existing materials and decreases the new
resources needed. Of course, entire construction can be reused, e.g. through rehabilitation
respectively strengthening, whether for the same or new use, saving both resources and money.
This approach can be pushed further by not assuming that new always performs better. Reusing
means also, that the material can be used for the same purpose and in the same matter as it was
used before deconstruction. Reuse is possible for all metals, which can be remelted.
Recycling, the third approach again conserves resources and diverts materials from landfills.
Demolition and renovation projects, present numerous opportunities for recycling. The most
sustainable form of recycling converts waste into new products, such as scrap to new steel or
asphalt into new paving. Additionally, finding alternative uses for waste is a form of recycling.
Inert waste, such as concrete and brick, can be crushed and used as alternative daily cover for
municipal landfills, substituting for dirt, or wood scrap can be burned as boiler fuel. Gravel of
concrete or stone is a 2
nd
level recycling product. Thus, the former concrete can not be used for
the same purpose but at least as foundation of streets or pit fillings of new buildings. Wood of
good quality can be assimilated to chipboards.
The use as energy source is the lowest level of recycling. Wood as well as plastics and roof
sheetings are used as energy source. Only, if the use as energy source is impossible or
contamination rate is too high, disposal is allowed.
4.4 Classification parameters of contaminations of solid mineral debris
While all metals generally are not contaminated besides apparent attachments the mineral debris
needs to be analyzed. Especially mineral debris from industrial sites may have contaminations.
Typical possible contaminations are hydrocarbons, PAH and PCB for the solid material as well
as PH-values out of range or the conductivity being to high.
Result of the analysis is the classification of the mineral debris (see Figure 8). In order of the
classification the debris can be used as recycling material to e.g. refill the excavation pit or the
debris has to be deposited on a dump site. While the debris classified as Z0 or Z1 can be used as
filling material, all gravel with classification level of Z2 or greater may not be used on-site or
need additional requirements as for example the use above the groundwater level only (LAGA,
1997).

General

365

Figure 8: Classification of mineral debris. Up to level Z2 reuse of recycled debris is possible. Gravel or
debris of higher classification level needs to be deposited on special dumpsites (level I level III)
Table 3: classification parameters of possible contaminations of solid mineral debris by LAGA
Type Parameter Dimension Z0 Z1.1 Z1.2 Z2
arsenic mg/kg 20 30 50 150
plumb mg/kg 100 200 300 1000
cadmium mg/kg 0,6 1,0 3,0 10,0
chromate mg/kg 50 100 200 600
copper mg/kg 40 100 200 600
nickel mg/kg 40 100 200 600
mercury mg/kg 0,3 1,0 3,0 10,0
H
e
a
v
y

m
e
t
a
l

zinc mg/kg 120 300 500 1500
hydrocarbons mg/kg 100 300 500 1000
PAH mg/kg 1 5 15 20
PCB mg/kg 0,02 0,10 0,50 1,00
U
s
a
g
e

EOX mg/kg 1 3 10 15
Table 4: classification parameters of possible contaminations of eluate by LAGA
Type Parameter Dimension Z0 Z1.1 Z1.2 Z2
PH-value - 7 12,5 7 12,5 7 12,5 7 12,5
conductivity S/cm 500 1500 2500 3000
chloride mg/l 10 20 40 150
O
t
h
e
r

sulfate mg/l 50 150 300 600
arsenic g/l 10 10 40 50
plumb g/l 20 40 100 100
cadmium g/l 2 2 5 5
chromate g/l 15 30 75 100
copper g/l 50 50 150 200
nickel g/l 40 50 100 100
mercury g/l 0,2 0,2 1 2
H
e
a
v
y

m
e
t
a
l

zinc g/l 100 100 300 400

The scale ends with classification as LAGA Z5, which covers dangerous waste (e.g.
radioactive contaminated mineral debris). For the analysis, first the raw material is analyzed (see
Table 1). Second, the eluate will be analyzed. Mostly, contaminations follow the former use of
the former building (see Table 3 and 4) and vary between Z0 and Z2.

366
5 HOW TO DECIDE BETWEEN ON-SITE REYCLING AND OFF-SITE RECYCLING
5.1 Source separation
The construction and demolition waste industry operates in an environment where commodity
prices (i.e., metal, paper, fuel, etc.) and competing landfill disposal costs constantly change,
making it difficult to provide accurate information about recycling rates and costs. This also
prevents one from suggesting a typical approach. However recycling targets have been
defined by European law (2006/12/CE): On the horizon 2020 reuse and recycling of non
dangerous construction wastes must be higher than 70%, by weight. Source separation generally
yields the highest recycling rate and the best price for materials (DDC, 2003). The contractor
locates separate containers on the job-site, sorting out recyclable materials as they are collected,
and delivering each container to the processor when full. The best opportunities for source
separation tend to occur during the demolition and excavation phases of a project, when more
C&D waste is generated and one contractor is scheduling the work. Additionally, demolition
and excavation work can generate relatively homogenous waste streams. For example, asphalt
millings may be generated as a single material stream. During excavation, containers of clean
fill may be generated. There may be opportunities for onsite segregation of metal or for the
grinding, screening and reuse of concrete. On-site recycling techniques may reduce costs for
transportation but are not always useful.
5.2 Mass balance as indicator for on-site recycling
In the case that contaminations are little, the regulations by law and ecological as well as
economical thoughts tend to recycle the debris. Costs of deconstruction may be reduced, when
the mineral debris is not brought to off-site recycling sites but recycling becomes a part of the
deconstruction project and will be done on-site (Figure. 5). The advantage is that no transport
costs occur. On the other hand, a mobile recycling machine is needed. Besides the costs, the
most important aspect is to figure out, if the recycled concrete gravel can be used after the
deconstruction. It is possible, that removing an old industrial site leaves building pits which
have to be filled before rebuilding something new. Then, there is the need for the material. But
if the pit is part of the new building and mostly the gravel is not needed, on-site recycling is a
bad choice in the end there are costs for recycling on-site and transport costs as well as waste
management costs. During planning of a project, the masses of mineral debris can be estimated
by:

md cc
V V r = (1)
with V
md
as the volume of the mineral debris [m], V
cc
the volume of the cubic contents of the
building [m] and r the rate of V
cc
/ V
md
in percent. Usually, r varies between 8% and 15%. With
a specific gravity of 2,0 to 2,2 t/m the mass of concrete is about:
2 2
md md
M , V = (2)
5.3 Economic calculation
Since the costs of transport and off site waste management are usually measured in /t, the main
calculation value is the mass of debris M
md
. To decide between on-site recycling and off-site
recycling we need to estimate the costs for both cases. Recycling on-site is the better choice, if:
fixed md load md trans
C M C M C + s (3)
In formula 3, C
fixed
are the general costs of a recycling machine as e.g. the transport to the
demolition site, C
load
the costs for each ton of debris, which should be recycled and C
trans
the
costs for each ton brought to an external recycling site. In general, C
load
is less than C
trans
. On-
site recycling is attractive mostly above 1.000tons of raw material to recycle on-site, if the mass
balance fits and the produced gravel can be used directly.

General

367
5.4 Recycling of steel and metals
The rate of recycled steel and metal is high in general. Since the early times of deconstruction,
steel construction parts were separated and sold to scrap metal merchants. Mainly, the economic
aspect of metal as a resource are still driving the effort to separate any metal from other
materials. In opposite, the main aspect of separating waste from mineral debris is to keep the
costs per ton of waste removal little, since pollution of mineral debris tends to higher fees and a
bad quality of recycled gravel.
At least, separation of steel construction parts is quite easy. Modern buildings have a huge
amount of steel used for reinforced concrete. The will to separate the steel from the concrete
structure is depending on the scrap metal prices (Figure 9). The price of If costs of separation
steel from concrete are higher than the income from selling the separated steel, the rate will
become less. From practice, the price of core iron is about 20/ton below the price for
construction steel elements. The point of even is about 180/ton, but depending on the local
costs for disposal of mineral debris including steel at recycling factories.

