Professional Documents
Culture Documents
Scientific Conference
on Hardwood Processing 2013
7th- 9th October 2013
Florence, Italy
PROCEEDINGS
Supported by:
CNR IVALSA
ISCHP 2013
ISBN 9788890927300
PREFACE
BSTRACT
Topics covered A
by
ISCHP 2013:
Hardwoodagency
Forestry
Practices
& Wood
Quality
FAO is a specialized international
of the
United Nations
leading
global efforts to eradicate
hunger, eliminate rural poverty
and promote
the sustainable
management of natural resources. In a
Hardwood
Processing
& Optimization
world with an expected population
of 9Product
billion people
by 2050, demand for food and other products and
Hardwood
Development
services from nature will be greater than ever. This pressure will change both the extent and the quality
Hardwood Market & Sustainability
of a substantial portion of the worlds forests. Understanding these changes is essential if investments
are to be made to improve forest management to both provide for a growing human population and to
maintain crucial forest biodiversity. FAO is committed to generating reliable information on forests and
forestry practices to promote the conservation and sustainable use of forest resources worldwide.
This year, ISCHP
wasadditional
organized
by Cnr-Ivalsa
and Timber
Institute,support,
Italy. All
reviewed
publications
Clearly,
investment
is neededTrees
in research,
whilst technical
and 37
fieldpeer
activities
are still
in the proceedings
were
subject
to
a
rigorous
one-sided
blind
peer
review
process
with
a
minimum
of
two
reviewers
needed to slow the loss and degradation of forest area, particularly in the tropics.
The Food and Agriculture Organization of the United Nations (FAO) is a specialized international
agency leading global efforts to eradicate hunger, eliminate rural poverty and promote the sustainable
management of natural resources. FAO serves its 192 member countries3 as a neutral forum, retaining the
following core functions: 1) facilitate and support countries in the development and implementation of
normative and standard-setting instruments such
as international agreements, codes of conduct, technical
EDITORS
standards and others; 2) assemble, analyse, monitor and improve access to data and information, in areas
related to FAOs mandate; 3) facilitate, promote and support policy dialogue at global, regional and country
S.country
Berti and regional level to prepare, implement, monitor
levels; 4) advise and support capacity development at
A.
Achim
and evaluate evidence-based policies, investments and programmes; 5) advise and support activities that
Fioravantitechnologies and good practices in the areas of
assemble, disseminate and improve the uptake M.
of knowledge,
Lihra
FAOs mandate; 6) facilitate partnerships for foodT.and
nutrition security, agriculture and rural development
V.
Loewe
between governments, development partners, civil society Muoz
and the private sector; 7) advocate and communicate
Marchal
at national, regional and global levels in areas of R.
FAOs
mandate.
J. Wiedenbeck
R. Zanuttini
FAO has 192 member states along with the European Union, Faroe Islands and Tokelau which are associate members
1
3
ACKNOWLEDGMENTS
1
Eduardo Mansur
, Alberto
Del ISCHP
Lungo requires
, Cesar Sabogal
The organization of an international
scientific
event like
effort from many individuals and
institutions.
2
2
Director,
Forestry Officer, Forest Assessment,
Forestry Officer, Forest Assessment,
Forest
Assessment,
Management
and
Thanks go to the members of the organizing
committee
who have
supported
us from and
theConservation
beginning. The Scientific
Management
and Conservation
Division,
Management
Conservation Division,
Forestry
Department,
Division,
Forestry
Department, The Editorial
Committee input
was
critical
in
assuring
the
quality
of
the
conference
and
in
securing
participation.
Forestry Department,
Food
and
Agriculture
Organization
of
the
Food
and
Agriculture
Organization
Committee
has
done
a
hard
work
organizing
the
peer
review
process
and
ensuring
the
scientificofquality of the
Food and Agriculture Organization of
United Nations, Rome, Italy
the United Nations, Rome, Italy
presented
thepapers.
United Nations, Rome, Italy
1
A special thank to the Cnr-Ivalsa as well as to the FCBA, FPInnovation, Virginia Tech and METLA who supported
the event.
ABSTRACT
FAO is a specialized international agency of the United Nations leading global efforts to eradicate
hunger, eliminate rural poverty and promote the sustainable management of natural resources. In a
Finally I would like to recall the fundamental work of my colleagues P. Burato, C. Capretti, E. Mele, P. Pestelli
world with an expected population of 9 billion people by 2050, demand for food and other products and
and L. Scalettiservices
to ensure
a good technical support to the conference.
from nature will be greater than ever. This pressure will change both the extent and the quality
of a substantial portion of the worlds forests. Understanding these changes is essential if investments
I hope this effort
will help to highlight the hardwood and to increase a sustainable management of our forests.
are to be made to improve forest management to both provide for a growing human population and to
maintain crucial forest biodiversity. FAO is committed to generating reliable information on forests and
forestry practices to promote the conservation and sustainable use of forest resources worldwide.
Clearly, additional investment is needed in research, whilst technical support, and field activities are still
Stefano Berti
needed to slow the loss and degradation of forest area, particularly in the tropics.
1. INTRODUCTION
The Food and Agriculture Organization of the United Nations (FAO) is a specialized international
agency leading global efforts to eradicate hunger, eliminate rural poverty and promote the sustainable
management of natural resources. FAO serves its 192 member countries3 as a neutral forum, retaining the
following core functions: 1) facilitate and support countries in the development and implementation of
normative and standard-setting instruments such as international agreements, codes of conduct, technical
standards and others; 2) assemble, analyse, monitor and improve access to data and information, in areas
related to FAOs mandate; 3) facilitate, promote and support policy dialogue at global, regional and country
levels; 4) advise and support capacity development at country and regional level to prepare, implement, monitor
and evaluate evidence-based policies, investments and programmes; 5) advise and support activities that
assemble, disseminate and improve the uptake of knowledge, technologies and good practices in the areas of
FAOs mandate; 6) facilitate partnerships for food and nutrition security, agriculture and rural development
between governments, development partners, civil society and the private sector; 7) advocate and communicate
at national, regional and global levels in areas of FAOs mandate.
FAO has 192 member states along with the European Union, Faroe Islands and Tokelau which are associate members
1
4
ABSTRACT
Scientific
Committee
FAO is a specialized international agency of the United Nations leading global efforts to eradicate
Stefano Berti - IVALSA, Italy
hunger, eliminate rural poverty and promote the sustainable management of natural resources. In a
Blanchet
- Universit
Laval,
Canada
world with an expected Pierre
population
of 9 billion
people by 2050,
demand
for food and other products and
services from nature willUrs
be greater
than ever.- VIRGINIATECH,
This pressure will change
both the extent and the quality
Buehlmann
USA
of a substantial portion of the worlds forests. Understanding these changes is essential if investments
Wolfgang Gard - University of Delft, The Netherlands
are to be made to improve forest management to both provide for a growing human population and to
Frederic
- FCBA,
Francereliable information on forests and
maintain crucial forest biodiversity.
FAO Rouger
is committed
to generating
forestry practices
to promote
and sustainable
use of forest
resources worldwide.
Jan-Willem
vanthedeconservation
Kuilen - University
of Munich,
Germany
Clearly, additional investment is needed in research, whilst technical support, and field activities are still
Erkki Verkasalo - METLA, Finland
needed to slow the loss and degradation of forest area, particularly in the tropics.
1. INTRODUCTION
Editorial Committee
The Food and Agriculture Organization of the United Nations (FAO) is a specialized international
Alexis Achim - Universit Laval, Canada
agency leading global efforts to eradicate hunger, eliminate rural poverty and promote the sustainable
3 Italy
MarcoFAO
Fioravanti
University
of countries
Florence,
management of natural resources.
serves its
192 member
as a neutral forum, retaining the
Torstenand
Lihra
- FPInnovations,
Canada
following core functions: 1) facilitate
support
countries in the
development and implementation of
normative and standard-setting Vernica
instrumentsLoewe
such as
international
agreements,
codes of conduct, technical
Muoz
- INFOR,
Chile
standards and others; 2) assemble, analyse,
monitor and
improve
access to data and information, in areas
Rmy Marchal
- CIRAD,
France
related to FAOs mandate; 3) facilitate, promote and support policy dialogue at global, regional and country
Jan Wiedenbeck - USDA Forest Service, USA
levels; 4) advise and support capacity development at country and regional level to prepare, implement, monitor
University
of Turin,
Italy and support activities that
and evaluate evidence-based Roberto
policies, Zanuttini
investments and
programmes;
5) advise
assemble, disseminate and improve the uptake of knowledge, technologies and good practices in the areas of
FAOs mandate; 6) facilitate partnerships for food and nutrition security, agriculture and rural development
between governments, development partners, civil society and the private sector; 7) advocate and communicate
at national, regional and global levels in areas of FAOs mandate.
FAO has 192 member states along with the European Union, Faroe Islands and Tokelau which are associate members
1
5
table of content
KEYNOTE SPEECHES
pag.
10
18
28
43
Pruning effects on the production of new epicormic branches: a case study in young stands
of pedunculate oak (Quercus robur L.)
G. Attocchi
44
Growth development of hardwood high value timber species in central south Chile, South America
V. Loewe, A. lvarez , and L. Barrales
50
Characterization of walnut hybrid plus trees for good quality wood plantations
I. Urbn, R. Licea, C. Gonzlez, F. Villar De Arcos, O. Santaclara, and E. Merlo
62
70
77
86
93
Studying the Time-Dependent Behavior of Sugar Maple by a Dynamic Mechanical Analyser (DMA)
A. Kaboorani and P. Blanchet
101
109
117
123
Solid state NMR evaluation of chemical changes on three hardwood species due to densification
I. Santoni, J.Sandak, A. Sandak, D. Pauliny, E. Callone, and S. Dir
131
pag.
139
140
148
Detection and classificaton of brittleheart in tropical hardwood in relation with timber strength
W. Gard, H. Kuisch, and J. W. van de Kuilen
157
Effect of compression and thermal modification on selected technical properties of European aspen and silver birch
V. Mttnen, M. Reinikkala, H. Herjrvi, and K. Luostarinen
165
Effect of pre-drying on colour and mechanical properties of thermally modified Silver birch timber
V. Mttnen and T. Krki
172
Influence of wood structure on the sorting of high quality walnut (Juglans nigra) timber
M. Hauptmann, M. Mller, H. Hanea, U. Mller, and C. Hansmann
180
185
Transient and Steady-State Heat Transfer through Platanus Orientalis, Fagus Orientalis
and Pinus Elderica: An Experimental Study
S. M. Rassoulinejad-Mousavi, A. Kharazipour, H. Hosseinkhani, and M. Ghorbanali
193
Mona Lisa wood Pavillon: optimization and parametric design by using poplar plywood engineered components
G. Callegari, G. Castro, R. MD Iasef, A. Spinelli and R. Zanuttini
202
210
219
220
227
Property variation in Black Ash (Fraxinus nigra M.), a case for its utilization in thermally modified timber (TMT) products 233
M. Leitch, B.Forbes, S. Miller, and J. Riffel
Investigations on the bondline discoloration of PVAc in wood assemblies
F. Chiozza, F. Toniolo, I. Santoni, and B. Pizzo
241
Bonding quality of ash (Fraxinus excelsior L.) for use in glued laminated timber
M. Knorz, S. Torno, and J.W. van de Kuilen
250
Adhesion and penetration of Polyvinyl acetate (PVAc) glues on some commercially valuable hardwood species
I. Santoni, B. Pizzo
258
266
276
283
291
pag.
POSTER
297
298
306
314
322
327
335
336
356
COMMUNICATION
Improving Hardwood Production Yields and Conserving Resources with Proper End Sealing
Norman E. Murray
339
342
347
352
354
359
363
364
Keynote Speaker
Director,
Forest Assessment, Management and
Conservation Division,
Forestry Department,
Food and Agriculture Organization of
the United Nations, Rome, Italy
ABSTRACT
FAO is a specialized international agency of the United Nations leading global efforts to eradicate
hunger, eliminate rural poverty and promote the sustainable management of natural resources. In a
world with an expected population of 9 billion people by 2050, demand for food and other products and
services from nature will be greater than ever. This pressure will change both the extent and the quality
of a substantial portion of the worlds forests. Understanding these changes is essential if investments
are to be made to improve forest management to both provide for a growing human population and to
maintain crucial forest biodiversity. FAO is committed to generating reliable information on forests and
forestry practices to promote the conservation and sustainable use of forest resources worldwide.
Clearly, additional investment is needed in research, whilst technical support, and field activities are still
needed to slow the loss and degradation of forest area, particularly in the tropics.
1. INTRODUCTION
The Food and Agriculture Organization of the United Nations (FAO) is a specialized international
agency leading global efforts to eradicate hunger, eliminate rural poverty and promote the sustainable
management of natural resources. FAO serves its 192 member countries3 as a neutral forum, retaining the
following core functions: 1) facilitate and support countries in the development and implementation of
normative and standard-setting instruments such as international agreements, codes of conduct, technical
standards and others; 2) assemble, analyse, monitor and improve access to data and information, in areas
related to FAOs mandate; 3) facilitate, promote and support policy dialogue at global, regional and country
levels; 4) advise and support capacity development at country and regional level to prepare, implement, monitor
and evaluate evidence-based policies, investments and programmes; 5) advise and support activities that
assemble, disseminate and improve the uptake of knowledge, technologies and good practices in the areas of
FAOs mandate; 6) facilitate partnerships for food and nutrition security, agriculture and rural development
between governments, development partners, civil society and the private sector; 7) advocate and communicate
at national, regional and global levels in areas of FAOs mandate.
10
FAO has 192 member states along with the European Union, Faroe Islands and Tokelau which are associate members
International
Scientific
Conference
International
Scientific
Conferenceon
onHardwood
Hardwood Processing
Processing (ISCHP
(ISCHP 2013)
2013)
FAOs policies and activities are guided by its member countries, mainly through the FAO Governing
and Technical Statutory bodies4.
2. CHANGING FORESTS RESOURCES
Forests are central to human life because of their capacity to continuously provide essential products
such as timber, fibre, fuelwood, medicines, food and other non-wood forest products; and crucial services for
biodiversity conservation, mitigation of climate change, soil and watershed conservation, natural disaster risk
reduction, and others. Forests also have fundamental social roles as many people depend on forests and forest
products for their livelihoods, having their survival strategies, cultural and recreational values deeply connected
with the existence of forest resources. Forests provide employment and income, contributing to local and
national economies and to reducing rural poverty.
Nevertheless, forests continue to be lost or degraded, especially in the tropics. FRA 2010 estimated the
global area of forest loss to be circa 13 million ha per year (FAO, 2010). While the pace of deforestation has
decreased in recent years in various countries, it continues to happen at an alarming rate. The causes of
deforestation and forest degradation are diverse and include land conversion (mainly for agriculture expansion),
fires and other natural hazards, as well as unsustainable and illegal forestry practices. Forests cover about one
third of Earths territory (around 4 billion ha), with wide variation from country to country. It is generally agreed
that the global forest area needs to increase and be sustainably managed in order to guarantee the continuing
generation of forest related benefits to meet the worlds needs.
While considerable efforts have been made, especially during the last decade, to improve forest
assessment and reduce deforestation and forest degradation, and while research has produced adequate methods
and tools for the sustainable management of forest resources, it is disturbing to notice that only a small part of
production forests are sustainably managed. FAOs Global Forest Resources Assessment indicates that 2.18
billion ha of forests (about 54% of the global forest area) are defined as production or for multiple uses (FAO,
2010). However, current estimates (through certification praxis) indicate that only about 19% of the forests being
used can be considered as sustainably managed, and this figure is substantially lower in the tropics (Sarre and
Sabogal, 2013).
This poses a critical problem for the sustainability of the resources, in a world where population is
expected to grow from 7 to 9 billion people in a generations time (by 2050) and where per capita consumption
patterns are changing, increasing demand for food and other products and services from nature. This will
increase the pressure over the planets limited natural resources: water, land, marine and terrestrial biodiversity.
The increasing demand for food, fibre and fuel can trigger substantial unplanned land use changes, including
large-scale forest clearance. While forests alone cannot solve the main global challenges related to food, energy
and climate change, they are a fundamental element for a long term solution and require continuing monitoring,
sustainable management and conservation practices. For these reasons, FAO is strongly committed to support
member countries implementing and delivering actions towards sustainable forestry.
3. THE GLOBAL FOREST RESOURCES ASSESSMENT
Measuring and monitoring the extent and quality of forest resources is one of the main pillars of
sustainable forest management (SFM). Since 1948, through the Forest Resources Assessment (FRA), FAO is
committed to monitoring and reporting the extent of the worlds forests.
Collecting and maintaining information on the extent of forest resources is an important task that
informs both policy and investment decisions by a wide variety of actors. Governments, donors, private
4
The Conference and the Council, the main Governing bodies, guide the overall policies and regulatory frameworks of the
Organization. The Committee on Forestry (COFO), six Regional Forestry Commissions and three Technical Commissions: the International
Poplar Commission, Silva Mediterranea and the Advisory Committee on Sustainable Forest-based Industries are the Technical Statutory
bodies that advise and support the Forestry Department in addressing forestry policies and activities..
11
International
Scientific
Conference
International
Scientific
Conferenceon
onHardwood
Hardwood Processing
Processing (ISCHP
(ISCHP 2013)
2013)
companies and individuals all make forest investment decisions and those that have reliable information on
forest change are most likely to succeed in maintaining healthy forests. Governments are also able to use forest
data submitted through the FRA in reporting to the Convention on Biological Diversity to assess progress
towards the Biodiversity Targets, and the United Nations Framework Convention on Climate Change to estimate
the forest carbon stock changes. The proportion of land area under forests is used as one of the indicators for the
Millennium Development Goals (UN 2006, UN 2008).
The FRA is currently undertaken every five years, working with countries in a transparent
documentation process. FRA 2010 included information on 233 countries and territories for 1990, 2000, 2005
and 2010. The next assessment will be released in 2015 and, for the first time ever, it will be produced using a
common questionnaire, harmonized by six major global or regional forest data collection and analysis
5
CFRQ: Collaborative Forest Resources Questionnaire is the result of the harmonization of six major global or
regional forest data collection and analysis processes from: 1) the International Tropical Timber Organization
(ITTO),2) Forest Europe, 3) the United Nations Economic Commission for Europe (UNECE), 4) the
Observatory of Central African Forests (OFAC), 5) the Montreal Process, and 6) the FAO Forest Resources
Assessment (FRA).
12
Production forests: Forest area designated primarily for production of wood, fibre, bio-energy
and/or non-wood forest product;
Multiple use forests: Forest area designated primarily for more than one purpose and where none
of these alone is considered as the predominant designated function.
When relating to protection and social functions of forests, the FRA classification gets more complex,
including:
-
Protection of soil and water: Forest area designated or managed for protection of soil and water,
with the following sub-categories: Protection for production of clean water, protection for coastal
stabilization, protection for desertification and control, protection for avalanche control,
protection for erosion, flood protection or reducing flood risks, protection for other soil and water
related purposes;.
Ecosystem services, cultural and spiritual values. Forest areas designated or managed for these
purposes may overlap with other designated or management functions. In FRA 2015 they will
include the following sub-categories: Public recreation, Carbon storage or sequestration, Spiritual
or cultural services;
Forest areas protected and designated primarily for conservation on biodiversity form another
category, including but not limited to forests within protected areas. They include forests in
protected areas under IUCN Categories I-IV but exclude IUCN Categories V-VI.
Another forest classification in FRA 2015 is for forest areas intended to be in permanent forest land use,
when they are highly unlikely to be converted to other land uses. This category includes both the Permanent
Forest Estate (PFE), i.e. forest areas designated by law or regulation to be retained as forest and private forest
lands that are likely to stay in forest land use.
FRA 2015 will also report on progress towards sustainable forest management at both the enabling
environment/policy scale and at the operational scale. Since there is a wide range of views on what constitutes
SFM, the FRA uses a series of questions that will allow users of FRA data to make their own informed
judgements about the degree to which forests are likely to be sustained. It will include forest areas that have a
long-term management plan and the degree to which governments monitor these plans. FRA 2015 will also
report both international and domestic forest management certification schemes with published standards and
independently verified by a third-party.
FAO-FRA used to report on forest cover according to the main species composition up to year 2000.
Then this classification was abandoned since few countries were able to collect this information and the
resulting assessment was considered too weak and unreliable.
6. TRENDS IN FOREST AREAS
FRA 2010 reported a total forest cover area of 4 billion hectares, equivalent to 31% of total land area.
The total cover area corresponds to an average of 0.6 ha per capita. Among the countries assessed, FRA reported
that ten countries or areas have no forest at all and an additional 54 have forest on less than 10 percent of their
total land area.
As previously mentioned, although there are signs of decreasing deforestation in several countries,
around 13 million hectares of forests are still converted to other land uses or lost through natural causes each
4
13
year. Forests are particularly threatened in the tropics and in arid or semi-arid countries. For example, in
Australia severe drought and forest fires have exacerbated the loss of forests in the period since 2000 - 2010.
Afforestation, reforestation and natural expansion of forests in some countries and regions have reduced
the net loss of forest area at the global level. The net change in forest area in the period 20002010 is estimated
at -5.2 million hectares per year (an area about the size of Costa Rica), down from -8.3 million hectares per year
in the period 19902000.
Afforestation, reforestation and sustainable forest management are key actions needed to guarantee the
generation of forest benefits, including wood and non-wood forest products, mitigation of climate change and
reduction of carbon emissions caused by deforestation and forest degradation. Figures from FRA 2010 indicate
an increase in forest area in countries where management plans are used. FAO estimates that around 18.5 million
people are directly employed in forest management and conservation but many more are directly dependent on
forests for their livelihoods.
7. IMPLEMENTING SUSTAINABLE FOREST MANAGEMENT
Countries and forest actors aim to achieve SFM to meet multiple needs such as the production of timber
and non-wood products, food security, poverty alleviation, biodiversity conservation, soil and water
conservation, and the development of rural communities, as well as to fulfil their international commitments (e.g.
the Non-Legally Binding Instrument on All Types of Forests, the International Tropical Timber Agreement, the
Convention on Biological Diversity, the Bonn Challenge for Forest Restoration, the Millennium Development
Goals, etc.).
There is renewed international recognition of SFM as an important component of sustainable
development; however SFM implementation remains a major complex challenge, as much as a huge opportunity.
8. DEFINING SFM
Sustainable forest management deals with sustainable use and conservation of existing forests and
their expansion in a way that maintains and enhances the multiple forest values through human intervention.
People are at the centre of the SFM as it attempts to contribute to societys diverse needs in perpetuity.
Forest management is the process of planning and implementing practices for the stewardship and use
of forests and other wooded land targeted at specific environmental, economic social and cultural objectives.
Forest management deals with the overall administrative, economic, legal, social, technical and scientific aspects
related to natural and planted forests. It implies various degrees of deliberate human intervention, ranging from
actions aimed at safeguarding and maintaining the forest ecosystem and its functions, to favouring specific
socially or economically valuable species or groups of species for the improved production of goods and
services.
The most widely internationally agreed-on definition of SFM is provided by the United Nations General
Assembly in the Non-legally Binding Instrument on All Types of Forests (NLBI) as follows: [a] dynamic and
evolving concept [that] aims to maintain and enhance the economic, social and environmental values of all types
of forests, for the benefit of present and future generations. This statement of intent makes it clear that SFM
will change over time but that its purpose, at a minimum, is to maintain all forest values in perpetuity. SFM is a
multidimensional concept as it incorporates the three pillars of sustainability, i.e. economic, social, and
environmental aspects.
Seven thematic elements of SFM have been
identified in the NLBI as a reference framework for
SFM (see Box 1). These elements, used to define the
Criteria and Indicators processes for SFM, allow
forest owners and stakeholders to assess SFM in
specific country or local conditions, i.e. defining what
the objectives of SFM are and how forests should be
managed to achieve them, always respecting the basic
14
BOX 1
Thematic elements of SFM
The seven thematic elements of SFM, as adopted in the UN
Forest Instrument on All Types of Forests, are:
1. Extent of forest resources
2. Forest biological diversity
3. Forest health and vitality
4. Productive functions of forest resources
5. Protective functions of forest resources
6. Socio-economic functions of forest resources
7. Legal, policy and institutional framework.
5
principle of perpetuity in the maintenance and enhancement of forest values that is embedded in the concept of
sustainability.
9. THE CPF FACTSHEETS
The Collaborative Partnership on Forests (CPF)6 has sought to tackle misperceptions and incorrect information
about SFM and highlight its relevant contributions to sustainable development. A voluntary arrangement
consisting of 14 international organizations, institutions and secretariats, in 2012 CPF released eight factsheets
underlining the important role of forests and sustainable forest management (SFM) for human well-being,
sustainable development and a green economy.
The factsheets focus on eight topics and are available at the CPF webpage (www.cpfweb.org/)
o
o
o
o
o
o
o
o
The 14 members of the CPF are: Center for International Forestry Research (CIFOR), Convention on Biological Diversity
(CBD), Food and Agriculture Organization of the United Nations (FAO), Global Environment Facility (GEF), International
Tropical Timber Organization (ITTO), International Union for Conservation of Nature (IUCN), International Union of Forest
Research Organizations (IUFRO), United Nations Convention to Combat Desertification (UNCCD), United Nations
Development Programme (UNDP), United Nations Environment Programme (UNEP), United Nations Forum on Forests
(UNFF), United Nations Framework Convention on Climate Change (UNFCCC), World Agroforestry Centre (ICRAF) and
World Bank.
15
International
Scientific
Conference
International
Scientific
Conferenceon
onHardwood
Hardwood Processing
Processing (ISCHP
(ISCHP 2013)
2013)
FAO is developing a comprehensive technical package of knowledge tools, best practices, processes
and their examples of application to facilitate the implementation of SFM under varied contexts, hereafter
referred to as the SFM Toolbox. Its target audience are primarily forest managers, as well as land use planners
interested in the role of forests in sustainable development, and those who support them including civil-society
organizations and non-governmental organizations and private sector associations.
The SFM Toolbox has the following objectives:
Increase capacities to effectively implement SFM by providing practical SFM-related knowledge, tools
and experiences at local and landscape levels;
Support forest managers and other stakeholders in progressing towards SFM in a more integrated and
cohesive way, with solutions tailored to their specific needs, contexts and major areas of interest, while
taking into account the general socio-economic and environmental contexts in which their SFM action
is developed, offering a more comprehensive approach to achieving SFM;
Connect the different areas of FAO expertise related to SFM to facilitate integration of SFM in a
landscape approach to natural resources management;
Identify and address knowledge and tool gaps, as well as possible incoherencies and conflicts.
The SFM Toolbox will consist of a web-based platform that is user-friendly, interactive and responsive
to users interests and needs. It will include: knowledge modules, database of SFM tools, database of SFM cases,
and guidance for national, sub-national and local SFM planning to advance its implementation.
The scope of the Toolbox is global (for all types of forests); however it will be an evolving set that will
also include regional and national SFM resources.
The Toolbox will make use of knowledge references, tools and cases produced by FAO, members of
the Collaborative Partnership on Forests, other partners and member countries.
12. FORESTS IN THE LANDSCAPE CONTEXT
Sustainable forest management cannot work in isolation or in constant competition with other land use
patterns. It must be a practice understood and accepted by society, adopted by local communities as part of their
regular activities and concern, and supported by the local and national authorities. SFM must generate income
for those directly involved in its practice, while maintaining the other benefits required for long-term production
and conservation of the resources. While forestry alone cannot solve the major global challenges related to
sustainable development, they are certainly part of the solution.
Hence the importance of bringing forests and forestry to other fora and debates to make sure other
sectors understand their contribution to economies, as well as for climate change mitigation, food security, clean
energy generation and less-energy intensive construction materials (wood or bamboo in replacing iron or
concrete), for conservation of biodiversity and for disaster risk reduction, and various other aspects that require
forests to be part of the territorial planning towards sustainability. An example of this effort can be identified in
the so called Model Forests, where individuals and groups supporting a wide range of forest values agree to
collaborate to devise and demonstrate ways to manage forest-based landscapes and natural resources using
approaches that are both locally acceptable and nationally relevant. In Model Forests, different people form
partnerships to manage their own natural resources according to their history, their economic situation and their
identity. The International Model Forest Network (IMFN) was launched by Canada in the early 1990s to
integrate the principles of sustainable development into forestry. IMFN includes more than 55 sites in over 24
countries on five continents, with thousands of partner organizations. The Model Forest of the Florentine
Mountains is so far the only existing Model Forest in Italy.
13. CONCLUSIONS
Despite the many difficulties associated with the concept of SFM, the management of some forests
today is consistent with it. Some forests have been managed for more than one hundred years; while it cannot be
said definitively that such forests are under SFM, the fact that they still exist and are still productive is prima
facie evidence of this.
16
Considering the prerequisites for SFM, one which is definitely crucial is that forest management
(particularly in the tropics) must become more profitable, which may involve better prices for wood and nonwood products, greater use of currently unmarketable species, payments for ecosystem services, subsidies, or
some other financing mechanism. Other necessary prerequisites include:
-
SFM embodies a process that is the best alternative available for maintaining and increasing the
contributions of forests to global well-being. The risks posed by resource degradation and depletion and by
climate change make SFM imperative: more than ever, humanity will need the products and ecosystem services
provided by forests.
There is a considerable body of knowledge and experience which can be shared and adapted to local
conditions to facilitate progress towards SFM. This knowledge and experience, from varied contexts, are
however quite diverse, not always accessible, and often focused on limited elements of SFM. Further research,
technical support, advice and exchange of knowledge are still needed to enhance the implementation of SFM.
REFERENCES
UN 2006 - Report of the Secretary-General on the work of the Organization, Document, A/61/1, New York,
United Nations
UN 2008 - The current official MDG framework. Official UN website for the MDG indicators
(http://mdgs.un.org/unsd/mdg/Host.aspx?Content=Indicators/About.htm)
FAO 2010 Global Forest Resources Assessment 2010, FAO Forestry Paper 163, Rome, FAO
2011 - Strategic Plan for the Florentine Mountains Model Forest 2011 - 2016
FAO 2012a Global Forest Land Use Change 1990-2005, FAO Forestry Paper 169, Rome, FAO
FAO 2012b FRA 2015 Terms and Definitions, Forest Assessment Working Paper 180, Rome, FAO
FAO 2013 - Sarre, A. and Sabogal, C, Is SFM an impossible dream? Unasylva 240, vol. 64 2013/1, Rome,
FAO
17
Holzforschung
Technische Universitt Mnchen
Mnchen, 80797, Germany
ABSTRACT
Tropical hardwoods are important species for high end structural applications where both mechanical resistance
and high natural material durability are required. In this paper, some historical data is presented about species
and applications. As markets require secure and reliable products, the issue of the importance of strength grading
is discussed. Grading is assessed both visually as well as by machine for which economic solutions exist nowadays.
Structural applications are presented as examples on how tropical hardwoods can be used in high value
applications. Such applications are also promising for developing countries, as the variety of infrastructure that
can be realized with tropical hardwoods is large. It allows countries to become more self-sufficient and less
dependent on other building materials that are less sustainable.
1. INTRODUCTION
Tropical hardwoods show some important advantages when compared with softwoods, traditionally being the
most widely used species in Europe. Tropical hardwoods generally offer higher biological resistance as well as better
mechanical properties. These are advantages with regard to technical applications. At the same time, questions have
been asked and are continuously being asked, about the sustainability of tropical hardwoods with respect to forest use.
Still, in many countries the value of certified wood with respect to origin and chain of custody is being questioned, by
pressure groups and the general public. In countries where building with tropical hardwoods has a longer tradition, the
discussions seem to have lessened slightly, with both the general public, pressure groups and timber traders, agreeing
upon the importance of both forest and chain-of-custody certification.
Whereas most tropical hardwoods were imported into Europe from Africa, some trading relations also go back to
French Guyana, British Guyana and Surinam and partly also Indonesia.
More recently, the introduction of FSC certified timber has resulted in
a large increase in species from other South-American countries as
Brazil, but also Peru and Bolivia. Species recently introduced on the
market comprise Angelim Vermelho (Dinizia exelca), Massaranduba
(Manilkara bidentata) and Cumaru (Dipteryx odorata), to name just a
few species used for heavy construction. A typical and challenging
application of these hardwoods is a mitre gate as shown in figure 1.
Timber sizes in such an application are typically around 350 x 350 mm,
sometimes even more.
Traditionally, such doors were made of Oak, treated for durability
with tar. During the last century such doors were increasingly made
from Azob (Lophira alata), which remains one of the most popular
tropical hardwood species used throughout Europe for heavy duty
applications. Other traditionally applied hardwoods used in heavy
construction are Demerara Greenheart (Tabebuia spp.) and Basralocus
(Dicorynia spp.), whereby the later normally is used in the form of
18
Edyn
(1)
4 f 2A2 U
For a large number of tropical hardwoods, the value of the dynamic modulus of elasticity has been determined and
compared against the static modulus of elasticity in a four point bending test (Ravenshorst, van de Kuilen, 2009).
Whereas the measurement of Edyn can be performed using both stress waves and ultrasound techniques, ultrasound is
less favourable. Tropical hardwoods are normally measured with a moisture content above the fibre saturation point.
Above FSP however, ultrasound wave speed is dependent on the moisture content, whereas stress waves are not
affected. A simple equation valid for the conversion of Edyn to Estat for all wood species reads:
E stat |
Edyn
(2)
1.15
Figure 2. Cross grain failure in a bending test of a sheet pile wall board
19
Trade name
Andira (sucupira
vermelho)
Angelim vermelho
Azob
Bangkirai
Basralocus
Bilinga
Botanical name
Andira spp.
Country/Region
Brazil
Dinizia excels
Lophira alata
Shorea spp.
Dicorynia spp.
Nauclea didernichi
Dypteryx spp.
Goupia glabra
Astronium lecointei Ducke
Brazil
West Africa
Indonesia
Surinam
West
and
Africa
Brazil
Brazil
Brazil
Tabebuia spp.
Milicia excels
Mizilaurus itauba
Lecythis spp.
Eucalyptus marginata
Eucalyptus diversicolor
Eucalyptus diversicolor
Koompassia malaccensis
Intsia spp.
Qualea paraensis D.
Guyana
Central Africa
Brazil
Brazil
Australia
Australia
South Africa
South-East Asia
South-East Asia
Brazil
D60
D40
D40
D40
D40
D50
D35
D60
D60
D40
Brazil
Cameroon
Honduras
Ghana(Cameroon)
Brazil
Brazil
D60
D40
C24
D50
D40
D50
Sapucaia
Tali
Manilkara spp.
Autranella congolensis
Terminalia spp.
Cylicodiscus gabunensis
Caryocar villosum
Aspidospermum
desmanthum
Lecythis Pisonis
Erythrophleum spp.
D50
D60
Tali
Teak*
Uchi torrado
Vitex
Erythrophleum spp.
Tectona grandis
Sacoglottis guianensis
Vitexcofassus spp.