Figure 9: Scrap metal prices from 2008 to March 2010. The higher the prices of scrap metal the higher
the rate of reuse or recycling (data from "Wirtschaftsvereinigung Stahl" average crap metal prices for
Germany).
6 CONCLUSIONS
An overview of the different strategies for deconstruction and of demolition practice was
presented with the main focus on keeping the recycling rate as high as possible during
deconstruction. Sorting the waste materials is hereby evident in all.
Four phases defined for deconstruction. For those sites that can accommodate source
separating of materials, a number of facilities that accept segregated streams generally exist in
the perimeter. For example, scrap yards accept only metal, and clean fill processors accept
only concrete, dirt and other inert materials. These facilities recycle virtually 100% of the
material they handle, and their fees for accepting segregated materials are much lower than
those charged by mixed C&D processors.
The recycling possibilities are limited by contaminations, the use of the material as a recycled
product and regulations by law. Mineral debris is usually processed to gravel of concrete. This

368
gravel needs to be classified by the LAGA-system below Z2 to reuse the material. Producing the
gravel on-site is only reasonable, if the outcome of the mass balance is stable as well as
economically beneficial. Off-site recycling may be better for small projects. Some materials
such as pumice stone can not be recycled even if not contaminated, because there is no need for
the recycled product.
REFERENCES
2006/12/CE, 2008. European Directive 2006/12/CE, modified 2008.
AMCS , 2009. ArcelorMittal, Commercial Sections: Bridges with rolled sections, report.
DDC, 2003. Department of Design and Construction of New York: Construction and Demolition Waste
Manual, May 2003.
D&D, 2008, Portioli, F., Ungureanu, V.: Demolition and deconstrcution of buildings, COST Action C25,
Sustainability of Constructions Integrated Approach to Life-time Structural Engineering,
Proceedings of Seminar Dresden.
Dorsthorst, B.J.H & Kowalczyk, T. 2002. Design for recycling. CIB Publication 272, Paper n.8. Proc. of
the CIB TG39 Deconstruction Meeting, Design for Deconstruction and Materials Reuse, Karlsruhe,
Germany. Rotterdam: CIB.
FRANCE, 2009. Ministre de l'Ecologie, de l'Energie, du Developpement Durable et de l'Amnagement
du Territoire: Code de lenvironnement, directive 98/8/CE, articles L.522-1 et suivants, 2009.
HKBD, 2004. Hong Kong Buildings Department: Code of Practice for Demolition of Buildings.
Hurley, J. & Hobbs, G. 2003. CIB TG39 - UK Country Report on Deconstruction, REPORT 9.
Knpple, H.J. 2007. Die neue Nachweisverordnung. Mll und Abfall 39(1), S. 2528 (2007), ISSN 0027-
2957.
Kristinsson, J., Hendriks, Ch. F., Kowalczyk, T. and Dorsthorst, B.J.H. 2001. Reuse of secondary
elements: Utopia or Reality. Proc. CIB World Building Congress, Paper n.230. Wellington, New
Zealand.
KWTBau. 2005. Monitoring-Bericht Bauabflle, Erhebung 2002. Arbeitsgemeinschaft
Kreislaufwirtschaftstrger Bau. Berlin, 31. Oktober 2005.
LAGA, 1997. Merkbatt: Anforderungen an die stoffliche Verwertung von mineralischen
Reststoffen/Abfllen - Technische Regeln. Mitteilung der Lnderarbeitsgemeinschaft Abfall (LAGA)
Nr. 20.
Lippok, Jrgen and Korth, Dietrich. 2004. Abbrucharbeiten. Hrsg. Deutscher Abbruchverband e.V.,
Verlagsgeselschaft Rudolf Mller GmbH &mp;mp;mp; Co. KG, Kln.
Rbenack, K.D., Lipok, J., Korth, D. 2007. Abbrucharbeiten: Grundlagen, Vorbereitung, Durchfhrung.
Jrgen Lipok, Dietrich Korth (eds). Verlagsges. Mller. ISBN 978-3-481-02417-8.
SCI, 2002. SCI, ECSC project LCA for steel construction, Final report.
SYMONDS Group. 1999. Report to DG XI. Construction and demolition waste practices and their
economic impact. UK.
Wei, A. 2002. Rckbau von Plattenbauten und umweltgerechte Wiederverwendung von
Betonfertigteilen am Beispiel eines 16-geschossigen Wohnhochhauses vom Typ PH 16 in Leipzig-
Grnau, Garskestrae 5. final report. http://edok01.tib.uni-hannover.de.
WMU, 2000. European Commission, Waste Management Unit of DG Environment. DG ENV.E.3.
Working Document n1: Management of Construction and Demolition Waste. Brussels.

Buildings

369
P. Kamrath
Paul Kamrath Abbruchtechnik und Entsorgung GmbH, Dortmund, Germany
Deconstruction of buildings: Masses and types
1 OVERVIEW
1.1 Introduction
The size of a building to deconstruct is measured by its total volume [m]. The total volume is
defined by the outer measures of a building. Putting the total mass of concrete, bricks, wood,
steel and plastics in tons in accordance to the total volume gives the possibility to compare
different buildings with each other. In general, the total mass differs between 15% up to 40%.
Beside the mass per volume unit other factors have to be used for categorization. Those
parameters are for example the age of a building to find out, whether land filling waste rate has
increased over the last decades, the number of floors, kind of roof etc.. With this data analysis
one can find out, if the aspects of recycling and waste minimization as important values of
sustainability are adequately considered or not today.
1.2 Deconstruction wase
Demolition work in general produces a lot of waste. The amount of waste can be reduced
through recycling of construction waste. Assorted raw materials such as steel, zinc and other
metals can be fed to the raw material cycle again. Demolition work belongs to the life cycle of a
building; hence the importance of raw materials and waste in the consideration of sustainability
must be taken into account, such as the consumption of CO2. With the exception of
contamination, there are four major types of waste: construction debris, wood, metal and trash.
The waste consists of no more usable products such as plastic, PVC, polystyrene, rubbish. For
an overall analysis of the sustainability of a building is therefore of central interest incurred in
the extent to which various substances.
1.3 Deconstruction techniques
Deconstruction of buildings is best done by the use of a four-phase-concept (Kamrath &
Hechler, 2010). In the scheme of the four-phase concept, decontamination is done before the
general deconstruction of other materials.
Table 1: Main deconstruction techniques in Germany (2002, KWTBau)
Main techniques Usage Quota
Deconstruction by excavators with shears etc. 82%
Exploding 4%
Deconstruction with wrecking balls 3%
Other machines 3%
Precussion, abrasion, heating etc. 3%
Robots 0.3%
Other 1.7%

370
In general, separation of waste is important and the key to the possibility of high recycling
rates. The demolition industry has undergone major transformation within the last 20 years.
More recently, mechanized processes with specialized equipments are replacing manual labor.
Most deconstruction work is done by excavators with about 82%. Today, excavators are
effective, flexible due to the possibility to use different equipment parts. Thus, demolition by
wrecking balls becomes already a less used technique with only 3% use in Germany.
On the one hand, wrecking balls are very effective; on the other hand the deconstruction is
much uncontrolled. Unfortunately, the more uncontrolled the technique is, the less possibilities
for separation are given. Wrecking balls might still be a good choice, if no neighboring
buildings, streets or places are next to the deconstruction object and if the object mainly is build
of concrete such as silos or bunkers (Figure 1 left).
An alternative for the use of wrecking balls is the use of an excavator with an applied
cracking shear. The shear leads to a much more controlled deconstruction process, with only
little noise emission, the possibility to deconstruct close to other buildings, up to heights of 50m
(together with the use of longfront excavators). Figure 1 (right) shows the use of a longfront
excavator with attached cracking shear during the deconstruction of a cart house in North-
Rhine-Westphalia, Germany.
Another controlled way for deconstruction of brick-parts of buildings is the use of pulling /
pushing arms (Figure 2) or pulling / pushing in general. Separation is often done by gabbing
tools. These grabbing devices can be used for the deconstruction of wood-structures or roof-
structures as well (Figure 3).
Powerful but noisy are hydraulic breakers, which are mainly used for underground-
constructions (fundaments). The extreme noise emission limits the use of breakers so, that they
are mostly used for thick structures where shears become ineffective.
Besides the huge role of excavators for deconstruction, all other forms of demolition and
deconstruction play only a side-role. Even exploding structures is only 4 times used out of 100
deconstruction projects. While the aspect other machines describes the deconstruction by
radlader or similar machines which are not typically used for deconstruction but may be used.

Figure 1: Deconstruction by use of wrecking balls (left). Use of cracking shears (right). While wrecking
balls are effective but lead to a uncontrolled deconstruction, shears are highly precise and have less noise
emission.

Buildings

371
Other points to generally different techniques as reusing of structures for example. With less
than 1% usage quota those possibilities are not common yet but may play a role in the future.

Figure 2: Pulling and / or pushing as effective technique for the deconstruction of brick walls.

Figure 3: Pulling and grabbing of parts. Especially older roof constructions can easily be deconstructed

372

Figure 4: Hydraulic breakers are used, if no possibilities to use shears are given. Effective but noise
emission.
2 FINDING KEY NUMBERS FOR DECONSTRUCTION
2.1 Calculating masses
When something will be deconstructed, the exact masses are mostly unknown. One reason for
the lack of exact information is the absence of exact drawings of the building. Then, it does not
make sense to make exact drawings and measurements of a building only to deconstruct it.
Thus, it is evident to make a guess of the masses. To be quite exact, data of preliminary
deconstruction projects should be used to extrapolate from this data. The goal is to calculate the
masses of concrete, bricks, wood and waste with help of the volume of the building. A couple of
categorization procedures help to find correction factors to multiply with the volume, so that the
masses can be calculated as:
i building i
m V f (
In this formula, m
i
describes the calculated mass of material i, V
building
the over all volume of
the building in m and f
i
the correction factor. The estimation of the correction factor is difficult,
since two buildings never are the same. But, a rough categorization is possible. f
i
depends on the
size of the building the number of floors for example. While older buildings have the ceiling
made out of wood newer buildings tend to have the floor made of concrete. Heat insulation
(walls, roof) produces waste that has to be counted. The former use of the building gives
information about the masses as well: Old apartment houses with mainly long-term lodgers tend
to have several overlapping carpets and / or PVC-floors but little amount of hard plaster walls.
Gypsum in general has to be removed before the deconstruction. Buildings of private owners
and different generations living inside tend to have several renovation phases and thus a lot of
gypsum plaster used.
(1 )

Buildings

373
2.2 Data for industrial buildings
In the late 1990s some research of different German universities tried to figure out, if gravel of
concrete is applicable for new concrete again (BMI 1998). One aspect was to figure out, how
much concrete belongs to the sizes of halls and industrial buildings. In general, the examined
buildings were mainly out of concrete solely and data was taken from the coal industry in the
eastern part of Germany. In general, the mass correction factors laid in the interval of 0.07 and
0.25 in m of mass /m of volume. In dependence to the numbers of floors and the height of each
floor the correction factors are as given in table 2:
Table 2: Correction factor f
i
by literature data from the BMI project.
All buildings Correction
Range of most buildings (95%) 0.070 0.250
Buildings of chemical industry (given data) Correction
77% of all buildings 0.070 0.200
Steel frame with brick-filling, no intermediate ceilings 0.040 0.090
General steel frame and brick-wall buildings 0.090 0.250
General buildings with intermediate ceilings, height > 7.0m Correction
Most cases 0.180 0.200
All cases 0.150 0.250
Halls, no intermediate ceilings, height > 7.0m Correction
Steel frame with brick-filling 0.040 0.090
Brick-wall 0.070 0.110
All cases including ground floor 0.110 0.150
Small buildings no intermediate ceilings, height < 7.0m Correction
All buildings 0.100 0.270

2.3 Typical buildings examples
While accessible data covers only industrial buildings, most deconstruction covers general
buildings (housing) and offices. Besides construction materials such as concrete, bricks and
steel typical relevant materials to take into account are wood and general waste, which is not
covered by concrete / brick, metals and wooden structures.
Some typical examples of general buildings will help, to find relevant house types and their
typical mass-volume rate as the percentage of the total volume which describes the amount of
debris.

Figure 5: General building, 1930, brick-
walls, wood-floor, wooden roof. No heat
insulation.

Type of building Housing 1 or 2 families
Structure Brick-walls, wooden roof,
wooden floors, floor above
cellar: concrete
Year build 1930
Total volume 702 m
Total waste 272 to
Total waste in m 123 m
Mass-volume correction
factor f
i

0.175
Wood 5.16 to
General waste 2.38 to

374

Figure 6: Electricity supply building,
1950, brick-walls, wooden roof, one
floor, no heat insulation.

Figure 7: Huge building, 1940, brick-
walls, wooden roof, wooden floor

Figure 8: Small building. Wood, Brick-walls

Type of building General building
concrete floors, floor
above cellar: concrete
Year build 1950
Total volume 3504 m
Total waste 1218 to
factor f
i

0.160
Wood 6.50 to
Type of building Housing, more families
cellar: concrete
Year build 1940
Total volume 3014 m
Total waste 950 to
factor f
i

0.150
Wood 49.64 to
Type of building Housing
cellar: concrete
Year build 1950
Total volume 7138 m
Total waste 392 to
factor f
i

0.180
Wood 11.46 to

Buildings

375

Figure 9: Appartementhouse, 2 floors,
concrete.

Figure 10: House for governmental uses.
Build in the 1960s, brick-walls, concrete-
floors.
2.4 Mass-volume correction factor fi for general buildings
In general, all values for the mass-volume correction factor f
i
are inside the interval of 0.110
0.200. The research data from the early 90s showed results inside an interval of 0.07 up to
0.250. Lower values have been made for halls, which have no walls inside. Higher values are
possible for concrete structures of industrial sides. In general, the analysis shows, that f
i

becomes higher for concrete structures instead of general brick-wall structures and lower, the
bigger the building was. The amount of wood is about 3 to 4 times higher than the amount of
general waste.
3 CONCLUSION
For the planning and calculation of deconstruction the most important data is the total volume of
the building and the total mass of the building. The volume can be calculated by digital air
photography or given overview plans for example. Since detailed plans are missing mostly and
creating new plans for calculation of the deconstruction only is not done, good guessing of the
total masses is needed. The masses are calculated by the total volume and a mass-volume
correction factor f
i
. Some research for industrial buildings was done in the late 1990s by a
German research group. Depending on the numbers of floors and the mass of a tall building is
about 7% and 25% of the total volume. The analysis of data of general houses in this paper was
able to shrink the interval down to values between 11% and 20% as typical percentage for the
mass-volume relationship.
Type of building Housing, youth meeting
Structure concrete
Year build 1970
Total volume 2250 m
Total waste 988 to
factor f
i

0.200
Wood 31,62 to
General waste 2 to
Type of building Governmental use
Structure Brick walls, floors:
concrete, roof: wood,
insulated
Year build 1960
Total volume 10017 m
Total waste 2410 to
factor f
i

0.110
Wood 20,84 to
General waste 11,23 to

376
The analysis of data helps, to make efficient calculations for the mass of concrete or brick-
stone waste for general houses. But, a range for f
i
from 0.11 to 0.20 can achieve deviation of
about 100% from calculation to reality. Mostly such big errors are avoided by the knowledge of
the engineer. Thus, there is further research needed, to create a structured and detailed
categorization framework. Within this framework, the values for the factor f
i
should depend on a
categorization procedure. After categorization, the calculated value for f
i
should be inside an
interval of maximum 3% to avoid miscalculations of masses.
Nevertheless, any analysis of data as done in this paper is helpful, to become a better
knowledge about mass calculation with the help of the total volume.
REFERENCES
BMI, 1998, Datensammlung und Datenauswertung zur Ermittlung von Kennzahlen; appllicable through
the internet: http://www.b-i-m.de/.
Kamrath P., Hechler O., 2010, Demolition and recycling of demolition rubble after deconstruction,
publishing in progress.
KWTBAU, 2007, 5. Monitoring-Bericht Bauabflle (Erhebung 2004) der Arbeitsgemeinschaft
Kreislaufwirtschaftstrger Bau (mo-nitoring of C&D-waste, source 2002). Arbeitsgemeinschaft Bau,
Berlin, appllicable through the internet:
http://www.bmu.de/files/pdfs/allgemein/application/pdf/5_monitoring_bericht_kwtb.pdf

Bridges

377
F. Portioli, M. DAniello
E. Cadoni
University of Applied Sciences of Southern Switzerland
R. P. Borg
Faculty for the Built Environment, University of Malta, Malta
O. Hechler
Demolition and deconstruction of bridges
1 INTRODUCTION
Demolition and deconstruction methods are becoming important issues for the maintenance and
rehabilitation of transportation infrastructure. Different methods can be used for bridge
demolition and deconstruction. The selection of the most appropriate technique depends on
several factors.
The location is one of the most important factors to be considered for the demolition
planning. It is clear that a demolition project of a bridge located in a busy city centre will be
different form a demolition project in a remote site. Furthermore, different demolition
techniques and access facilities will be used for bridges spanning over a waterway or a
motorway.
Another important issue to be considered for the planning of demolition operations and
techniques is the assessment of the structural behaviour of the bridge and the characterization of
the material properties (Borg 2008). For the evaluation of the structural response in different
demolition steps several methods exist in the literature. Demolition analysis prior to the
demolition and/or deconstruction of bridges is becoming increasingly important. Recently, new
modeling approaches have been developed for the analysis of progressive collapse of structures
under exceptional events. In case of explosive or deliberate demolition methods, the applied
element method (AEM) can be used for this type of analysis as it allows to simulate and analyze
the debris resulting from a demolition and predict the effect on neighbouring structures (Figs. 1,
2).
As for recycling, selective demolition techniques can be used, even if higher costs and more
time have to be considered. Transportation could represent one of the most critical activities in
the removal of a bridge. Removal and disposal costs of bridge sections have to be carefully
considered in the design process.
The aspects of sustainability involve both the deconstruction process with its impact on the
environment and the aspects concerning recycling of materials (Hechler & Hauf 2009). The
demolition or deconstruction process of bridges has thus to be well planned to minimize the
risks of causing damage to persons and properties of the public, endangering the health and
safety of site personnel and damaging the neighbourhood environment, see Table 1.