Brazil
Cameroon
Kongo-Brazzaville
Ghana
South-East Asia
Brazil
Central Africa
Cumaru
Cupiuba
Gonalo Alves
(muiracatiara)
Demerara greenheart
Iroko*
Itauba
Jarana
Jarrah*
Karri (Australian) *
Karri (South African)
Kempas*
Merbau*
Mandioqueira (sucupira
amarelo)
Massaranduba
Mukulungu
Nargusta
Okan/Denya
Piquia
Piquia marfim
Strength class
D30
Central
D50
D70
D50
D24
D35
D60
D35
D40
D40
D40
D40
D30
STRENGTH
The density is generally considered as important for joints and connections, since the embedding strength of timber
is based on the characteristic density. However, an analysis of the relationship between board density and joint
strength reveals that the correlation is about zero. A number of tests have been performed. In the first series of tests, 4
meter long boards were measured for density and boards were matched in the sense that boards with similar density
were paired for use as side and middle members of joints respectively. In this way, series of joints could be
manufactured with almost the same density for all three members. Of course, in this case, differences in density with a
20
Figure 3 Relationship between density and load carrying capacity of joints for spruce (left) and bolted joints made with azob
(right)
For design purposes, the density requirement for strength class assignment may be lifted, for all species. This would
be important especially for hardwoods, as the density of hardwoods cannot be estimated on the basis of growth rings,
as is common practice for softwoods. A minimum density per species may be specified and the load carrying capacity
of joints could be determined directly. For the calculation of the embedding strength, a recent study showed that a
simple equation can be used for all species reading [Sandhaas et al., 2013]:
f emb
0.082 U
(3)
Such an equation is easy for the structural engineer as a change in fastener diameter during the design stage does not
change the embedding strength, making the design process more smoothly. Due to this high embedding strength, high
load carrying capacity connections can be achieved, reducing the amount of fasteners and timber cross sections at the
same time. This saves time and money during manufacturing.
2.3. DURABILITY AND BIOLOGICAL HAZARDS
In marine applications the use of tropical hardwoods can be advantageous. Two types of marine borers are common
almost around the world: shipworm, (Teredo spp. / Bankia spp.,) and gribble (Limnoria spp.), see figure 4. The risk for
marine borers is influenced by water temperature (minimal 5C) and salt level (minimal 7%). In addition, the pollution
Figure 4. Left: Shipworm (Courtesy Rijkswaterstaat), Right: Gribble and its decay pattern
21
22
WALLS
Along
canals
and
waterways, thousands of
kilometers of load bearing
sheet pile walls are installed.
These walls can be made of
timber boards up to 8 meters
long, but generally the longest
board length is around 6
meters. A typical board may
have dimension of 80 x 300 x
6000
mm,
giving
a
length/thickness ratio of 75
(This is not the span to depth
ratio). The relatively slender
boards are used in flatwise
bending and widths may vary
between 100 and 300 mm. An
Figure 6. Installation of sheet pile wall boards by means of vibration in weak soil
example of a sheet pile wall
during the construction stage
is shown in figure 6. It is a
typical application that can be used in wetlands with weak soils in estuaries and along rivers.
The upper side of the wall is generally fixed by an
anchor in the soil. Both chemically treated pine as well as
tropical hardwood is used. A specific character of the
boards is that they are coupled by means of a tongue and
groove system, allowing load sharing between boards, see
figure 7, with a maximum size of the tongue being 24 mm
(NEN 5493). The generally accepted allowable spring, 12
mm per 2m board length, cannot be applied for this
application and more stringent requirements need to be
applied. Installation of the sheet pile wall by vibration
becomes impossible when the boards do not fit together
over the full length of 6 meters. As a consequence, for this
specific application the accepted allowable spring is
maximized at 8 mm related to the full length of the board
instead of the internationally accepted reference length of
2 meters.
For this application, special load sharing factors have
been derived. (Van de Kuilen et al, 1996, Van der Linden
et al., 1994). Load sharing factors increase the design
strength as in the case of sheet pile walls, boards with low
strength are supported by neighbours with higher
strength, thus increasing the reliability of the system. This
leads to material savings. Whereas Eurocode 5 specifies a
Figure 7 Sheet pile wall boards with tongue and groove
general load sharing factor of kls = 1.1, for sheet pile walls
this factor has been determined at 1.33 and 1.15 for pine
and tropical hardwoods respectively. These values are higher than the one given in Eurocode 5 because of the plastic
behaviour of timber in compression in saturated conditions. This allows for much more plasticity and thus load
transfer between boards in a sheet pile wall application. The load sharing for walls made of softwoods is higher than
for tropical hardwoods because of the larger coefficient of variation in bending strength of softwoods as compared to
hardwoods.
23
Figure 9. Tensile joint in sawn timber member with slotted in steel plates and 10 dowels + bolt
(Sandhaas, 2012). The high quality craftsmanship allows for nice architectural applications. However, the materials
used are available throughout the world allowing for developing countries the use of tropical hardwoods in road
bridges, without the need of bringing non-sustainable
building materials to remote building sites.
3.3. MOORINGS AND FENDERS
Moving ships need space and guidance when docking and
entering near harbours and waterworks. Mooring and
fendering structures such as shown in figure 15 allow for this,
but need to be designed for impact loads, which can be very
high in the case of ships. Even though the time frame of
moving ships are in the order of the time frame where the
short term strength of timber is determined (seconds to
minutes), there is a great need for energy dissipation. The
energy of moving ships need to be dissipated by elastic
deformation of the structure. Here wood has the advantage of
24
Vehicle type
TB 11
Impact speed
(km/h)
100
900
Car
TB 51
70
20
13000
Bus
Test
The curved steam bent element is made of Brazilian wood species angelim vermelho and is meant to dissipate
energy from small cars that hit the rail and to guide them along the rail.. It is connected to the piles through two bolts
and a teflon spacer of 20 mm thickness, to allow for more energy dissipation at car impact and to allow dirt to wash
away to the ground. The steaming process compresses the wood fibres parallel to the grain and upon stretching avoids
the wood fibres from breaking instantly. This is important as a key requirement of the highway authority is that pieces
detaching from the rail during an incident needs to be avoided as far as possible. As a result of the steam treatment, the
element shows large deformability and thus energy dissipative capacity. The deformation and energy dissipating
capacity is shown in figures 12 and 13. The area under the curves shown in figure 13 is the energy diisipative capacity
and is used in numerical simulations to assess the car passengers safety. The steam bending technology is an
established technology, traditionally used in furniture making.
25
Figure 12. Three stages of bending of the steam bent element with car rail, connected to the pile with two bolts and
white coloured spacers.
Load-displacement behaviour of joint
12
Oak-1
Oak-2
Oak-3
Angelim-1
Angelim-2
Angelim-3
Angelim-4 (spacer)
Angelim-5 (spacer)
10
Load (kN)
8
6
4
2
0
0
50
100
Displacement (mm)
150
200
26
27
EU FLEGT Facility
European Forest Institute
Barcelona, Spain, 08025
EFINORD
European Forest Institute
Copenhagen, Denmark,
1958
EFICEEC-EFISEE
European Forest Institute
Zagreb, Croatia, 10000
ABSTRACT
Implemented in March 2013, the EU Timber Regulation is affecting hardwood exporters and importers. The EUTR
requires proof of timbers origin and legality to ensure that no illegal timber is imported into the EU. The EUTR is
part of the EUs Forest Law Enforcement, Governance and Trade (FLEGT) Action Plan, with its specific goal to
end illegal logging, thereby improving sustainability of forest resources. To this end the EU intends to block
imports of any wood or wood product which comes from unknown sources. Certification of SFM will help EU
importers minimize risk, which is an essential part of their required due diligence system. Monitoring organizations
are established to assist trade associations and businesses to construct comprehensive due diligence systems.
National competent authorities are designated to follow the trade of the new FLEGT-licensed timber and timber
products.
In 2013, the first year of the EUTR, there are positive impacts, of which the most important is awareness of the
disastrous situation with illegal logging, driven by exports of illegal timber. Another positive development is
tropical timber exporters documenting the legality of their wood exports. Yet another positive feature is
establishment of due diligence systems by EU importers.
However, there are considerable problems to ensuring legal trade. For example, China exports primary- and
secondary-processed hardwood products to Europe, based on imported logs which most often do not have
comprehensive documentation of their origin and legality. Other problems exist too, and the implementation of the
EUTR may eventually be defined similar to the US Lacey Act Amendment, i.e. by the courts of law.
Senior Timber Trade and Policy Analyst; Tel. +358 50 375 2685; E-mail: Ed.Pepke@efi.int
Trainee; Tel. +46708196160; E-mail: Alexandru.Giurca@efi.int
3
Senior Researcher; Tel. +390332786592, Fax +39 0332 785230; E-mail: Ragnar.Jonsson@jrc.ec.europa.eu
4
Manager EFI Central-East and South-East European Regional Office; Tel. +385 1 2431 517; E-mail: Marko.Lovric@efi.int
2
28
1. INTRODUCTION
The importance of legal and sustainable trade of forest products is increasingly recognized by governments,
international organizations, non-governmental organizations, trade associations, timber traders and the wood
processing industries. The illegal trade of wood has a detrimental effect on the forest sector as it undermines the legal
trade by reducing prices and profits. Furthermore, illegal trade deprives governments of legitimate revenues. Too
much of the illegal trade stems from illegal logging which is disastrous for the worlds forests. Illegal logging is but
one cause of deforestation, which remains far too high, even though it fell from 16 million hectares per year from
1990-2000 to 13 million hectares per year in 2000-2010 (FAO, 2011). Illegal logging and trade tarnish the image of
the forest sector.
Estimates of the extent of illegal logging and the trade of illegal wood products vary and can never be accurate as
there will never be statistics on this hidden commerce. However, the World Bank estimated the economic loss from
illegal wood trade to be US$10 billion annually, and an additional US$5 billion in tax evasion and royalties (World
Bank, 2012).
Illegal logging has severe environmental, social and economic consequences. Its negative impacts include:
biodiversity loss, destruction of animal and human habitats, disruption of water cycles, increases of carbon emissions,
climate change, undermining rural livelihoods, reducing competitiveness of legitimate forest industry, lost revenues
for Governments, undermining the rule of law and fostering corruption.
This paper for and its presentation at the International Scientific Conference on Hardwood Processing 2013 cover
the hardwood (temperate and tropical) market trends, the European Union policies affecting the markets and the
impacts on hardwood markets. A network analysis of EU-27 trade flows of all forest products reveals clustering of
member countries in four groups depending of their centrality. Longitudinal analysis of trade flows shows statistically
significant impact of FSC and PEFC certification schemes on the volume of trade, as well as moderate level of
correlation in the trade flow patterns between continental and tropical roundwood and sawnwood.
1.1 European Union Forest Law Enforcement, Governance and Trade Action Plan
The European Union (EU) established the Forest Law Enforcement, Governance and Trade (FLEGT) Action Plan
in 2003 to fight against illegal logging and associated trade, which have devastating effects on the worlds forests and
the people who depend upon them (European Union, 2012). The FLEGT Action Plan was a breakthrough in
recognizing the shared responsibilities of producing countries and of consuming countries in illegal logging and trade.
It proposes combined demand- and supply-side measures to overcome these issues. By focusing on the root causes of
illegal logging, particularly governance and enforcement issues, it promotes legal timber. Before the FLEGT Action
Plan there was no strategic legislation that focused on halting the placement of illegally harvested timber on the EU
market.
The Action Plan focuses on seven broad areas:
1.
2.
3.
4.
5.
6.
7.
Support to timber exporting countries, including action to promote equitable solutions to the illegal logging
problem
Activities to promote trade in legal timber, including action to develop and implement Voluntary Partnership
Agreements between the EU and timber exporting countries
Promoting public procurement policies, including action to guide contracting authorities on how to deal with
legality when specifying timber in procurement procedures
Support for private sector initiatives, including action to encourage private sector initiatives for good practice
in the forest sector, including the use of voluntary codes of conduct for private companies to source legal
timber
Safeguards for financing and investment, including action to encourage banks and financial institutions
investing in the forest sector to develop due care procedures when granting credits
Use of existing legislative instruments or adoption of new legislation to support the Plan, including the EU
Timber Regulation
Addressing the problem of conflict timber.
The cornerstone of the Action Plan is the Voluntary Partnership Agreement (VPA) which is a legally binding
agreement between the EU and partner countries, by which all timber exported is verified as legal through a legality
29
assurance system (LAS). This LAS consists of legality definition established through a multi-stakeholder process that
clarifies what is legal timber, procedures for legality verification, tracking system, issuing of FLEGT licenses and an
independent audit. Under the VPA, countries agree to control and license timber exports as legal, and the EU agrees to
import only licensed timber from those countries.
1.2 EU Timber Regulation
One essential part of the FLEGT Action Plan was the introduction of the EU Timber which came fully into effect
in March 2013Regulation (European Union, 2012). It reinforces the VPA initiative and complements it. It focuses on
the changes of behaviors of business operators1 in the European Union that must establish due diligence systems
(DDS) to be able to prove that all timber products they handle are legal. DDS must contain comprehensive information
about operators supply of timber, assessment of risk and its mitigation, thorough recordkeeping, availability for
inspection by EU Competent Authorities (European Union, 2013). These have important consequences for the entire
timber trade, especially the hardwood trade since the VPAs are being established with tropical timber producing
countries.
The EUTR requires operators that place timber on the EU market, e.g. the first importer, to exercise due diligence
to minimize the risk of illegal timber imports (European Union, 2013). It applies to imported and timber harvested
within the EU, hence, the Regulation is in accord with the World Trade Organization rules.
The EUTR applies to a wide range of timber products, from roundwood to primary-processed products such as
sawn hardwood, to secondary-processed products such as wooden furniture and to paper products. While FLEGT
products are commonly thought to be industrial roundwood, sawnwood, plywood and veneer, the EUTR also covers
fuelwood, particleboard, fiberboard, packaging such as pallets, builders joinery and carpentry and prefabricated
wooden buildings. A full listing of products covered appears in the EUTR annex (European Union, 2010).
FLEGT-licensed timber will be produced by the VPA countries. This license will verify that the timber is legal,
and will be easily imported into Europe, and other countries requiring proof of legality. The first FLEGT-licensed
timber is expected to be traded in 2014. Imports of FLEGT-licensed timber will comply with the EUTR, minimizing
due diligence requirements. Trade of FLEGT-licensed timber is designed to combat illegal logging and promote
sustainable management of tropical forests. Under the EUTR, wood carrying a permit by the Convention on
International Trade in Endangered Species of Wild Fauna and Flora (CITES) is also considered to be compliant.
Implementation of the EUTR will be controlled within each EU Member State by a designated Competent
Authority. For example, in Italy where this conference is being held, the Competent Authority is the Ministry of
Agriculture, Food and Forestry. The European Commission is accrediting Monitoring Organizations to assist
companies and trade associations in establishing and maintaining due diligence systems (European Union, 2013).
1.3Similar legislation in other countries
The EU is not alone in its quest to ensure legal timber trade. The United States amended the Lacey Act in 2008 to
prevent illegal logging by banning imports of illegal timber products (EU FLEGT Facility, 2012). The Lacey Act
Amendment (LAA) differs from the EUTR in that the LAA applies uniform requirements on all operators, including
US timber importers, traders, processors, middlemen, wholesalers and retailers. Like the EUTR, for the LAA,
certification of sustainable forest management is not sufficient to prove legality. The LAA prohibits any person from
importing, exporting, selling, receiving, acquiring or purchasing any plant (e.g. wood or paper product), knowing that
it was taken, possessed, transported or sold in violation of the underlying laws or regulations. The prohibition extends
to operators abroad. There is also a precedent available for prosecuting non-US citizens under the LAA. The penalties
and sanctions in the LAA vary according to the severity of the offence, the value of the goods and the type of operator,
with larger corporations paying the highest penalties.
Australia has a relatively new law to protect its imports from any illegality. The Illegal Logging Prohibition Act
(Australian Government, 2012) has the same intent as the EUTR to curb illegal logging by blocking imports of illegal
timber. Illegal logging is defined as harvested in contravention of laws either in the country of harvest, or in Australia.
Importers are required to practice and prove due diligence. Australian wood processors must also be able to prove due
diligence in the sourcing of their wood products, and must make an annual statement to the Department of Agriculture,
Fisheries and Forestry. Like the EUTR, this legislation applies to imported and domestic timber, thus it will conform
The EU defines operators as those who first place timber products on the EU market. They carry the main responsibility for the
EUTR requirements. Traders who buy or sell timber products already within the EU are only required to keep records of from
whom they buy and to whom they sell.
30
to World Trade Organization requirements. Failure to comply is punishable by fines and incarceration as well as
forfeiture of the wood products.
New Zealand has a voluntary code of practice (MPI 2010a). Under this code, not only are importers expected to act
responsibly when importing timber or timber products, but also consumers are encouraged to act responsibly and to
buy only those products that are labeled as legal under the code.
In 2006, Japan launched a green purchasing policy that covers legality of wood, after the discussion on illegal
logging among the G8 countries in 2005 (Government of Japan 2007). The main objective of this law is to encourage
the State, independent administrative institutions and local governments to buy eco-friendly goods, including legal
wood.
Other importing countries are building upon these experiences to be protect their markets from any illegal timber.
1.4 Exporters responses to the EU Timber Regulation
On the other hand, exporting countries are establishing legality verification schemes to meet EUTR requirements
to prove their exports legality. All the VPA countries are developing such systems. China, the worlds greatest
exporter of furniture, exceeding Italy and Germanys combined exports since 2005 (CINTRAFOR, 2011), has
established the China Timber Legality Verification Scheme (CTLVS) to provide a suitable approach to meet the new
timber legality requirements of international markets, especially the US Lacey Act Amendment and the EUTR
(Proforest, 2011). The CTLVS is based on Chinas domestic timber management and control system. Domestically
produced timber could be fully qualified as verified legal timber by Chinas three-permit system. This system includes
a forest harvesting permit, a timber shipment permit and a timber processing permit. The scheme requires
China-authorized organizations to issue timber legality certificates. The Chinese customs office will check the
certificates when Chinese companies export their timber products. A wood tracking system is needed for both schemes
once timber products enter the country. There are still many challenges to be overcome in the progress of
implementing the CTLVS.
2. MARKET DEVELOPMENTS FOR HARDWOODS
2.1 Global hardwood trade
As the EUTR impacts Europes trade, this section places Europes and the VPA countries trade into the global
context. And as the conference is dedicated to hardwoods, this section examines the importance of temperate and
tropical hardwood trade in relation to total timber trade.
The global trade of wood and paper products grew steadily over the past 20 years, doubling from 1992 to 2007,
until the global economic crisis in 2008 (figure 1). Since 2009, trade is growing again. These trends correlate with
growth in gross domestic product and population.
Roundwood equivalent in m3
900
800
700
600
500
400
300
200
100
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
31
Based on value, primary-processed tropical timber, of which most but not all is hardwood, accounted for 34% of
global trade in 2012 (figure 2). Tropical timber generally carries a higher unit value leading its market share to be
higher than when measured in volume. Globally temperate timber, hardwood and softwood, has a greater share of the
volume, which was 83% according to ITTO; 17% was tropical by volume (ITTO, 2013).
50
45
40
US$ billion
35
30
25
20
15
10
5
0
2003
2004
2005
2006
2007
Temperate
2008
2009
2010
2011
2012
Tropical
Demand for hardwood (temperate and tropical), using the sawnwood consumption as an example, fell starting 20
years ago in 1994 and continued downward until 2003. The weakening demand resulted from a combination of
factors, including substitution by wood, e.g. reconstituted panels, and non-wood, e.g. melamine-coated flooring.
Additionally, political problems including the economic crisis of 1998 and infra-structure problems constricted supply
in some tropical countries. Restrictions on exports of roundwood to encourage value-added processing within tropical
countries, along with reductions in harvests, either to regain sustainable production or due to deforestation, also
reduced exports. However, the markets strengthened until the global economic crisis in 2008 (figure 3). Demand
began rising again in 2010 and has recovered to 124 million m3, approximately the level of 1994 (FAO, 2013).For
reference, sawn hardwood consumption in 2012, at 124 million m3, was 44% of the sawn softwood consumption that
same year.
32
140
120
Million m3
100
80
60
40
20
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Using the most commonly traded hardwood product, sawnwood, according to the ITTO, tropical imports rose in
volume until 2004 and were falling well before the global economic crisis, i.e. from 2005 (figure 4). In part the decline
was due to the factors mentioned above, but to an extent concern about illegality and deforestation caused buyers to
substitute temperate hardwoods. The lack of tropical timber certified for sustainable forest management also caused
traders, retailers and consumers to use alternative woods. Furthermore, some countries, e.g. Ivory Coast, suffered from
deforestation and were unable to supply prior volumes. In 1994, the earliest ITTO statistics on tropical sawnwood
trade show that tropical sawnwood accounted for 37% of the traded volume. But that slipped to a low of 29% in 2002,
before rising back to approximately one-third (34%) in 2012. Sawn hardwood markets remained stronger in value,
maintaining market strength through 2007, before the 2008 downturn; since 2009 the trade values of sawn hardwoods
has been rising (figure 5).
35
30
Million m3
25
20
15
10
5
Temperate
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
Tropical
33
10
9
8
US$biillion
7
6
5
4
3
2
1
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
US$ billion
20
15
10
5
0
2003
2004
2005
2006
2007
Temperate
Note: Based on 4 primary products.
Source: ITTO, 2013.
34
2008
2009
2010
Tropical
2011
2012
Most of the EUs imports of tropical timber is in the form of sawnwood in terms of volume (figure 7). Industrial
roundwood, which is sawlogs and veneer logs, have been decreasing, in large part as tropical producers strive to
increase value-added processing.
4500
4000
3500
3000
2500
2000
1500
1000
500
0
2001
2002
2003
2004
2005
Industrial roundwood
Source: ITTO, 2013.
2006
2007
Sawnwood
2008
Veneer
2009
2010
2011
2012
Plywood
Currently in 2013 the EU has signed, or is in the process of negotiating VPAs with 14 countries: Cameroon,
Congo, Democratic Republic of Congo, Gabon, Ghana, Guyana, Honduras, Indonesia, Ivory Coast, Liberia, Malaysia,
Thailand and Vietnam. Eventually all of these countries could export FLEGT-licensed timber to the EU. These
countries accounted for approximately two-thirds of the EUs tropical primary-processed timber imports in 2012
(figure 8).
3000
Million euros
2500
2000
1500
1000
500
0
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Non-VPA tropical countries
14 VPA countries
It should be noted that the EUs value of the trade of secondary-processed products, e.g. wooden furniture, exceeds
that of primary-processed products. In 2012, EU imports of secondary-processed products were 30.3 billion euros,
exceeding primary-processed products at 20.1 billion euros by approximately 50% (figure 9). Note that this trade
includes imports within Europe.
40
30
35
International
Conference on products
Hardwood
Processing
(ISCHP plywood
2013) and veneer).
Note: 4 Scientific
FLEGT primary-processed
(roundwood,
sawnwood,
Source: Comext via Eurostat, 2013.
It should be noted that the EUs value of the trade of secondary-processed products, e.g. wooden furniture, exceeds
that of primary-processed products. In 2012, EU imports of secondary-processed products were 30.3 billion euros,
exceeding primary-processed products at 20.1 billion euros by approximately 50% (figure 9). Note that this trade
includes imports within Europe.
Billion euros
40
30
20
10
0
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Secondary
Primary
Core countries: Austria, Belgium, Finland, France, Germany, Italy, Netherlands and Sweden
Core dependent countries: Bulgaria, Cyprus, Denmark, Greece, Luxembourg, Malta Poland, Portugal and
Spain
Independent periphery: Czech Republic, Estonia, Hungary, Latvia, Lithuania, Romania, Slovakia and
Slovenia
This grouping is presented in Figure 10, where the closeness of countries and the thickness of lines both represent
the strength of bilateral trade, size of the arrowheads represents strength of export, and the size of the circles represents
the volume of production of all forest products. The more central a country is, the closer it is to the centre of the
graph.
36
There is an observed tendency of higher shares of forestry in GDP to increase export of forest based products
(p<0.01), whereas the share of forestry in GDP of the destination countries plays no role (p0.1).Having more FSC
and PEFC certificates positively influences export, whereas the FSC influences the export marginally more than PEFC
(by 9% at p<0.01). However, having more FSC certificates has no effect on import, whereas number of PEFC
certificates has positive effect on import from countries that have similar number of PEFC certificates (p<0.01)..
Imports of continental and tropical roundwood have no significant impact on exports, but they do show moderate
correlation (0.337); the same situation is with continental and tropical sawnwood, but the correlation is here stronger
(0.608).
There is a tendency (p<0.01) of increasing the volume of trade, where it is preferred that the future partners are
strongly engaged at bilateral trade (effect reciprocity at p<0.001) and that they are engaged within small clusters of
intensive trade (effect transitive triplets at p<0.01). It is very unlikely that EU-27 countries will choose trading partners
with high corruption perception index (p<0.001), and there is no grouping of countries depending on the time in which
they have joined the EU (EU membership effects have p0.1).
Data on bilateral trade of all forest products was gathered from export data of EFI`s Forest Product Trade Flow
Database. Data on production of all forest products as well as for total imports of different product groups (roundwood
and sawnwood from continental and tropical sources) was gained from FAOSTAT. Data on share of forestry in GDP
was gained from UNECE database, sixed onto the state in 2000, and expressed in per mil. Data on import of illegal
timber was gained from WWF, fixed for state in 2004, and expressed in millions cubic meters roundwood equivalent
(RWE) volume. EU enlargement process was shown in a six-point ordinal scale ranging from European Coal and Steel
Community in 1952 up to the wave of EU enlargement in 2007. Cross-sectional analysis was performed in UCINET,
NetDraw and KeyPlayers software, with 1000m3 as a unit of change. Longitudinal analysis was performed in R
software, packages SIENA and NETWORK, with 10,000m3 as a unit of change.
3. IMPACT ON HARDWOOD MARKETS
3.1Impact on hardwood trade
37
timber was gained from WWF, fixed for state in 2004, and expressed in millions cubic meters roundwood equivalent
(RWE) volume. EU enlargement process was shown in a six-point ordinal scale ranging from European Coal and Steel
Community in 1952 up to the wave of EU enlargement in 2007. Cross-sectional analysis was performed in UCINET,
NetDraw and KeyPlayers software, with 1000m3 as a unit of change. Longitudinal analysis was performed in R
International
ScientificSIENA
Conference
on Hardwoodwith
Processing
(ISCHP
2013)
3
software, packages
and NETWORK,
10,000m
as a unit
of change.
3. IMPACT ON HARDWOOD MARKETS
EU Timber Regulation Impacts Global Hardwood Markets
The EUTR is intended to have positive impacts on the hardwood trade, especially the tropical hardwood imports
into Europe. If tropical timber exporting countries are to benefit from sustainable management of their forests, through
planned legal harvesting, processing and exports, then there must be a strong market demand. As shown in the
previous section, the demand for hardwoods, especially tropical hardwoods, has been weak since the 2008 global
economic and financial crisis. The continuing recessions in many EU countries and in the euro zone, mean demand is
low. This is evident in the most important demand generator for hardwoods, i.e. housing construction. All construction
is stagnant in Europe with housing starts at almost half of its 2007 level of 2 million units (single- and multi-family
units combined) (Euroconstruct, 2013).Following a 4.3% downturn in residential construction in 2012, Euroconstruct
forecasts a 2.2% downturn in 2013 and then a small improvement in 2013 (1.1%) and 2014 (2.3%).According to
Euroconstruct, the reasons for the depressed housing market are high unemployment and lack of consumer
confidence.
But there are other important reasons for the decline in tropical timber trade. Consumers and retailers are concerned
and confused about the issues of tropical forests degradation and deforestation1. As a result, the demand for tropical
timber has weakened. However, this concern has bolstered the market for certified forest products which give retailers
and consumers the assurance that the wood products they buy and sell are sustainably produced.
3.2 Possible substitution
Substitution is another reason for the downturn in the demand for tropical timber. Temperate hardwood is often a
viable option in certain uses for tropical timber. Some tropical hardwood species have natural durability and have
traditional markets in applications where exposed to water and weather. However, thanks to research and development
of surface coating, heat treating and chemical impregnation, temperate hardwoods and softwoods perform equally in
many applications. Over the last few years oak has consolidated its dominant market position in the European flooring
and joinery sectors while tropical hardwoods have continued to lose market share. Indeed, results from econometric
analysis indicate a significant positive cross-price elasticity, implying that oak and tropical lumber are substitutes
(Giurca, 2013).
Choice of hardwood species is often dictated by fashion. Architects and designers often influence fashion, and
therefore are a key target market for organizations promoting temperate or tropical hardwoods (AHEC, 2013). The
need to protect reputations of architects, designers and their clients, e.g. governments, has weakened demand for
timber which risks being unsustainably or illegally produced.
3.3 Impact on exporters to EU
Under the EUTR, FLEGT-licensed timber products will be easily imported into Europe. It is the objective to
strengthen the tropical timber imports in order to drive change in VPA countries. Those changes are to be in legislation
and its enforcement, as well as economic change derived from profitable exports of sustainable, legal timber. Thus, the
EUTR is poised to increase the tropical timber trade between VPA and EU countries.
The EUTR will also impact non-VPA countries exporting to Europe. One case would be for countries processing
imported timber, and then exporting value-added products, e.g. China. As mentioned above, Chinese manufacturers
are developing their documentation of the source of imported and domestic timber in order to offer assurance of
legality to their customers in the EU. Another case would be for countries which have forest resources, such as Russia,
which has the worlds largest softwood resource, but also a rich hardwood resource. One of Russias main export
markets is the EU, and Russian exporters will need to be able to prove the legal source of the products they sell.
3.4 Impact on companies and their associations
Potential unintended impacts of the EUTR on tropical hardwood markets can be discussed in terms of ambiguity2.
Hence, as the effects and requirements of the EUTR are not yet fully known or understood, ambiguity can arise.
Empirical evidence in the economic literature shows that agents normally display aversion to ambiguity. This implies
that, when they take economic decisions, they tend to focus on worst case scenarios (cfr. Ellsberg, 1961; Camerer and
Weber 1992; Etner, Jevela and Tallon, 2009). This was supported by higher EU imports of primary-processed tropical
timber in the two months preceding the EUTR, i.e. January and February 2013, (up by 12% to 281,000 m3) than in the
1
Reduction of deforestation is a key objective of the program on Reducing Emissions from Deforestation and forest Degradation
(REDD+).
2
It is beyond the scope of this paper to provide an accurate treatment of preferences incorporating ambiguity. It suffices to say that
Knight (1921) describes as ambiguous an economic context wherein the lack of accurate information prevents an agent to uniquely
derive a probability distribution over the reference state space, so that also the expected values of the variables under consideration
are compatible with more than one probability distribution.
38
following two months (243,000 m3); normally EU imports of tropical wood tend to be evenly distributed in the first
four months of the year (ITTO MIS, 2013).
Thus, concerns around documentation and licenses could result in ambiguity among operators in the EU. In doubt
about the legality of tropical timber, and hence of its real value, this could push them away from tropical hardwood
imports. They could instead decide to trade with partners coming from countries with (perceived) more reliable
documentation, such as temperate hardwood or softwood exporters. Hence, EU timber importers (operators), traders
and processors, and their trade associations, are establishing DDS. In order to minimize risk, they must have assurance
of the legality of their wood. This is relatively easy for timber obtained within the EU, as well as some other sources,
e.g. the United States. It is not as easy for wood products obtained with long series of transactions, where
chain-of-custody does not exist. As a result, EU businesses will choose safe sources, which could lead away from
less-documented tropical and temperate timber (both hardwoods and softwoods), or even substitute non-wood based
commodities for wood. Further, exporters of tropical timber, uncertain about the costs of complying with legal
requirements, could opt to trade with partners with less stringent legislation.
Many European timber associations have strongly supported the EUTR as evidenced by announcements on their
websites and newsletters. Other countries associations have also published in favor of the EUTR objectives. Within
Europe and outside associations are assisting member companies to develop DDS or to prepare satisfactory proof of
legality. Individual companies, e.g. Cor in Italy, have expounded the benefits of the EUTR and incorporated the
requirements into their business practices (Cor, 2013). In a SWOT analysis, Cor cites these opportunities of
implementing the EUTR:
As part of the SWOT analysis, Cor also presents the EUTR weaknesses, strengths and threats. They conclude that
we all need a level playing field of competition, both in producing countries and in the EU. [From] the business point
of view: illegal logging and trading products of illegal logged forests represents unfair competition.
4. CONCLUSION
While the EUTR only came into effect this year in 2013, it is already impacting the global timber trade. The
greatest effects to date are on the EU trade as companies and their trade associations establish DDS to mitigate the risk
of dealing with illegal timber. To ensure legality of their sources, they are requiring their customers to provide
documentation of the chain-of-custody of the timber back to the area of harvest this means significant changes for
suppliers of timber to the EU. As the EUTR applies to timber traded within Europe, adequate documentation is
required between buyers and sellers of EU member states.
Eventually the EUTR, combined with similar legislation in other countries, should improve the markets for
hardwood timber. Having assurance of legality, and precluding illegal timber from the market, will strengthen markets
by raising prices and profits. Governments will receive greater revenues which can be returned to develop the forest
sector in producing countries.
Potential unintended impacts of the EUTR can result from concerns around documentation and licenses and
ensuing ambiguity. Operators in the EU could be dissuaded from importing tropical hardwood, instead deciding to
trade with partners with (perceived) more reliable documentation. Hence, EU businesses could choose safe sources
rather than less-documented tropical and temperate timber (both hardwoods and softwoods), or even substitute
non-wood based commodities for wood. Further, exporters of tropical timber, uncertain about the costs of complying
with legal requirements, could choose to trade with partners with less demanding legislation.
4. LITERATURE CITED
American Hardwood Export Council (AHEC). www.americanhardwood.org. Accessed 9 August 2013.
39
ensuing ambiguity. Operators in the EU could be dissuaded from importing tropical hardwood, instead deciding to
trade with partners with (perceived) more reliable documentation. Hence, EU businesses could choose safe sources
rather than less-documented tropical and temperate timber (both hardwoods and softwoods), or even substitute
non-wood based commodities for wood. Further, exporters of tropical timber, uncertain about the costs of complying
International
Conference
on Hardwood
(ISCHP
2013)
with legal Scientific
requirements,
could choose
to trade Processing
with partners
with less
demanding legislation.
4. LITERATURE CITED
EU Timber Regulation Impacts Global Hardwood Markets
Products
Industry.
database.
European Union. 2010. Regulation 995/210 of the European Parliament and of the Council. Official Journal of the European
Union. http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2010:295:0023:0034:EN:PDF. Accessed 31 July 2013.
European Union. EU Timber Regulation.2010. http://ec.europa.eu/environment/eutr2013/index_en.htm. Accessed 31 July
2013.
European Union. The EU FLEGT Action Plan. www.euflegt.efi.int/portal/home/flegt_intro/flegt_action_plan. Accessed 31
July 2013.
European Union. Procedural rules for the recognition and withdrawal of recognition of monitoring organizations as provided
for in the EUTR. http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:32012R0363:EN:NOT. Accessed 31 July 2013.
European Union. Detailed rules concerning the due diligence system and the frequency and nature of the checks on monitoring
organizations
as
provided
for
in
the
EUTR.
http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:32012R0607:EN:NOT. Accessed 31 July 2013.
European Union. Competent authorities listing. http://ec.europa.eu/environment/eutr2013/contacts/index_en.htm. Accessed
31 July 2013.
European Union FLEGT Facility. 2012. Literature review: US Lacey Act Amendment and the EU Timber Regulation Focus
on their communications. Unpublished report.
European Union FLEGT Facility. 2013. FLEGT-licensed timber will help importers meet the due diligence requirements of the
EUTR.
Fact
sheet.
www.euflegt.efi.int/files/attachments/euflegt/publications_2013/fact_sheet_on_flegt-licensed_timber-en-april13-v3.3_ap_rcv.pdf.
Accessed 31 July 2013.
Food and Agriculture Organization. ForesStat online database. http://faostat.fao.org/site/626/default.aspx#ancor. Accessed
20 June 2013.
Food and Agriculture Organization. Forest Resources Assessment 2010. www.fao.org/forestry/fra/fra2010/en. Accessed 31
July 2013.
Giurca, A. 2013. Market based and regulatory/enforcement mechanisms- assessment of impacts on timber trade between
South-east Asia and Europe; M.Sc. Thesis; Southern Swedish Forest Research Centre, Swedish University of Agricultural
Sciences: Alnarp, Sweden.
Knight, F.H. 1921. Risk, Uncertainty, and Profit; Hart, Schaffner & Marx: Boston, USA, pp.315.
ITTO Market Information Service (MIS). www.itto.int/market_information_service. Accessed 31 July 2013.
ITTO Statistical Database. www.itto.int/annual_review_output. Accessed 6 August 2013.
and
40
Ministry for Primary Industries (MPI). 2010. Code of practice for importing and selling legal and sustainable tropical timber
timber
products
in
New
Zealand.
www.mpi.govt.nz/portals/0/documents/forestry/forestry-nz/illegal-log/cop-for-importing-selling-legal-sustainable-tropical-timber
.pdf. Accessed 30 June 2013.
Proforest. 2011. Update on China: Timber legality verification scheme. http://forest-trends.org/documents/files/doc_2777.pdf.
World Bank. Justice for forests: improving criminal justice efforts to combat illegal logging. Available via:
www.illegal-logging.info/item_single.php?it_id=1272&it=document. Accessed 20 July 2013.
41
Session I
43
International
Scientific
Conference on Hardwood Processing (ISCHP2013)
International Scientific Conference on Hardwood Processing
(ISCHP
2013)
1. INTRODUCTION
Oak plantations are usually managed to produce high quality timber. There are several standards to define the
quality of a log where commonly accepted characteristics are identified, namely diameter, straightness, health,
absence of shakes and cracks, ring width and number of epicormic branches. This study focuses on the last feature.
Epicormic branches deteriorate the timber quality producing defects on timber, even though few epicormics are
accepted even in best grade logs without depreciation. Nevertheless due to the high capital investment and the
intensive management required for high quality silviculture, it is important to prevent the occurrence of epicormics,
in order to guarantee a successful result at the end of the rotation.
There is a general agreement about the origin of epicormic branches in oak (from adventitious or proventitious
buds) and an exhaustive classification on how they form and develop is well documented (Fontaine et al. 1999,
Fontaine et al. 2004, Colin et al. 2008, Colin et al. 2010). On the contrary, the physiological processes and the
environmental factors that promote or suppress their growth and development are poorly understood (Harmer 1990).
Previous research showed that epicormic branches may be a response to a stress (disease, competition, pruning) or to
a sudden release (thinning) and light conditions. Light especially has been investigated (Roussel 1978, Evans 1982,
Wignall and Browning 1988, Hibbs et al. 1989, Spiecker 1991) not without contrasting results, consequently its
influence remains still unresolved. The fact that light regulation, through thinnings, plays a fundamental role in the
epicormics survival can be observed in the French system of using light crown thinning during long rotation (180
year or more), with the main purposes of obtaining a constant radial growth and shading the stems to limit the
epicormic branches development. Managing an understorey of shade tolerant species is another silvicultural method
adopted to control the light in later stages of the stand rotation by shading the lower part of the trunk. This has been
proved to be beneficial against the development of epicormic branches (Henriksen and Sanojca 1983, Courraud
1987).
A direct method to control the production of epicormic branches is pruning, but it just removes epicormic
branches and their initial sprout is not prevented (Evans 1982). Kerr and Harmer (2001) investigated time of the year
44
Corresponding author: Tel.: (+46) 040 41 -5127; Fax: (+46) 040 46-2325; E-mail: giulia.attocchi@slu.se
International
Scientific
Conference on Hardwood Processing (ISCHP2013)
International Scientific Conference on Hardwood Processing
(ISCHP
2013)
and frequencies of pruning over four years, and recorded an inconsistent short-term production of new epicormic
branches and confirmed the advice of constant annual pruning for a long-term control. Pruning has gained more
interest due to the contemporary idea of free growth of oak (Stern 1973, Jobling and Pearce 1977, Kerr 1996),
consisting in very heavy thinning to increase stem growth and shorten the rotation. As a consequence, artificial
pruning becomes essential. This somehow contrasts the traditional light thinning regimes used in France (BaryLenger and Nebout 1993) and Germany (Kenk 1993), where pruning is not widely practiced.
A project on management of oak for high-quality wood production is on-going at the Southern Swedish Forest
Research Centre, Alnarp to provide a scientific basis for modifying contemporary oak silviculture to optimise the
production of high-quality timber. In this paper I present preliminary results related to an investigation on the effect
of pruning on the production of new epicormic branches in two young even-aged plantations of pedunculate oak,
accounting for the spatial distribution along the stem.
2. MATERIALS AND METHODS
2.1. EXPERIMENTS AND STAND DESCRIPTION
This study was based on trees sampled in four selected plots of two statistically designed thinning experiments in
two young stands of pedunculate oak in Denmark (Table 1). The experimental design is identical for both
experiments and includes eight treatments with different combinations of pre-commercial thinning grade and time of
first pre-commercial thinning. All thinning treatments were replicated twice within each experiment in a randomized
block design. For the part of the study reported here, only plots thinned to residual stem density of 1000 trees ha -1
were investigated.
Experiment 1516 is located in the forest Haslev Orned of Bregentved Estate. The stand was established in spring
1989 by sowing on a former meadow, using 9095 kg of acorns ha-1. Experiment 1517 is located in Brendstrup
forest of Aarhus Block East of the experiment was established in the late autumn of 1990 by planting of two-yearold seedlings on former agricultural land. Block West of the experiment was planted 1991 on former agricultural
land using two-year-old seedlings.
Table 1: Description of the plots and sample size. Age referes to the time of inventory.
Experiment
Location
1516
Bregentved
1517
rhus
Plot
Origin
Year of
experiment
installation
Seed
2002
Seedling
2003
2
1
2
Age
Unpruned
22
10
17
22
10
10
23
10
10
22
12
14
For testing of the research hypothesis pruning was applied to 100 evenly distributed crop trees ha-1 during early
spring 2011 in the selected plots. The selection of crop trees was based on qualitative criteria as well as spatial
distribution. Here, primary and secondary branches present on the lower 6 m of the stem were removed using a
manual pruning saw mounted on a six-meter telescopic pole. For all pruned trees the following data were collected:
height above ground of the lowest living branch before and after pruning (LLB and NLB respectively), a count of
the number of primary, living branches (Nb) and epicormic branches (Ne) which were pruned. In July 2011, newly
flushed epicormic branches were counted on the first six 1-meter stem sections on samples of pruned and unpruned
trees (Table 1). The reason for the unbalanced sampling design was that some trees were measured as part of another
investigation. On each 1-m section, different types of new epicormic branches were counted using a three-class tally
list: new single branches (Nes), number of branches originating from a cluster (Nec) and number of clusters (Pec),
hereafter called points of origin. Epicormic buds were not considered in this study.
2.2. STATISTICAL ANALYSIS
45
the number of primary, living branches (Nb) and epicormic branches (Ne) which were pruned. In July 2011, newly
flushed epicormic branches were counted on the first six 1-meter stem sections on samples of pruned and unpruned
trees (Table 1). The reason for the unbalanced sampling design was that some trees were measured as part of another
investigation. On each 1-m section, different types of new epicormic branches were counted using a three-class tally
International
Scientific
Conference
Hardwood
Processing
(ISCHPfrom
2013)
list: new single
branches
(Nes), on
number
of branches
originating
a cluster (Nec) and number of clusters (Pec),
hereafter called points of origin. Epicormic buds were not considered in this study.
2.2. STATISTICAL ANALYSIS
Linear mixed models (West et al. 2007, Zuur et al. 2009) were used to investigate the research hypothesis, under
the usual assumption of identical, normally distributed errors with zero mean and common variance, using complex
nesting to take into account the three levels of split-plot design.
The response variable was the total number of new epicormic branches, obtained as sum of new single branches
and points of origin, while as fixed effects, pruning (pruned, unpruned), stem section (0-1 m, 1-2 m, etc.), DBH
(diameter at breast height - 1.3 m) and H (total tree height) were used. Due to the nested structure of data, plot and
tree identity (nested in plot) were specified as random effects to correct for non-independence by imposing 2-way
nested correlation (within trees and within plots) in the model. Inspection of the residuals indicated violation of
variance homogeneity which was recovered modifying residual variance structures in each section.
The optimal variance structure was estimated based on restricted maximum likelihood (REML) (West et al.
2007), using Akaikes information criteria (AIC) and likelihood ratio test. Model reduction was performed by
backwards stepwise elimination of non-significant terms, using maximum likelihood ratio test (ML), based on an
approximate 2 distribution. The analyses were performed using the R statistical and programming environment,
version 2.14.2 (R Core Team 2012) and the nlme package for linear and nonlinear mixed effects models (Pinheiro
et al. 2012).
3. RESULTS AND DISCUSSION
A preliminary analysis was performed mainly to assess whether the number of pruned primary branches and the
foliage removed influenced the overall production of new epicormic branches. The total number of new epicormic
branches per tree increased with pruning; the analysis showed that both primary branches and epicormic branches
removed with pruning were significant (L=8.842, df=1, p<0.05 and L=5.227, df=1, p<0.05, respectively). DBH and
pruning efficiency (given by difference between the total tree height and the height of the lowest living branch after
pruning) were not significant at the 5% level. The coefficients of the final model = 2.407 + 3.322
indicate the relative importance of removal of epicormic branches as well as of primary branches, both of
which led to production of additional, new epicormic branches. According to the research hypothesis pruning
reduces the number of new epicormic branches as compared to unpruned oak trees. In this study, an overall negative
effect of pruning was detected in the preliminary analysis.
The main analysis consisted on comparing pruned and unpruned trees, where pruning clearly influenced the
composition of epicormic branches (classified as single branches or clusters) along the stem. Pruning had a
significant effect on the number of new epicormic branches, and the interaction with section was also significant
(L=18.257, df=4, p<0.05). Due to collinearity DBH and H were tested independently and were not significant
components of the model at 5% level. The average number of all types of epicormic shoots was higher in unpruned
than pruned trees on the three lower stem sections below 3 m, and vice-versa for the three upper sections above 3 m
(Figure 1). Unpruned trees presented a consistent pattern across plots, increasing from below, peaking at the section
1-2 m and thereafter decreasing with increasing height along the stem. Pruned trees had a similar pattern along the
stem, but fluctuated somewhat during the decrease towards 6 m. This pattern was common for both experiments
where not only a significant effect was detected, but also a similar pattern in the number and location of new
epicormic branches across sites and plot replications.
46
14
10
8
6
4
10
8
6
4
0
1
Section
Unpruned
Pruned
12
12
Exp. 1517
14
Unpruned
Pruned
Section
Figure 1: Average number of epicormic branches for pruned and unpruned trees in experiments 1516 and 1717.
This provides a strong basis for drawing conclusions, even though it is difficult to outline consequences and
implications for forest management because of the significant interactions encountered.
The significant interaction between pruning and height along the stem makes it difficult to give recommendations
for forestry practice. Similar results were reported by Spiecker (1991) who for both pedunculate and sessile oak
found that pruning leads to an increase in the number of epicormic branches, and therefore discouraged the practice.
In contrast Kerr and Harmer (2001) reported for pedunculate oak a general decrease in the production of epicormic
branches with pruning compared to the control plot. A possible explanation for this contrasting result could be due to
genetics (Jensen, 2000), age or different site, since it is known that environmental factors affect the production of
epicormic branches (Evans 1982).
Recent investigations have demonstrated that past epicormic emergence present at a given time correlate
positively with the subsequent frequency of epicormic branches in sessile oak (Fontaine et al. 2001, Morisset et al.
2011, Morisset et al. 2012). For the purpose of this study, data on epicormic buds were not recorded; therefore a
possible correlation between the presence of dormant buds and pruning could not be established. However, the
suggestion that buds may burst vigorously when a branch above is pruned, due to the release of internal
physiological processes (Colin et al. 2012), may be potentially important for the interpretation of the results. For
example, pruning weakens the tree because of the reduction in foliage mass (Meier et al. 2012) and it stresses and
exposes more meristematic tissue to increased levels of light (Fink 1984) and these may trigger the development of
epicormic branches.
The second part of this study aimed to investigate the spatial distribution along the stem of epicormic branches in
relation to pruning. As already reported above, the number of new epicormic branches generally decreases with
increasing height along the stem, except for section 1 for both pruning treatments and section 4 of only pruned trees
in experiment 1517. This is an indication of the positive shading effect due to the size of individual crowns, resulting
in a lower number of epicormic branches in the stem sections (4-6 m) close to the canopy. This is in disagreement
with previous investigations of Spiecker (1991), Henriksen and Sanojca (1983), Smith (1965) and Ward (1966), who
all had an increase of the number of epicormic branches with increasing height above ground. A possible reason is
that they had much older trees and that the location effect could be age dependent. Moreover, they had no replicates
and while Henriksen and Sanojca (1983) worked with pedunculate oak, both Smith (1965) and Ward (1966)
investigated American oaks.
Due to the significant interaction between pruning and section, pruned trees have more epicormic branches closer
to the canopy in comparison with the unpruned trees. There are several possible explanations for this. More light
penetrates to the upper stem sections of pruned trees or hormonal regulation triggers the emergence of more
epicormic branches near the lower canopy branches (Meier et al. 2012). Here, more primary branches were removed,
creating callus tissue, and consequently more epicormic branches emerged. Most of these were probably of
47
International
Scientific
Conference on Hardwood Processing (ISCHP2013)
International Scientific Conference on Hardwood Processing
(ISCHP
2013)
adventitious origin. The interaction also highlighted the contrast between the within-plot competition (quantified in
the second part of the analysis) and the positive shading effect due to the size of individual crowns, resulting in a
lower number of epicormic branches in the stem sections (4-6 m) close to the canopy.
Surprisingly, DBH and H were not significant components of the model. This is in contrast to results reported by
Nicolini et al. (2001), but in agreement with Harmer (1990), Colin et al. (2008) and Morisset et al. (2012) stating that
the occurrence or emergence of epicormic branches is unrelated to the size of individual trees. This might be due to
individual tree variation (hence the significant random component in the model) or to the different diameter classes
investigated in other studies. Furthermore, certain individual trees consistently bear more epicormic branches over
time independently of the thinning practice, as reported by Henriksen and Sanojca (1983), Ward (1966), and
Spiecker (1991). Even though tree size variables were not significant in predicting new epicormic branches, it cannot
be inferred that tree social status is of no importance as data were not collected to test this hypothesis. Consequently,
there is a missing link between the relationship among tree size, social status and the composition of epicormic
branches.
The completion of the overall project will seek to provide valuable information on the pruning effect as well as
of the thinning practice in the emergence of epicormic branches and other qualitative aspects of high quality oak
timber production.
4. CONCLUSION
In conclusion for this part of the investigation, pruning has an overall negative effect on the emergence of new
epicormic branches. The same analysis with additional information about spatial variation along the stem revealed
an interaction between pruning and the height along the stem, indicating that the number of new epicormic shoot
changes with increasing height along the stem and pruning. A new hypothesis should state that pruning leads to the
formation of more epicormic branches, and it should furthermore account for the spatial variation in the emergence
of epicormic branches along the stem. The new hypothesis should be tested for a wider range of site types and oak
genetics and throughout a full rotation.
REFERENCES
Bary-Lenger, A., and J.-P. Nebout. 1993. Le chne. Editions du Perron, Alleur-Lige. 604 pp.
Colin, F., Mechergui, R., Dhte, J.-F. and F. Fontaine. 2010. Epicormic ontogeny on Quercus petraea trunks and thinning effects
quantified with the epicormic composition. Ann. For. Sci. 67(8):813.
Colin, F., N. Robert, J.L. Druelle, and F. Fontaine. 2008. Initial spacing has little influence on transient epicormic branches in a
20-year-old sessile oak plantation. Ann. For. Sci. 65(5):508517.
Colin, F., A. Sanjines, M. Fortin, J.D. Bontemps, and E. Nicolini. 2012. Fagus sylvatica trunk epicormics in relation to primary
and secondary growth. Ann. Bot. 110(5):9951005.
Courraud, R. 1987. Les gourmands sur les chnes rouvre et pdoncul. Fort Entreprise 45:2033.
Evans, J. 1982. Free growth and control of epicormics. In: Malcolm, D.C., J. Evans, and P.N. Edwards (Eds.). Broadleaves in
Britain. Institute of Chartered Foresters, Edinburgh. pp. 183190.
Fink, S. 1984. Some Cases of Delayed or Induced Development of Axillary Buds from Persisting Detached Meristems in
Conifers. Am. J. Bot. 71(1):4451.
Fontaine, F., F. Mothe, F. Colin, and P. Duplat. 2004 Structural relationships between the epicormic formations on the trunk
surface and defects induced in the wood of Quercus petraea. Trees Struct. Funct. 18(3):295306.
Fontaine, F., F. Colin, P. Jarret, and J.L. Druelle. 2001. Evolution of the epicormic potential on 17-year-old Quercus petraea trees:
first results. Ann. For. Sci. 58(5):583592.
Fontaine, F., E. Kiefer, C. Clement, M. Burrus, and J.L. Druelle. 1999. Ontogeny of the proventitious epicormic buds in Quercus
petraea. II. From 6 to 40 years of the tree's life. Trees Struct. Funct. 14(2):8390.
Harmer, R. 1990. The timing of canopy and epicormic shoot growth in Quercus robur L. Forestry 63(3):279283.
Henriksen, H.A. and K. Sanojca. 1983. Bgeunderskovs betydning for udvikling af vanris p stilkeg (Quercus robur L.). Forstlige
Forsgsvsen i Danmark 39(1):93119.
Hibbs, D.E., W.H. Emmingham, and M.C. Bondi. 1989. Thinning red alder: effects of method and spacing. For. Sci. 35(1):1629.
Jensen J.S. 2000. Provenance variation in phenotypic traits in Quercus robur and Quercus petraea in Danish provenance trials.
Scan. J. Forest Res. 15:297308.
48
International
Scientific
Conference on Hardwood Processing (ISCHP2013)
International Scientific Conference on Hardwood Processing
(ISCHP
2013)
Jobling, J. and M.L. Pearce. 1977. Free growth of oak. Forestry Commission Forest Record 113:117.
Kenk, G.K. 1993. New perspectives in German oak silviculture. Ann. For. Sci. 50(6):563570.
Kerr, G., 1996. The effect of heavy or 'free growth' thinning on oak (Quercus petraea and Q. robur). Forestry 69(4):303317.
Kerr, G. and R. Harmer. 2001. Production of epicormic branches on oak (Quercus robur): Effects of frequency and time of
pruning. Forestry 74(5):467477.
Meier A.R., M.R. Saunders, and C.H. Michler. 2012. Epicormic buds in trees: a review of bud establishment, development and
dormancy release. Tree Physiol. 32:565584.
Morisset, J.B., F. Mothe, B. Chopard, D. Francois, F. Fontaine, and F. Colin. 2011. Does past emergence of epicormic branches
control current composition of epicormic types? Ann. For. Sci. 69(2):139152.
Morisset, J.B., F. Mothe, and F. Colin. 2012. Observation of Quercus petraea epicormics with X-ray CT reveals strong pith-tobark correlations: Silvicultural and ecological implications. For. Ecol. Man. 278:127137.
Nicolini E., B. Chanson and F. Bonne. 2001. Stem growth and epicormic branch formation in understorey beech trees (Fagus
sylvatica L.). Ann. Bot. 87:737750.
Pinheiro, J., D. Bates, S. DebRoy, D. Sarkar, and the R Development Core Team. 2012. nlme: Linear and Nonlinear Mixed
Effects Models. In: R package version 3.1-103.
R Core Team. 2012. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna,
Austria. ISBN 3-900051-07-0, URL http://www.R-project.org/.
Roussel, L. 1978. Lumiere, gourmands et rejets de souches. Revue forestire franaise. 3:186-200.
Smith, H.C. 1965. Effects of clearcut openings on quality of hardwood border trees Liriodendron tulipifera, Quercus rubra,
Prunus serotina. J. For. 63(12):933937.
Spiecker, H. 1991. Controlling the diameter growth and the natural pruning of Sessile and Pedunculate oaks (Quercus petraea
(Matt.) Liebl. and Quercus robur L.). Selbstverlag der Landesforstverwaltung Baden-Wrttemberg. Diss. University of
Freiburg, Stuttgart. 135 pp.
Stern, R.C. 1973. The production of high value hardwoods. Timber Grower 48:1121.
Ward, W.W. 1966. Epicormic branching of black and white oaks. For. Sci. 12(3):290296.
West, B.T., K.B. Welch and A.T. Gaecki. 2007. Linear mixed models : a practical guide using statistical software. Chapman &
Hall/CRC, Boca Raton. 353 pp.
Wignall, T.A., and G. Browning. 1988. The effects of stand thinning and artificial shading on epicormic bud emergence in
pedunculate oak (Quercus robur L). Forestry 61(1):4559.
Zuur, A.F., E.N. Ieno, N.J. Walker, A.A. Saveliev, and G.M. Smith. 2009. Mixed Effects Models and Extensions in Ecology with
R. Springer New York, New York. 574 pp.
49
ABSTRACT
Considering the interesting markets that timber of several hardwood species have, and the productive potential of Chile,
in the central southern area of the country, in the last 20 years several experimental trials were established, pure or
mixed, located between the Araucana and Los Lagos regions. Trials include exotic species, such as Sweetgum
(Liquidambar styraciflua), Ash (Fraxinus excelsior), Black walnut (Juglans nigra), Chestnut (Castanea sativa), Yellowpoplar (Liriodendron tulipifera), Black alder (Alnus glutinosa), Cherry tree (Prunus avium), Red American Oak (Quercus
rubra) and European Oak (Quercus robur); and the natives Chilean hazelnut (Gevuina avellana) and Raul (Nothofagus
alpina). Since establishment (between years 1994 and 2001), a semi intensive quality arboriculture management was
applied, and periodic measurements were performed. The experimental design used in the experimental plots corresponds
to completely randomized blocks.
The work presents both in quantitative and graphical form values for dasometric parameters and growth evolution of
7 experimental trials 11 to 19 years after the establishment, evaluated performing analysis of variance (ANOVA) and
Fisher's least significant differences test (LSD). According to species and sites, the average annual Diameter at breast
height (1.3 m) (DBH) growth ranges between 0.5 and 1.8 cm and the average annual height growth between 0.38 and
1.07 m.
1.
INTRODUCTION
The analysis of world trends seems to indicate the need to have a diversified forest sector, both considering
species composition (quantity of cultivated species) as well as their geographic distribution. Reasons that
stimulate the diversification enhancement are found in elements of economy and forest politics, because it
allows limiting biotic and abiotic risks, distributing them in space and time; it also allows containing
economic risks, facing eventual market fluctuations; maximizing the use of sites; cultivating species of
economic interest different from the used traditionally; and finally, it is pertinent since Chile has a wide range
of environments and there exists the possibility of accessing high value products market niches.
Then it turns out advisable to diversify the productive spectrum for strategic, ecological and economic
reasons; of stability to face better the international markets; ecologically, to generate landscapes recognized
by its beauty and suitability for recreation, and to reduce losses due to plagues or diseases, as well as other
Vernica Loewe M.: Tel.: (56-2) 23667111; Fax: (56-2)23667131; E-mail: vloewe@infor.cl
50
benefits such as water production, soil conservation/improvement, carbon fixation, fauna hosting, etc.; and
economically, since interesting productive alternatives can be generated.
One technology that allows diversifying the agriculture and forest sector is arboriculture, which deals with
broadleaved species known as "noble" or minor, because generally do not form pure forests, in order to fulfill
economic, aesthetic and environmental purposes, enhancing the generation of multiple products, which could
increase the environmental value and the productive sector, that can be applied by small, medium and big
owners or companies (Paris et al., 1999).
More than twenty years ago the first R&D projects related to arboriculture oriented to harvest high value
timber were initiated in Chile. Among the achieved results are adjustments of European technologies, such as
pruning, considering the local reality (higher radiation and thus higher vigor/growth), and some technological
developments suitable to the national reality (Loewe, 2003b).
This work presents results of growth parameters obtained 11 to 19 years after the establishment of seven
experimental trials designed to test the applicability of quality arboriculture to obtain high quality timber from
eleven hardwood species in central south Chile.
2.
Between 1994 and 2001, seven experimental trials were established to assess arboriculture techniques, both in
pure and mixed plantations. These trials are located between the Araucana and Los Lagos regions (Table 1),
a zone characterized by its temperate weather, with mild temperatures and water availability through rainfall
distributed in most of the year. They test the exotic species Liquidambar styraciflua (Sweetgum), Fraxinus
excelsior (Ash), Juglans nigra (Black walnut), Castanea sativa (Chestnut), Liriodendron tulipifera (Yellowpoplar), Alnus glutinosa (Black alder), Alnus cordata (Italian alder), Prunus avium (Cherry), Quercus rubra
(Red American oak) and Quercus robur (European oak); and the natives Gevuina avellana (Chilean hazelnut,
Figure 5) and Nothofagus alpina (Raul). Trials experimental design corresponds to completely randomized
blocks with 3 repetitions.
Table 1: Experimental trials site characterization
Site
Region
County
Age
Type
Initial
Site location
density (m)
Latitude
Longitude
3x3
72 36' 41.05''W 37 42'' 56.47'' S
Renaico
Araucana
Angol
15
Mixed
Rinconada Araucana
Cunco
14
Pure,
Provenances
3x3
Carrizal*
Araucana
Toltn
16
Pure,
Provenances
3x3
Loncoche
Ros
Loncoche
19
Pure, density
2x2 and
3x3
Los Lagos
Ros
Los Lagos
12
Mixed
4x4 and
8x8
Management activities
Sub-soiling, sporadic
weed control, fertilization
(age 0, 1, 3), pruning
(age 1, 2, 3, 4, 5, 6, 7, 8),
thinning (age 7 and 12),
irregular summer
irrigation (frequency of
3-5/year).
72 14' 5.76'' W 38 55'' 48.54'' S
Sub-soiling, sporadic
weed control, fertilization
(age 0, 1, 2), pruning
(age 1, 2, 4, 7).
73 8' 1.32'' W 39 12'' 39.63'' S Ploughing, sporadic weed
control, fertilization (age
0, 1, 2), pruning (age 2,
4, 6).
72 39' 45.51''W 39 18" 59.70'' S Ploughing, sporadic weed
control, fertilization (age
0), pruning (age 0, 2, 3,
5, 6, 7, 9, 11), thinning
(age 11).
72 54' 59.98''W 39 9'' 4.98'' S Ploughing, sporadic weed
control, fertilization (age
51
Region
County
Age
Type
Initial
density (m)
Villarrica
Ros
Villarrica
15
Pure,
Provenances
3x3
Osorno
Lagos
Osorno
15
Pure,
Provenances
3x3
Site location
Latitude
Longitude
Management activities
Among the mixed trials, Renaico associates Chestnut, Ash, Black walnut and Sweetgum (Figure 3); in Los
Lagos nine treatments were experimented (T1: Pure Cherry; and the associations T2: Cherry, American Red
Oak, Chestnut, European Oak; T3: Cherry, American Red Oak, Chestnut, European Oak, Black Alder; T4:
Cherry, American Red Oak, Chestnut, European Oak, Chilean Hazelnut (Gevuina avellana); T5: Cherry,
American Red Oak, Chestnut, European Oak, Notro (Embothrium coccineum); T6: Cherry, American Red
Oak, Chestnut, European Oak, Piche (Fabiana imbricata); T7: Cherry, American Red Oak, Chestnut,
European Oak, Black Alder, Piche; T8: Cherry, American Red Oak, Chestnut, European Oak, Chilean
Hazelnut, Piche; and T9: Cherry, American Red Oak, Chestnut, European Oak, Notro, Piche).
From the establishment, a semi intensive management according to quality arboriculture principles (Loewe,
2003a) was applied, including soil preparation, weed control, pruning and in some cases, thinning (Table 1).
Experimental plots (treatment unit) ranged from 12 to 64 trees/plot according to density and association
design (higher complexity includes lower quantity of main species trees). In mixed plantations species were
mixed tree-by-tree. Each plot has a buffer of at least two rows. Within a site, all treatments were subjected to
the same planting pattern, spacing, and tending.
Each tree measurement included DBH and total height, as well as other parameters related to quality, such as
straightness and vigor. The evaluation considered analysis of variance (ANOVA) and Fisher's least significant
differences test (LSD).
3.
RESULTS
Results obtained for the studied species are presented in Table 2, indicating type of plantation (pure (P) or
mixed (M)). Table 3 presents growth variables for the 3 species planted both in monoculture and in
association; and Table 4 shows the effect of two densities on growth variables.
Table 2: Growth variables for the studied species by sites
Species /Sites
DBH
(cm)
Sweetgum
Renaico (M)
Rinconada (P)
Black walnut
Renaico (M)
Chestnut
Height
(m)
Parameter
DBH MAI
(cm)
Height MAI
(m)
10.95
9.21
12.84
6.37
0.86
0.66
0.73
0.45
12.76
11.26
0.85
0.75
52
Parameter
DBH MAI
(cm)
1.83
1.22
0.78
0.80
DBH
(cm)
27.47
14.59
14.83
15.12
Height
(m)
14.17
8.51
10.30
13.78
Height MAI
(m)
0.94
0.71
0.54
0.73
23.20
13.15
1.55
0.88
17.73
14.93
1.19
1.00
13.79
9.82
0.86
0.61
15.28
12.26
0.96
0.77
9.01
7.36
0.75
0.61
12.64
8.66
1.05
0.72
6.69
11.48
13.22
5.52
9.79
12.34
0.56
0.60
0.70
0.46
0.52
0.65
14.65
6.13
0.94
0.41
16.28
11.93
0.86
15.79
12.99
0.83
MAI: Mean annual increment
0.63
0.68
Table 3: Growth variables for several species in monoculture and mixed plantation
Species
DBH MAI
SE(cm)
Height MAI
SE (m)
Sweetgum
Pure
0.66 0.06 (a)
0.45 0.01 (a)
132
Mixed
0.86 0.04 (b)
0.73 0.03 (b)
83
Cherry
Pure
0.56 0.07 (a)
0.46 0.06 (a)
117
Mixed
0.65 0.07 (b)
0.59 0.06 (b)
99
Chestnut
Pure
1.22 0.10 (a)
0.71 0.04 (a)
154
Mixed
1.14 0.07 (a)
0.74 0.04 (a)
192
Different letters indicate statistically significant differences (p 0.05)
MAI: Mean annual increment
Table 4: Growth comparison for three species established with two densities at Loncoche site
Species /density
Chestnut
3x3 m
DBH (cm)
14.83 (a)
Height (m)
10.03 (a)
0.78 (a)
0.54 (a)
53
Chestnut
Pure
1.22 0.10 (a)
0.71 0.04 (a)
154
Mixed
1.14 0.07 (a)
0.74 0.04 (a)
192
Different
letters indicate
statistically
significant
International Scientific Conference
on Hardwood
Processing
(ISCHP
2013) differences (p 0.05)
MAI: Mean annual increment
Table 4: Growth comparison for three species established with two densities at Loncoche site
Species /density
DBH (cm)
Height (m)
14.83 (a)
10.03 (a)
0.78 (a)
0.54 (a)
2x2 m
15.12 (a)
13.78 (b)
0.80 (a)
0.73 (b)
Cherry
3x3 m
2x2 m
11.48 (a)
13.22 (b)
9.79 (a)
12.34 (b)
0.60 (a)
0.70 (a)
0.52 (a)
0.65 (a)
Raul
3x3 m
2x2 m
16.28 (a)
11.93 (a)
0.86 (a)
0.63 (a)
15.79 (b)
12.99 (b)
0.83 (a)
0.68 (a)
Different letters indicate statistically significant differences (p 0.05)
Figures 1 and 2 compare visually the effect of species associations in growth for Sweetgum and Cherry.
1,0
0,8
B
A
0,6
0,4
0,2
0,0
Pure
Mixed
DBH (cm)
Height (m)
Figure 1: Mean annual DBH and height growth for Sweetgum in pure and mixed plantations
1,0
0,8
0,6
0,4
0,2
0,0
Pure
DBH (cm)
Mixed
Height (m)
Figure 2: Mean annual DBH and height growth for Cherry in pure and mixed plantations
54
55
DISCUSSION
Species
Even though growth parameters change enormously according to species, sites, silviculture treatments and
whether species are grown under a pure or mixed scheme (Table 2), variable growth rates were found in all
studied plantations, with DBH annual growth varying from 0.56 to 1.22 cm without irrigation, and up to 1.83
cm with irrigation. For height, annual growth ranged between 0.41 to 1.0 m with and without irrigation
respectively.
For Cherry the observed height growth (0.46-0.65 m/year) with semi intensive management was similar than
results obtained in cherry intensive agroforestry 8 year old plantations in five locations in Languedoc, France,
which showed annual height increments ranging from 0.31 to 0.69 m/year (Balandier and Dupraz, 1999).
Cisneros et al. (s/f) forecast for cherry intensive plantations in Castilla y Len, Spain, average DBH mean
annual increment at age 10 ranging from 1.4 cm/y in high quality sites, 0.95 cm/y in medium quality sites and
0.5 cm/y in lower quality sites. At age 20 this values are 1.4; 1.0 and 0.6 cm/y respectively. Our results show
a lower performance when compared to those values, considering Los Lagos as a medium quality site, and
Loncoche a low quality site. Negrini (1994) report for arboriculture plantations in Emilia Romagna region,
Italy, height mean annual increment at age 4 of 0.94 m/y for cherry, also higher than those found in Chile.