378
Table 1.Requirements for demolition and deconstruction (Hechler & Hauf 2009).
Requirements Action measures
Damage to persons and
properties of the public
needs is to be strictly
avoided.
- Suitable hoarding required to enclose the site during demolition
- Scaffold erected along the frontage abutting the highway and other such places as
may be required.
Safety measures for site
access need to be
considered.
- A person carrying on demolition operations should make sure that the
construction which is partly demolished is so far as is reasonably practicable,
properly secured or closed against entry at all times.
Health and safety of site
personnel should be
assured during the whole
deconstruction process.
- job safety trainings of deconstruction workers
- communication on the demolition planning to the workers
- training the equipment
- protection and site installation need to be appropriate for the demolition works
- Appropriate insurances need to be provided
Environmental impact to
the neighborhood
environment is restricted.
- Impacts are air pollution, Noise, Water, Hazardous materials and Traffic
disturbance

AIR:
- As for air pollution, concrete breaking, handling of debris and hauling process are
main sources of dust from demolition. - Dust mitigation measures complying with
the Air Pollution Control (Construction Dust) Regulations of the specific country
shall be adopted to minimize dust emissions.
- Burning of waste shall not be allowed. Diesel fumes generated by mechanical
plant or equipment shall be subject to the control of the Air Pollution Control
(Smoke) Regulations.

NOISE:
- Pollution arising from the demolition works including, but not limited to, the use
of equipment (such as pneumatic breakers, excavators and generators, etc.),
scaffolding, erection of temporary works, loading and transportation of debris, etc.
- Noise pollution affects the workers, and the sensitive receivers in the vicinity of
the demolition site.
- Silent type powered mechanical equipment shall be used to reduce noise impact
as much as practicable.
- Demolition activity shall not be performed within the restricted hours as
established.

WATER:
- The discharge of wastewater from demolition sites requires a valid discharge
license.
- Effluent shall be treated to the standards as stipulated in the licence before
discharge.
- The demolition contractor shall maintain proper control of temporary water
supply and an effective temporary drainage system.

HAZARDOUS MATERIALS:
- measures are strictly to be taken to avoid the escape of hazardous materials in the
environment as dust or suspensions.

TRAFFIC DISTURBANCE:
- deconstruction needs to take place without essential interference to the traffic
under the bridge
- The increase in accident potential has to be addressed, which however is
controlled due to adequate traffic control measures.
- User costs due to deconstruction have to be considered. They represent the public
damage due to loss in time, route diversions and changed accidental volume for
roads and the costs for operation interference for railways. They can multiply the
deconstruction costs to a greater extend if the economical loss due to traffic
interference by e.g. traffic jams are taken into account. Therefore temporary
closures of lanes may e.g. be considered for night work only.

Bridges

379
Figure 1. Explosive demolition of a truss
bridge (www.jaycashman.com)
Figure 2. Progressive collapse analysis of a bridge by
the applied element method (ASI 2010).
Of course, the main goal of demolition and deconstruction is to reduce all impacts to the
environment the waste from deconstruction incorporates generally the major impact.
The aspects concerning the reduction of demolition waste are common to other structural
typologies. Of course, the type of demolition and deconstruction technique influences the rate of
material recycling or reuse.
2 DEMOLTION AND DECONSTRUCTION TECHNIQUES FOR BRIDGES
In general, the different factors influencing the choice of demolition methods can be
summarized as follows: financial and client constraints; time limits; strength and quality of
materials; the shape and size of the structure; location and accessibility; the amount of material
to be removed; environmental concerns, including noise, dust, vibrations, and debris; worker or
public safety and health; recycling of materials; and removal, transport, and disposal of debris
(Abudayyeh et al. 1998).
All these factors should be used as different criteria in decision making methods for the
selection of the most suitable technique (Anumba et al. 2003).
An overview of different demolition and deconstruction techniques with specific reference to
bridges is presented in this section.
As reported in BS 6187 Code of practice for demolition, structural demolition methods can
be identified as follows: 1. Progressive demolition; 2. Deliberate collapse mechanism; 3.
Deliberate removal of elements. Progressive demolition should be considered to be the
controlled removal of sections of the structure, whilst retaining the stability of the remaining
part and avoiding collapse of the whole part of the building to be demolished. Deliberate
collapse mechanism should be considered to be the removal of key structural members to cause
complete collapse of the whole or part of the building or structure. The deliberate removal of
elements is a demolition method used to remove selected parts of the structure by dismantling or
deconstruction. Prior to the removal of any parts of the structure any potential instability or
collapses must be assessed. In case of unstable structures, for example, personnel within the
structure could not be used or temporary support may be necessary.
In general, with reference to the bridge structural parts, different demolition techniques can
be used for the decks, beams and piers. As for the decks, traditional demolition techniques with
excavators can be used. If it is not possible to reach the deck from the bottom, the demolition
can be carried out from the top of the deck on the adjacent spans (Figs 3, 4). The beams of the
bridge are generally dismantled with cranes or launching girders (Figs. 5, 6). For piers, the
demolition methods that are generally used are: machine-mounted attachments such as crushers;
diamond-impregnated wire saw that wraps around the pier to cut it into blocks; and explosive
demolition (Fig. 7).
For concrete bridges the following techniques can be used: machine-mounted demolition
attachments; hydrodemolition; blasting and miniblasting; sawing and cutting; ball and crane;
splitting; jackhammers; thermal demolition.
A comparison of different techniques in terms of applications, advantages and disadvantages is
provided in Table 2.

380
Table 2. Bridge demolition techniques (Abudayyeh et al. 1998).
Method Application Advantages Disadvantages
Machine-mounted
demolition
attachments

Hydraulic
hammers
Demolition of
bridge decks,
peers, slabs, and
pavements
High production rate,
greater mobility, operable in
inclement weather
Noise, dust, and vibration
Whiphammers Bridge deck
removal
High production rates High energy input
Crushers Full and partial
bridge removal
No dust, low noise, no
vibrations, great mobility,
operablein inclement
weather, rapid and safe
cutting of rebar

Hydrodemolition
Nonabrasive
water jet
cutting
Partial removal of
deteriorated
concrete in bridge
decks
Minimum labor, low noise,
no dust, high production
rate, no vibration, remaining
concrete surface irregular
allowing good bonding to
new concrete
Rebar shadow problems,
cost, needs large quantities
of water, and disposal of the
water that is mixed with
debris
Abrasive water
jet cutting
Partial removal of
deteriorated
concrete in bridge
decks
No dust, low noise, minimal
vibration, and very accurate
cutting
Cost, dangerous due to the
high pressures used, large
quantities of water are
needed, and disposal of the
water that is mixed with
debris
Blasting and
miniblasting
Full and partial
bridge
removal
Speed, short durations of
noise and dust
Dust, noise, vibrations,
flying debris, and dangerous
Concrete sawing
and cutting
Partial removal of
deteriorated
concrete
No dust, no vibration, and
produces clean edges
Difficulties arise around
rebar, cost
Splitting
Mechanical
splitters

Full and partial
bridge removal
No vibration, inexpensive,
little dust, remaining
concrete undamaged, and
can be used underwater
Time consuming and
requires the use of breakers
to expose rebar

Chemical
splitters
Full and partial
bridge removal
No vibration, no noise, safe,
and non-explosive
More expensive than
mechanical, requires more
time
Jackhammers Partial removal of
deteriorated
concrete in bridge
decks
Easy to use Slow, noise, dust, and
remaining concrete and
rebar may be damaged
Thermal Boring and
Cutting
Method is suitable
in the partial
removal of
concrete and steel
No vibration, low noise, can
be used in places that are
not easily accessible, and
can be used underwater.
Cost, fire hazard, and
generates large amount of
fumes
Ball and crane Bridge removal Safety of project workers,
simplicity of method
Control of the swing, dust,
noise, and vibration
In case of steel bridges, the girders can be cut into manageable sizes to facilitate demolition,
and recycled which is a clear benefit in terms of sustainability. Steel beams are either partially
or totally flame cut or, alternatively cut using shears attached to a modified excavator (Fig. 8).
Bolts and rivets are rarely removed prior to recovery. Methods to promote an increase in the
amount of steel to be reused are likely to involve removal of bolts from areas where access and
space is restricted. This is likely to involve a greater risk of injury to operatives where machines
are not available.