For Chestnut, according to Bourgeois et al. (2004), DBH mean annual increment at age 17 ranges between
0.56 cm/y without thinning, and 0.75 cm/y thinned at age 12; our results are higher both for experimental
plots with (0.78, 0.80 and 1.83 cm/y) and without thinning (1.22 cm/y).
56
In the case of Black walnut, Schlesinger and Williams (1984) indicate for pure plantations DBH mean annual
increment ranging 0.22 to 0.9 cm/year according to site whereas our results reached 0.85 cm/y in the mixed
plantation trial that contain this species. Pedlar et al. (2006) report growth of Black walnut plantations in
Ontario, Canada, indicating height mean annual increment ranging from 0.14 to 0.28 m/year in pure
plantations in different soil types, and between 0.18 and 0.40 m/year in mixed plantations also for different
soil types; for DBH, values varies from 0.33 and 0.38 cm/year in pure plantations, and 0.25 and 0.62 cm/year
in mixed plantations. Buresti et al. (1997) report mean diametric growth values for arboriculture plantations
located in central Italy, of 0.63 cm/year. Our results are higher than values reported by Pedlar et al. (2006),
due perhaps to the better climatic conditions existing in the south of Chile, and higher to the ones reported by
Buresti et al. (1997) probably due to the irrigation that the trial had received.
For Ash, Cisneros et al. (s/f) report for pure plantations in the Segovia Province, Spain, average DBH mean
annual increment of 0.79 cm/y, and 0.57 m/y for height. Minotta (1996) report height mean annual increment
ranging 0.08 to 0.29 m/y for different soil preparation techniques; our results in a mixed plantation reached
the values of 1.55 cm/y and 0.88 m/y respectively, higher in both cases, but lower than reported by Negrini
(1994) for arboriculture plantations in Emilia Romagna region, Italy, who indicate height mean annual
increment at age 4 of 1.1 m/y.
Considering European Oak, we got growth values similar to the ones reported in Italy by Negrini (1994) in
arboriculture plantations in Emilia Romagna region, with height mean annual increment at age 4 of 0.8 m/y;
and by Becciolini and Pelleri (2006) who indicate growth for 18 years old European oak arboriculture
plantations thinned at age 12, located in central Italy, average DBH mean annual increment of 0.94 cm/y for
pure plantations, and between 0.98 and 1.02 cm/y for mixed plantations with European alder (Alnus cordata),
showing a positive effect of this N-fixing companion species. Our results are better than those reported by
Mason and Baldwin (1995) in Scotland, that reach average annual DBH growth ranging between 0.51 and
0.55 cm/y in 19 year old pure plantations; and similar to the DBH annual growth reported by Buresti et al.
(1999) in a 17 years-old plantation thinned at age 13, that reached 1.01 cm/y. Our results outgrowth values
obtained in plantations managed with silviculture that mimic nature located in the same region, where DBH
annual growth for best trees reached 0.7-0.9 cm/y (Siebert, 2006).
As per Sweetgum, Davis and Trettin (2006) report annual growth values of DBH of 0.78 cm/y, and 1.0 m/y
for height in a pure plantation intensively managed established on former agricultural land in North Carolina,
USA. This values are higher than ours regarding pure plantation, and similar for height in mixed plantations;
even though their results are higher, mortality reached 88% while in our case was 10 to 38%; it is worthy
considering than our experience includes semi intensive management. With the same species, Kaczmarek et
al. (2012) tested different densities and fertilization schemes in intensively managed bioenergy plantations in
the southeastern USA, obtaining mean annual height growth of 1.4-1.5 m/y without fertilization and 1.5-1.6
with fertilization; and mean annual diametric growth of 1.5-1.7 cm/y without fertilization and between 1.6-2.0
cm/y with fertilization, all of them superior to our experience.
For Red oak, our results are considerable higher than those reported by Carvalho et al. (1997), who obtained
in a 3 year-old plantation located in northwest Portugal height annual growth of 0.09 m/y with weeding, and
0.15 m/y with weeding and fertilization.
Yellow-poplar growth parameters are high, and consistent with the rapid growth that characterizes the species
(Hart et al., 2010). Clatterbuck (2004) in plantations aged 9-18 years measured annual height growth of 1.211.34 m/y, and annual DBH growth ranging 1.1 to 1.6 cm/y, similar to those found in the Osorno experimental
unit.
Among both Alder species, Italian alder got the best performance in the difficult conditions experimented,
including stagnant water all over the site for several weeks, up to 8 month some years; maybe this is why our
57
results are lower than those reported by Buresti and Frattegiani (1994) in mixed 10 years-old plantations with
European oak, in which Italian alder showed annual growth of 1.2 cm/y in diameter and 0.96 m/y in height.
In the case of Black alder, our findings agree with the values reported by Atasoy et al. (2003) for15 years-old
plantation of Black alder established in different sites in Turkey at 3x3 m, where annual DBH growth ranged
0.56 to 0.85 cm/y, and annual height growth between 0.54 and 0.7 m/y; and also with Daz-Vaz (1993) who
report Black alder DBH growth in the Araucania region, Chile, ranging 0.8-2.0 cm/y.
Hawe and Short (2012) indicate that ash is extremely exacting as to site conditions, whereas cherry and
European oak are relatively indifferent to site conditions; we couldnt confirm these observations not having
enough trials for all species.
Regarding native species, height annual growth in our Rauli and Chilean hazelnut trials were superior than
values reported by Alvarez and Lara (2008) for strip plantations located in similar areas aged 4 (0.2-0.3 m/y
for Chilean hazelnut and 0.5-0.6 m/y for Rauli).
In synthesis, out of the twelve species studied, 3 performed worse, 5 similar, and 4 better than cited works by
other authors; nonetheless, it must be kept in mind that the conditions of our experimental units and the others
to which were confronted are not always completely comparable. These growth rates show the aptitude of the
area for high timber value species cultivation, as it was stated by Loewe (2003b), due mostly to soil quality
and rain frequency, elements that favor sustained tree development.
Plantation type (monoculture/associations)
When comparing the effect of mixed plantations, we observed that differences in height and DBH growth
between pure and mixed plantations were marked (Table 3) for Sweetgum (Figure 1) and Cherry (Figure 2),
reaching 23.0 and 13.8% for DBH mean growth, and 38 and 22% for height mean growth respectively; these
findings agree with Gabriel et al. (2005), whose studies have shown that mixed plantations could reach
heights up to 24% higher than pure plantations. In Los Lagos, pure and mixed plantations including Notro and
Piche showed lower performance in height (Loewe et al., 2013).
Results were not statistically significant in the case of Chestnut, however, for this species, the association
brings important advantages for high quality timber production, such as better trunk form, lower occurrence
of epicormic branches and lower forking, already quantified in Chile by Loewe et al. (2005) and Loewe et al.
(2013). These characteristics were also marked in the case of pure Yellow-poplar (Figure 4), which has
marked form deformations in most trees, which will reduce timber quality. Even though this can be measured,
it wasnt done in this study. For this we agree with Hawe and Short (2012) in the sense that Chestnut performs
better in mixed plantations.
In the case of Cherry, our findings agreed with information from Tani et al. (2006) and Loewe (2003b), who
found that Cherry trees grown with companion trees perform better, especially if associated with N-fixing
species; these positive effects are probably due to many co-occurring favorable causes, being remarkable the
higher availability of nitrogen. Mohni et al. (2009) indicated that growth in dense plantation with companion
species has often demonstrated greater height and diameter growth compared to pure stands. According to
Schlesinger and Williams (1984), companion species may alter a site through soil influences, microclimate
changes, vegetation changes and/or through pest habit changes.
Density
The positive effect of a higher density in height growth observed at Loncoche site for three species is
normally exploited in pure plantations for improving stem form, even though this doesnt guarantee
necessarily a higher quality timber. The impact of density in mean annual height and DBH growth could be
hide by the thinning performed at age 11.
58
Since it is well known that there exist treatment interactions over time, we think that this issue is a topic
worthy to study through longer observations.
5.
CONCLUSIONS
Our results indicate that most of the studied species show satisfying growth performances, being interesting
productive alternatives under arboriculture management for the central south area of Chile. However, results
presented are at early stage, and thus preliminary, so it would be important to carry on measurements at a later
age.
Our results showed superior height (up to 38%) and DBH (up to 23%) in mixed plantations when compared to
pure plantations, as well as better stem form regarding straightness and forking absence, even though these
data were not quantified in this study. Then our findings show as several other studies have already done
both in Chile (Siebert and Loewe, 2002; Loewe and Gonzlez, 2006; Loewe et al., 2008) and abroad (Buresti,
1995, Binkley, 2003, Forrester et al., 2004, Gabriel et al., 2005, and Hung et al., 2011)-, that mixed
plantations have advantages for growth parameters as well as others related to quality, such as forking and
branching, all of them of economic importance.
Companion species can play an important role in high value timber species cultivation; however, it is
necessary to select and match companion and main species growth characteristics for different site conditions.
Choosing the right species associations for each site can favor species development, high quality timber
production and reduce intensity of required management, making a better use of plantations potential.
6.
REFERENCES
Alvarez, C., Lara, A. 2008. Crecimiento de una plantacin joven en fajas con especies nativas en la
Cordillera de Los Andes de la provincia de Valdivia. Bosque 29(3): 181-191.
Atasoy, H., Kucuk, M., Ulu, F. 2003. Spacing trial son Black alder (Alnus glutinosa) in Eastern Black Sea
region of Turkey.
Balandier, P., Dupraz, C. 1999. Growth of widely spaced trees. A case study from young agroforestry
plantations in France. Agroforestry Systems 43:151-167.
Becciolini, R., Pelleri, F. 2006. Consociazione tra farnia e ontano napoletano, valutazione degli effetti in un
impianto prima e dopo un diradamento. Sherwood 119.
Binkley, D. 2003. Seven decades of stand development in mixed and pure stands of conifers and N-fixing red
alder. Canadian Journal of Forest Research 33 (2003), pp. 2274-2279.
Bourgeois, C., Sevrin, E., Lemaire, J. 2004. Le Chataignier, un arbre, un bois. Institut pour le Developpement
Forestiere, 347 p.
Buresti, E. 1995. Walnut trees in mixed stands with shrubs and trees. In: European development of Walnut
and fruit production as an alternative and extensive system to agricultural crops. Workshop of EU
AIR/Walnut Project, March 27-30, 1995. Thessaloniki, Greece.
Buresti, E., Frattegiani, M. 1994. Impianti misti in arboricoltura da legno. Primi risultati in un impianto di
farnia (Quercus robur L.) e ontano napoletano (Alnus cordata Loisel). Ann. Ist. Sper. Selv.: 183-199 XXIII.
10
59
Buresti, E., De Meo, I., Pelleri, F. 1999. Criteri e risultati di un diradamento in un impianto di arboricoltura
da legno di farnia (Quercus robur L.). Annali Istituto Sperimentale Selvicoltura Arezzo, Vol. 29:29-40.
Buresti, E., De Meo, I., Falcioni, S., Frattegiani, M. 1997. Lutilizzo del noche comune, del noche nero e del
noche ibrido in arboricoltura da lego. Primi risultati di una prova comparativa in impianti misti di 8 anni di
eta. Annali Istituto Sperimentale Selvicoltura Arezzo, 243-260 XXV e XXVI.
Carvalho, J., Pacheco, M., Torres, F. (1997). Growth response of Quercus rubra to different establishment
and management techniques in an agroforestry system in northwestern Portugal. Proc. International
Workshop of Agroforestry for Sustainable Land-Use. Montpellier, France, 23-29 June. Pp. 119-125.
Cisneros, O., Martnez, V., Montero, G., Alonso, R., Turrientes, A., Ligos, J., Santana, J., Llorente, R.,
Vaquero, E. Plantaciones de frondosas en Castilla y Len, cuaderno de campo. Junta de Castilla y Len, 74
p.
Clatterbuck, W.K. 2004. Growth and development of yellow-poplar plantations on three sites ranging from 9
to 18 years. In: Connor, Kristina F., ed. 2004. Proceedings of the 12 th biennial southern silvicultural research
conference. Gen. Tech. Rep. SRS71. Asheville, NC: U.S. Department of Agriculture, Forest Service,
Southern Research Station. 594 p.
Davis, A.A., Trettin, C.C. 2006. Sycamore and sweetgum plantation productivity on former agricultural land
in South Carolina. Biomass and Bioenergy 30(2006)769-777.
Daz-Vaz, J.E. 1993. Aliso (Alnus glutinosa), ficha forestal. Lignum (11):24.
Forrester, D.I., Bauhus, J., Khanna, P.K. 2004. Growth dynamics in a mixed-species plantation of Eucalyptus
globulus and Acacia mearnsii. Forest Ecology and Management 193 (2004). Pp. 81-95.
Gabriel, K., Blair, I.F., Mason, W.L. 2005. Growing broadleaves trees on the North York Moors, results after
nearly 50 years. Quarterly Journal of Forestry 99:21-30.
Hart, J.L., Austin, D.A., Van de Gevel, S.L. 2010. Radial growth responses of three co-occurring species to
small canopy disturbances in a secondary hardwood forest on the Cumberland Plateau, Tennessee. Physical
Geography 31(3): 270-291.
Hawe, J., Short, I. 2012. Poor performance of broadleaf plantations and possible remedial silvicultural
systems, a review. Irish Forestry 69 (1&2): 126-147.
Hung, T.D., Herbohn, J.L., Lamb, D., Nhan, H.D. 2011. Growth and production varies between pair-wise
mixtures and monoculture plantations in North Viet Nam. Forest Ecology and Management 262 (2011), pp.
440-448.
Kaczmarek, D.J., Wachelka, B.C., Wright, J., Steel, V.C., Aubrey, D.P., Coyle, D.R., Coleman, M.D. 2012.
Development of high-yielding Sweetgum plantation systems for bioenergy production in the southeastern
United States. USDA. 9th Biennial short rotation woody crops operations working group conference. Oak
Ridge, Tennessee.
Loewe, M.V. 2003a. Arboricultura para produccin de madera de alto valor. INFOR-FIA, 56 pp.
Loewe, M.V (Ed.). 2003b. Perspectivas para el desarrollo de la arboricultura para produccin de madera de
alto valor en Chile. INFOR-FIA, 289 p.
11
60
Loewe, M. V., Gonzlez, O.M. 2006. Plantaciones Mixtas, Un modelo productivo con potencial para Chile.
INFOR-FIA, 299 p.
Loewe, M. V., Gonzlez, O.M., Balzarini, M. 2013. Wild Cherry tree (Prunus avium L.) growth in pure and
mixed plantations in South America. Forest Ecology and Management, pp. 31-41. DOI
10.1016/j.foreco.2013.06.015
Loewe, M.V., Siebert, W. H., Gonzlez, Q.Y., Gonzlez, O. M. 2005. Castao y Ciprs en plantaciones mixtas.
Revista Chile Forestal 313: 39-42.
Loewe, V., Siebert, H., Gonzlez, Y., Gonzlez, M. 2008. Consociazione e caratteristiche stazionali.
Sperimentazioni in impianti di raul in Cile. Revista Sherwood N 140. Pp. 48-49.
Mason, W.L., Baldwin, E. 1995. Performance of Pedunculate Oak after 40 years in mixture with European
Larch and Norway Spruce in southern Scotland. Scottish Forestry 49 (1) 5-13.
Minotta, G. 1996. Arboricoltura da legno: techniche di impianto e prime colturali. Pubblicazione N 1236,
Dipartimento di Colture Arboree, Universita di Bologna.
Mohni, C., Pelleri, F., Hemery, G.E. 2009. The modern silviculture of Juglans regia L: a literature review.
Die Bodenkultur 59 (Sondernummer): 19-32.
Negrini, C. 1994. Il bosco torna in pianura, 15 anni di esperienza. Atti Consorzio della Bonifica Reno-Palata.
Pp. 7-12.
Paris, P., Pisanelli, A., Buresti, E., Cannata, F. 1999. Agroforestry in Italy: Tradition of the practice and
research indications on new models. En: Forestry and Agroforestry for Environmental Protection and Rural
Development. SINO-ITALIAN Workshop. 1999.
Pedlar, J.H., McKenney, D.W., Fraleigh, S. 2006. Planting black walnut in southern Ontario: midrotation
assessment of growth, yield and silvicultural treatments. Can. J. For. Res. 36: 495-504
Schlesinger, R.C., Williams, R.D. 1984. Growth response of Black Walnut to interplanted trees. Forest
Ecology and Management 9 (1984) 235-243.
Siebert, H. 2006. Experiencias silvcolas con las especies Quercus robur y Quercus rubra. Gua da
silvcola.8 p.
Siebert, W.H.; Loewe, M.V. 2002. Gestione forestale compatibile con lambiente. Lesperienza del Cile.
Revista Sherwood N 77. Pp. 21-24.
Tani, A., Maltoni, A., Mariotti, B., Buresti Lattes, E. 2006. Juglans regia L. Tree plantations for wood
production in mining area of S. Barbara (AR). Evaluation of N-fixing accessory trees effect. In: Proceedings
of the V SISEF Congress: Forests & Society - Changes, Conflicts, Synergies , Edited by: E. Lingua, R.
Zano, G. Minotta, R. Motta, A. Nosenzo, G. Bovio.
12
61
62
63
64
65
66
67
International
Scientific Conference on Hardwood Processing (ISCHP
2013)
68
69
Department Territorio e Sistemi Agro Forestali, University of Padova Legnaro, Padova, Italy
C.R.A. Consiglio per la Ricerca e la Sperimentazione in Agricoltura, Forestry Research Centre, Arezzo,
Italy
3
IVALSA - Trees and Timber Institute, Sesto Fiorentino, Firenze, Italy
2
Abstract
Secondary woodland has recently attained notable importance in terms of its extension, increasing the wooded area of Central
Europe, and by raising new questions about its management. In the Italian pre-Alps, ash-lime and ash-sycamore woodlands cover
more than 150,000 ha. Usually, these new woodlands occupy privately-owned, abandoned farmland. They are generally
unmanaged or managed as coppice stands for firewood.
In order to determine the possible uses of ash timber from secondary woodlands, we carried out physical and mechanical tests on
this wood from two study areas in north-eastern Italy. The timber characteristics of two ash stands, aged 37 and 60 years
respectively, were determined by analysing twenty-five tree trunks. We calculated the wood density at a moisture content of 12%,
wood basic density, radial and tangential shrinkage, the ratio between tangential and radial shrinkage, and module of rupture and
module of elasticity. The presence of black heart was estimated in the transverse part of each sample. The timber from the two
stands presented good mechanical and physical properties, suitable for sawn timber and veneer. A high incidence of black heart,
especially in the older stand, changes the woods physical properties and strongly reduces its aesthetic value.
Single-tree-oriented silviculture was employed, which could be an interesting management system as it increased the speed of tree
growth, producing trees of marketable size (60 cm diameter at breast height) in less than 70 years, thereby also reducing the
incidence of black heart in the stems.
1. INTRODUCTION
Over the past fifty years, the wooded area in Italy has increased at an annual rate of 15,000 ha (Piussi 2006). This is
mainly due to the recolonisation of abandoned farmland and pastures by woodland in a secondary succession, as
confirmed by Del Favero et al. (1998). Sycamore (Acer pseudoplatanus L.), small-leafed lime (Tilia cordata Mill.)
and common ash (Fraxinus excelsior L.), in particular, are typical species characterizing pre-Alpine secondary
woodland. In fact, according to the IFNC (National Inventory of Forests and Forest Carbon Sinks), sycamore-lime and
ash woods cover 153,904 ha, and 88% of these woods are located in northern Italy (IFNC 2007). These post-pioneer
species will colonise fertile, newly-abandoned land very quickly (Bernetti 1995, Dobrowolska et al. 2011), and
nowadays the resulting woodland is left unmanaged to follow a secondary succession. Eventually, the woods may be
coppiced to produce firewood. However, this management system does not fully make use of the species remarkable
potential for high-quality timber production. In particular, non-intervention ash stands are very dense; there is strong
competition between trees, and dominant trees increase in girth only very slowly. Under these growing conditions, ash
will reach a marketable size only after more than 60 years of age and the tree trunks will frequently be affected by
*
70
black heart (Thill 1970, Uzielli and Petrucci 1984, Duflot 1995, Kerr 1995, Kerr 1998, Del Favero 1998, Kadunc
2005).
The aim of this study is, first of all, to evaluate the physical and mechanical properties of ash timber and to check
for the presence of black heart in secondary ash woodlands. Hence, traditionally managed (thinning from below) and
unmanaged stands of different ages were examined. Secondly, we tested a different management approach in order to
find out whether it is possible to produce ash timber of marketable size (with a diameter at breast height, dbh, of 60
cm) in an acceptable time-period. For this purpose, a management system (Duflot 1995, Bastien and Wilhem 2000,
Wilhem 2003, Spiecker et al. 2009) was employed that is characterized by an early selection of a limited number of
target trees (50-100 per ha), around which frequent thinnings from above were carried out. This new silvicultural
approach (Single Trees Oriented Forest Management - STOFOM), which is rather common in Central Europe but
almost unknown in Italy, has recently been employed in some pre-Alpine woodlands (Pelleri 2000, Pividori 2002,
Pelleri and Fontana 2004, Pelleri et al. 2009, Giulietti et al. 2009).
2. MATERIALS AND METHODS
The study was carried out in the Recoaro Terme Municipality of the Agno-Chiampo forest district (Veneto
pre-Alps, north-east Italy), where ash-sycamore secondary woodland is common. Due to the oceanic influence, the
areas climate is temperate-cold, with a high annual rainfall (Recoaro Terme, 2149 mm); the mean annual temperature
is 10.4 C. The geology of the area is very variable: on the higher-lying part of the Agno river basin, metamorphic
rocks prevail, whereas dolomite rocks abound near the Little Dolomites. The soils vary in composition and depth, and
their pH ranges from sub-acid to sub-alkaline (Pelleri and Fontana 2004).
The two plots examined are in high-forest ash secondary woodlands of different ages (37 and 60 years). In these
stands, the silvicultural approach was oriented to improving their structure and stability by traditional management
practices, characterized by thinning from below.
Unit
-1
Trees
dmg
Hmg
HD
BA
V
no. ha
cm
m
m
m2 ha-1
m3 ha-1
ash
sycamore
other species
no. (in %)
no. (in %)
no. (in %)
2nd thinning
before felled after
1020
368
652
18.2
17.8
18.4
25.8
25.5
26.0
33.3
26.56
9.16
17.40
299.1
101.6
197.5
77
8
15
97
3
0
71
12
17
Unit
Trees
dmg
Hmg
HD
BA
V
no. ha-1
cm
M
M
m2 ha-1
m3 ha-1
ash
other species
no. (in %)
no. (in %)
preparatory cutting
before felled
after
552
234
318
23.7
20.3
25.9
26.4
24.9
27.2
33.4
24.36
7.57
16.79
245.7
79.8
165.9
80
20
93
7
70
30
The younger stand (= MS, aged 37 years) is a pure ash secondary woodland on former fertile farmland, with a high
tree density and a remarkable height of dominant trees (HD). This stand was thinned in 2001 and in 2005 (Table 1a).
The older stand (= LR, aged 60 years) is an ash woodland on an abandoned wet pasture, characterized by sycamore,
grey alder (Alnus incana Vill.) and by an irregular field layer of common hazel (Corylus avellana L.). In 2005, a first
thinning from below was carried out in the whole stand; already existing gaps in the canopy were enlarged and new
small gaps created. The aim of this management (preparatory cutting) was to prepare the hitherto unmanaged stand for
regeneration (Table 1b).
Another young stand (= COV) was examined as it had previously been managed by the STOFOM system (Giulietti
et al. 2009). COV is a dense, young-growth ash-sycamore stand on formerly fertile farmland. In 2001, when the stand
was in the thicket-pole stage, a first selective thinning was carried out. Another two thinnings took place in 2004 and
71
2010, around the best 100 target trees (TT) per ha. These TTs had been selected from vigorous dominant trees, whose
stems were without defects, with a thick and well-shaped crown. Using the same criteria, the best 100 target trees per
ha were also selected in an unmanaged control plot. In total, the COV plot was thinned three times, at 14, 18 and 24
years of age (Table 2).
Table 2: Stand and target-tree characteristics in plot COV during the thinning operations
COV
Unit
Trees
BA
100 target trees
dbh
Height
Crown insertion
no. ha-1
m2 ha-1
cm
m
m
72
73
ID
Uf
Ueq
Df
De
D0
BD
Bl
Bt
Br
C
MOEd
MOES
MOR
Sd
HB
Unit
%
%
g/cm3
g/cm3
g/cm3
g/cm3
%
%
%
MPa
MPa
MPa
MPa
MPa
Mean
48.2
11.9
0.86
0.74
0.70
0.58
0.38
12.70
6.24
47.44
14117
11096
102.00
17.74
4.03
SD
10.02
0.33
0.08
0.06
0.06
0.04
0.33
2.54
1.52
4.43
2123
2294
7.99
1.60
0.48
No.
373
375
373
376
376
375
377
377
376
85
85
85
85
85
85
Samples without visible signs of black heart and those with 100% black heart are compared in Table 4.
Table 4: Comparison of the physical characteristics of ash wood with and without black heart
ID
Uf
Ueq
Df
De
D0
BD
Bl
Bt
Br
Unit
%
%
g/cm3
g/cm3
g/cm3
g/cm3
%
%
%
0% black heart
Mean SD
No.
46.77 8.39 346
11.95 0.32 345
0.85
0.07 346
0.73
0.06 345
0.69
0.06 346
0.58
0.04 346
0.37
0.32 346
12.53 2.47 346
6.15
1.47 346
P
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
< 0.05
< 0.01
> 0.05
All the physical properties analysed showed a statistically significant difference in the mean values of samples
with and without black heart, except radial shrinkage. In general, ash wood affected by black heart has a higher
moisture content in fresh wood, and consequently higher fresh wood density. Both moisture content and wood density
at equilibrium differed less in absolute terms, showing that the relatively higher values of those variables in fresh wood
are due to a higher water content and not to a modification at cell wall level. In any case, the basic density value that
specifies the dry mass in relation to fresh plant material is again higher in samples with black heart. A possible
explanation for this is the presence of extractives in wood affected by black heart. These substances are possibly the
same as those that modify wood colour.
The dimensional stability of ash wood also differs in relation to the incidence of black heart. We observed higher
values of shrinkage in all three anatomical directions (only longitudinal and tangential being significantly different).
This is also noted when comparing the ratio of tangential and radial shrinkage, which is lower in wood without black
heart (2.03), compared with wood with black heart (2.27).
74
4. CONCLUSIONS
This study showed that wood affected by black heart contains more water; it is heavier and less stable during
variations of its moisture content. We were unable to compare the mechanical properties of wood with and without
black heart, but mechanical test results for wood samples without defects suggest similar mechanical properties as
those of European ash from other provenances.
We suggest that ash wood affected by black heart should be used where extreme dimensional stability is not
required. Furthermore, thermal treatment of ash wood affected by black heart may overcome aesthetic problems of
discolouration, and stabilize the wood.
In order to reduce the presence of black heart in ash and to obtain the best quality timber, tree trunks of 60 cm
diameter should be produced in less than 70 years, employing a management system that ensures dbh increments
around 1 cm. Single tree oriented management will maintain these growth rates.
REFERENCES
Bastien, Y. and G.J. Wilhelm. 2000. Une sylviculture darbres pour produire des gros bois de qualit. (traduzione italiana a cura di
Silvia Bruschini In: Sherwood-Foreste ed alberi oggi, 86: 5-13) Rev. For. Fr., LII (5): 407-424.
Bernetti, G. 1995. Selvicoltura speciale. UTET Torino: 150-160.
Del Favero, R. 1999. Primo contributo sulla conoscenza del frassino maggiore in alcune formazioni del nord-est dell'Italia: la
qualit degli alberi in piedi e le caratteristiche dell'anello legnoso. LF.M. (6): 289-307.
Del Favero, R., E. Abramo, and A. Zanella. 1996. La stima della qualit degli alberi in piedi: il caso della rovere e del frassino
maggiore. I.F.M., n 6, 367-387.
Del Favero, R., L. Poldini, P.L. Bortoli, C. Lasen, G. Dreossi, and G. Vanone. 1998. La vegetazione forestale e la selvicoltura nella
Regione Friuli-Venezia Giulia. Regione Autonoma Friuli-Venezia Giulia, Direzione Regionale delle Foreste, Udine. 490 pp.
Dobrowolska, D., S. Hein, A. Oosterbaan, S. Wagner, J. Clark, and J.P. Skovsgaard. 2011. A review of European ash (Fraxinus
excelsior L.): implications for silviculture. Forestry 84: 133-148.
Duflot, H. 1995. Le frne in libert. LD.F., Paris. 192 pp.
Giordano, G. 1987. Tecnologia del legno. I legnami del Mondo. UTET: Torino.
Giulietti, V., F. Ferretti, and F. Pelleri. 2009. Thinning trials in secondary growth ash-maple stands carried out in the
Agno-Chiampo district (VI): results after the second thinning. Annali CRA-SEL, 35, 2007-2008: 87-100.
INFC. 2007. Le stime di superficie 2005. Prima parte. Autori: Tabacchi, G., F. De Natale, L. Di Cosmo, A. Floris, C. Gagliano, P.
Gasparini, L. Genchi, G. Scrinzi, and V. Tosi. Inventario Nazionale delle Foreste e dei Serbatoi Forestali di Carbonio.
ISO 3129. 1975. Wood - Sampling methods and general requirements for physical and mechanical tests.
ISO 3130. 1975. Wood - Determination of moisture content for physical and mechanical tests.
ISO 3131. 1975. Wood - Determination of density for physical and mechanical tests.
ISO 3133. 1975. Wood - Determination of ultimate strength in static bending.
Kadunc, A. 2005. Factors influencing brown heart formation in common ash (Fraxinus excelsior L.). Centralblatt Forstwesen,
Austrian Journal of Forest Science 122(4): 205-220.
Kerr, G.1995. Silviculture of ash in southern England. Forestry 68: 63-70.
Kerr, G.1998. A review of black heart of ash (Fraxinus excelsior L). Forestry 71: 49-56.
Manetti, MC. 2002. Tree ring growth by core sampling at the CONECOFOR Permanent Monitoring Plots. The deciduous oak
(Quercus cerris L.) type. J. Limnol. 61 (suppl. 1):55-61.
Manetti, MC., G. Bertini, and M. Pichi. 2003. S.M.I.L. 3.1-Sistema di Misura degli Incrementi Legnosi . un programma per
l'indagine incrementale di sezioni trasversali e per l'analisi del fusto. An. Ist. Sper. Selv (32): 65-71.
Pelleri, F. 2000. Growth analysis and thinning trial in a newly formed ash stand. An. Ist. Sper. Selv. (XXIX 1998): 17-28.
Pelleri, F. and S. Fontana. 2004. Thinning trial in secondary growth ash-maple stands carried out in the Agno-Chiampo district
(Eastern Prealps). An. Ist. Sper. Selv (31): 39-54.
Pelleri, F., M. Pividori, and V. Giulietti. 2009. Cultural practices in ash-sycamore secondary forests in Northern Italy. Atti III
Congresso Nazionale di Selvicoltura, Taormina, 16-19 ottobre 2008: 887-893.
Piussi, P. 2006. Close to nature forestry criteria and coppice management. Nature-based forestry in central Europe: alternatives to
industrial forestry and strict preservation. Edited by Jurij Diaci. Ljubljana: 27-37.
75
76
Holzforschung
Technische Universitt Mnchen
Mnchen, 80797, Germany
ABSTRACT
In European countries tropical hardwoods have been successfully used for load bearing applications for centuries.
Some countries have their own visual strength grading system either exclusively for tropical timber or included in
more general grading guidelines for sawn timber. These grading systems use partly altered grading characteristics
or different permissible values. Within the framework of determining and assigning strength class of tropical wood
species, these standards have been reviewed and assessed. None of the standards consider the influence or
interference of combinations of visual characteristics. From experience it seems that tropical timbers which have
been traded on the European market and have been used to determine the strength class, often do not include
specimens with the lowest allowable values of the grade characteristics. This means that the assigned strength
class often does not comply with the lower limit of visual grading characteristics. In the past, strength classes of
tropical wood species were often determined on small clear test specimens and adjusted for structural sizes
applying some factors based on experiences. It is concluded that a broad European standard for visual strength
grading of tropical hardwood would encourage trade and use of tropical timber for load bearing structures in
Europe. Moreover introducing machine strength grading systems for tropical hardwoods would increase the yield.
Furthermore sampling aspects might be altered in order to make it more applicable.
1. INTRODUCTION
Import of tropical timbers from non-sustainable forests into the European market has been banned by the EU
Timber Regulation (EUTR). About 1 million cubic meter of sawn tropical timber enters the European market each year.
Main importers in Europe are Belgium and The Netherlands (ITTO 2013). These countries are also acting as a
distributer for tropical timber in Europe. Primarily, tropical timber is used for non-load bearing applications such as
decking, cladding, furniture and fences. A considerable amount of timber from sustainable forests from Africa, South
America and Asia is used in load bearing structures. The advantages of these wood species in comparison to coniferous
species and European grown hardwoods are both the high strength properties and the remarkable biological durability
against micro-organisms.
Important application fields for tropical timber in load bearing structures are hydraulic works such as lock gates,
mooring poles, jetties, fenders, sheet piles and bridges (pedestrian, bikers) (Kuilen 2008). The new developments in the
Netherlands have generated high valuable applications such as guard rails (motorway level H2) and glued laminated
beams for roof structures.
During the last decades several new wood species from the tropics have been entering the European market.
Almost 80% of these wood species belongs to a group of lesser-known species. That means that the identification
characteristics and/or wood properties are hardly known by the market participants.
Following the European design codes (EN 1995), timber for load bearing structures has to be strength graded to
assign characteristic strength values. In principle, there are two strength grading methodologies widely accepted:
Firstly, visual strength grading where visual characteristics of the timber such as knots could be related to strength
77
properties and secondly, machine strength grading where intrinsic properties such as dynamic modulus of elasticity
would be a predictor for strength. Presently for hardwood (incl. tropical) there are no approved machine grading
systems on the market. But visual grading systems for hardwood have been established in several European countries
for decades. In both cases, settings for strength classes have to be established by developing models based on
destructive timber testing.
To determine characteristic strength values for lesser known wood species require extensive testing programs
according the European Standard EN 384. In some cases, such a massive test program does not pay for wood species
which are available in small amounts only. The standard EN 384 is quoted in the product standard (EN 14081) which
forms the basis for the CE marking for sawn timber as construction material. Since January 2012 CE marking is
obligatory in the European market.
In this paper current visual strength grading systems for tropical timber will be discussed with regard to the effect on
strength class assignment.
2. VISUAL STRENGTH GRADING STANDARDS
Visual strength grading is based on single characteristics and features on the surface of the timber which could be
related to the characteristic values of strength properties. The principles of visual grading for soft- and hardwoods are
laid down in the European Standard EN 14081-1:2005+A1:2011. This standard states the features that should be
considered in standards for visual strength grading. In table 1 these characteristics are listed.
Table 1: Visual characteristics for strength grading standards (according EN 14081-1:2005+A1:2011)
Characteristics
Specification
Knots
Maximum dimensions of knots or knot holes shall be specified. In certain sizes, knot groupings
affect the timber strength and shall be taken into account.
Slope of grain
The grading standard should give limitations on the slope of grain for each grade specified.
Local fibre deviations around knots shall be disregarded in measuring slope of grain.
The grading standard shall contain a requirement for either density or rate of growth.
Fissures
Where fissures have a significant effect on strength, e.g. shear strength of a beam, they shall be
limited. Otherwise they shall be disregarded. In that standard guidance is given with regard to
the maximum total of length of fissures in a piece of timber.
Wane
Criteria for the wane limitations with references to the width, thickness and length of the piece
shall be given. The maximum wane permitted shall not reduce the edge and face dimensions to
less than 2/3 of the basic dimensions of the piece.
Warp
The limitations of maximum permitted values of warp for bow, spring and twist shall be given.
Maximum warp (in mm) over 2 m of length is listed in this standard.
Degradation by biological
organisms
Standards shall include requirements that limit fungal and insect damage to timber. Soft rot shall
not be allowed in any grade.
Reaction wood
Standards for softwood species shall take account of compression wood and tension wood for
hardwoods.
EN 14081 has been drafted for structural size timber but not for small clears. Depending on the level of severity of
the characteristics on the surface of the timber, different strength grades may be established and added to the standard
EN 1912. In this standard, wood species are assigned to strength classes based on a visual grading system.
2.1. NATIONAL STRENGTH GRADING STANDARDS
The UK, France and the Netherlands are main users of tropical timber for structures. Main use is in marine
structures, or other structures exposed to weather so durable wood species are of most concern. Consequently, these
countries have national standards for visual strength grading of hardwood standards where tropical hardwoods have
been included.
78
National grading standards have been drafted based on the experience and wood qualities available on the market.
These standards have been reviewed during the last few years by national committees. The most current strength
grading standards are the British Standard BS 5756:2007+A1:2011 Visual grading of hardwood, the French Standard
NF B 52-001-1:2011 Visual classification for the use of French softwood and hardwood species in structures Part
1: Massive wood. and the Dutch Standard NEN 5493+C1:2011 Quality requirements for hardwoods in civil
engineering works and other structural applications. Each one of these standards gives a set of visual characteristics
with only one set of corresponding threshold values. This means that only one quality of tropical hardwoods has been
defined leading to one strength class per wood species.
The three standards contain the required characteristics laid in EN 14081-1 except Density and rate of growth. In
principle tropical hardwoods do not have distinguishable growth rings. Therefore the density is not a visible
characteristic for tropical hardwood species.
The current national grading standards comprise more or less the same set of characteristics regarding strength
properties but often differ on the threshold levels of the single characteristic. Moreover, the measurement
methodologies regarding visual features are not always comparable between the standards, such as the measurement of
fissures and derived limits from that.
The Dutch standard is a comprehensive standard which considers not only strength related features but also aspects
in conjunction with a variety of applications. Different categories of timber use have been defined such as decking,
hand railings for bridges, sheet piling and other hydraulic applications with large cross sections. Guidelines for
specification and dimensioning of e.g. sheet piles are included. The British and French standard consider exclusively
strength related characteristics.
Table 2 gives an overview of the 3 national standards regarding the characteristic related thresholds.
Table 2: Grading features and the related thresholds of national grading standards
Great Britain
BS 5756:2007+A1:2011
France
NF B 52-001-1:2011
The Netherlands
NEN 5493+C1:2011
General
Visual strength
class labeling
Threshold <0.25
Threshold <0.25
Threshold 0.2
Slope of grain
1:10
Wane
French Guyana:
< 10 % in thickness
Other tropical hardwoods:
See BS 5756
Not permitted
Fissures
No remark
Boxed heart
79
Great Britain
BS 5756:2007+A1:2011
France
NF B 52-001-1:2011
The Netherlands
NEN 5493+C1:2011
Brittleheart
Not permitted
Not permitted
Not permitted
Tension wood
No remark
Not permitted
No remark
Sapwood
No remark
French Guyana:
One surface and edge :< 50% of
thickness, <50% of width
All other tropical hardwood:
No remark
Not permitted
Minor restrictions
No remark
Not permitted
Distortion
Bow:
Spring:
Twist:
Cup:
Bark in pockets
Biological damage and other defects are considered in the three standards and have been restricted on different
natures and amounts. It seems difficult to assess these restrictions in relation to the effect on the strength properties as
this applies also to for instance barks in pockets, resin and gum pockets and kind of mechanical damages. It is likely
that these features and restrictions are established from history, experience, application and appearance of the timber
which has been traded on the domestic market. For example the restriction of sapwood has no relation with the strength
but with natural durability of the timber and thus has an application aspect. Restrictions on distortions could come from
national/European design codes where these have been taken in to account in design calculations and/or practical
reasons which are related to the production process of the structure or their elements.
Slope of grain
It is believed that Slope of grain and Knot values are the major strength related visible parameters for tropical
hardwoods. Depending on the quality of the tree, its timber could be knot free. Finally, the only decisive characteristic
would be the slope of grain. In order to recognize and measure the slope of grain on tropical hardwood the surface has
to be planed, otherwise difficulties in interpretation may occur. The progression of the slope of grain in tropical timber
is sometimes fuzzy on the surface without a clear slope. This may lead to unreliable measurements. Interlocked grain
might affect strength related properties but only the British standard takes this into account but without clear definition
and measurement instruction.
Wane
Since sapwood is not allowed according to the Dutch standard, consequently wane is not permitted. In contrast to
this, the British and French standards accept a certain amount of wane. The cross section in thickness and width may be
reduced up to 1/3 of the side by the presence of wane. This may cause a strength loss of 15% in bending. In this regard
the Dutch standard may lead to higher strength class allocation of the same wood species.
Less visible characteristics
Tropical hardwood may contain brittleheart (Kuisch et al. 2012) which is not permitted by any of the visual grading
standards (Table 2). However brittleheart is almost invisible to the naked eye, as is tension wood. Tension wood may
reduce the strength properties to a certain extent but brittleheart in its most severe form could reduce the strength by
about 90%.
Interdependency of grading features
All grading features have been treated separately with regard to the allocated thresholds. No interdependencies have
been considered in the standards. It is not clear if specific combinations of characteristics impair the strength properties
stronger than a single feature. It could be questioned if a prediction model for allocating the strength class has taken this
into account. At the same time, the probability that more than one grade determining property occur at the same
instance seems small and so far not supported by practical experiences.
80
Wood species
Origin
Number of sub-samples
Number of specimens
Size of specimen
In this paragraph two aspects with regard to sampling are highlighted because of their key importance.
Identification of Wood species
The sample should represent a population which should be defined by a wood species, its origin and the growing
conditions. Tropical hardwoods are normally graded after they have entered the European market. First challenge is
the identification of the timber, second to find out the country/location of origin.
Usually timber is delivered largely in batches under a trade name. Trade names in particular are assigned by
individual traders or common names are introduced by the wood industry of single countries.
After the tree is harvested, the trunk will be transported to a dispersal area or straight to a sawmill where the primary
conversion takes place. Normally, different wood species from different origins are processed at the same workstation.
From this stage it is possible that the timber could be mixed because of the same secondary visual characteristics. The
timber gets a trade name which is used during the following process and transport stages until it arrives in Europe.
Several attempts have been made to constitute a tabulated overview of common timber trade names in standards with
the accompanying scientific name of the wood species (EN 13556). It seems that those lists have a limited validity
because of the dynamics of the trade market which changes trade names of wood species and merges wood species
under one trade name. The reason for this could be for example the use of secondary identification characteristics such
as colour, density and grain texture of the timber but also the decrease of availability of certain wood species and
qualities.
During transportation of the timber in the country of origin and/or regions the timber can get different trade names.
For example the trade name cumaru is used in Brazil for an assortment of at least eight wood species, whereas in
Venezuela the trade name sarrapia is used (Wiselius 2010). From these species, only two, (Dipteryx odorata
(Aubl.)Willd. and Dipteryx oleifera Benth.), can be distinguished by the width of the rays (Figure 2) with the aid of a
light microscope. In practice however, identification may be not possible because the anatomical features are not
distinctive enough or the needed skills at the site are not available. In several cases wood species cannot be identified at
species level but be determined at a higher hierarchical level of the plant systematic such as Family or Genus. To
a great extent this applies to assortments which consist of less known tropical hardwood species (Ravenshorst et al
81
2012). Cumaru should not be identified as a wood species (botanical name) but as timber species (trade name), which
could include different wood species. These different wood species should then be included in the testing however.
Figure 2 shows their small differences between two wood species. The variation of the wood anatomical elements
on the slices of the cross-section is not significant. The most identifying feature is the formation of the rays. Even this
feature can vary in intensity within a single species, whereby the identification could be doubtful.
82
In EN 1912 a lot of tropical hardwood species have been tabulated with their strength class assignment. Many
assignments have been based on test results of small clear samples. Not in all cases the transformation factor from small
clear samples to structural sizes is known or documented. Recent test results, from tropical hardwood species, derived
from samples of structural sizes do not comply with the declared values in EN 1912. For that reason a study was carried
out with various timber species in order to determine a correction factor for small clear specimens following EN 384.
Lanvin et al (2009) presented preliminary results of this study and proposed a correction factor of 0.35 derived from the
regression line. Figure 3 shows, from the same research, the regression line for the merged data of five wood species
expressed as Y= 0.31x. However, this value of Y is the prediction of the mean 5%-values of the full size data. This
means that for this prediction line 50% of the predicted 5%-values for individual wood species will be too high. To
make sure that all 5%-values are safely predicted, a prediction equation of Y= 0.21x needs to be applied. In the French
standard NF B 52-001-1:2011 a transformation factor of 0.3515 (Y=0.35x) was introduced.
Figure 3: Relation between mean bending strength of small clear samples and characteristic bending strength of full size samples.
fm,mean: mean bending strength, fm,k: characteristic bending strength, regression line Y=0.31x (solid line), safety regression line
Y=0.21x (lower dashed line)
4. DISCUSSION
The extensive use of tropical hardwoods in load bearing structures is limited to a few European countries. In history
strength properties of these wood species were based on test procedures differing from current standards, and on
experience in certain applications. Countries such as The Netherlands, France, Belgium and UK had established their
own grading rules and strength class system. Because of the harmonization of national building product regulations,
and the free trade of building products throughout the European domestic market, it is desirable to increase consistency
for assessment procedures and definition of terms. Most of the relevant testing procedures with regard to strength
grading have been harmonized in Europe.
Weakness of the visual strength grading system
At the moment tropical hardwood is restricted to visual strength grading procedures. These are edited nationally
and use partly different characteristics and threshold values as described above. When settings are determined for a
grade, the testing sample has to cover the whole range of characteristics and threshold values which are described in the
standard for visual grading. Usually, tropical timber which enters the European market is already pre-sorted even
though many times information on this pre-sorting is lacking. Low grade timber regarding appearance normally
remains in the country of origin. The established grading standards do not take this into account.
It might happen that a lesser known tropical timber species would be used in load bearing structures. Therefore a
strength class necessarily have to be determined. The timber has to be graded according to an existing visual strength
grading standards such as NEN 5493. The timber sample should represent the quality of the timber from their origin.
Moreover the sample should cover the whole range of characteristics with their accompanying threshold values. At
least the sample should include enough specimens where the grading characteristics have values close to the threshold.
In Figure 1 the threshold border is indicated by the horizontal solid line. On the right side of the line, specimens do not
comply with the requirements and will be rejected. Samples of batch 2 fulfill the requirements of the standard because
83
there are enough specimens with values of the characteristics close to the threshold line. Batch 1 (squared markers)
comprises mainly specimens with low threshold values, means high strength values. This batch does not comply with
the requirement of the standard. Hence the characteristic strength may not to be determined. However, it would be
possible to obtain a strength class if the threshold line would be shifted to the left in Figure 4 (horizontal dashed line).
Through this the specimens of batch 1 cover the whole range of the new threshold line. Another possibility is to
degrade the characteristic strength value by a factor in order to compensate the missing specimens at the lower quality
level, close to the threshold line of the standard. Such a factor could be derived from similar wood species which have
been tested in considerable amounts.
Stapel et al (2011) has shown that the minimum allowable number of specimens per sample according to EN 384 is
clearly too small to represent a certain quality of a wood species including its origin. Sample size and number of
samples have a significant influence on the characteristic strength values which are assigned to certain species and
grades. This may apply particularly for tropical hardwood species with a widespread growth area such as Brazil.
Figure 4: Principle of obtaining two different strength classes by applying one visual grading set of characteristics and threshold
levels (vertical lines) on the a wood species.
5. CONCLUSIONS
It has been shown that current national standards for visual strength grading are not consistent in the definitions of
characteristics and threshold levels which might lead to diverse interpretations and finally to different strength classes,
obtained from the same timber batch. Important visual characteristics such as knots are often not present or hardly to
identify. The latter applies also for slope of grain.
One European standard for visual strength grading for tropical timber would be an advantage for both supplier as
well as users. More knowledge about the interdependency of visual strength characteristics would be helpful to
optimize the threshold settings.
Two major aspects have been identified for tropical hardwood regarding sampling:
1. The identification of timber species needs attention because the common procedure of species identification might
not be applied in the field because of its complexity. A wood species independent strength grading system for
tropical hardwoods is recommended.
2. Confining the location of origin of the timber species by country as done for softwoods in Europe seems not
reasonable because the area of Brazil is twice that of the EU. Other parameters such as climate, growth, forest
management etc. are governing the strength properties more than national borders.
84
REFERENCES
BS 5756:2007. Visual grading of hardwood Specification. BSI British Standards Institution, London, UK.
EN 13556:2003. Round and sawn timber-Nomenclature of timbers used in Europe. European Committee for Standardization,
Brussels
EN 14081-1:2005+A1:2011. Timber structures - Strength graded structural timber with rectangular cross section - Part 1:
General requirements. European Committee for Standardization, Brussels.
EN 1912:2012. Structural Timber - Strength classes - Assignment of visual grades and species. European Committee for
Standardization, Brussels.
EN 1995-1:2004. Eurocode 5: Design of timber structures - Part 1-1: General -Common rules and rules for building European
Committee for Standardization, Brussels.
EN 338:2009. Structural timber - Strength classes. European Committee for Standardization, Brussels
EN 384:2010. Structural timber - Determination of characteristic values of mechanical properties and density. European
Committee for Standardization, Brussels
EN 408:2010+A1:2012. Timber structures - Structural timber and glued laminated timber- Determination of some physical and
mechanical properties. European Committee for Standardization, Brussels.
Geach, J. and Thomas, D. 2011. Applied Genetics for Forest Conservation and Sustainable Trade.-The state of DNA technology
for trees and wood products. DoubleHelix Tracking Technologies Pte Ltd, Singapore, Report July 2011.
Hltken, A.M., Schrder, H., Wischnewski, N., Magel, E., Degen, B. and Fladung, M.. 2012. Development of DNA methods to
identify CITES-protected timber species: A case study in the Meliaceae family. Holzforschung, Vol. 66, pp. 97104.
ITTO 2013. Tropical Timber Market Report. Volume 17 Number 7, 1st 15th April 2013.
Kuilen, JWG van de. 2008. The use of timber in hydraulic structures. In proceedings COST E53: End users needs for wood
material and products. Delft, The Netherlands.
Kuisch, HP. , Gard, WF. , Botter, E. and Kuilen, JWG van de. 2012. Brittleheart as a critical feature for visual strength grading of
tropical hardwood - approach of detection. In P Quenneville (Ed.), World conference on timber engineering (pp. 20-39).
Lanvin, JD., Reulin, D., Rouger,F., Kuilen, JW van de, Ravenshorst,G., Reinbol, G.; Bourguignon, H., Grar, J., Guibal, D.,
Boilley, E. and Verna, M. 2009. Simplified strength properties assessment for tropical hardwoods in view to CE marking. In
Preceding: ISCHP 09 International Scientific Conference on Hardwood processing. Paris, France
NEN 5493:2011. Quality requirements for hardwoods in civil engineering works and other structural applications. Nederlands
Normalisatie-instituut, Delft, The Netherlands.
NF B 52-001-1:2011. Regulations governing the use of timber in structure Visual classification for the use of french softwood
and hardwood species in structures Part 1: Massive wood. Association Franaise de Normalisation (AFNOR), La Plaine
Saint-Denis Cedex, France.
Ravenhorst, G.J.P., Gard, WF and Kuilen, JWG van de 2012. Classification and sampling of tropical wood species for strength
and durability assignments. In P Quenneville (Ed.), World conference on timber engineering.
Stapel, P., Kuilen, JWG van de and Ravenshorst, G. 2011. Influence of sample size on assigned characteristic strength values. CIB
Conference working commission W18 - timber structures, CIB-W18/44-17-1, Alghero, Italy
Wiselius, S.I. 2010. Houtvademecum. Centrum Hout Almere/Netherland.
85
1. INTRODUCTION
In the temperate hardwood forests of North America, forest practices have, for a long period, favoured the
harvesting of the largest trees. Diameter-limit cutting was applied with a preference for vigorous trees, which resulted
in residual stands mainly composed of poor-quality stems and low-value species, which lacked genetically superior
seed sources (Nyland 1992). As a result, the two most economically important industries of the forestry sector sawn
wood and veneer had to adjust to a raw material supply of lower quality (Majcen 1994).
In an attempt to improve the situation and control forestry operations on public land, a new forest management
system was introduced in Quebec (Canada) in 1986, which focused on the improvement of forest productivity and the
protection of the natural regeneration. Under this new regime, selection cutting was introduced in the temperate
hardwoods forests, and in less than a decade, diameter-limit cutting was phased out. The selection cutting system was
designed to prioritize the harvesting of good quality non-vigorous trees, i.e. those that were likely to die before the next
harvesting cycle, estimated at 20 years (Majcen 1994). The second harvest priority was low-quality trees that were also
likely to die before the next harvesting cycle. This was thought to more closely reflect natural forest dynamics, and was
devised to improve the future quality of the resource by protecting the understory, leaving a larger number of vigorous
trees and promoting biodiversity (Bdard and Majcen 2003).
In a wide-scale study monitoring forest growth after such selection cuttings, Bdard and Brassard (2002) showed
that the industrial application of this system was deficient. They attributed poor growth after harvest to inadequate
estimations of tree vigour. To alleviate this situation, the Ministry of Natural Resources of Quebec (MRN) developed a
new tree vigour classification system, which is used to determine harvesting priorities (Boulet 2007). In this system,
standing trees are categorized in four harvesting priorities based on visual defects such as fungal attack, stem checking,
crown defects, etc. The classification system, which is still in use in the province, prioritizes the harvest of dying or
unproductive trees, and aims at preserving trees with vigorous growth. Thus, forest industries had to adapt to a new
situation where the bulk of the timber supply was based on low-vigour (i.e. dying and unproductive) trees.
* Corresponding author: Tel.: (418) 656-2131 ext: 14833; Fax: (418) 656-2091; E-mail: mariana.hassegawa.1@ulaval.ca
86
To counteract the resulting loss of quality in wood supplies, Pothier et al. (2013) proposed that silvicultural
prescriptions should aim to prioritize the harvesting of small, low-vigour trees because these are more likely to be of
high quality. The authors observed that the probability of occurrence of low-vigour/low-quality trees increases directly
with stem diameter at breast height. Because of their characteristics, these individuals are not suitable for conversion
into veneer or lumber. Therefore, a proportion of larger low-vigour/low-quality trees should be replaced by an
equivalent volume of smaller low-vigour/high-quality stems.
However, Pothier et al. (2013) acknowledged that their assessment did not take into consideration either the product
assortment obtained from those trees or its monetary value. To address this issue, models were developed for predicting
volume by log grade in a tree (Fortin et al. 2009), and for estimating the merchantable value of lumber contained in a
log of a given grade and size (Petro and Calvert 1976). However, the latter model was developed almost four decades
ago, which casts a degree of doubt as to its current applicability. Therefore, the objective of this study was to improve
the predictions of lumber product recovery for logs.
2. MATERIAL AND METHODS
2.1. SAMPLING
To evaluate the current applicability of the model, 32 yellow birch (Betula alleghaniensis Britt.) and 64 sugar
maple (Acer saccharum Marsh.) trees were sampled from multiple stands in two regions of the province of Quebec
(Canada). They were measured and categorized, applying a widely used sawlog classification system to estimate the
sawn wood potential from a log (Fortin et al. 2009, Hanks 1976, Petro and Calvert 1976, Vaughan et al. 1966).
According to this classification, logs can be arranged in three categories, namely F1, F2 and F3, in descending order of
quality. A fourth grade (F4) was also included, as there is the possibility of obtaining sawn wood from the smaller logs.
Depending on the log size, the material was either transported to the sawmill at the School of Forestry in Duchesnay
or to Laval University, in Quebec, to be converted into boards. Independently of their dimensions, logs were sawn to
maximize high grade lumber (i.e. knot-free and sapwood) (Richards et al. 1979, Steele 1984). Sawn boards were 2.5 cm
thick, with variable widths and lengths, ranging from 7.6 to 35.6 cm, and 1.2 to 3.6 m, respectively. All boards were
graded according to the National Hardwood Lumber Association standards (NHLA 2011), before and after being kiln
dried. Then, the lumber grade-color characteristics were evaluated to enable the classification of the material according
to standards established in the Hardwood Market Report (HMR).
2.2. LOG VALUE MODEL
The method currently used to predict the lumber merchantable value of a log is based on a model developed in the
1960s (McCauley and Mendel 1969, Mendel and Smith 1970, Petro and Calvert 1976), shown in Equation 1. Despite
the fact that this method is found in different publications, we took Petro and Calverts (1976) publication as the basis
for our comparisons, henceforth referred to as P&C.
LMV QI P V
(1)
87
The QI is a numeric value that indicates the relative value of a log in terms of yield of the 25 mm (4/4) lumber. It
represents the sum of the sawn wood yield by board class multiplied by the relative price by board class (Equation 2):
QI Yc RPc
(2)
Figure 1: Price variation for board category 1C, for sugar maple and yellow birch.
As observed, prices did not follow a predictable path. However, this variation does not negatively influence the
prediction, as the model was conceived to account for the price variation through time.
The uncertainty associated with variations in price can be reduced by using the relative price. Petro and Calvert
(1976) stated that the latter variable would remain reasonably constant over time. When their work was first published,
the authors observed constancy between the relative prices of different lumber classes. They stated that the relative
price for lumber quality FAS was always 1.33 to 1.54 compared to the base quality; the quality 2C was 0.69 to 0.73 and
the quality 3C was 0.38 to 0.47 of 1C. They advocated that these relationships should hold for the decade that followed
the publication. To verify if those numbers were still valid, we calculated the relative lumber prices for the 5-year
period up to and including 2012. The average values for both species can be found in Table 1.
88
Table 1: Average relative prices for sugar maple and yellow birch.
Lumber quality
1960-1970
FAS
1C
2C
3C
1.33 1.54
1.00
0.69 0.73
0.38 0.47
2008-2012
Yellow birch
Sugar maple
1.81
1.65
1.00
1.00
0.70
0.69
0.25
0.37
Adapted from HMR
As observed, the relative price for lumber quality FAS has increased when compared to the 1970s. The relative
price for the class 3C, on the other hand, has decreased. This shows that the market prices for each board quality class
do not follow the same trend. As the relative price is one of the variables used for obtaining the QI, the variation
observed above will definitely have an influence on this key element.
We compared the predicted P&C quality indices with the ones obtained using our samples (Figure 2).
Figure 2: Observed and predicted (P&C) quality indices for sugar maple and yellow birch.
The predicted values for QI are underestimated by P&C, especially for lower grade logs. This may be due to
changes in the yield of sawn lumber or to the variation in the relative price. Either way, the quality index, being an
important variable for estimating lumber merchantable value, should be re-evaluated.
The prediction of lumber value of a log uses, along with the market price for board 1C and the QI, the sawn wood
volume, which is the total board volume that can be obtained from a log. Stem defects and intrinsic characteristics, like
taper or the presence of decay, have a negative influence on lumber yield. Figure 3 presents the sawn wood volume
predicted by P&C against the observed volume for both species.
The predictions from P&C seem to be underestimating the actual sawn wood volumes for both species, especially
for log grades F2 and F3. The tendency for logs F1 is not very clear due to the small number of observations in our
database. In the 1970s, the industry did not consider log grade F4 as having sawlog potential, so we assigned a zero
value to those predictions. This discrepancy in the values demonstrates the need for re-assessing the sawn wood
volume.
89
International
Scientific Conference on Hardwood Processing (ISCHP2013)
International Scientific Conference on Hardwood Processing
(ISCHP 2013)
Figure 3: Observed and predicted (P&C) sawn wood volumes for sugar maple and yellow birch.
The lumber monetary value was calculated using Equation 1 with both the 1976 and 2012 input parameters. The
market price for sawn wood for the reference quality was the same in both cases, for comparison purposes. The other
parameters (QI and sawn wood volume) were as given by Petro and Calvert (1976). The comparison between P&C and
the observed results for both species can be seen in Figure 4.
Figure 4: Lumber monetary values for sugar maple and yellow birch using 1976 (predicted) and 2012 (observed) parameters.
P&C underestimated the lumber monetary value of a log when compared to the current parameters. The excess
zeros are due to the exclusion of log grade F4 in their method. After verifying the imprecision in the estimation of the
LMV, we decided that an update of the model was necessary.
3.2. RECALIBRATION OF THE LMV MODEL
After analyzing the equations proposed by Petro and Calvert (1976), we verified that the concept for calculating the
lumber monetary value was correct and still applicable to the current situation. However, the underestimation of the
lumber values was noticeable due to changes in the quality index and sawn wood volume throughout the years.
90
Therefore, a recalibration of the LMV model, through the modelling of the quality index and the sawn wood volume,
was necessary. This re-evaluation would provide predictions that would be not only more accurate, but also that would
be less sensitive to variations in internal wood properties As we intended to propose a practical way of estimating
lumber merchantable value, we tried to eliminate the use of proxy variables.
The proposed quality index was developed as a function of log grade and log small end diameter (Equation 3):
QI D LG
(3)
Estimates
-6.305757
3.874347
3.592827
1.077617
0.240565
Having the predictions for QI and sawn wood volume, we proceeded with the estimation of LMV using Equation 1.
We used an average market price (P) for quality 1C for the last five years. The comparison between the lumber
merchantable values predicted by the re-assessed model and the observed values can be seen in Figure 5.
Figure 5: Predicted and observed LMV for sugar maple and yellow birch.
After the recalibration, the LMV model gave better estimates of the lumber merchantable value of a log. The
predicted values were closer to the observed values, even though there was some bias for the most valuable logs.
4. CONCLUSIONS
The methodology for estimating lumber merchantable value of a log is not new; some of its concepts were created
over 50 years ago. Although some of the parameters have changed throughout the years, the structure of the model
91
developed by Petro and Calvert (1976) is still applicable. Taking this in consideration, a recalibration of the original
model was necessary to properly predict LMV.
The P&C model requires several variables which are not easily obtained. Some of these depend on predictions from
other sources (e.g. the sawn wood yield is derived from volume tables), which may not be up-to-date. Based on these
observations, two models were built to estimate quality index and sawn wood volume, using surrogate variables. The
LMV was obtained using the equation described in Petro and Calverts (1976) publication. The predictions using the
recalibrated model were much closer to the observed values, showing its efficiency in predicting LMV. The QI and
sawn wood yield predictive models are dynamic, and should perform well under the new silvicultural guidelines.
REFERENCES
Bdard, S. and F. Brassard. 2002. Actual effects of selection cuttings in public forests of Quebec in 1995 and 1996. (Les effets rels
des coupes de jardinage dans les forts publiques du Qubec en 1995 et 1996.) Ministre des Ressources Naturelles.
Qubec. [In French]
Bdard, S. and Z. Majcen. 2003. Growth following single-tree selection cutting in Quebec northern hardwoods. For. Chron. 79(5):
898-905.
Boulet, B. 2007. External defects and indices of tree decay: a field guide. (Dfauts et indices de la carie des arbres: Guide
d'interprtation.) Quebec. 2nd ed. Les Publications du Qubec. 317 pp. [In French]
Fortin, M., F. Guillemette and S. Bdard. 2009. Predicting volumes by log grades in standing sugar maple and yellow birch trees
in southern Quebec, Canada. Can. J. For. Res. 39(10): 1928-1938.
Hanks, L.F. 1976. Hardwood tree grades for factory lumber. USDA For. Serv., Res. Pap. NE-333. 81 pp.
HMR. Hardwood Market Report. Lumber Values for the years 2004 to 2012.
Majcen, Z. 1994. History of selection cutting in uneven-aged forests in Quebec. (Historique des coupes de jardinage dans les forts
inquiennes au Qubec.) Rev. For. Fran. 45(4): 375-384. [In French]
McCauley, O.D. and J.J. Mendel. 1969. Adjusting quality index log values to represent commercial sawlog product values. USDA
For. Serv., Res. Pap. NE-149. 29 pp.
Mendel, J.J. and W.H. Smith. 1970. Quality index tables for some Eastern hardwood species. USDA For. Serv., Res. Pap. NE-167.
25 pp.
NHLA. 2011. Rules for the measurement and inspection of hardwood and cypress. v 1.1. 101 pp.
Nyland, R.D. 1992. Exploitation and greed in eastern hardwood forests: will foresters get another chance? J. For. 90(1): 33-37.
Petro, F.J. and W.W. Calvert. 1976. How to grade hardwood logs for factory lumber.(La classification des billes de bois franc
destines u sciage.) Forestry Technical Report 6F. NRCAN. Can. For. Serv., Forintek Canada Corp. Ottawa, Ontario. 66
pp. [In French]
Pothier, D., M. Fortin, D. Auty, S. Delisle-Boulianne, L.-V. Gagn and A. Achim. 2013. Improving tree selection for partial cutting
through joint probability modelling of tree vigour and quality. Can. J. For. Res. 43(3): 288-298.
Richards, D.B., W.K. Adkins, H. Hallock and E.H. Bulgrin. 1979. Simulation of Hardwood Log Sawing. USDA For. Serv., Res.
Pap. FPL-355. 9 pp.
Steele, P.H. 1984. Factors determining lumber recovery in sawmilling. USDA For. Serv., Gen. Tech. Rep. FPL-39. 8 pp.
Vaughan, C.L., A.C. Wollin, K.A. McDonald and E.H.Bulgrin. 1966. Hardwood log grades for standard lumber. USDA For. Serv.,
Res. Pap. FPL 63. 54 pp.
92
1. INTRODUCTION
In the late 1960s, the Canadian wood industry started the improvement of poplar natural resource
productivity through sylviculture, conventional tree breeding, and biotechnology programs (Richardson 1998).
Although these programs have resulted in great advances (improved growth, stem form, adaptability, and
disease resistance), certain areas including wood utilization require further investigation.
Bending solid wood could be a good option to generate more value-added products using poplar hybrid
clones. Bent wood pieces allow creating curved objects that are generally stronger and stiffer than objects made
with pieces shaped by sawing. When wood is bent, the interior face (concave) is subjected to compressive
strains while the exterior face (convex) is exposed to tensile strains. Softening treatment greatly increases wood
compressibility, but in contrast has a little effect on tensile properties (Wangaard 1952). To restrict elongation of
the convex side a tension strap normally is used. Without restrain, tension and compression stresses are balanced
on opposite sides of a quasi-centrally located neutral axis. Moreover, the addition of a steel strap with endblocks will provoke a shift of the neutral axis of the fibres towards the convex side. As a result, the portion of
the wood in tension will be reduced and the one in compression increased (Stevens and Turner 1970; Hoadley
2000; Sandberg and Navi 2007). A successful bend is achieved when permanent wood deformation is obtained
without obvious wood failure (Rice and Lucas 2003).
During bending two kinds of deformation appear: plastic and viscoelastic deformations. Softening treatments
drive the wood to a viscoelastic state. After bending, the wood is still in the viscoelastic state. Thus, the relative
amount of elastic deformation will spring-back; while the relative amount of plastic deformation will stay in the
reached shape. To avoid spring-back, the bent workpiece has to be fixed to its position until the elastic
deformation changes to plastic deformation. This change is considered as stress relaxation (Eggert 1995).
There are two principal ways to bent wood: using heat, humidity and pressure, known as steam bending, or
using chemical products. In steam bending, wood pieces are placed in a steaming machine at about 100C, bent
and then kept in a bending plate until cooling is completed and the moisture content (MC) required for final use
is achieved. In general, steam bending is conducted mainly in hardwoods that have initial MC between 12% and
25%. The level of curvature and deformation followed by the pieces after bending is a function of the initial MC
and the setting conditions. Jorgensen (1965) suggested an initial MC between 12% and 20%, Stevens and
Turner (1970) 25% MC, and Kollmann and Ct (1975) between 17% and 25% MC.
The suggested steaming times also vary among studies. The most common estimates are 1 hour per 25.4 mm
of thickness for dry wood (12% MC or less) and 30 min per 25.4 mm of thickness for green wood (Wangaard
*Corresponding author: Tel: (418) 656-5852; Fax: (418) 656-2091; E-mail: Roger.Hernandez@sbf.ulaval.ca
93
1952; Peck 1957). Lemoine and Koch (1971) steamed at atmospheric pressure southern pine pieces of 12.7 mm
and 25.4 mm thick at 17% MC during only 10 and 20 minutes, respectively, with adequate results. On the other
hand, the maximum curvature that can be obtained when bending a given species for a particular thickness can
be roughly estimated using the ratio of radius to thickness, called the thickness ratio. According to Kubler
(1980), this ratio should be 30 for dry wood at ambient temperature, 13 with steamed wood, and 3 when using a
steel strap.
The present study evaluates the steam bending behavior of wood pieces coming from seven poplar hybrid
clones grown in three sites, after two wood drying treatments. Bending quality was assessed using quantitative
and qualitative methods. The first method evaluated the curvature recovery or most known as spring-back of the
wood pieces. The second method classified visually the wood pieces by type and degree of ruptures. The results
would provide knowledge on steam bending performance of wood from poplar hybrid clones for the
development of value-added wood products in Canada.
2. MATERIALS AND METHODS
Material for this study came from the experimental sites of Platon, St-Ours, and Windsor, which were
established by the Ministry of Natural Resources and Wildlife of Quebec. The sites, originally abandoned
agricultural lands, represent typical soil types available for planting hybrid poplar clones in southern Quebec.
Detailed description of the experimental sites is given elsewhere (Pliura et al 2007).