Bridges

381
Figure 3. Selective demolition of a steel bridge. Figure 4. Demolition of a reinforced concrete
bridge from the top of the deck. (Viarenghi
2009).

Figure 5. Deconstruction of a steel-composite bridge
using launching girders (Viarenghi 2009).
Figure 6. Lifting of prestressed concrete
beams by cranes (Viarenghi 2009).

Figure 7. Common pier demolition techniques: crushing, sawing and cutting and explosive demolition
(Viarenghi 2009).

For what concerns lifting and moving, little or no dismantling is required by modern
construction equipment for the removal of a structure, depending on the size of the structure and
the extent of the movement. Often this will allow the disassembly of a fairly large stream-
crossing structure on dry land adjacent to the site. In general, the lifting of bridges is done by
cranes or lunching girders (Fig. 9).
Although the ability to move an intact bridge allow to simplify dismantling operations, some
questions should be taken into account, as the state of stress induced during lifting and
movement. With this regard, the inspection and the required structural analysis must be

382
rigorous. After being moved, the structure must be placed on level temporary bearing surfaces
on firm material to facilitate its disassembly (McKeel 2006).

Figure 8. Flame cutting of a steel truss bridge. Figure 9. Lifting 106-foot truss bridge using cranes
(McKeel 2006).
3 CASE STUDIES
In this section, few examples of demolition and deconstruction of bridges are presented. The
first one is the Tenza bridge, a dismissed highway bridge in Campania region. The second
example is a suspended steel arch deck bridge near Augsburg (South Germany). Due to the span
and size of the superstructure these two examples show the difficulties and efforts for the
deconstruction. Further the successful consideration of the goals for deconstruction is outlined.
The third example is a suspended steel arch deck bridge near Augsburg (South Germany). The
last case study is a truss railway bridge in Italy.
3.1 Demolition of bridges: The Tenza Bridge
The Tenza bridge (Fig. 10), located in southern Italy, was built in the sixties as part of Salerno-
Reggio Calabria highway and was open to traffic till few years ago. Recently the highway
owner society planned to change the geometry of the route, since it does not respect anymore
the current safety standards. Therefore the bridge was closed to traffic as it belonged to a
substituted portion of the highway.
The bridge structure of the Tenza Viaduct consists of 3 different structures, one main and two
approach spans. The main span is an open spandrel arch structure that is 120 m long and 50 m
deep. The bridge deck and its wall piers are supported by a ribbed, solid slab, and fixed-fixed
arch. Each approach span is 30 m long and is supported by multiple wall piers of different
heights. Each individual pier is made of two RC columns; those external are connected over
their entire height by an RC wall.
The bridge was built during the economic growth after the Second World War and although
no record exists on the method used for its construction, wood formworks were probably used,
as that was a very common practice at that time. In the nineties, the bridge was strengthened to
resist seismic loads. The retrofitting consisted of RC encasement of piers and arches cross-
sections.
The design of the demolition was addressed through a sustainable approach. Hence, the main
goal was to limit the economical cost and guarantee low environmental and social impacts of the
operations. Given the huge volume of the structure, a controlled demolition by explosions was
considered. The way to proceed in order to design such a demolition is to define an unstable
configuration of the bridge to be achieved through local damage induced to structural elements
by the explosions. Such configuration will activate a collapse mechanism that will lead to the
demolition of the bridge. In this case, the main constraints related to the demolition regarded the
access to the valley to remove the debris. In fact, the access to the valley of the bridge is not

Bridges

383
easy and all the area is under the control of the local environmental authority. Hence, the
demolition path was designed so to have small debris elements, collapsed far from the center of
the valley, where access of trucks is easier.

Figure 10. Tenza Viaduct view. Figure 11. Set up for high strain rate
tests.
Hence, particular attention was posed to the explosion definition. In literature, a number of
empirical formulations provide the amount of explosive necessary to induce fracture to
reinforced concrete elements. Such formulations are based on an energy balance between the
explosion energy and the fracture energy of concrete. However, in order to have particular
precision in such predictions, dynamic properties of steel internal reinforcement and of concrete
are necessary. At this aim, relationships and procedures presented in TM 5-1300 (TM 1990) and
CEB Bulletin 187 (CEB 1988) were used. However, in order to calibrate such procedures,
dynamic tests on materials were analyzed to have fracture energy values of concrete and steel.
Such data allowed having a more reliable prediction of the amount of explosive to install in
order to have the desired collapse mechanism. It should be mentioned that also collapse
mechanism activation needs for particular attention, since it is strictly related to mechanical
properties of the structure, in terms of elasto-plastic behavior. Hence, in order to predict well the
collapse behavior of the structure after the initiation induced by the explosion, collected data
from structural characterization were analyzed.
3.2 Deconstruction of Bridges: Lechbrcke BAB A8 close to Augsburg
Lechbrcke BAB A8 close to Augsburg The motorway BAB A8 between Munich and Stuttgart
is one of the main traffic axes from the west to the east in Germany. 56,000 vehicles per day are
on the road (with only two lanes in each direction). The estimates numbers will grow up to
60,000 vehicles per day in the next years. Therefore the highway was and will be extended in
sections to a six-lane highway between Stuttgart and Munich. The old existing Lechbrcke at
Augsburg was a suspended arch bridge built in 1950 and consisted of two independent arches
for both traffic directions. The span of the arch was 93.5m. The bridge was a steel construction,
steel arch and steel deck, with an additional concrete layer of 15cm on the roadway. Due to the
importance of the road link traffic, the main focus for the reconstruction of the bridge over the
river Lech laid on the maintenance of the traffic - for the surroundings there is no possibility to
intercept the traffic. Based on theses conditions a concept was developed to minimize the impact
on the traffic but also to reduce structural measures for the river Lech. Figure 12 shows that
before the deconstruction was started the first new suspended arch bridge was erected next to
the old ones. The traffic lanes were now moved together on the new bridge so that the
deconstruction could be started.

384

Figure 12. Traffic on the new Lechbrcke
preparation of the deconstruction.
Figure 13. Starting of lifting of the arch.
Time reduction also played an important role as well as the environment protection of the
river Lech. Therefore it was decided to lift off the arch bridge. Due to the dead load and the
overhang of the cranes it was not possible to lift the arch in one piece. So the concrete deck was
demolished from the deck and after additional stability measures the bridge, already connected
with cables to the crane could be divided at the midspan. For the lifting very heavy cranes were
needed, the weight of the half arch was about 225 tons (Figure 13).
Other scenarios for the deconstruction were also discussed. But the deconstruction into the
river was denied due to the anger of the pollution as well as that the bridge can cause backwater
or even a flood. In this case the applied deconstruction method was the most expensive
alternative, but the best solution for the environment and impact on nature and traffic.
3.3 Neckarbrcke AK Viernheim BAB 6 near Mannheim
The original bridge was built in 1934/1935 and had a total span of 400m with 6 arches of single
spans between 49.5 and 66.45m (Figure 14). The bridge consisted out of two superstructures for
each driving direction. The last arch of both superstructures was spanned over the canalized and
navigable Neckar. Due to upgrading of the four-lane to a six-lane motorway the replacement of
the bridges has been necessary. It has been decided to keep the concept of separate
superstructures for each driving direction. To minimize the traffic impact on the highway the
following reconstruction steps has been chosen: 1) Construction of the first separate
superstructure next to the existing bridge, 2) Bypass of the traffic from the original bridge over
the new superstructure, 3) Deconstruction of the original bridge, 4) Construction of the second
new superstructure.
For the deconstruction several conditions had to be considered. In transverse direction under
the existing bridges a gas-pipeline and the new superstructure were not allowed to move or to be
damaged during the deconstruction time. The Neckar channel for the vessel traffic could only be
stopped for several hours at the weekend for the deconstruction of the arches. The area round
the bridges is a landscape conservation area which also had to be considered. Before the
deconstruction was started the water authority required to capture the fish in the Neckar and set
them out again several hundred meters downstream. Based on these guidelines a deconstruction
concept was developed and it was decided to use a blast deconstruction for the first five arches
and a lifting process for the arch over the Neckar channel. This concept of course caused
additional costs, but the lifting of the arch over the channel was the only way to have no impact
on the vessel traffic. Before that the concrete deck was removed to reduce the weight of the
concrete superstructure. A special attention was set on no falling concrete pieces from the deck
in the Neckar channel. For the blast, the difficulties were to ensure that the superstructure will
cause no damage to the new bridge built next to the existing, but also from the falling of the
concrete mass the gas pipeline had to be protected. Therefore mounds were built up under each
arch to reduce the fall height of the concrete mass as well as the vibrations for the ground. Due
to the blasting and the missing horizontal counterforce at the last abutment before the channel
arch the analysis for the stability had to be done and the security factor was reduced to 1.0. In
the following figures the blasting of the superstructure (Figure 15) and the lifting of the channel

Bridges

385
arch (Figure 16) can be seen. The concrete of the superstructure was crushed and brought to a
dumpsite. The Arch over the Neckar channel was cut in several slices with a total weight of 100
tons. Due to the span of the arch of 53.8m four heavy cranes on both abutment sides were
needed to lift up the arch that was divided in the crown into two parts (Figure 17). The example
of the deconstruction of the Neckarbrcke BAB 6 near Mannheim shows exemplary aspects to
be considered. Not only the stability of the constructions plays an important role, but also
conditions that of course are not considered when a bridge is built, as the protection of the
environment and the existing infrastructure around the construction site. The significance of the
impact on the rolling traffic on the highways and roads during deconstruction has here to be
highlighted.