Seven clones, each represented by five trees from each site, were selected for this study (Table 1). These
clones are mainly recommended for southern Quebec, based on growth rate, bolt form, adaptability, and disease
resistance (Valle 1995). From July to early September 2007, 105 trees were harvested. Trees from St-Ours and
Windsor were cut after 15 growing seasons. Trees from Platon were cut after 17 growing seasons, except for
those of clone 915508, which were felled after 13 growing seasons. Each tree was cross-cut in three segments: a
butt log of 1.50 m followed by two logs of 2.45 m long. Specimens for bending tests were all obtained from the
2.45-m log farthest to the butt. Boards of 34 mm thick were sawn with a WoodMizer (Indianapolis, IN) portable
bandsaw and stored in a freezer at approximately -4C before drying.
Table 1: Selected clones for the study
Clone
Hybrid
915508
DN x M
3230
TxD
319.6
3586
DxN
3570
DxN
131
DxN
4813
DxN
3565
DxN
323.3
325.6
331.0
331.4
344.0
349.5
To study the effect of two drying schedules on wood bending, boards that were kiln dried in a previous study
(Hernndez et al 2011) were used. Boards were dried to a target moisture content of 8% in a 2.5 m3
experimental kiln using two treatments: a conventional schedule (dry bulb temperature [Tdb] between 60 and
82C), and an elevated-temperature schedule (Tdb between 60 and 90C) with top restraint loading (7.5 kN/m2)
in each case. After drying, boards were evaluated for drying quality, planed to 25 mm thick and stored in a
conditioning room at 20C and 40% RH to reach nominal equilibrium moisture content of 8%.
Some of the material was not suitable to bending tests due to drying defects and presence of knots in the
boards. Consequently, 130 boards were selected and just one parallel to the grain oriented sample of 15(R) x
30(T) x 560(L) mm could be obtained from each board. A block of 15 x 30 x 50 mm was then cut from each
sample to obtain the basic density (oven-dry weight by green volume ratio) (Table 1). The final dimensions of
steam bending samples were 15(R) x 30(T) x 500.5(L) mm. The ends of each sample were sealed with two coats
of varnish to prevent excessive absorption of moisture by the ends and minimize end checking and splitting
(Peck 1957). Twenty samples of red oak (Quercus rubra L.), a wood species commonly used for making bent
members (Peck 1957; Stevens and Turner 1970), were also prepared to be used as reference.
94
The experiments began with the measurement of the initial weight of specimens. Steaming was then
performed with a steaming cylinder GHEbavaria during 40 minutes (based on preliminary tests) at a pressure of
0.2 bar and temperature of about 105C. After steaming, each sample was immediately weighted and inserted
into a steel strap with fixed end-stops with the growth rings facing up. The assembly was then fixed in the
bending machine GHEbavaria, model Stuzama II, by two clamps, one at each end. Bending was performed
using a circular-shape mould of 200 mm of radius and with an arm pressure of 4 bar. The steel strap containing
the bent sample was then cramped with a metal clamp. Finally, this assembly was placed in a conditioning room
at 20C and 40% RH for 8 days (setting time) for MC stabilization.
2.2 QUALITATIVE ASSESSMENT
To make a successful bend, the wood must be deformed without provoking visible failures either in tension
or in compression. Thus, all bent wood samples were evaluated according to five quality grades. The grades
were established according to the degree of failure in the tangential faces. Samples without signs of failure were
classified as grade 1 (excellent), those with superficial compression failures no longer than 3 mm were in grade
2 (good), samples with compression failures that did not go through across the width of the sample were in
grade 3 (acceptable), those with compression ruptures that go through across the width of the sample were in
grade 4 (poor), and finally samples with rupture in tension were considered as grade 5 (very poor).
2.3 QUANTITATIVE ASSESSMENT
SPRING-BACK
After the initial setting time (8 days), the chord (C) of the bent sample was measured before (C0) and
immediately after removal from the cramped plate (C8). The nominal chord of the circular-shape mould was
374.5 mm. Chords of un-cramped samples were then successively measured over time at days 9, 11, 18, 30, and
60 (C9 to C60). Therefore, the memory recovery was evaluated according to two parameters: instantaneous
spring-back and spring-back over time. The instantaneous spring-back was the difference between C8 and C0
(both measured at day 8). The spring-back over time was the difference between the chord measured over time
(C9, C11C60) and C8. The sample weight was also taken at days 9 to 60 to assess the variation in MC. The
samples remained permanently in the conditioning room.
2.4 STATISTICAL ANALYSIS
A binary analysis was applied on the qualitative assessment to find the probability to get wood pieces of
good bending quality using a glimmix procedure. Also, the influence of the sources of variation (clone, site and
drying time) on this probability was evaluated. Results of instantaneous spring-back were analyzed with a mixed
procedure in a split-plot design. The clones and sites were the sources of variation as principal plots and the
drying treatment was the source of variation as sub-plot. The spring-back over time was analyzed in a split-plotplot design considering the storage time as a source of variation as sub-sub-plot. Given the changes in springback over time, this data was analyzed as repeated measures design. Moreover, the MC was added as a covariable in both spring-back analyses. When a source of variation was statistically significant, multiple
comparison tests with Bonferonni corrections were carried out to determine which factor level differed from the
others. These comparisons were done at the comparison-wise error rate of 0.05/c, where c is the total number of
independent comparisons. The normality of data was verified using Shapiro-Wilk test. Statistical analyses were
performed using SAS software, version 9.3 (SAS Institute 2010).
3. RESULTS AND DISCUSSION
3.1 QUALITATIVE ASSESSMENT
The qualitative analysis for poplar hybrid clones showed that defect-free bent samples (grade 1) were absent.
Grade 2 represented 13%, grade 3 39%, grade 4 36%, and grade 5 12% of the total samples. Hence, 88% of the
poplar hybrids samples (grades 2 to 4) showed failures by compression and 12% (grade 5) presented failures by
tension. Failures occur if the steamed wood is stressed beyond its tensile or compressive strength limit.
Compressive failures are principally due to high stress, defects or zones of weakness. Tensile failures occur
when the strap and end blocks are not applying sufficient pressure to keep the tension of the convex side below
the limit (1 to 2% of wood piece length, according to Peck 1957). In the present work, pieces of poplar hybrids
presented compression failures in form of a crosswise fold or wrinkles, sometimes extending across the width on
the concave side. Some pieces also presented crushing at the ends and tensile failures in the convex side.
The binary analysis was applied to find the probability to get wood pieces of good bending quality. Thus, the
four grades were grouped into two classes. Class 0 corresponded to the grades 3, 4, and 5, while class 1
95
represented grade 2. Pieces of grade 2 were absent in Platon and therefore this site was eliminated from the
analysis. Results showed that the probability to get wood pieces of good bending quality was not affected by the
sources of variation (site, clone, and drying treatment). However, the statistical power of this analysis was low
due to the number of pieces removed.
Consequently, to classify the seven clones according to their performance on steaming bending, a different
approach, based on Ozarska and Daian (2010), was applied. First, the frequency distribution of the visual
ranking was expressed as percentage of samples in different ranking groups (Table 2). Then, the cumulative
frequency distribution was used to establish the lowest level at which the bends pieces of a clone could fail at a
95% confidence level (Table 3). According to this table, only the clone 3570 bent satisfactorily. This means that
at least 95% of the samples passed the criteria described for this level. The rest of the clones were rated as wood
having bad bending quality.
Table 2: Frequency distribution by visual ranking
Visual ranking (%)
1
Number of
samples
3570
19
43
33
21
3586
20
45
25
10
20
3230
10
35
45
10
20
3565
18
41
29
12
17
131
27
53
13
15
915508
27
55
18
22
4813
20
53
20
15
Total (%)
13
39
36
12
130
Clone
decent bend
just satisfactory
bend
bad
bending quality
1+2
1+2+3
1+2+3+4
1+2+3+4+5
3570
19
62
95
100
3586
20
65
90
100
3230
10
45
90
100
3565
18
59
88
100
131
34
87
100
915508
27
82
100
4813
20
73
80
100
Clone
96
effect. Hence, this could indicate that either a higher initial MC or a longer steaming should improve the
bending performance of wood pieces. On the other hand, it could also indicate that the remaining rigidity of
wood increases as final MC decreases.
The instantaneous spring-back is a natural reaction of wood to the release of compressive stress imposed
during the bending operation. To limit this tendency, the piece should be held in its clamp until it has cooled and
dried. The results show that the setting time for poplar clones was not adequate. The pieces should have
remained cramped for a longer period in order to reach the nominal equilibrium moisture content (EMC) and
reduce the instantaneous spring-back effect. The MC of pieces was in average 9.6% at day 8 and continued to
decrease afterwards. This confirms the importance of the total stabilization of MC after bending (during the
setting time), in order to limit the instantaneous spring-back.
Table 4: Means of instantaneous spring-back for seven poplar hybrid clones, compared to red oak.
Clone
3570
8.6(0.8)1A2
4813
8.9 (1.2) A
3230
9.9 (0.8) A
3586
10.4 (0.8) AB
131
11.6 (1.1) AB
3565
11.7 (0.9) AB
915508
12.9 (0.7) B
Red oak
7.1
97
23
21
19
Spring-back (mm)
17
Platon
15
Windsor
13
11
St. Ours
Red oak
7
5
10
20
30
40
50
60
70
9,7
9,6
9,5
9,4
9,3
9,2
9,1
9,0
8,9
8,8
10
20
30
40
50
60
70
Figure 2: Progression of the wood moisture content of poplar clones during the storage time at 20C and 40% relative
humidity.
98
Red oak samples showed an instantaneous spring-back of 7.1 mm and a spring-back over time (from day 8
to day 60) of 3.9 mm, for a total of 11 mm (Fig 1). Poplar hybrid samples showed a total spring-back of 21.5
mm, that is an instantaneous spring-back of 10.6 mm and a spring-back over time of 10.9 mm (Fig 1). The
memory effect was hence two times greater in poplar samples than in red oak samples (21.5 mm versus 11 mm
of total spring-back). Again, performance respect to spring-back of studied poplars could be improved if only
the best clones are considered (3570 and 4813). In this case their total spring back was of 19.7 mm, but still
higher than red oak values.
Red oak probably performed better than the poplar clones due to its higher wood density. Davis (1942)
indicated that heavy woods (those with a basic density of 500 kg/m3or over) bent better than light woods (those
with a basic density of less than 400 kg/m3). Furthermore, when growth of poplar trees is accelerated, the
presence of tension wood is considerably increased in juvenile wood. Tension wood is difficult to machine
smoothly, is commonly lower in mechanical properties, and is also more prone to collapse (Bendtsen 1978).
Furthermore, the initial moisture content of 8% used in this experiment could have been too low for steam
bending operations. It is known that plasticity of wood increases with moisture content (Sandberg and Navi
2007). MCs between 12% and 25% have been normally suggested by several authors. This implicates that more
research is needed in order to obtain optimum steaming, drying, and setting conditions for steam bending.
4. CONCLUSIONS AND RECOMMENDATIONS
The wood of the poplar hybrid clones studied had poor steam bending properties. They showed a large
memory effect and a poor surface quality after bending. Among the factors studied, only the clones showed a
significant effect on bending performance. Also, the moisture content variation of the bent pieces had a
significant influence on the spring-back. The sites and kiln drying schedules did not affect differently the steam
bending properties of specimens. Results also showed that the clone (P. x euramericana x P. maximowiczii)
915508 had the worst performance. In contrast, the Euramericana clones (P. deltoides x P. nigra) 3570 and 4813
had the best response to steam bending. Further studies on these clones could be conducted in order to optimize
the parameters of steam bending.
REFERENCES
Bendtsen B.A. 1978. Properties of wood from improved and intensively managed trees. Forest Prod J 28(10):61-72.
Davis E.M . 1942. Machining and related characteristics of southern hardwoods. Technical Bulletin 824. USDA Forest
Service, U.S. Gov. Washington DC. 42 pp.
Eggert O.T. 1995. Stress relaxation in solid wood bending. In: Proc of the 12th International Wood Machining Seminar.
Tokyo, Japan, pp. 203-214.
Jalaludin Z., H.W. Samsi, and P. Elham. 2007. Properties of bent rubberwood in different storing conditions. Born Sci (J Sci
Tech) 20(1):56-64.
Jorgensen R.N. 1965. Furniture wood bending: An inquiry into the theory and practice of bending wood. Furniture Des
Manuf 37(12):60-61.
Hernndez R.E., S. Constantineau, and Y. Fortin. 2011. Wood machining properties of poplar hybrid clones from different
sites following various treatments. Wood Fiber Sci 43(4):394-411.
Hoadley, R.B. 2000. Understanding wood. A craftmans guide to wood technology. The Taunton Press, Newtown, 280 pp.
Kollmann F.F.P. and W.A Ct Jr. 1975. Principles of wood science and technology. Solid Wood. Heidelberg-Berlin,
Springer Verlag-New York.
Kubler H. 1980. Wood as building and hobby material. John Wiley & Sons Inc. New York. 256 pp.
Lemoine T.J., and P. Koch. 1971. Steam-bending properties of southern pine. Forest Prod J 21(4):34-42.
Murakami R., F. Tanaka, and M. Norimoto. 2002. Relationship between bending quality and wood species. Wood Res 89:2122.
Ozarska B., and G. Daian. 2010. Assessment of microwave bending capabilities for Australian wood species. Forest Prod J
60(1):64-68.
Peck E.C. 1957. Bending solid wood to form. Agric Handb 125 USDA Forest Service, U.S. Gov. Printing Office. Washington
DC, 37 pp.
Pliura A. S.Y. Zhang, J. Mackay, and J. Bousquet. 2007. Genotypic variation in wood density and growth traits of poplar
hybrids at four clonal trials. Forest Ecol Manag 238:92-106.
Rice R.W., and J. Lucas. 2003. The effect of moisture content and bending rate on the work required to bend solid red oak.
Forest Prod J 53(2):71-77
Sandberg D, P. Navi. 2007. Introduction to thermo-hydro-mechanical (THM) wood processing. School of Technology
Design. Reports, No. 30. Vxj University, Vxj, Sweden, 167 pp.
99
SAS Institute. 2010. SAS/Stat Users Guide, release 9.3 ed. SAS Institute Inc, Cary, NC.
Stevens W.C., and N. Turner. 1970. Wood bending handbook. Woodcraft & Supply Corp., Parkersburg, WV, 109 pp.
Richardson J. 1998. Industrial potential of poplars in Canada. In: Proc of Annual Meeting of the Poplar Council of Canada.
September 21-24, 1998. Quebec, Canada, 94 pp.
Valle G. 1995. Projet du MRNQ sur lamlioration gntique des peupliers dans la rgion du Saguenay Lac Saint-Jean.
In Proc Annual Meeting Conseil du peuplier du Canada, Chicoutimi, Quebec, 181 pp.
Wangaard FF (1952) The steam bending of beech. J. of Forest Prod Res Soc 2:35-41.
100
Research Leader Secondary Manufacturing, FPInnovations, 319, rue Franquet, Qubec, QC, G1P 4R4,
Canada.
Abstract
Viscoelastic and mechano-sorptive creep is a time-dependent characteristic of wood that plays an important role in the
mechanical integrity of structural wood products. Measuring this property is always a time-consuming and challenging
operation. To mitigate this difficulty, a Dynamic Mechanical Analyzer (DMA) was used to obtain a quick creep
measurement. Testing was conducted on sugar maple (Acer saccharum) wood specimens in the radial direction under
different relative humidity (RH) conditions. Viscoelastic creep measurements (creep under constant RH) showed that
creep depended on time, RH and load level. Specimens exposed to high RH experienced high total and plastic
deflections. Under the applied stress levels (5, 25, 35 and 45% of modulus of rupture (MOR)); wood exhibited linear
viscoelastic behavior. As RH increased, its effects on creep intensified. Evidence of a mechano-sorptive effect was
provided by the fact that RH change during the loading period resulted in very high deflections. The findings of this
research were in complete accordance with the creep measurements reported by other authors using traditional
methods.
Introduction
Deflection is a key parameter in timber design. It results from two factors, elastic deflection and creep.
Elastic deflection is often determined with a stress-grading machine.
Creep, the second contributing factor, increases deflection over time under a sustained load. Because wood
is a viscoelastic material under normal stress, temperature and moisture content (MC) conditions (Van Der
Put 1989), it is susceptible to creep. Creep of wood members may produce largely irrecoverable
deformation, and result in serviceability and strength reduction problems over time. And the sustained load
causing failure does not have to be larger than the ultimate static load (Schaffer 1972). Creep is the
continued deformation under a constant stress while stress relaxation is the reduction in stress with time
under a prescribed strain.
Creep research is critical wherever the prediction of total deflection is necessary for durable structural
design in timber structures. Creep occurs at even very low stresses and continues over a period of years.
Studies on the creep behavior of wood have been conducted for decades (Clouser 1959; Kingston and
Armstrong 1951). Most early studies focused on small, clear wood specimens subjected to constant
environmental conditions.
Developments in timber engineering have created a need for increased research activities aimed at a better
understanding of the mechanical behavior of wood subjected to changing environmental conditions. It has
been shown in the literature (Bodig and Jayne 1982; Gerhards 1985) that the creep behavior of wood is
affected by the level of applied loads, and the ambient environment. Researchers have found that creep
increases with time and loads (Clouser 1959; Kingston and Armstrong 1951; Bodig and Jayne 1982;
Gerhards 1985, 1988, 1991; Pentoney and Davidson 1962; Szabo and Ifju 1970).
Under constant moisture conditions, the non-linearity of creep response depends mainly on stress level. It is
also related to MC changes and to the sorption history undergone by the material. The strain increment
observed in humidity cycling could result, not from a change in the elastic modulus, but from continuous
breaking and reforming of hydrogen bonds through the movement of water (Navi et al. 2002).
101
High MC and large MC changes have a significant effect on creep. An EMC greater than 16% over a long
term may result in more than the expected 50% increase in creep deflection. The amount of additional
deflection seems to depend on the degree of moisture change. Members in the roof of an unheated shed will
experience greater seasonal creep than those in a climate-controlled building.
It is known that wood exposed to MC changes exhibits much greater deformation than wood under constant
climate conditions. The simultaneous action of load on the element and changing MC is called the
mechano-sorptive effect (MSE). The effects of MSE are quite different from those observed under constant
RH conditions: MSE is not directly dependent on time, but on MC changes; creep increases regardless of
the sign of MC change (sorption, desorption). MSE deformation is not recovered when the load is removed.
The most evident manifestations of MSE in structural applications are increased deflection of beams and
transverse deformation in support bearing zones of wooden beams exposed to climate variation.
Independent reports on MSE can be found in the literature (Armstrong and Kingston 1960; Armstrong and
Christensen 1961) in the late 1950s and early 1960s.
In this study, a Dynamic Mechanical Analyzer (DMA) was used to measure the creep of wood under
constant and variable MCs. The DMA presents several advantages over traditional creep measurement
methods. Wood is a biological material with large variability. To minimize the effects of natural defects of
wood such as knots, irregular grain and pitch pockets, small specimens are preferred. In DMA tests,
specimen size is reduced to a minimum, so that the effects of variability on creep may be minimized. As
measurements are conducted in a small chamber, controlling and monitoring temperature and humidity is
much easier than with traditional methods. Moreover, the time needed to conduct the test is shorter, and
load and deflection can be precisely controlled and recorded. Accuracy is very important because
distinguishing differences between total creep and its three separate components is essential to
understanding of creep behavior in wood.
The main objective of this study was to develop a new method based on DMA to measure creep in wood
using sugar maple, and to compare the results to existing data available in the literature. Successful
development could provide engineers with a rapid and reliable tool to measure wood creep. In this study
creep in the radial direction was measured under constant temperature and humidity conditions
(viscoelastic creep) and also under constant temperature and variable humidity conditions (mechanosorptive creep).
Materials and Methods
Sugar maple (Acer saccharum) wood blocks were visually selected for freedom of any defects, i.e. perfect
growth ring orientation and no grain deviation. They were dried in a dryer until they reached constant mass
at 54.4C and 70.9% RH. They were subsequently conditioned at 20C and 65% RH. After they had
reached EMC, the test specimens were cut to final sizes, i.e.: 40 mm in length x 5 mm in width x 1 mm in
thickness. Prior to creep measurement, the specimens were conditioned for one week to the RH and
temperature of the creep tests (30C and 37.0, 67.0 and 83.9% RH, corresponding to 7, 12 and 17%
EMC).Three specimens were run for each set of combinations.
Creep measurement
Before creep measurement, three-point static bending tests were carried out to determine modulus of
rupture (MOR) values, which were then used to determine stress levels in the creep measurements. As for
the creep tests, MORs were measured at three EMC levels, i.e.: 7, 12 and 17%.
Creep measurement was conducted in three-point bending mode with a DMA instrument (TA Instruments
Q800 DMA), capable of controlling humidity, temperature and load.
Viscoelastic creep
Viscoelastic creep measurements were also carried out at three constant RH levels (37, 67 and 83.9%) at a
temperature of 30C. The stress applied during testing was variable and a ratio of MOR. The stress levels
were 5, 25, 35 and 45% of MOR at the corresponding EMCs. Both loading and recovery times were 60
minutes. Instantaneous deflection (i), total deflection (t), elastic deflection (e), viscoelastic deflection
(v) and plastic deflection (p) were determined for all specimens. Instantaneous deflection (i) was
determined at the time of loading. Total deflection (t) was determined before unloading. Elastic deflection
(e) was measured at the time of unloading. Viscoelastic deflection (v) and plastic deflection (p) were
determined at the end of the tests. Plastic deflection (p) is the permanent deflection of wood. Viscoelastic
deflection (v) is the difference between calculated viscoelastic deflection (v) and plastic deflection (p).
Three tests were run for each set of stress and RH level.
Mechano-sorptive creep
102
Mechano-sorptive creep measurements were designed in five phases, including a wet-dry cycle (Figure 1).
In each phase, humidity changes were designed to simulate the effects on wood creep under service
conditions. During all phases, the temperature remained at 30C. In Phase 1, the specimens were
conditioned at 67% RH for 2 h. In Phase 2, the RH was increased from 67 to 83.9%, raising EMC from 12
to 17%. The specimens were kept at this humidity for 5 h. In Phase 3, RH was reduced from 83.9 to 37%,
causing the wood to dry down to 7% EMC. The specimens were then conditioned at 37% RH for 5h. In
Phase 4, the RH was raised from 37 to 67%, increasing wood EMC from 7 to 12%. The specimens were
kept at this RH for 2h. The stress levels in the mechano-sorptive creep test were set at 5, 10 and 15% of
MOR.
Statistical analyses
A one-way analysis of variance model was used to study the effect of RH on the creep of wood at different
RH levels. The GLM procedure of the SAS software was used and pairwise comparisons were then made
using protected Fisher LSD (Least Significant Difference).
Results
Modulus of rupture (MOR) is an accepted criterion of strength, although it is not a true stress because the
formula by which it is calculated is valid only within the elastic range. MOR reflects the maximum loadcarrying capacity in bending and is proportional to the maximum moment borne by the specimen. The
MORs of the wood specimens in the radial direction at the various RH levels are given in Figure 2. An
increase in RH clearly lowered MOR values. Raising RH from 37% to 83.9% reduced MOR by more than
30%. Drow (1945), James (1964), Leontev (1960), Sulzberger (1953) and Wilson (1932) also reported that
the bending strength of wood decreased as RH increased.
Creep
Viscoelastic creep
The deflection values observed at different stages, loading and recovery periods in the creep tests
conducted at 67% RH and the various stress levels (5, 25, 35 and 45%) are summarized in Table 1.
Instantaneous deflection progressively increased as applied stress was raised. The increase was more
pronounced as stress levels increased from 5 to 25%. The absolute values of elastic deflection, viscoelastic
deflection and plastic deflection also increased with increasing applied stress. But when the proportion of
elastic deflection in total deflection was used as a criterion for comparison purposes, it was found to
decrease with increasing applied stress. For example, although an increase in stress level from 5% to 45%
caused the absolute values of elastic deflection and total deflection to increase (from 0.27 to 1.57 and from
0.32 to 2.37, respectively), the proportion of elastic deflection in total deflection decreased (from 0.84 to
0.66). An opposite trend was observed for plastic deflection. Both the absolute values of plastic deflection
and their proportion in total deflection increased with increasing applied stress. An increase in applied
stress from 5 to 25% led to a very significant increase in the proportion of plastic deflection but the
proportion decreased with further increases in stress level (from 25 to 35 and 45%). On the whole, as
103
applied stress increased, plastic deflection played a more pronounced role in total deflection. Explaining
changes in the proportion of viscoelastic deflection in total deflection is somewhat complex and beyond the
scope of this work.
80
70
MOR [MPa]
60
50
40
30
20
10
0
7%EMC
12% EMC
17% EMC
It has been reported that creep increases with time and applied load (Clouser 1959; Kingston and
Armstrong 1951; Bodig and Jayne 1982; Gerhards 1985, 1988, 1991; Pentoney and Davidson 1962; Szabo
and Ifju 1970). Gerhards (1985) measured deflections in Douglas-fir 2 by 4 beams at three different load
levels for up to 220 days. He found that the ratio of elastic deflection to total deflection decreased with time
and increasing stress level.
Table1: Total deflection values and their three components at 7%, 12% and 17% equilibrium moisture contents
(EMCs)*.
Applied stress
Total deflection t Elastic deflection e
Viscoelastic deflection
Plastic deflection
(% of MOR)
(mm)
(mm)
v (mm)
p (mm)
7%EMC
0.26
0.21
0.04
0.01
5
(10.17)
(17.16)
(22.91)
(20.00)
1.46
1.28
0.12
0.06
25
(11.04)
(8.99)
(16.39)
(42.03)
1.76
1.4
0.19
0.17
35
(13.60)
(21.90)
(13.03)
(67.11)
2.07
1.45
0.36
0.26
45
(7.59)
(14.98)
(7.87)
(34.16)
12%EMC
0.32
0.27
0.03
0.02
5
(15.87)
(11.18)
(41.69)
(39.28)
1.63
1.28
0.25
0.1
25
(14.71)
(11.09)
(28.28)
(27.03)
1.80
1.37
0.29
0.14
35
(9.95)
(10.42)
(7.52)
(13.75)
2.37
1.57
0.43
0.37
45
(11.84)
(4.92)
(24.52)
(26.76)
17%EMC
0.34
0.25
0.07
0.02
5
(18.08)
(19.39)
(23.28)
(52.14)
1.80
1.31
0.38
0.11
25
(10.00)
(11.88)
(15.91)
(27.06)
2.40
1.42
0.54
0.44
35
(7.73)
(7.09)
(8.50)
(12.94)
3.30
1.83
0.69
0.78
45
(6.18)
(9.23)
(3.23)
(7.39)
*Coefficients of variation (COVs) are shown in brackets (%)
104
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
Figure 3 provides a better representation of the effects of stress level on total deflection over time. The
linear proportionality of applied stress and total deflection confirmed the linear viscoelastic behavior of
sugar maple wood in a bending test for stress levels below 45%.These results were fully consistent with
those of Segovia et al. (2013), who reported that, in creep tests at stress levels below 45% of bending
strength, the viscoelastic behavior of sugar maple wood was linear.
Several researchers have studied the linear viscoelastic behavior of wood in creep tests (Drow 1945;
Moutee et al. 2010; Schniewind 1968; Nielsen 1984; Schaffer 1972; Le Govic et al. 1989; Keith 1972; Navi
and Heger 2005). Moutee et al. (2010), Segovia et al. (2013), Nielsen (1984), Schniewind (1968) and Keith
(1972) found that wood behaves in a linear viscoelastic manner for loads below 45% of bending strength.
Schaffer (1972) described the behavior of wood as nonlinear regardless of load level but stated that, at low
stress, linearity could be a good approximation. Le Govic et al. (1989) set a limit for linear behavior at 3035% of the short-term strength for bending and up to 50% of the short-term strength for tension.
The results of creep tests conducted at 37% RH at the four stress levels are given in Table 1. The response
of wood to the loads was very similar to that at 67% RH, although total deflections were slightly lower than
at 67% RH. Deflection increased with time. Total, plastic and viscoelastic deflection values at a given
stress level were lower for 37% RH than for 67% RH.
Figure 3: Relationship between stress and total deflection during loading period at 12% EMC.
Table 1 shows the results of creep tests conducted at 83.9% RH and 30C. Total, elastic, viscoelastic and
plastic deflection values were significantly higher than at 37% and 67% RH. The proportions of plastic
deflection in total deflection were much higher at this RH as well. However, the elastic and viscoelastic
deflections showed an opposite trend, representing a smaller portion of the total deflection. It is well known
that water is a plasticizing agent for wood. The replacement of hydrogen bonding within the amorphous
components by water-carbohydrates links enhances the flexibility of the polymer network. As a result,
wood with high MC is less rigid than dry wood (Kelley et al. 1987; Obataya et al. 1988). Increases in creep
due to moisture could be explained by the fact that water in the cell walls increases creep by lubricating the
slip interface. Creep is accelerated while water is moving into or out of the wood (Bach 1965; Gnanaharan,
Haygreen 1979; Ellwood 1954).
Mechano-sorptive creep
The deflection values at the different stages of the mechano-sorptive creep measurements are shown in
Table 2. At the 5% stress level, the total deflection at the final stage was more than three times the level
observed at the original 12% EMC. Such a huge increase in creep is indicative of the mechano-sorptive
effect (MSE). As stress increased, the impact of moisture change on creep became more pronounced. The
highest deflection was detected at the 15% stress level. Other researchers have reported that, when the MC
of the loaded wood increased, deflection increased rapidly, but final deformation depended on moisture
level, the duration of the process having little effect. Whereas viscoelastic creep depends on the duration of
loading, MSE at a constant stress level is not time-dependent (Armstrong and Kingston 1962; Schniewind
1968). Several reasons have been mentioned to explain MSE behavior. MSE has been attributed to
increased molecular mobility during moisture transport (Bazant 1985; Ericksson and Noren1965;
Gibson1965); increased molecular mobility during moisture change (Van Der Put 1989; Bethe 1969; Ranta
105
Maunus 1975; Hunt1986; Hanhijarvi 1995); sorption-induced physical aging (Wang et al. 1992; Padanyi
1991); material-specific inter-fiber mechanisms (Soremark and
Fellers 1993; Haslach 1994);
thermodynamic ratcheting actions (Haslach 2000) and sorption-induced stress gradients (Caulfield 1994;
Pickett 1942; Selway 1992; Habeger, Coffin 2000).
The creep of wood was found to depend on load level and time. Relative humidity is the most influential
factor shaping the creep pattern. The impact of humidity becomes crucial when a loaded wood member is
exposed to a concurrent RH change. Changes in humidity during loading increased creep tremendously.
This phenomenon, so called mechano-sorptive creep, is very important in load bearing members as
mechano-sorptive creep is not much dependent on time. Hence, humidity change could impose a
substantial creep on wood members in short time. Such a creep can endanger the integrity of wood
structure if it is not taken into account during designing the structure. The results of this study are in
complete accordance with the findings of Segovia et al. (2013) working on measuring viscoelastic and
mechano-sorptive creep of sugar maple by a conventional method.
Table2: Total deflection values at the different steps of the mechano-sorptive creep measurements*.
Applied load
t (mm)
t (mm)
t (mm)
t (mm)
(%)
12% EMC
12% to 17% EMC
17% to 7% EMC
7% to 12% EMC
0.37
0.98
(14.48)
(26.29)
0.79
1.7
10
(13.70)
(18.04)
1.1
2.8
15
(12.46)
(11.26)
* Coefficients of variation (COVs) are shown in brackets (%)
5
1.1
(17.66)
1.9
(9.43)
3.41
(10.46)
1.14
(16.09)
2.31
(12.72)
3.61
(8.30)
Conclusions
The results of this study showed that a Dynamic Mechanical Analyzer (DMA) can be used to measure the
viscoelastic and mechano-sorptive creep of sugar maple solid wood. Moisture was found to control the
behavior of wood under load. In viscoelastic creep, moisture affected total deflection and its three
components (elastic, viscoelastic and plastic creep) as well as their relative proportions. Plastic deflection
made up a higher proportion of total deflection as moisture increased. Stress level was another factor
controlling wood creep. Total and plastic deflections increased with increasing stress but elastic deflection
played a minor role at high stress levels. In mechano-sorptive creep measurements, a major increase in
deflection was observed even at a low stress level. As mechano-sorptive creep is not controlled by time,
changing moisture in load-bearing members can cause a significant deflection in a very short time. The
creep patterns and their fluctuations detected by DMA had been recognized by conventional methods and
the results obtained in this study are fully consistent with the literature, which confirms the validity of the
method under evaluation.
Literature Cited
1.
2.
3.
4.
5.
6.
7.
8.
106
Armstrong LD and Kingston RST 1960. Effect of moisture changes on creep in wood. Nature
185(4718):862863.
Armstrong LD and Christensen G N .1961. Influence of moisture changes on deformation of wood under
stress. Nature, 191(4791):869870.
Armstrong LD and Kingston RST.1962. The effect of moisture content changes on the deformation of wood
under stress. Aust J App Sci 13(4):257276.
Bach L.1965.Non-linear mechanical behavior of wood in longitudinal tension. Ph.D. Dissertation, State
Univ. College of Forestry at Syracuse University, NY, USA, 228 pp.
Bazant ZP .1985. Constitutive equation of wood at variable humidity and temperature. Wood SciTechnol
19:159-177.
Bethe E .1969. Strength properties of construction wood stored under changing climates and mechanical
load. Holz Roz-Werkstoff, 27(8):291 303.
Bodig J and Jayne B.1982.Mechanics of wood and wood composites. Van Nostrand Reinhold Company.
New York, NY, USA.
Caulfield DF.1994. Ester cross-linking to improve wet performance of paper using multifunctional
carboxylic-acids, butanetetracarboxylic acid and citric-acid. Tappi Journal 77(3):205215.
Clouser WS.1959. Creep in small wood beams under constant bending load. Report No.2150, U.S. Dept. of
Agric. Forestry Service, Forest Products Lab., Madison, WI, USA.
Drow JT.1945. Effect of moisture content on the compressive, bending, and shear strengths, and on the
toughness of plywood. U.S. For. Serv. Rep. No. 1519. U.S. Dep. Agric., For. Serv., For. Prod. Lab.,
Madison, WI, USA.
Ellwood EL .1954. Properties of American beech in tension and compression perpendicular to grain and
their relation to drying. Yale Univ. School of Forestry Bull. No. 61. New Haven: Yale University, USA 82
pp.
Ericksson L and Noren B.1965.The effect of moisture changes on the deformation of wood with tension in
fibre direction. Holz Roz-Werkstoff 23(5): 201209.
Gerhards CC .1985. Time-dependent bending deflections of Douglas-fir 2 by 4's. For Prod J 35(4):18-26.
Gerhards CC .1988. Effect of grade on load duration of Douglas-fir lumber in bending. Wood and Fiber Sci
20 (1):146-161.
Gerhards CC .1991. Bending creep and load duration of Douglas-fir 2 by 4s under constant load. Wood and
Fiber Sci 23(3):384-409.
Gibson E .1965. Creep of Wood: Role of Water and Effect of a Changing Moisture Content. J Nature
206:213215.
Gnanaharan R and Haygreen J .1979. Comparison of the behavior of a basswood waferboard to that of
solid wood. Wood and Fiber 11(3):155-170.
Habeger CC and Coffin DW .2000. The role of stress concentrations in accelerated creep and sorptioninduced physical aging. J Pulp Pap Sci 26:145157.
Hanhijarvi A .1995. Deformation kinetics based rheological model for the time-dependent and moisture
induced deformation of wood. Wood Sci Tech 29: 191195.
Haslach HW.1994. The mechanics of moisture acetated tensile creep in paper. Tappi 77(10):179186.
Haslach HW.2000. The moisture and rate-dependent mechanical properties of paper: A review. Mech TimeDependent Mater 4:169210.
Hunt DG .1986. The mechano-sorptive creep susceptibility of two softwoods and its relation to some other
materials properties. J Mater Sci 21: 20882096.
James WL.1964. Vibration and static strength and elastic properties of clear Douglas-fir at various levels of
moisture content. For Prod J 14(9):409-413.
Keith CT.1972.The mechanical behavior of wood in longitudinal compression. Wood Sci 4:234-244.
Kelley S, Rials T, Glasser W.1987. Relaxation behavior of the amorphous components of wood. J Mater Sci
22(6): 617-624.
Kingston RST and Armstrong LD.1951.Creep in initially green wooden beams. Austr J of Appl Sci 2(2):306325.
Leontev NL .1960.The strength of wood at varying moisture content. Derev Prom 10:17-18.
Le Govic C, Felix B, Benzaim A , Rouger F.1989.Creep Behavior of Wood as a Function of Temperature:
Experimental Study, Modelling and Consequences for Design Codes,. Proceedings of the Second Pacific
Timber Engineering Conference, University of Auckland, New Zealand, 28-31 August, 273-277.
Moutee M, Fortin Y, Laghdir A, Fafard M. 2010.Cantilever experimental setup for rheological parameter
identification in relation to wood drying. Wood Sci Tech 44: 31-49.
Navi P., Pittet V., Plummer C.J.G.2002. Transient moisture effects on wood creep. Wood Sci Tech, 36:447
462.
Navi P and Heger F.2005.Comportement thermo-hydromcanique du bois. Presses Polytechniques et
universitaires romandes, CH-1015 Lausanne, Switzerland.
Nielsen LF .1984. Power Law Creep as Related to Relaxation Elasticity, Damping, Rheological Spectra, and
Creep Recovery - with Special Reference to Wood. IUFRO Timber Engineering Group Meeting, Xalapa,
Mexico.
Obataya E, Norimoto M, Gril J .1998. The effects of adsorbed water on dynamic mechanical properties of
wood. Polymer 39(14): 3059-3064.
Padanyi ZV .1991.Mechano-sorptive effects and accelerated creep in paper. International Paper Physics
Conference, Kona, Hawaii,USA, 397411.
Pentoney RE and Davidson RW .1962. Rheology and the study of wood. For Prod J 12(5) 243248.
Pickett G .1942.The effect of change in moisture-content of the creep of concrete under a sustained load. J.
Amer. Concrete Inst 13(4):333355.
Ranta Maunus, A.1975. The viscoelasticity of wood at varying moisture content. Wood Sci Tech 9: 189205.
Schaffer EL .1972.Modeling the creep of wood in a changing moisture environment. Wood Fiber 3:232235.
Schniewind A .1968. Recent progress in the study of the rheology of wood. Wood Sci Tech 2:188-206.
Selway J W.1992. Kirkpatrick, J. Proc. Cyclic humidity effects on paperboard packaging, Madison, WI,
USA. 3147 pp.
107
108
International
Scientific Conference on Hardwood Processing (ISCHP 2013)
Experimental investigation on colour and chemical modifications of normal and red heartwood beech (Fagus sylvatica L.)
ABSTRACT
The focus of this study is to investigate the changes occurring at different exposure times on artificially
photo-irradiated surfaces chemistry and colour of beech normal and red hearted wood, in order to understand the
mechanisms that cause the changes. Fourier Transform Infrared (FT-IR) spectroscopy and reflectance
spectrophotometry were used to assess artificial sunlight influence. Colour changes in wood surfaces during
photo-degradation were monitored by measuring CIELAB parameters (L*a*b*) and then calculating their
differences and the total colour change E*. The accelerating ageing of the wood samples was performed in a
Model 1500E Solar Box, equipped with a UV filter that cuts off the spectrum at 280 nm. Colour was measured
using an X-Rite CA22 reflectance spectrophotometer. Chemical surface changes were monitored using a Nicolet
Avatar 360 Fourier transform infrared spectrometer.Colour monitoring allowed to find that wood surface colour
undergoes an important variation due to photo-irradiation, occurring within the first 24-48 hours. FT-IR
spectroscopy allowed to investigate the rate of photo-degradation of wood surface due to lignin oxidation. The
most important result is that a correlation of the colour changes may be derived with the photo-degradation of
lignin obtained by FT-IR analysis, suggesting the possibility to choose the reflectance spectrophotometry as a
non-invasive standard method to monitor the state of preservation of the wood surfaces.
1. INTRODUCTION
The transition to a less intensive use of the forests, due to a territorial protection policy and to an improvement of the
socio-economic conditions of the mountain people, led to the conversion of beech (Fagus sylvatica L.) coppices into
transitional high forests. The annual cut, allowed in public forests, is an economic resource that cannot be renounced by
the forest owners (Coppini and Hermanin 2007). Beech sometimes forms red heartwood (Knoke 2002, Torelli 1984,
Vek et al. 2013). This abnormality gives a reddish colour to the wood, induced by oxygen penetration that causes a
degradation of the soluble carbohydrates and starch into coloured phenolic substances (Albert et al. 2003). The red
heartwood restricts the applications to low value products, due to aesthetic reasons, but also due to technological
problems including hard impregnation and problems in drying processes (Phler et al. 2006, Vek et al. 2013). In order
to investigate the changes occurring at different exposure times on artificially photo-irradiated surfaces chemistry and
colour of beech normal and red hearted wood, reflectance spectrophotometry and infrared spectroscopy were applied
to both kinds of wood. The obtained data were correlated to check the relationship between the variation of colour and
chemical components.
* Corresponding author: Tel.: (0039) 0761357401; Fax: (0039) 0761357; E-mail: lomonaco@unitus.it
109
Experimental
investigation
on colour
and chemical(ISCHP
modifications
of normal and red heartwood beech (Fagus sylvatica L.)
International Scientific
Conference
on Hardwood
Processing
2013)
2. METHODOLOGY
Wood samples were obtained by a single board of beech. The samples (dimensions: 25x5x0.8 cm) were stored in
darkness in a conditioned room to reach the 12% of moisture content. To perform the FTIR analysis directly on the
wood surface, slices with a size of 10 (diameter) x 2 (thickness) mm were obtained from the specimens of beech. The
dimensions of the slices were suitable for the FT-IR diffuse reflectance accessory.
The accelerated ageing of the samples was performed in a Model 1500E Solar Box (Erichsen Instruments) and the
colour was measured using an X-Rite CA22 reflectance spectrophotometer. More information are detailed in Lo
Monaco et al. (2011). Measurements were taken at the following hour intervals: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 24,
48, 72, 96, 120, 144, 168, 216, 312, 408 and 504 h. To guarantee that successive measurements were made in the same
area a graduated mask was used for wood samples. Sixty colour measuring points were chosen: thirty for red heartwood
and thirty for normal beech wood. Three measures for each point were performed, according to Normal 14/83, so that
one hundred and eighty measurements were obtained at each exposure time, then average values and standard
deviations were calculated.
Infrared spectra were obtained using a Nicolet Avatar 360 Fourier transform spectrometer. More information are
detailed in Pelosi et al. (2013). FT-IR spectra were recorded at the following time intervals: 0, 6, 12, 24, 48, 72, 96, 120,
144, 168, 216, 312, 408 and 504 hours. Peak heights were measured using OMNIC software according to the method
described in the literature (Pandey 2003).
Differences between treatments were checked with the standard paired t-test, with ANOVA and MANOVA
analysis. Post-hoc tests were conducted with Tukey HSD test method. Linear and non-linear regression analysis was
used to develop prediction models.
3. RESULTS AND DISCUSSION
Wood colour, expressed as L*a*b* coordinates, exhibits positives values both in normal and red heartwood beech,
before and after the irradiation in Solar Box. The normal wood coordinates before ageing (L*=82.060.60;
a*=4.920.15; b*=17.360.53) are similar to those found by other authors for beech (Todorovi et al. 2012) but quite
different from those of Japanese beech (Fagus crenata, L* =67.20 1.18; a*=6.48 0.41; b*=20.100.66) (Mitsui and
Tsuchikawa 2005). The red heartwood coordinates before ageing (L*=66.851.33; a*=10.470.49; b*=21.280.76)
are similar to those found by other authors for heartwood beech (Todorovi et al. 2012). By comparing the chromatic
coordinates, it can be observed that L* value before irradiation is clearly smaller in red heartwood than in normal wood
whereas the values of a* and b* parameters are larger in the red heartwood so enhancing the yellow and red hue in this
kind of wood. Concerning the L*a*b* changes as function of irradiation time, lightness (L*) clearly decreases markedly
in normal wood varying from 82.06 to 68.99 after 504 hours of exposure. The greatest L* change occurs within the first
48 h of exposure, indicating a clear darkening of the wood surface. After this time, L* reaches an almost constant value
(Figure 1). The rapid decrease of lightness observed for the normal wood within the first 48 h of exposure can be
ascribed to photodegradation processes. The photodegradation is primarily related to the decomposition of lignin due
to the chromophore groups absorbing energy, especially in the UV range of the sunlight spectrum (Pandey 2005,
George et al. 2005, Hon and Shiraishi 2001). In red heartwood, lightness undergoes a little decrease in the first 12
hours. After this time L* coordinate increase reaching the final value of 69.45 at 504 hours of exposure (Figure 1).
The chromatic coordinate a* undergoes an increase as function of exposure time in normal wood, varying from 4.92
to 9.49 after 504 h of irradiation. The greatest changes occur within the first 12 hours of irradiation in Solar Box. In red
heartwood sample a* coordinate exhibits little and fluctuating changes ranging from 10.47 at time 0 to 8.49 at 504
hours of exposure. At last, the values of the chromatic coordinate b* increase both in normal wood and in red
heartwood, during the irradiation times, varying from 17.36 (0 hours) to 29.57 (504 hours) for normal wood and from
21.28 to 31.05 in red heartwood. The greatest changes, also in this case, occur within the first hours of exposure and
indicate a clear yellowing of wood surface. It is interesting to note that the L*a*b* coordinates in normal wood and in
red heartwood wood are tending to similar values after 504 hours of exposure in Solar Box suggesting that the colour of
the two kinds of wood become the same.
Colour changes can also be analysed using E* variations as function of exposure times (Figure 2). In this way the
effect of the Solar Box exposure is even more evident. After 48 hours of exposure in Solar Box a rapid colour change
in normal wood can be observed (see Figure 2 curve A).
110
International
Conference
Hardwood
Processing
(ISCHP
ExperimentalScientific
investigation
on colour andon
chemical
modifications
of normal
and red2013)
heartwood beech (Fagus sylvatica L.)
Figure 1: Changes in L*a*b* values due to exposure in Solar Box of the normal wood and red heartwood over 504 h.
Figure 2: Changes in E* values due to exposure in Solar Box of the normal wood (A) and red heartwood (B) over 504 h.
The value of E* vary from 1.48 (after 1h of exposure) to 16.23 (after 48h of exposure), followed by little increases
to reach the final value of 18.46 at the end of the irradition. Concerning the red heartwood, E* values increase during
the time varyign from 1.27 (after 1h) to 10.11 (after 504h). The greatest total colour changes occur within the first 12
hours, as can be observed in Figure 2. By comparing the values of E* in normal wood and red heartwood, it can be
stressed that in borth cases an increasing trend has been observed and that the greatest variation occur within the first
24-48 hours. In both cases, after 168 hours of exposure, a plateau has been reached in the E* values, even if normal
wood undergoes greater colour changes in comparison with red heartwood, as shown in Figure 2.
The MANOVA test was applied to the colorimetric data of normal and red hearted wood, before exposure, in order
to evaluate the difference between the two kinds of wood. The results (p level <0.001) demonstrate that there is a highly
statistically significant difference between normal wood and red heartwood.
To further evaluate the differences between normal wood and red heartwood, the statistical analysis was applied
also to the the single L*, a* and b* coordinates before exposure of the samples in Solar Box. The results of the Tukey
test applied to the L*a*b* coordinates at time=0 demonstrate that normal wood and red heartwood constitute two
groups concerning every chromatic coordinate. The statistical difference between normal and red hearted wood is
highly significant. The time variations for the two studied cases were then examined by means of MANOVA test. This
test demonstrates that, from 0 to 504 hours of exposure, differences statistically significant are present as concerns both
the time intervals and also the types of wood (normal and red hearted). Furthermore, the simultaneous variation of time
and type of wood is statistically significant. The relationships between the chromatic coordinates and the exposure
times were examined by applying statistical regression analysis. The overall results are shown in Table 1. The
equations shown in Table 1 for the variables L*, a* and b* as function of the exposure times, give the second degree
polynomial curves of the chromatic variables during the time. All constants of the polynomials are characterized by
highly statistical significance (p<0.01). Wood colour varies with the exposure time, in particular, lightness (L*) values
111
Experimental
investigation
on colour
and chemical(ISCHP
modifications
of normal and red heartwood beech (Fagus sylvatica L.)
International Scientific
Conference
on Hardwood
Processing
2013)
decrease for normal wood (darkening) and increase for red heartwood (lightening). The same trend can be observed for
the coordinate a*. At last, b* coordinate increases both in normal wood and red heartwood (yellowing).
At last, for normal wood and red heartwood a regression analysis was applied to the time as function of colour
coordinates (Table 2). The results shown in Table 2 underline the highly statistical significance of the obtained
measures regarding the dependent variable time (t) as function of the colour coordinates (L*, a* and b*). This result
suggests that the colorimetric parameters can be used to derive the exposure times of the wood samples, in a totally
non-invasive procedure.
Table 1: Regression analysis for the dependent variables L*a*b* of normal wood and red heartwood as function of the time (t).
Normal wood
Variable coefficient
Intercept
t
t2
Red heartwood
p-level
L*
76.615
-0.045
0.00005
Variable coefficient
<0.01
<0.01
<0.01
Radj.=0.541; p<0.01
L*= 76.615-0.045t+0.00005t
Intercept
t
t2
a*
<0.01
<0.01
<0.01
Radj.=0.489; p<0.01
a*=6.963+0.020-0.00002t
Intercept
t
t2
<0.01
<0.01
<0.01
Radj.=0.495; p<0.01
L*= 64.515+0.021t-0.000002t2
a*
6.963
0.020
-0.00002
p-level
L*
64.515
0.021
-0.00002
Radj.=0.684; p<0.01
b*=22.19+0.053t-0.00006t2
<0.01
<0.01
<0.01
Radj.=0.467; p<0.01
a*=10.067-0.008t+0.00001t2
b*
22.19
0.053
-0.00006
10.067
-0.008
0.00001
b*
<0.01
<0.01
<0.01
22.954
0.057
-0.00006
<0.01
<0.01
<0.01
Radj.=0.848; p<0.01
b*=22.954+0.057t-0.00006t2
Concerning infrared spectroscopy on the samples, the spectra were collected in DRIFT modality directly on wood
slices in order to avoid analysing also the unaltered wood (Tolvaj et al. 2011). In fact, wood is a good light absorber for
infrared, visible and ultraviolet light (Hon and Shirashi 2001, Kataoka and Kigushi, 2001).
Band assignment was made according to literature references (Pandey 2003, Chang et al. 2002, Moore and Owen
2001, Colom et al 2003).
The irradiating-time dependent diffuse reflectance FT-IR spectra of normal wood and red heartwood are shown in
Figures 3-4. It can be observed that the intensity of the bands at 1507 cm-1, 1594 cm-1 and 1465 cm-1, associated to
lignin, decrease during photo-degradation. This is accompanied by an increase in the intensity of the band at 1739 cm-1,
due to carbonyl absorption. The intensities of the peak associated to carbohydrates at 1375 cm-1 is not significantly
affected by irradiation, so this band has been used as internal reference to evaluate the lignin decay.
In order to determine the rate of lignin decay and carbonyl formation, the intensities of the lignin band at 1507 cm-1,
the carbonyl band at 1739 cm-1 and the carbohydrate reference band at 1375 cm-1 were calculated (Pandey 2005). Then
the relative change in ratio of lignin/carbohydrate bands at different exposure times were derived and statistically
evaluated. The lignin/carbohydrate (I1507/I1375) intensity ratio decreased quickly with the exposure times both for
normal wood and red heartwood and after 48h of irradiation it decreased to about 50% of its original value. As the
carbonyl groups increase during the photo-ageing whereas the lignin band decreases, the ratio of intensity of the bands
at 1507 cm-1 and 1739 cm-1 undergoes a fast and significant decrease after few hours of exposure. In fact, after 12 hours
of exposure it decreased to about 50% of its original value.
112
International
Conference
onchemical
Hardwood
Processing
(ISCHP
2013)
ExperimentalScientific
investigation
on colour and
modifications
of normal
and red
heartwood beech (Fagus sylvatica L.)
Table 2: Regression analysis for the dependent variable t (time) as function of the L*a*b* coordinates, in normal wood and red
hearthwood.
Variable coefficient
Intercept
L*
a*
b*
L*2
a*2
b*2
p-level
Normal wood
10525.09
-256.27
121.18
-111.61
1.67
-11.27
3.03
p<0.01
p<0.01
p<0.01
p<0.01
p<0.01
p<0.01
p<0.01
R adj.=0.891; p<0.01
t=10525.09-256.27L*+121.18a*-111.61b*+1.67L*2-11.27*2 +3.03b*2
Red heartwood
Intercept
L*
a*
b*
L*2
a*2
b*2
9925.968
-137.215
-509.234
-218.429
1.004
24.137
4.599
p<0.01
p<0.01
p<0.01
p<0.01
p<0.01
p<0.01
p<0.01
R adj.=0.889; p<0.01
t=9925.968137.215L*-509.234*-218.429+1.004L*2+24.137*2+4.599b*2
Figure 3: Diffuse reflectance spectra of normal wood slices measured at chosen exposure times.
The trends of the peak ratios put in evidence that the photodegradation process is similar in normal wood and in red
heartwood. To verify this statement, the inference statistics was applied to the I1507/I1375, I1739/I1375 and I1507/I1739 peak
ratios as function of the time. In particular, the T-test for dependent samples was applied to the peak ratios as function
of the time in order to verify if statistically signifant differences can be found between normal wood and red heartwood.
The results shown in Table 3 demonstrate that no statistically significant differences exist between the two types of
wood, in fact the p values are always >0.05.
According to this result, the experimental data will be statistically examined in a single group, without distintion
between normal wood and red heartwood.
113
International Scientific
Conference
on Hardwood
Processing
2013)
Experimental
investigation
on colour
and chemical(ISCHP
modifications
of normal and red heartwood beech (Fagus sylvatica L.)
Figure 4: Diffuse reflectance spectra of red heartwood slices measured at chosen exposure times.
Table 3: T-test for dependent samples on normal wood and red heartwood.
Average
I1507/I1375
Normal wood
0.856
Standard
deviation
0.475
I1507/I1375
Red heartwood
0.812
0.512
I1739/I1375
Normal wood
I1739/I1375
Red heartwood
3.162
0.961
3.087
0.752
I1507/I1739
Normal wood
I1507/I1739
Red heartwood
0.354
0.338
0.316
0.311
Diff.
14
-0.044
-0.670
Degree of
freedom
13
p-level
14
-0.075
-0.379
13
>0.05
14
-0.038
-1.976
13
>0.05
>0.05
Successively the relationships between the FT-IR peak ratios and the exposure times were examined by applying
statistical regression analysis. The results are shown in Table 4.
The equations shown in Table 4 for the variables I1507/I1375, I1739/I1375 and I1507/I1739 as function of the exposure
times, give the second degree polynomial curves (for I1507/I1375 and I1507/I1739) and the line (for I1739/I1375). All constants
of the polynomials and of the straight line are characterized by highly statistical significance (p<0.01).
At last the colour changes of wood during irradiation were correlated with lignin decay and the formation of
carbonyl groups produced by the photo-degradation process. In fact, as the colour changes of the surfaces are mainly
due to the liberation of C=O groups due to photo-degradation of lignin, there would be a correlation between the colour
coordinates and the relative intensity of the infrared peaks associated to lignin. The results of the regression analysis put
in evidence that only the I1507/I1375 peak ratio has a statically significant relation with the chromatic coordinates, in
particular with L* and a* (see Table 5).
The polynomial curve derived by the application of the regression analysis exhibits high statistical significance
(p<0.01). Only the chromatic coordinates L* and a*, that represents the lightness and the green-red axes, are
statistically influent in relation to the peak ratio suggesting that they can be used to evaluate the photodegradation of
wood.
114
International
Conference
Hardwood
Processing
(ISCHP
ExperimentalScientific
investigation
on colour andon
chemical
modifications
of normal
and red2013)
heartwood beech (Fagus sylvatica L.)
Table 4: Regression analysis for the dependent variables I1507/I1375, I1739/I1375 and I1507/I1739 as function of exposure time (t).
Variable coefficient
p-level
I1507/I1375
Intercept
t
t2
1.431
-0.006
0.000007
<0.01
<0.01
<0.01
Radj.=0.777; p<0.01
I1507/I1375 = 1.431-0.006t+0.000007t2
I1739/I1375
Intercept
t
<0.01
<0.01
1.45
0.004
Radj.=0.412; p<0.01
I1507/I1375 = 1.45+0.004t
I1507/I1739
Intercept
t
t2
1.26
-0.004
0.000004
<0.01
<0.01
<0.01
Radj.=0.613; p<0.01
I1739/I1507 = 1,26-0,004t+0,000004t2
Table 5: Regression analysis for the dependent variables I1507/I1375 as function of the chromatic coordinates.
Variable coefficient
p-level
-21.948
<0.01
2.95
<0.01
0.002
<0.01
a*2
-0.163
<0.01
Intercept
a*
L*
Radj.=0.892; p<0.01
I1507/I1375 = -21.948+2.953a*+0.002L*2-0,163a*2
4. CONCLUSIONS
The study of colour changes in normal wood and red heartwood showed that the surface colour undergoes an
important variation due to photo-irradiation. The statistical difference between normal wood and red heartwood is
highly significant. At last, for normal wood and red heartwood the regression analysis underlines the highly statistical
significance of the obtained measures regarding the dependent variable time as function of the colour coordinates. This
result suggests that the colorimetric parameters can be used to derive the exposure time of the wood samples, in a totally
non-invasive procedure.
FT-IR spectroscopy allowed to investigate the rate of photo-degradation of wood surface due to lignin oxidation.
The T-test for dependent samples, applied to the peak ratios as function of the time demonstrated that no statistically
significant differences exist between the two types of wood.
115
Experimental
investigation
on colour
and chemical(ISCHP
modifications
of normal and red heartwood beech (Fagus sylvatica L.)
International Scientific
Conference
on Hardwood
Processing
2013)
The regression analysis put in evidence that the differences between the values at the various irradiation times are
statistically significant, apart from the I1737/I1376 ratio, suggesting the possibility of using the FT-IR peak ratios to
evaluate the ageing time of the wood surface.
The results of the regression analysis demonstrated that only the I1507/I1375 peak ratio has a statically significant
relation with the chromatic coordinates, in particular with L* and a*. This finding showed that non-invasive colour
measurements can be used to evaluate the photo-degradation of wood.
REFERENCES
Albert, L., Hofmann, T., Nmeth, Zs. I., Rtfalvi,T., Koloszr, J. Varga, Sz., Csepregi, I. 2003. Radial variation of total phenol
content in beech (Fagus sylvatica L.) wood with and without red heartwood. Holz als Roh- und Werkstoff 61:227230
Chang, H. T., Yeh, T. F., Chang, S. T. 2002. Comparisons of chemical characteristic variations for photodegraded softwood and
hardwood with/without polyurethane clear coatings. Polymer Degradation and Stability 77:129-135.
Colom, X., Carrillo, F., Nogus, F., Garriga, P. 2003. Structural analysis of photodegraded wood by means of FTIR spectroscopy.
Polymer Degradation and Stability 80:543-549.
Coppini, M., Hermanin, L. 2007. Restoration of selective beech coppices: A case study in the Apennines, Italy. Forest Ecology and
Management 249:1827
George, B., Suttie, E., Merlin, A., Deglise, X. 2005. Photodegradation and photostabilisation of wood - the state of the art. Polymer
Degradation and Stability 88:268-274.
Hon, D. N. S, Shiraishi, N. 2001 Color and Discoloration and Weathering and photochemistry of wood. Wood and Cellulose
chemistry. Marcel Dekker Inc., New York, 385-442, 513-546.
Kataoka, Y., Kigushi, M. 2001. Depth profiling of photo-induced degradation in wood by FT-IR Microspectroscopy. Journal of
Wood Science 47:325-327.
Knoke, T. 2002. Value of complete information on red heartwood formation in beech (Fagus sylvatica). Silva Fennica 36(4):
841851.
Lo Monaco, A., Marabelli, M., Pelosi, C., Picchio, R. 2011. Colour measurements of surfaces to evaluate the restoration materials.
In: Proceedings of SPIE 8084. pp. 1-14.
Mitsui, K., Tsuchikawa, S. 2005. Low atmospheric temperature dependence on photodegradation of wood. Journal of
Photochemistry and Photobiology 81:84-88.
Moore, A. K., Owen, N. 2001. Infrared spectroscopic studies of solid wood. Applied Spectroscopy Reviews 36:65-86.
Pandey, K.K. 2003. FTIR studies of the changes in wood chemistry following decay by brown-rot and white-rot fungi.
International Biodeterioration and Biodegradation 52:151-160.
Pandey, K. K. 2005. Study of the effect of photo-irradiation on the surface chemistry of wood, Polymer Degradation and Stability
90:9-20.
Pelosi, C., Agresti, G., Calienno, L., Lo Monaco, A., Picchio, R., Santamaria, U., Vinciguerra, V. 2013. Application of
spectroscopic techniques for the study of the surface changes in poplar wood and possible implications in conservation of
wooden artefacts. In: Proceedings of SPIE 8790-41. pp.1-14.
Phler, E., Klingner, R., Knniger, T. 2006. Beech (Fagus sylvatica L.) Technological properties, adhesion behaviour and
colour stability with and without coatings of the red heartwood. Annals of Forest Sciences 63:129137.
Todorovi, N., Popovi, Z., Mili, G., Popadi, R. 2012. Estimation of heat-treated beechwood properties by colour change.
Bioresources 7(1):799-815.
Tolvaj, L., Mitsui, K., Varga, D. 2011. Validity limits of Kubelka-Munk theory for DRIFT spectra of photodegradated solid wood.
Wood Science and Technology 45:135-146.
Torelli, N. 1984. The ecology of discoloured wood as illustrated by beech (Fagus sylvatica L.). IAWA Bulletin 5(2):121-127.
Vek, V., Oven, P., Poljanek, I. 2013. Content of Total Phenols in Red Heart and Wound-Associated Wood in Beech (Fagus
sylvatica L.). Drvna Industrija 6 (1):25-32.
116
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
) Institute of Wood Science and Technology Department of Material Science and Process Engineering,
University of Natural Resources and Life Science, Konrad Lorenzstrasse 24, 3430 Tulln, Austria,
ulrich.mueller@boku.ac.at
) Competence Centre for Wood Composites and Wood Chemistry, Wood K plus, 4040 Linz, Austria
1. INTRODUCTION
Lots of scientific research has been performed on the discolouration of wood surfaces caused by light irradiation
and technological processes such as wood drying, steaming or modification (Teischinger et al. 2012). On the
other hand several studies investigating the chemical photo reaction are available (Scott 1990, Rabek 1995,
Andrady et al. 1998, Derbyshire et al. 1995, Kiguchi and Evans 1998, Hon and Minemura 2001). Other studies
investigated the specific absorption of different macromolecules and extractives in wood (Vorreiter 1949,
Norrstrom 1969, Gierlinger et al. 2004, Moya et al. 2012).
However, colour effects of solid wood in a native state as well after wetting or oil coating are hardly
investigated. In this paper we propose that the reflection of light on a wood surface f(E) corresponds the
following equation:
f(E) = f(R) + f(S) + f(Rf) + f(A)
where f(R) represents the surface roughness, f(S) the wood structure, f(RF) refection index of the wood cell wall
and f(A) light absorption of the wood cell wall.
1 117
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
Surface roughness determines the gloss of the surface. Increasing roughness results in a decrease of
gloss and a colour shift in direction of white (Hecht 2001). The internal structure (i.e. surface roughness of the
inner cell walls, lumen diameter and light conduction along the wood cells) influences the light refraction and
translucency. Lower density means higher translucency and an increase of light refraction (Op cit.). Light
absorption determines the wood colour, light intensity, and light penetration. In this study effects of roughness,
structure and absorption of thin wood slides were investigated.
2. MATERIAL AND METHODS
For the study poplar (Populus spec.), walnut (Juglans regia), and ebony (Diospyros spec.) were used. To
produce specimen with very light wood colour poplar was bleached with hydrogen peroxide. Wood slides (0.02
to 1.60 mm) were produced by using a sledge microtome and a circular saw. The surface of all samples were
sanded with grid of 280. Surface roughness after oil and wax coating was measured by means of an Alicona
Infinite Focus device (Alicona Imaging GmbH, Raaba/Graz, Austria). To study the effect of surface roughness
and change in light refection the slides were: poplar untreated (N=126), poplar moistened with distilled
water (N=126), poplar saturated with distilled water (N=126), poplar coated with boiled linseed oil (N=126),
castor oil (N=126), and walnut oil (N=126), poplar bleached and linseed oil coated (N=126), walnut untreated
(N=64), and ebony untreated (25). These and additional samples (walnut bleached (N=61), walnut bleached and
linseed oil coated (N=61), ebony bleached (N=25) and ebony bleached and linseed oil coated (N=21)) were used
to measure light intensity and colour saturation.
To measure the transparency of coatings, coatings are applied on black and white standard surfaces.
Difference in colour can be related to the transparency of coatings. The same procedure was used to determine
the transparency of the thin wood slides. Wood slides (0.02 mm to 1.60 mm, in total more than 800 specimens)
were placed on white and black standards. For colour measurements a Phyma CODEC 400 Vis
spectrophotometer (spectra between 400 to 700 nm were collected) was used. The CIELAB colour system was
applied (DIN 6174, DIN EN ISO 11664-4) to determine the CIELAB colour difference. The colour difference
was plotted over the thickness of the specimens resulting in an exponential decrease of colour difference with
increasing thickness of the wood slides. An exponential function was fitted to the data points by using a nonlinear-least-square fit. Comparing the different samples a one way ANOVA was used.
For gloss measurements a BYK-Gardner micro-TRI-gloss device was used. Measurements were
performed at 20, 60, 80 (in longitudinal and transversal direction). For viscosity measurements of the different
oils a rotation rheometer with a cone-plate measuring system was used. To further characterize the oils, their
molecular weight was determined by using the average composition and average molecular weight of the
individual fatty acids was used. The penetration behavior of the various oils in wood has been studied using
microscopy; prior to microscopy, oils were stained with Sudan III solution and then measured.
Roughness on a nano-scale was measured by means of an AFM, electron microscopy (Quanta 250 FEG,
FEI, Hillsboro, OR, USA). Because it was not possible to perform AFM scans on solid wood due to the
macroroughness, coated Thypha fibers (Typha latifolia) were used as an alternative. The Typha fiber provided a
relatively large surface without any major cavities, but it is also a lignocellulosic material like wood (Hauptmann
et al.). Natural fibres (Typha latifolia) were dipped into the oil and thereafter the surface roughness of the fibres
was determined and compared to the surface of untreated fibres. AFM measurements were performed on a
Bruker Dimension Icon (Bruker, Santa Barbara, CA, US) in standard tapping mode (Op cit.).
118
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
Newmann 2013), the maximum path of the noticeable light in the wood substrate is in the range of 0.5 mm for
poplar untreated, 0.8 mm for poplar oil treated, 0.2 mm walnut untreated, and 0.03 mm for ebony untreated,
respectively. For poplar samples bleached a colour difference of E = 2 was observed at a sight higher thickness
of approx. 0.83 mm. In general oil treatment caused an increase in the transparency of the wood slides. For
natural poplar and poplar bleached an increase of the slide thickness of approx. 0.3 mm was observed. In
Figure 2 the light reflection depth of the different wood species with and without water and oil treatment are
presented.
Fig. 1: CIELAB colour difference of thin wood slides (0.02 mm to 1.60 mm) which were placed on white and black
standards; a) poplar after bleaching, b) natural ebony wood.
Fig. 2: light reflection depth of the different wood species with and without water and oil treatment
3 119
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
This means that the diffuse reflection of solid wood surfaces is influenced by the transparency of the shallow
wood cell walls. With increasing brightness of the wood colour a higher proportion of the light is reflected from
deeper cell layers. Impregnation of the wood slides with water leads to much higher transparency (up to 3-fold
higher transparency for bleached poplar samples). Coating with different oils also results in an increase of the
transparency of bright wood species, but is getting negligible small for darker wood species. Furthermore, the
natural colour of the used oils has no effect on the appearance of the coated wood. It was also found that the
penetration of the various oils is significantly different. This could be a reason for the various resulting
roughness values.
On the other hand coating of solid wood with different oils increases the gloss (see Figure 3c). With
increasing viscosity and higher molecular weight also smoother surfaces can be observed for oil coated solid
wood specimen (Figures 3a and 3b). In general a reduction of brightness was observed with increased gloss.
Fig. 3: Optical 3D surface measurement of wood surfaces coated with a) sunflower oil and b) shellac wax proving much
smother surface for the wax treatment (violet and red indicate a level of -60 and +20 m, respectively). c) Higher gloss
values of poplar wood after oil treatment was observed.
This effect can be observed on a micrometer level, as well as on a nanometer level. AFM measurements
clearly showed that oil treatment results in a lower roughness on the nano-meter scale in comparison to untreated
120
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
fibre-surfaces. On the other hand surface roughness correlates with decreasing brightness of wood surfaces
treated with the same oils (Figure 4) (Hauptmann et al. 2013). The results correspond to studies performed to
evaluate the effect of varnishes of paintings. The gloss, as well as the colour saturation, is influenced by the
roughness of the varnishes (de la Rie 1987). On a rough surface, light is not only specularly reflected, but also
diffused. Scattered light from a rough surface appears whiter (the reflection spectrum in the reflection minimum
increases) (de la Rie 1987, Hecht 2001). Similar optical effects were also observed for hair care products
(Watanabe et al. 2007).