Figure 14. Old bridge Neckarbrcke before
deconstruction.
Figure 15. Blasted superstructure and arch.

Figure 16. Lifting of the arch 5 over the Neckar
channel.
Figure 17. Lifting of the separated arch over the
Neckar channel.
3.4 The Manuel Dimech Bridge, Malta.
The Manuel Dimech Bridge forms part of the TEN-T Network in Malta, see figure 18. The
Bridge is located on Route No. 1, Mikiel Anton Vassalli Road. It was built between 1967 and
1971 across the Wied Ghomor valley and over two roads running on the sides of the valley.
Two separate similar bridge structures were built, one for each carriageway. The bridge
structure consisted of a gerber girder construction. The edge spans as well as the cantilever
beams of the mid span consisted of prestressed triple-celled box girders haunched at the location
of the intermediate supports, and made of cast in-situ concrete. Five precast prestressed concrete
beams were hung in-mid span. The edge spans were 36.9 m long while the mid span was 54.9 m
long. The cantilever length of the box girders in mid span was 9.15 m on either side, such that
the remaining span between the gerber hinges was 36.6 m. The minimum construction height of
the triple-celled box girder was 1.98 m, and the maximum construction height was 2.75 m. The
webs were 25 cm thick which widened towards the intermediate supports. Vertical stiffening
plates were located at each third of the edge spans. The construction height of the suspended
precast elements in mid span was 1.98 m; the bottom chord width was 76 cm and the top boom
width was 66 cm; the web thickness was 18 cm. The precast beam elements were also combined

386
by plates provided at each third of the span and at the supports, as well as by the cast in place
reinforced concrete deck.
The superstructure was entirely prestressed in the longitudinal direction. The prestressing
steel tendons were reported to consist of 12 or 36 wires of 7 mm diameter each of steel grade
1550/1750. It was further reported that each 3 of them were anchored by a wedge. On either
side of the piers 2 tensioning recesses were situated underneath each web in the bottom slab.
The suspended beams were each prestressed by 8 prestressing tendons; 4 tendons of 36 wires D
= 7 mm and 4 tendons of 12 wires D = 7 mm each were provided. The tensioning recesses for
the 36-wire-tendons were located in the enlarged beams ends. The tensioning recesses for the
12-wire-tendons were located in the top boom, underneath the subsequently poured deck.
The clearance between the kerbs of each carriageway was 9.15 m, and the carriageway of the
edge spans varied.

Figure 18. The Manuel Dimech Bridge before
Demolition (2006).
Figure 19. Reinforced Concrete elements resulting
from demolition of the Bridge. (2006).
The structure was elastically supported on elastomeric bearings. The abutments consisted of
gravity walls, and the piers were designed as hammer head piers. Both abutments and piers had
shallow foundations, on rock. Each superstructure had 4 expansion joints, 2 at the abutments
and 2 at the gerber hinges.
It was reported that the eastern superstructure sloped in the eastern direction, and the western
superstructure sloped in the western direction, therefore affecting driving unfavourably. Both
superstructures were provided with a gravity surface drainage. No maintenance measures to the
bridge had been reported or recorded since the date of construction. A comprehensive inspection
and investigation of the structure of the old bridge superstructure and substructure were carried
out, intended primarily to assess its the damages and structural integrity, and to serve also as
the basis for a decision as to whether the bridge structure should be rehabilitated or entirely or
partially reconstructed (KHP 2005). The assessment referred to the bridge superstructure, the
bridge substructure, and bridge equipment, including the bearings, the expansion joints, the road
surface, shoulders and parapets, crash barriers and water drainage system.
The old bridge structure exhibited various signs of damage, particularly spalling of concrete
and exposed corroding reinforcement in both the superstructure and the substructure. The
longitudinal gradient was discontinuous in the location of the gerber joints. However other
deficiencies were reported. The proposals for reinstatement of the bridge structure were based
on the damage assessment, and also on other criteria which were specific to the particular
project. The proposals for the reinstatement of the bridge depended on the assessment of the
damage carried out, and also on other conditions which included economic considerations,
timeframes for a reinstatement intervention, and traffic considerations. In particular it was
required to retain one of the superstructures open to traffic during reinstatement works. Hence
the criteria adopted for the selection of the options for reinstatement were influenced by specific
requirements (KHP 2005).
The solution proposed included the demolition of the existing bridge superstructure, and
reconstruction of the superstructure of the bridge as a 3-span cast in place box girder. The

Bridges

387
demolition of the bridge superstructure was carried out in accordance to a demolition plan. The
proposal included also the rehabilitation of the bridge substructure, through the removal of the
carbonated concrete cover using high pressure water jetting, and applying a 10cm shotcrete
layer. The reconstruction and repair exercises included also additional tasks.
In the demolition of the bridge superstructure, a scaffolding erected over the whole length of
the bridge was proposed. This was intended to simplify the demolition process. The demolished
material could be laid down on the underlying scaffolding, resulting in a relatively quick
demolition process. Furthermore the scaffolding system could be used at a later date to support
the formwork of the new superstructure box girder.
Demolition had to be carried out, taking into consideration existing adjacent buildings at the
north end of the bridge, and the existing piers, abutments and adjacent bridge structure.
Demolition of the existing bridge was carried out in two phases, to allow for continued traffic
flow on the second bridge while the first bridge was being demolished. The performance of the
structure during demolition was taken into consideration throughout the different phases of
demolition.
The bridge equipment was removed in the preliminary phase. The bridge structure was cut in
pieces of varying sizes, depending on weight and location within the structure, see Figure 19. In
particular diamond wire and saw cutters were utilised to cut the existing bridge in pieces, in
accordance to a pre-established demolition plan. The bridge elements were then lifted away.
Mechanical equipment was further utilised to break the concrete bridge elements into smaller
elements. The main tasks at demolition stage included also the separation of the steel from the
concrete, and the crushing of the concrete. Hydro-demolition was used for the surface removal
of concrete in the substructure of the bridge, in view of the rehabilitation exercise.
Various waste materials were considered for recycling, including in particular the steel
reinforcement, and the concrete. The waste concrete from the bridge structure, was considered
for the production of recycled aggregate concrete. Large concrete elements were broken into
smaller elements, which could eventually be crushed for the production of aggregate for
concrete. The assessment of various engineering properties of the recycled concrete aggregate
was considered necessary for classification, and for eventual use of the product in concrete
mixes (Borg 2008).
3.5 The railway bridge at Cavarzere (Italy)
In the second part of the 19th century, the Italian Government developed a railway network in
the northern part of Italy in order to get easy the transportation of goods and the military control
at the borders of new unified kingdom of Italy. In 1884 the construction of the railway line
Adria-Mestre started from Padua. This railroad was completed in 1916, during the first world
war. Therefore it became a crucial arterial line for the military transportation. From 1913 to
1916 a large number of railway bridges were constructed along this network with spans 10 to 60
m. Among them there is the bridge on Adige river built in the town of Cavarzere. The bridge
was damaged during the bombing attacks of the second world war. After the war, it was
repaired (1948). Recently, owing to the need to updating the rail line and to reduce the
maintenance costs, RFI (owner of Italian railway network) decided to demolish the bridge and
to substitute with a new one. Therefore, the dismantling occurred from 28th February to 20
th

March 2009. The bridge consisted by three simple supported trussed beams rested on masonry
piles (Figure 20). Each span length was 34.80m for a total length of 104.40m. It consisted of a
symmetric structure about the middle length of the intermediate span. Per span, the main girders
are two riveted trusses 3.50 m high and 3.30 m apart, consisting of combined plates and L and C
sections, as shown in Figure 21. In addition, the main trusses were connected with riveted built-
up transverse secondary beams, a horizontal bracing system, and longitudinal secondary trussed
beams, where the railway superstructure transmits the train loads.
The trussed beams of the bridge were dismantled with launching girders and cranes to
remove the parts cut into manageable sizes to facilitate demolition, and recycled which is a clear
benefit in terms of sustainability. The demolition was carried out to the following phases:
1.The trussed spans were uplifted by means of hydraulic jacks in order to position the
launching girders at the abutment and on piles (Fig. 22).

388
2.The girders were pulled on the launching girders toward one abutment of the bridge (Fig.
23). To do this it was necessary to position some tracks under the main girders to allow the
bridge sliding.
3. To avoid the out-of-alignment of the trussed beams and to facilitate the sliding of the
girders, it was decided to join together the spans by means of a new steel frame designed to
provide flexural continuity to the bridge (Fig. 24)

Figure 20. The bridge of Cavarzere as it was. Figure 21. The rail superstructure of the bridge
of Cavarzere as it was.

Figure 22. The use of hydraulic jacks to uplift the trussed span (a) and the position of the launching
girders (b).

Figure 23. The trussed beams pulled on the launching girders (a); a particular of the track positioned
under the girder to allow the bridge sliding (b).

4. When the trussed beams were pushed on the river side they were flame cut into smaller
trunks (Fig. 24b).

Bridges

389
5. the cut parts were uplifted by cranes and moved on the demolition site to disassemble into
smaller elements (Fig. 25).

Figure 24. The new steel frame welded to the trussed beams (a). Cutting of the main girder (b).

Figure 25. The uplift and the final movement on the adjacent demolition site.
4 CONCLUSIONS
The demand on demolition and deconstruction of bridges is constantly increasing in Europe. As
a consequence, also the impact of demolition and deconstruction is growing mainly the traffic
disturbance and the increase in tonnage of wastes are here to be noted. The minimization of
traffic disturbance for deconstruction and reconstructing is already considered in todays
practice. The reduction of construction and demolition wastes comes more and more in the
focus. Recycling has been identified to be an appropriate solution it is already happening
driven by cost savings for contractors and processors. Opportunities for more aggressive reuse
and re-cycling will improve as the potential for cost savings increases and as processors
improve their capacity to handle and market the recoverable materials, and as the construction
industry be-comes more aware of savings that could accrue from more aggressive recycling. As
a result, the designer should considered deconstruction during design. Combined with
innovative recycling procedures and use of recycled materials, to be further developed, the
economical motivation for the reduction of landfill will increase. Thus the today prediction, that
wastes are just resources in the wrong place in the future will be hardened. Overall a well-
defined strategy for deconstruction is not only reducing the environmental impact but also
bargain for the deconstruction company. The examples in this paper show, how a sophisticated
approach of demolition fulfill the todays demand for sustainability.

390
REFERENCES
Abudayyeh, O., Sawhney, A,. El-Bibany, H. Buchanan, D. 1998. Concrete bridge demolition methods
and equipment. Journal of Bridge Engineering, Vol. 3, No. 3, August 1998, pp. 117-125.
Anumba, C., Abdullah, A., Fesseha, T. 2003. Selection of demolition techniques: a case study of the
Warren Farm Bridge. Structural Survey, Vol. 54, No. 1, pp. 36-48.
Anumba, C., Marino, B., Gottfried, A., Egbu, C. 2004. Health and safety in refurbishment involving
demolition and structural instability. Research Report 204. Health and Safety Executive-funded
research project, UK.
Applied Science International 2010. Extreme loading for structures (ELS). Advanced nonlinear dynamic
analysis of structures.
Borg, R.P. 2008. Assessment of bridge structure, demolition and recycling of materials. In Proc. of COST
C25 Seminar Sustainability of Constructions, Dresden 6,7 October 2008. (pp. 3.17-3.29). Eds. L.
Bragana et al. Bruxelles: COST.
British Standard Institute. 2000. BS 6187 Code of practice for demolition, UK.
Cadoni, E., Asprone, D., Prota, A., Manfredi, G. 2009. Criteria for optimized design of demolition of a
bridge: the Tenza bridge. In Proc. of COST C25 Workshop: Sustainability of Constructions: Integrated
approach to life-time structural engineering. Timisoara, 23-24 October 2009. (pp. 211-216) Eds. L.
Bragana et al.
European Commission Directorate-General for Energy and Transport (2009), European Union Energy
and Transport in Figures.
Hechler, O & Hauf, G. 2009. Deconstruction of Bridges Examples. In Proc. of COST C25 Workshop:
Sustainability of Constructions: Integrated approach to life-time structural engineering. Timisoara, 23-
24 October 2009. (pp. 191-198) Eds. L. Bragana et al.
Hechler, O. & Schaur, C. 2007. Demolition and reconstruction of bridges. In Proc. of COST C25
Workshop Sustainability of Constructions, Lisbon 19-15 September 2007. (pp. 3.77-3.84). Eds. L.
Bragana et al. Bruxelles: COST.
KHP, 2005, Manwel Dimech Bridge, St Julians, Malta, Bridge Inspection, Evaluation of Existing Bridge,
Proposal for Reinstatement, Konig, Heunisch und Partner; Malta Transport Authority, Malta.
McKeel, W.T., Miller, A.B., Clark, K.S., Saufley, R.W., Bushman, W.H. and T.L. Lester. 2006. Best
Practices for the Rehabilitation and Moving of Historic Metal Truss Bridges. Report No. VTRC 06-
R31. Virginia Transportation Research Council.
Petzek, E., Bancila, R., Schmitt, V. 2009. Deconstruction of Bridges Examples. In Proc. of COST C25
Workshop: Sustainability of Constructions: Integrated approach to life-time structural engineering.
Timisoara, 23-24 October 2009. (pp. 199-210) Eds. L. Bragana et al.
Obrien, E., Enright, B., Caprani, C. 2008. Implications of Future Heavier Trucks for Europes Bridges. In
Proc. of Transport Research Arena Europe 2008, Ljubljana.
Sustainable Bridges: Assessment for Future Traffic Demands and Longer Lives, Sixth Framework
Programme. 2003-2007. D1.3 European Railway Bridge Demography.
Viarenghi, M. 2009. Quando lo scarto diventa risorsa. Analisi delle tecniche di demolizione selettiva e
indifferenziata. Rifiuti da costruzione e demolizione e terre e rocce da scavo nel D.Lgs. 152/2006.
Verbania 16 Dicembre 2009, Italy.

Bridges

391
F. Portioli
O. Hechler
Conclusions
The contributions presented in this chapter provide a state-of-the-art report on demolition and
deconstruction of building and bridge structures and serve as a basis for best practices and
recommendations.
The different methods and techniques for demolition and deconstruction have been discussed
on the basis of various sustainability issues.
With special reference to design for deconstruction, the numerous references given in the
literature with recommendations for the dimensioning of building elements with implications
for their practical performance in terms of conversion and reuse have been organized in order to
provide a general guidance. In particular, a set of recommendations has been developed based
on literature and own experience. The recommendations were differentiated in general
principles, principles in the choice of materials, structural design principles, principles for the
connections and for the foundations.
An overview of the different strategies for deconstruction and of demolition practice was also
presented with the main focus on keeping the recycling rate as high as possible during
deconstruction. The limitation of recycling possibilities due to contaminations, the use of the
material as a recycled product and regulations by law were widely discussed.
General issues concerning building and bridge demolition and deconstruction were discussed
as well. The number of case studies presented showed how sophisticated approaches of
demolition fulfill the todays demand for sustainability.
To evaluate the applicability of most common decision making methods, the types of
alternatives and criteria that usually are taken into account in the literature for considered
problems were showed. The survey showed that different approaches are used to solve partial
problems mainly concerning retrofitting or demolition and deconstruction. Applications
including all possible alternatives available to decision makers could represent further
development of current studies, also in terms of simplified criteria.

392

393
List of all contributions from COST C25
Proceedings of the First Workshop. Sustainability of Constructions. Integrated
Approach to Life-time Structural Engineering. Lisbon 13-15 September 2007. Cost
Action C25. Editors: L. Braganca, H. Koukkari, R. Blok, H. Gervasio, M. Veljkovic, Z.
Plewako, R. Landolfo, V. Ungureanu, L.S. Silva. ISBN: 978-989-20-0787-8. Published
by Multicomp, Lda. COST Publication.

Introduction to Life-time structural engineering
R. Landolfo

Service life methodologies
W. Trinius & C. Sjstrm

Modelling of corrosion damage for metal structures
R. Landolfo, L. Cascini & F. Portioli

Fatigue of steel and composite bridges
U. Kuhlmann, H.-P. Gnther, J. Raichle & M. Euler

Durability assessment modeling of reinforced concrete elements
C. Giarma, A. Kudzys, O. Lukoeviien, J. Radic, J. Bleiziffer & Sz. Woliski

Systematic monitoring of civil structures
T. Rauert, B. Hoffmeister, C. Heinemeyer, J. Radic, J. Bleiziffer & O. Hechler

Dynamic behaviour of short span ballasted railway bridges
C. Rebelo, C. Rigueiro, L. Simes da Silva & M. Pircher

Sustainable bridge construction through innovative advances
A.E. Long, R.K. Venables & J.D. Ferguson

Maintenance, reconstruction, repair, strengthening and rehabilitation of existing
masonry buildings
Z. Lj. Bozinovski, M. Secer & O. Bozdag

Demolition and reconstruction of bridges
O. Hechler & C. Schaur

Innovative forms of construction for sustainable bridges
A.E. Long, R.K. Venables, S.E. Taylor, A. Gupta & J. Kirkpatrick

Sustainable mixed building technologies applied to residential buildings: some
Romanian examples
D. Dubina, V. Ungureanu & M. Mutiu

Data management, structural maintenance and life cycle performance for the
CargoLifter airship hangar / Tropical Islands Dome
H. Pasternak, T. Bretschneider, G. Mosler & O. Schemmel

Steel end-plate connection with thermal-insulating layer
Z. ulcov & Z. Sokol & F. Wald

394
Proceedings of the Seminar. Sustainability of Constructions. Integrated Approach to
Life-time Structural Engineering. Dresden 6-7 October 2008. Cost Action C25. Editors:
L. Braganca, H. Koukkari, R. Blok, H. Gervasio, M. Veljkovic, Z. Plewako, R.
Landolfo, V. Ungureanu, L.S. Silva, P. Haller. ISBN: 978-3-86780-094-5. COST
Publication.

Life-time structural engineering
R. Landolfo & V. Ungureanu

Durability and Service Life Prediction Methodologies

State-of-the-art report on Service Life Design methods

Life time response of masonry structures using degrading model force-displacement
Z.Lj. Bozinovski

Probabilistic approach to service life design of reinforced concrete structures
S.Z. Wolinski

Risk based approach to service life assessment of building structures
S.Z. Wolinski

Bond strength and durability of textile reinforced wood
R. Putzger & P. Haller

Life Cycle Management Tools using LIFECON Procedures and Calculation Methods
E. Vesikari

Maintenance, repair and rehabilitation of RC-buildings
O. Bozda & M. Seer

On the rehabilitation of historical buildings- Sustainability aspects
O. Ccen & E. Efthymiou

Life-Cycle Assessment of the Spear Building
P. Negro, E. Mola, L. Cascini, R. Landolfo & F. Portioli

Seismic structural design integrated with life cycle cost analysis
M. Seer & O. Bozda

Evaluation of life cycle costs of bridges considering the construction process
U. Kuhlmann & G. Hauf

Blasting Technology for Demolition and Deconstruction
P. Lwe & T. Loose

Demolition and deconstruction of building structures
F. Portioli & V. Ungureanu

Deterioration and systematic condition investigation of rendered facade
J.M. Lahdensivu & J. Mattila

395
Proceedings of the Second Workshop. Sustainability of Constructions. Integrated
Approach to Life-time Structural Engineering. Timisoara, Romania, 23-24.10.2009.
Cost Action C25. Editors: L. Braganca, H. Koukkari, R. Blok, H. Gervasio, M.
Veljkovic, Z. Plewako, R. Landolfo, V. Ungureanu, L.S. Silva. ISBN: 978-973-638-
428-8. Published by Orizonturi Universitare. COST Publication.

Durability and service life of wood structures and components State of the art

Service life design of reinforced concrete structures Alternative probabilistic
approaches
Sz. Wolinski

Degradation models of concrete structures
E. Vesikari, C. Giarma & J. Bleiziffer

Degradation modes and models for masonry structures
Z.Lj. Bozinovski, J. Lahdensivu & E. Vesikari

Criteria for the assessment of existing railway bridges
E. Petzek, R. Bancila & V. Schmitt

Performance based evaluation of seismic retrofitting techniques
A. Dogariu & D. Dubina

Conservation of suburban concrete buildings
J.M. Lahdensivu & J.S. Mattila

Effect of water repellent coatings on the corrosion rate of reinforcement in carbonated
concrete facades and balconies
J.S. Mattila

Evaluation of performance of some metal based retrofitting techniques applied to
masonry walls
A. Dogariu & D. Dubina

Life time response of masonry structures using degrading model force displacement
Z.Lj. Bozinovski

A sustainability approach in structural restoration: Application of LCA analysis in a
steel-based intervention project
I. Zygomalas, .N. Ccen & E. Efthymiou

Evaluation matters and upgrade Iron Gate I of the Danube Hydropower Station
D. Dubina, D. Grecea & A. Dogariu

Guidelines for strengthening of railway bridges
L. Elfgren, B. Tljsten, A. Carolin & J. Eriksen

GFRP emergency bridges. An ecological lightweight solution
L. Blaga

Latest Software Development for Life Cycle Management of Bridges in Europe
E. Vesikari, A. Borrmann & K. Lukas

396
Condition assessment of steel structures. Recommendations for Estimation of
Remaining Fatigue Life
B. Khn & O. Hechler

Monitoring of a bridge with integral abutments
J. Eriksen, M. Veljkovic & M. Nilsson

Systematic condition investigation of concrete buildings
J.M. Lahdensivu & J.S. Mattila

Deconstruction of bridges
O. Hechler & G. Hauf

Deconstruction of Bridges Examples
E. Petzek, R. Bancila & V. Schmitt

Criteria for optimized design of demolition of a bridge: the Tenza bridge
E. Cadoni, D. Asprone, A. Prota & G. Manfredi

Sustainability of Constructions Integrated Approach to Life-time Structural Engineering
COST Action C25
L. Bragana, H. Koukkari, R. Landolfo, V. Ungureanu, E. Vesikari, O. Hechler (Eds.)
2011 The authors and the editors. All rights reserved.

397

AUTHOR INDEX

Andersson, J. 296

Bleiziffer, J. 40
Bischoff, R. 142
Borg, R. P. 377
Borrmann, A. 223
Bozinovski, Z. Lj. 69
Brhwiler, E. 282

Cacace, F. 201
Cadoni, E. 327, 377
Cascini, L. 51, 106, 317
Ciutina, A. 235

DAniello, M. 377
Dogariu, A. 235
Dubina, D. 235

Euler, M. 61

Fabbrocino, G. 130
Feltrin, G. 142
Frangopol, D. M. 288

Geler, A. 296
Giarma, Ch. 40, 179
Gnther, H.-P. 61

Hauf, G. 117, 120, 214
Hechler, O. 117, 120, 169, 214,
221, 296, 310, 315,
339, 355, 377, 391

Kamrath, P. 355, 369
Khn, B. 169
Kuhlmann, U. 61

Lahdensivu, J. 69
Landolfo, R. 1, 2, 17, 39, 51,
106, 235, 317
Lang, T. 282
Larsen, O. P. 339
Leone, M. F. 154
Lukas, K. 223
Lchinger, P. 282

Makkonen, L. 81
Mazzolani, F. M. 235
Meyer, J. 142

Naumes, J. 296
Nielsen, S. 339

Ojanen, T. 81
Okasha, N. M. 288

Peuhkuri, R. 81
Portioli, F. 51, 106, 235, 315,
317, 377, 391

Raichle, J. 61
Rainieri, C. 130
Rauci, M. 201

Saukh, O. 142
Sjstrm, C. 10

Taushanov, A. 179
Teruzzi, T. 192
Toratti, T. 81
Trinius, W. 10

Ungureanu, V. 235, 317

Vesikari, E. 1, 2, 17, 39, 40,
69, 223
Viitanen, H. A. 81
Vogel, T. 282

Zuccaro, G. 154, 201

Wolinski, Sz. 28, 94

Sustainability of Constructions Integrated Approach to Life-time Structural Engineering
COST Action C25
L. Bragana, H. Koukkari, R. Landolfo, V. Ungureanu, E. Vesikari, O. Hechler (Eds.)
2011 The authors and the editors. All rights reserved.

398

Sustainability of Constructions - Volume 2 - Integrated Approach To Life-Time Structural Engineering

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Sustainability of Constructions - Volume 2 - Integrated Approach To Life-Time Structural Engineering

Uploaded by

Copyright:

Available Formats

COST Action C25

Cl ions as a function of concrete age. Solid line is

is the initial value of partial safety factor corresponding to a certain design

, -16.0 and -15.0 and for

=1. The typical decision matrix is then:

. As a consequence, in the AHP

for i=1,,m (13)

You might also like