4.CONCLUSION
Reduction of the refraction index yields an increase of diffusely reflected light
The smoother the surface (nano-scale as well as micro-scale) the darker and the higher the gloss of the wood
surface.
The darker the wood the lower the transparency and the light penetration into the wood substrate. This means
that the colour of dark wood species is mainly influenced by appearance of the cells directly on the surface.
Whereas, the colour of bright wood species is influenced by the volume of the transparent surface layers.
Oil treatment results in an increase of the transparency of wood. The brighter the wood colour the higher the
effect of increased transparency after treatment.
The oil colour has no significant effect on wood colour.
Different oils (molecular weight, viscosity) strongly affect the surface roughness on the micrometer as well
as on the nanometer scale.
Decrease in surface roughness results in higher gloss and reduction of brightness of the wood surface.
Different oils (molecular weight, viscosity) show different penetration behaviour into the wood substrate.
REFERENCES
Andrady AL, Hamid S H, Hu X, Torikai, A J (1998) Effects of increased solar ultraviolet radiation on materials, J.
Photochem. Photobiol. B: Biology 46: 96-103.
de La Rie, R. E. (1987). "The influence of varnishes on the appearance of paintings," Studies in Conservation 32(1), 1-13.
Derbyshire H, Miller E R, Turkulin H (1995) Investigations into the photodegradation of wood using microtensile testing
Part 1: The application of microtensile testing to measurement of photodegradation rates, Holz als Roh- und Werkstoff
53: 339-345.
DIN 6174 (1979). Colorimetric Evaluation of Colour Differences of Surface Colours According to the CIELAB Formula.
EN ISO 11664-4, 2012, CIE 1976 L*a*b* Farbenraum: Norm : En Iso ; 11664-4, v. 4, 11 S.p.
Gierlinger, N., Jacques, D., Grabner, M., Wimmer, R., Schwanninger, M., Rozenberg, P., and Pques, L. E. (2004). "Colour
of larch heartwood and relationships to extractives and brown-rot decay resistance," Trees - Structure and Function 18(1),
102-108.
Has, M., Newmann, T., (2013) Color Management: Current Practice and The Adoption of a New Standard,
http://www.color.org/wpaper1.html, accessdate: 2013-05-28
Hauptmann, M., Mller, U., Obersriebnig, M., Gindl-Altmutter, W., Beck, A., Hansmann, C., (2013) The Optical
Appearance of Wood Related to Nanoscale Surface Roughness, Bioresources (accepted for publication)
Hecht, E., 2001, Optik: Mnchen ; Wien, Oldenbourg, XV, 1040 S. p.
Hon D.N.S., Minemura, N., (2001) Colour and discoloration. In: Wood and cellulosic chemistry, 2 nd edition (Eds. Hon,
D.N.S., Shiraishi, N.) New York, Marcel Dekker, pp. 385-442.
Kiguchi M, Evans PD (1998) Photostabilisation of wood surfaces using a grafted benzophenone UV absorber, Polymer
Degradation and Stability 61: 33-45.
5 121
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
Moya, R., Fallas, R. S., Bonilla, P. J., and Tenorio, C. (2012). "Relationship between wood color parameters measured by the
CIELab system and extractive and phenol content in Acacia mangium and Vochysia guatemalensis from fast-growth
plantations," Molecules 17(4), 3639-3652.
Norrstrom, H. (1969). "The colour of unbleached sulphate pulp," Svenska Traforskningsinstitutet, Stockholm.
Rabek J F (1995) Polymer degradation, Chapman and Hall, London.
Scott GG, (1990) Polymer Degradation and Stabilization, Elsevier Applied Science, London.
Teischinger, A., Zukal, M.L., Meints, T., Hansmann, C., Stingl, R. (2012) Colour characterization of various hardwoods
In: R. Nemeth, A. Teischinger (eds.) University of West Hungary Press, Sopron. ISBN 978-963-9883-97-0, Hardwood
Science and Technology. The 5th Conference of Hardwood Research and Utilization in Europe, 2012, p. 180-188
Vorreiter, L. 1949. Holztechnologisches Handbuch Band 1, Verlag Georg Fromme & Co, Wien.
Watanabe, S., Sato, H., Shibuichi, S., Okamoto, M., Inoue, S., and Satoh, N. (2007). "Formation of nanostructure on hair
surface: Its characteristic optical properties and application to hair care products," Journal of Cosmetic Science 58(4),
283-294.
122
Assessment,
and conservation
of forest resources:
Wound management
closure and discolouration
of sycamore
(Acer
pseudoplatanus L.) wood
following
artificial
pruning:
A
case study
the FAO perspective
Director,
Forestry Officer, Forest Assessment,
Forestry Officer, Forest Assessment,
Forest Assessment, Management and
Management and Conservation Division,
Management and Conservation
Conservation Division,
Forestry ABSTRACT
Department,
Division, Forestry Department,
Forestry Department,
Food
and
Agriculture
Organization
of
the
Food
and Agriculture Organization of
Food and Agriculture Organization of
United
Nations,
Rome,
Italy
the
United
Nations,
Sycamore
is considered
a moderate-high self-pruning species when grown in dense
stands.
ThisRome,
is notItaly
the case in
the
United Nations,
Rome, Italy
low density stands managed dynamically (combination of thinning from above and pruning), targeting the
production of high quality logs pruned up to 6 (even 8) m.
In such stands a research project dealing with the speed of wound closure and wood discolouration following
ABSTRACT
artificial pruning was launched in 2006. It was carried
out in sub-compartments 75G and 81B% managed by
Kronstadt and uga Forest Districts respectively (South-East of Transylvania, Romania).
FAO
is a specialized
agency
of the26United
Nations
leading axis
global
to eradicate
The size
of branches
removedinternational
by pruning was
variable:
branches
with vertical
(Va)efforts
between
1.2 cm and
5.3 cmhunger,
in sub-compartment
and 20
Va between management
3.3 cm and 11.5
cm in resources.
sub-compartment
eliminate rural75G
poverty
andbranches
promotewith
the sustainable
of natural
In a
81B%.world with an expected population of 9 billion people by 2050, demand for food and other products and
In sub-compartment
75G, allwill
pruning
wounds
were
closed
maximum
3 years,both
85%the
of extent
them being
closed
after
services from nature
be greater
than
ever.
This after
pressure
will change
and the
quality
2 years;
the
ones
closed
after
3
years
were
either
large
wounds
(Va
4.1
and
4.5
cm)
or
small
wounds
(1.2
cm
Va)
of a substantial portion of the worlds forests. Understanding these changes is essential if investments
locatedare
on to
trees
with
low
diameter
increments.
be made to improve forest management to both provide for a growing human population and to
In sub-compartment
81B%,
only
8 branchesFAO
(40%),
with Va between
3.3 and reliable
8.1 cm were
closed after
6 years;
maintain crucial
forest
biodiversity.
is committed
to generating
information
on forests
andthe
remaining 12 branches (Va 3.3-11.5 cm) were partially closed (29 to 90% of Va).
forestry practices to promote the conservation and sustainable use of forest resources worldwide.
In both fully and partially closed wounds, the area of discoloured wood ranges between 3.5 and 43 sq.cm, this
Clearly, additional investment is needed in research, whilst technical support, and field activities are still
parameter being strongly correlated with the Va of pruning wound.
needed to slow the loss and degradation of forest area, particularly in the tropics.
1. INTRODUCTION
1.Sycamore
INTRODUCTION
(Acer pseudoplatanus L.) is one of the most valuable broadleaved tree species in Europe, covering
1.2% of continents
and providing
1.7% ofofannual
wood harvest
et is
al. a2009).
It is a very
promising
The Foodforest
and area
Agriculture
Organization
the United
Nations(Thies
(FAO)
specialized
international
species
for
multiple-objective
forestry
taking
into
accounts
some
of
its
major
characteristics:
agency leading global efforts to eradicate hunger, eliminate rural poverty and promote the sustainable
- one of the fastest-growing broadleaved tree species when grown on suitable sites (Bastien 1997, Boppe
management of natural resources. FAO serves its 192 member countries3 as a neutral forum, retaining the
1889, Evans 1984, Haralamb 1967, Hein et al. 2009, Savill 1991, 2013, Waters 1992);
following
core
functions:
1) facilitate
and uses
support
in the development
and implementation
of
- yields
a valuable
timber
for numerous
(e.g. countries
veneer, furniture,
musical instruments,
flooring, parquets,
normative
andpulpwood,
standard-setting
instruments
suchand
as Kldtke
international
technical
joinery,
even firewood)
(Becker
2009, agreements,
Evans 1984, codes
Hein etofal.conduct,
2009, Rusanen
and
standards
and 2003,
others;
2) assemble,
analyse, monitor and improve access to data and information, in areas
Myking
Savill
1991, 2013);
related
to FAOs
3) facilitate,
promote
support policy
dialogue
global, regional
and country
- tolerant
to amandate;
wide range
of climatic
and soiland
conditions
(Haralamb
1967,atNegulescu
and Svulescu
1965,
Leslie
2005,
Savill
1991,
2013,
Stnescu
1979,
Stnescu
et
al.
1997,
ofletea
and
Curtu
2007);
levels; 4) advise and support capacity development at country and regional level to prepare, implement, monitor
- evaluate
high ecological
(including
biodiversity)
and landscape
values (Allegrini
et al. and
1998,support
Hein et activities
al. 2009, Leslie
and
evidence-based
policies,
investments
and programmes;
5) advise
that
2005).
assemble, disseminate and improve the uptake of knowledge, technologies and good practices in the areas of
Sycamore usually exhibits low potential for producing epicormic branches, even though such shoots can develop
FAOs mandate; 6) facilitate partnerships for food and nutrition security, agriculture and rural development
vigorously especially when the trees are suddenly exposed to lateral light in excess following thinning, are pruned too
between governments,
civil society
theand
private
sector;
7) advocate
andCourraud
communicate
brutally/too
late or standsdevelopment
are kept toopartners,
dense (Bastien
1997,and
Franc
Ruchaud
1996,
Hubert and
1998,
at
national,
regional
and
global
levels
in
areas
of
FAOs
mandate.
Kadunc 2003, Hein 2009).
Sycamore silviculture targets the production of healthy individuals, with straight boles, and even growth rings of
minimum 4-5 mm per year (Hein and Spiecker 2009). The target diameter (d.b.h.) of individual trees is 40-60 cm
(Joyce et al. 1998, Kerr and Evans 1993) and such large trees should be produced at rotation ages of maximum 50-60
years, when the occurrence of brown heart (heartwood discolouration), a major defect of sycamore wood reducing its
1
3
Corresponding
author: Tel.: (40) 268-417898; Fax: (40) 268-417898; E-mail: nvnicolescu@unitbv.ro
FAO has 192 member states along with the European Union, Faroe Islands and Tokelau which are associate members
1 123
International
Scientific
Conference on Hardwood Processing (ISCHP2013)
International Scientific Conference on Hardwood Processing
(ISCHP
2013)
market value (Boudru 1989, Hein 2004, quoted by Becker and Kldtke 2009, von Wedel 1964, quoted by Hein et al.
2009, Kadunc 2003, 2007) is still not important.
A major issue related to the production of high-quality sycamore trees for important wood assortments (e.g.,
sliced veneer, A-grade saw timber) is the dynamics of self-pruning, as branchiness and related knottiness are the key
determinants of wood quality (Hein et al. 2009). In this respect sycamore is considered a rapid self-pruning species
when grown in dense stands due to its quick early height growth (Hein 2009, Hein et al. 2009) (figure 1).
Figure 1: Nicely self-pruned sycamore tree of good quality in a regular stand with undergrowth.
The potential of self-pruning is considered by other authors as moderate due to its medium light requirements
when young (Hubert and Courraud 1998, Joyce et al. 1998).
The high-moderate self-pruning potential of sycamore trees is not able to express itself in low density stands
managed dynamically (combination of thinning from above and artificial pruning) and targeting the production of
high quality logs pruned up to 6 (even 8) m. In such stands, the natural shedding of low branches on sycamore trees is
slow as a result of reduced shade within the stand so the need for artificial pruning becomes obvious (Duyck 2008,
Joyce et al. 1998).
In these circumstances and taking into account the interest to produce new silvicultural guidelines for young and
widely-spaced Romanian plantations with valuable broadleaved tree species, a research project dealing with pruning
wound closure and wood discolouration following artificial pruning of sycamore trees was launched in 2006. The
project has the following objectives:
- definition of the potential correlation between the duration of wound closure and the size of pruned branch.
- definition of the potential correlation between the size of discoloured wood and the size of pruned branch.
- other issues of practical importance related to pruning of sycamore trees.
The objectives take into account the existing literature on these topics showing that:
- the duration of pruning wound closure is positively correlated with the branch size (Wright et al. 1980, Winterfeld
1955, in Schtz 1990).
- the pruned branch is always associated with decay column (discolouration) (Schatz et al. 2008, Wardlaw and
Neilsen 1999).
- the colour defects spread mainly towards the pith (Schatz et al. 2008).
- there is a strong correlation between the size of discolouration and increasing branch size (Chiu et al. 2004, Ow et
al. 2013, Wardlaw and Nielsen 1999).
The project started in 2006 took into account some hypotheses:
- sycamore trees to prune should be young (maximum 10-15 cm d.b.h.) in order to minimise the amount of knotty
wood at rotation age;
- pruning of sycamore trees should be carried out in the spring time before the beginning of new growing season,
even though there is no strict indication of the preferred pruning time for this tree species as pointed out recently by
Fini et al. (2013);
- the natural target pruning (cut away without leaving any stub - from the branch collar Shigo 1979) and not flush
cut pruning (cutting through branch collar) should be used as the first is considered a better method to facilitate
effective wound closure (Ow et al. 2013).
124
Area
(ha)
Aspect
75G
1.9
Mean
slope
(g)
20
Mean
elevation
(m)
800
Forest
vegetation
type
4111
Age
(years)
60%EB10%SYC10%NS10%VB
17
S
81B%
2.3
N
8
780
5221
70%NRO10%SYC10%WC
16
4111 Mountain European beech stands of high productivity; 5221 Mixed sessile oak-European beech of moderate productivity;
EB = European beech (Fagus sylvatica); SYC = sycamore; NS = Norway spruce (Picea abies); VBS = various broadleaved
species; NRO = northern red oak (Quercus rubra); WC = wild cherry (Prunus avium)
The local climate in both regions is a low mountain elevation-type, with a mean annual temperature of 7-8 Celsius
degrees and a mean annual rainfall of 650-750 mm. Brown forest soils exist in both stands.
The 75G stand was naturally regenerated following a small-scale windfall occurring in 1996. The 81B% stand
was planted with an initial stocking of 3,900 trees/ha (1.6 x 1.6 m) in 1997.
In these stands, a manual saw pruning of sycamore branches up to 2 m was performed in 2006. The removed
branches showed an elliptical shape (horizontal axis Ha smaller than the vertical axis Va), with variable sizes:
- 26 branches with Ha of 0.5-4.1 cm and Va between 1.2 cm and 5.3 cm in sub-compartment 75G;
- 20 branches with Va between 3.3 cm and 11.5 cm in sub-compartment 81B%.
The sycamore trees on which these branches were cut have had a mean diameter of 8.96 cm (range 7.4-11.4 cm),
a mean height of 10.0 m (range 7.5-11.4 m) and a mean height of pruning wounds of 1.32 m (range 0.35-1.70 m).
In sub-compartment 75G the d.b.h. of all 19 pruned trees have been measured with a caliper (precision 1 mm)
annually between 2006 and 2009. Ha and Va of the uncalloused areas were also measured with a slide gauge
(precision 0.1 mm) annually during the same period. In addition, two pruned trees were cut and split so that the area
of discoloured wood in relation with small-sized pruning wounds was measured with a specially built grid (squares of
0.25 sq. cm) in December 2012.
In sub-compartment 81B% the measurement of d.b.h. in 15 pruned trees was carried out in fall 2012. These trees
showing 20 pruning wounds were cut and split and the following parameters were either measured or observed: Ha,
Va, duration of wound closure, radial increment before and after wound closure, vertical length and area of
discoloured wood, presence of rot and/or insect traces, etc.
Pictures were taken on split sycamore wood samples with pruning wounds fully or partially closed as well as
discoloured wood in winter 2012.
The field data were compiled during the office work phase and the outputs of this project on wound closure, wood
discolouration, various correlations, etc., are shown below.
3. RESULTS
3.1. WOUND CLOSURE
In sub-compartment 75G, the 26 pruning wounds were closed after maximum 3 years as follows:
- 6 wounds (23%) with small Ha (0.5-1.2 cm) and Va (1.2-1.8 cm) in 1 year;
- 16 wounds (62%) with Ha of 0.7-2.8 cm and Va between 1.2 cm and 5.3 cm in 2 years. The majority of these
wounds was small and found on trees with low diameter increments (up to 4 mm/year); a much smaller proportion of
wounds was quite large (Va over 3 cm) and located on trees growing quickly in diameter (up to 15 mm/year);
- 4 wounds (15%) in 3 years. The latter wounds were either large (Va of 4.1 and 4.5 cm) and located on trees with
fast diameter increments (up to 15 mm/year) or small (both 1.2 cm Va) and found on trees with low diameter
125
International
Scientific
Conference on Hardwood Processing (ISCHP2013)
International Scientific Conference on Hardwood Processing
(ISCHP
2013)
0
0
0,5
1,5
2,5
3,5
4,5
5,5
Figure 1: Correlation between the length of vertical axis Va and duration of wound closure.
Figure 2: Pruning wounds closed after one year (left, both wounds) and three years (right, upper wound)
in sub-compartment 75G.
In sub-compartment 81B%, only 8 branches out of 20 (40%), with Va between 3.3 cm and 8.1 cm were closed
after 6 years.
The remaining 12 branches, with the Va between 3.3 cm and 11.5 cm in 2006, were only partially closed (from 29
to 90% of Va, the majority 9 out of 12 wounds - being closed at least 70% of Va).
3.2. WOOD DISCOLOURATION AND ROT
In sub-compartment 81B%, in case of both fully and partially closed wounds, columns of discouloured wood have
been noticed, being similar to the ones described by Soutrenon (1990, 1991a, 1991b). The length of these columns
spreading from the pruning cut ranges between 2.0 cm and 10.3 cm. This parameter is strongly correlated with the
vertical axis of pruning wound (r = 0.75) as shown in Figure 3.
The area of discoloured wood column spreading down from the same pruning position into the main stem ranges
between 3.5 sq. cm and 43 sq. cm. The area of discolouration is also strongly correlated (r = 0.81) with the Va of
pruning wound (Figures 4 and 5).
In addition, the presence of decay (rot) was noticed in 90% of both fully and partially closed wounds regardless their
size or location. This result is something expected to occur as the sycamore wood, generally considered as moderate
hard, with fine and uniform structure (Allegrini et al. 1998, Bastien 1997, Boudru 1989, Joyce et al. 1998, Kerr and
Evans 1993, Savill 1991, 2013, Waters 1992) is also regarded as quite sensitive to white rot (Anonymous 1983).
126
0
2
10
11
12
16
50
45
40
35
30
25
20
15
10
y = 4,1254x - 10,222
R2 = 0,6556
5
0
0
9 10 11 12 13
Figure 5: The smallest (left, 3.5 sq. cm) and the largest (right, 43 sq. cm) discolouration areas in sycamore wood.
127
International
Scientific
Conference on Hardwood Processing (ISCHP2013)
International Scientific Conference on Hardwood Processing
(ISCHP
2013)
- best period: before the beginning of growing season (end of winter-beginning of spring time). But, according to
Savill (2013), sycamore pruning should be done in summer as winter or spring pruning is followed by profuse
bleeding.
- the pruning cut should be clean and protect the branch swelling (use of natural target pruning method), without
leaving any stub or wounding the bark mostly below the cut. The pruning tools should be disinfected before
intervention.
- if also economically feasible and when pruning is performed on trees with quite slow diameter increments, larger
pruning wounds (with diameters over 3 cm) should be protected with specialised dressing substances.
- always maintain a crown length after pruning of minimum 50% of tree height to avoid losses of diameter increment.
LITERATURE CITED
Allegrini, Ch., R. Boistot-Paillard, J-Y. Bouvet, A. Depierre, and J-M. Mourey. 1998. The valuable broadleaved tree species in
Franche-Comt (Les feuillus prcieux en Franche-Comt). Socit Forestire de Franche-Comt, Thise. 29 pp. (in
French; no English abstract).
Allegrini, Ch. and A. Depierre. 2000. The selection of future crop trees (La slection des arbres davenir). Centre Rgional de la
Proprit Forestire, Office National de Forts, Thise et Besanon. 17 pp. (in French; no English abstract).
Anonymous 1983. Forest species for planting (Les essences de reboisement). Ecole Nationale du Gnie Rural, des Eaux et des
Forts, Nancy. 108 pp. (in French; no English abstract).
Bastien, Y. 1997. Silviculture of broadleaved tree species (Sylviculture des feuillus prcieux). Ecole Nationale du Gnie Rural,
des Eaux et des Forts, Nancy. 13 pp. (in French; no English abstract).
Becker, G. and J. Kldtke. 2009. Wood properties and wood processing of valuable broadleaved trees demonstrated with
common ash and maple in southwest Germany. In: Valuable broadleaved forests in Europe (ed. H. Spiecker, S. Hein, K.
Makkonen-Spiecker, and M. Thies), Brill, Leiden-Boston. pp. 161-167.
Boppe, L. 1889. Handbook of Silviculture (Trait de Sylviculture). Berger-Levrault et Cie, Libraires-diteurs, Paris et Nancy.
444 pp. (in French; no English abstract).
Boudru, M. 1989. Forest and silviculture: applied silviculture (Fort et sylviculture: sylviculture applique). Les presses
agronomiques de Gembloux, Gembloux. 248 pp. (in French; no English abstract).
Chiu, C.M., Lin, C.J., Tang, S.L., Tang, S.C., and C.N. Lo-Cho. 2004. Effects of different chemical dressings on wound areas of
pruned Taiwan Zelkova (Zelkova serrata Hay.). Taiwan Journal of Forest Science 19(3):177-186.
COFORD 2002. Silviculture of broadleaves. Silviculture and Forest Management no. 6, COFORD, Dublin. 6 pp.
Duyck, D. 2008. Silviculture of fast-growing broadleaved species (Sylviculture des feuillus croissance rapide). 5 pp.
http://www.foretpriveefrancaise.com/data/info/222917_Eclaircie_feuillus_divers.pdf. Accessed January 5, 2013. (in
French; no English abstract).
Evans, J. 1984. Silviculture of broadleaved woodland. Forestry Commission Bulletin 62, HMSO, London. 232 pp.
Fini, A., Ferrini, F., Frangi, P., Piatti, R., Faoro, M., Amoroso, G., Huylenbroeck, J. van and K. van Laere. 2013. Effect of
pruning time on growth, wound closure and physiology of sycamore maple (Acer pseudoplatanus L.). Acta Horticulturae
990:99-104.
Franc, A. and F. Ruchaud. 1996. Autecology of valuable broadleaved species: common ash, wild cherry, sycamore, Norway
maple (Autcologie des feuillus prcieux: frne commun, merisier, rable sycomore, rable plane). CEMAGREF
Editions, Imprimerie Louis Jean, Gap. 170 pp. (in French; no English abstract).
Haralamb, At. 1967. Culture of forest species (Cultura speciilor forestiere). Ediia a III-a. Editura Agro-Silvic. Bucureti. 755
pp. (in Romanian; no English abstract).
Hein, S. 2009. Modeling natural pruning of common ash, sycamore and wild cherry. In: Valuable broadleaved forests in Europe
(ed. H. Spiecker, S. Hein, K. Makkonen-Spiecker, and M. Thies), Brill, Leiden-Boston. pp. 103-122.
Hein, S. and H. Spiecker. 2007. Comparative analysis of occluded branch characteristics for Fraxinus excelsior and Acer
pseudoplatanus with natural and artificial pruning. Canadian Journal of Forest Research 37(8):1414-1426.
Hein, S. and H. Spiecker. 2009. Controlling diameter growth of common ash, sycamore and wild cherry. In: Valuable
broadleaved forests in Europe (ed. H. Spiecker, S. Hein, K. Makkonen-Spiecker, and M. Thies), Brill, Leiden-Boston. pp.
123-147.
Hein, S., C. Collet, C. Ammer, N. leGoff, J.P. Skovsgaard, and P. Savill. 2009. A review of growth and stand dynamics of Acer
pseudoplatanus L. in Europe: implications for silviculture. Forestry 82(4):361-385.
Hubert, M. 1980. Wild cherry, a tree for wood production (Le merisier, arbre bois). Institut pour le dveloppement forestier,
Paris. 56 pp. (in French; no English abstract).
Hubert, M. and R. Courraud. 1998. Pruning and reshaping of forest trees (Elagage et taille de formation des arbres forestiers).
Second edition. Institut pour dveloppement forestier, Paris. 303 pp. (in French; no English abstract).
Hugon, D. 1994. High pruning (Llagage a grande hauteur). Silva Belgica 1:35-40 (in French; no English abstract).
Joyce, P.M., J. Huss, R. McCarthy, A. Pfeifer, and E. Hendrik. 1998. Silvicultural guidelines for ash, sycamore, wild cherry,
128
129
International
Scientific
Conference on Hardwood Processing (ISCHP2013)
International Scientific Conference on Hardwood Processing
(ISCHP
2013)
Wardlaw, T.J. and W.A. Neilsen. 1999. Decay and other defects associated with pruned branches of Eucalyptus nitens. Tasforests
11:49-57.
Waters, T.L. 1992. The ecology and silviculture of sycamore. In: OFI Occasional Papers n. 41, Oxford Forestry Institute,
Oxford. pp. 8-12.
Wright, J.W., Hart, J.H. and W.A. Lemmien. 1980. Response of an eastern white pine provenance test to pruning. Forest Science
26(4):606-608.
130
CNR-IVALSA
National Research Council of Italy
via
Biasi 75, 38010, San Michele allAdige, Italy
ABSTRACT
Densification is a common process applied to wood in order to improve specific properties such as dimensional
stability, hardness and other mechanical properties. During densification, pressure and temperature are applied
leading to chemical and physical changes. Degradation of some chemical components, deformation of cell wall,
dislocation of molecules inside the tissues, colour change also take place. The quality and quantity of such changes
depend on pressure, wood species, moisture content, temperature, time of treatment, among others. The goal of this
work was to employ NMR spectroscopy for non-destructive and deep characterization of changes of wood due to
densification process. Beech (Fagus silvatica), ash (Fraxinus excelsior L. ) and poplar (Populus tremula L.) were
selected as representative wood species for investigation. The densification was performed in a hydraulic press
equipped with hot plates. The densification cycle was composed of pre-heating (plasticization), densification, post
treatment and cooling. The hot press temperature was set to 180C. 13C solid state NMR analyses were carried out
and NMR spectra were acquired with cross-polarization (CP) pulse sequence. Cellulose changes slightly in term of
increase crystalline component with respect to amorphous cellulose. As cellulose, hemicellulose is slightly affected
by densification since the shoulder at 103ppm doesnt decrease. Increase of carbonyl assigned peaks at 172ppm
and 21ppm, commonly due to hemicellulose, instead occurred due to lignin new complex formation. Moreover
lignin undergoes to -O-4 bond cleavage in syringyl units process and at the same time condensation reaction of
guaiacyl units. Densification affects chemical component of: ash, poplar and beech in descending order. In all
cases densification may be attributable to a mild heat treatment and as consequence, desirable for densified wood,
few loss of mechanical properties could be expected.
1. INTRODUCTION
Wood densification is a traditional process applied to wood in order to improve selected properties, such as density,
hardness and mechanical resistance. Pressure and temperature are applied during densification leading to chemical and
physical changes of wood. An apparent colour change, reduction of the sample dimensions and in consequence an
increase in density are evident effects of wood densification. Deformation of cell wall, dislocation of molecules inside
the tissues and degradation of some chemical components is also taking place. The extent of such changes depends on
the grade of densification mostly based on dimensional deformation, wood species, moisture content, temperature, time
of treatment, among others. The cell-wall component is made of cellulose and hemicellulose, which constitute the
* Corresponding author: Tel.: (+39) 0461-660232; Fax: (+39) 0461-650045; E-mail: santoni@ivalsa.cnr.it
131
carbohydrate fraction, and lignin formed by phenyl propane units. Due to different chemical features, these components
react to treatment in different ways. The densification process includes a heat treatment, during which cell wall
components undergo to a series of autocatalytic reaction. During heating, the first released volatiles are organic acids
like acetic acid, formed by the cleavage of acetyl group in hemicellulose (Dietrichs et al. 1978; Bourgois and Guyonnet
1988, Fengel and Wegener 1989). Acid catalysed cellulose and hemicellulose depolymerisation produces monomeric
and oligomeric sugar units (Carrasco and Roy 1992). More reactive chemical species could participate to further
reactions producing aldehyde as furfural from pentose or hydroxymethylfurfural from hexose sugars (Ellis and Paszner
1994, Kaar et al. 1991). Even if lignin is considered more stable during heating some modifications also occur.
Demethoxylation reaction with -O-4 cleavage results in higher concentration of phenolic groups (Fengel and Wegener
1989). The increase of lignin reactivity leads to various condensation reactions between lignin units and other
molecules, such as those resulting from degradation processes.
13
C NMR spectroscopy has been extensively used in the past (Melkior et al. 2012, Maunu 2002, Bardet et al. 2009,
Bardet et al. 1997, Newman and Hemmingson 1995) for the structural characterization of natural wood and for
highlighting the structural modifications due to different types of processing. In particular solid state 13C NMR is a
powerful local probe for analysing wood structures while avoiding any sample manipulation. The goal of this work was
to exploit NMR spectroscopy as a unique analytical tool for characterizing wood chemical changes derived from the
densification process in an open system.
2. MATERIALS AND METHOD
Wood samples of three hardwood species: European beech (Fagus sylvatica L.), European aspen (Populus tremula
L.) and European ash (Fraxinus excelsior L.) were used to produce the densified wood. Untreated wood samples of the
same species were also analysed for comparison.
All boards selected for densification were without visible defects: such as knots, splits, reaction wood and pith. The
initial moisture content (MC) was approximately 12% prior to the densification process. Densification was performed
at the CNR-IVALSA laboratories, in an open system, using a multi-platen hot-press. The final cooling of the densified
samples was performed out of the press, assuring limitation of the sample flatness changes, by applying clamping. In
general, after densification, wood is post-treated to prevent it from recovering its original dimension after absorption of
water. The densification cycle follows the literature reference (Navi and Sandberg 2012) and was composed of:
pre-heating (plasticization) 10 min; densification 10 min; post treatment 5 min and cooling 5 min. The hot press
temperature was set to 180C.
Solid state NMR analyses were carried out at the Klaus Mller NMR Laboratory with a Bruker 300WB
instrument operating at a proton frequency of 300.13 MHz. Spectra were acquired with cross-polarization (CP) pulse
sequence under the following conditions: 13C frequency: 100.07 MHz, /2 pulse length: 3.5s, 1H decoupling pulse
power: 47 kHz, recycle delay: 3 s, contact time: 0.2 ms, 5k scans. The wood samples were packed in 4 mm zirconia
rotors, which were spun at 8 kHz under air flow. Adamantantane was used as external secondary reference. The
changes in peak intensity were referenced to the cellulose C-1 signal at 105 ppm (Sivonen et al. 2002).
3. RESULTS AND DISCUSSION
The 13C CPMAS NMR spectra of wood samples allow the identification of the main chemical components by
assignment of signals due to their functional groups. Besides the relevant signal overlapping, cellulose, lignin and
hemicellulose can be recognized without uncertainty from the chemical shifts of different signals, as shown in Table 1.
The peaks of cellulose are found in the range 60 -105ppm. Peaks at 84 and 89ppm are assigned to C4 of amorphous
and ordered cellulose, respectively (Bardet et al. 2009). The same assignment is valid for the signals at 65 and 62ppm,
respectively, due to C6 in ordered and disordered cellulose. C2,3,5 carbon atoms of cellulose give rise to the signals at
72 and 75ppm with some overlapping due to C of lignin. C1 produces a single sharp peak at 105 ppm.
The region between 105 and 160ppm is characterized by the typical weak and broad signals due to the aromatic
carbon atoms of lignin. The peaks at 153 and 147ppm are assigned to C3,5 atoms of lignin syringyl (S) unit, etherified
and non-etherified, respectively. Moreover, the signal at 147 ppm accounts also for C3 and C4 of lignin guaiacyl (G)
units. The complex signal between 130 and 140ppm is assigned to etherified and non-etherified C1 and C4 of S and C1
of G units. The signal of lignin methoxyl groups is found at 56 ppm.
132
Hemicellulose peaks are overlapped to those of cellulose, with the exception of two characteristic peaks at 21 (CH3
COO ) and 172ppm ( CH3 COO ) assigned to the acetyl groups.
Table 1: 13C CPMAS resonance assignments of wood (Bardet et al. 2009, Wikberg an Maunu 2004)
Chemical
shift (ppm)
172
Main assignments
-COO-R, CH3-COO-
152
Carbohydrates;
(hemicellulose)
Lignin;
147
Lignin;
140-130
Lignin;
120106
Lignin;
105
Carbohydrates;
C1
88
Carbohydrates;
C4 crystalline cellulose
83
Carbohydrates;
74
Carbohydrates;
C2,3,5; lignin C
72
Carbohydrates;
C2,3,5
64
Carbohydrates;
C6
61
Lignin;
56
Lignin;
OCH3
21
Carbohydrates;
(hemicellulose)
CH3-COO-
S C3 (e), S C5 (e)
The 13C CPMAS NMR spectra, recorded on both natural and densified wood samples belonging to the three
selected species are shown in Figures 2-4. The main changes due to the densification process are highlighted in the
figures.
Figure 2 presents the spectra of untreated and densified ash samples. The main differences after densification can be
identified in the intensity increase of signals due to acetyl groups (172 and 21ppm) commonly assigned to
hemicellulose and of the syringyl peaks at 153 ppm and 147ppm. The change in intensity ratio of the main signals at 74
and 72ppm (C2,3,5) is also well visible.
133
Figure 2: 13C CPMAS NMR spectra of ash. Ash not treated black, densified grey.
Figure 3 shows the NMR results obtained for poplar. The densified sample shows a behaviour similar to that
described in the case of ash with the relevant difference observed for the peaks between 160 and 130ppm due to
syringyl and guaiacyl lignin units, which present an unchanged intensity.
Figure 3: 13C CPMAS NMR spectra of poplar: Poplar not treated black, densified grey.
134
The different behaviour of beech wood in comparison with ash and poplar woods can be deduced from spectra
reported in Figure 4. In this last case, the densification process leads only to negligible changes, which affect the signals
at 153 and 147ppm (S and G groups of lignin) and at 74ppm (cellulose). It is noteworthy that the acetyl related signals
(172 and 21ppm) do not show any intensity variation after treatment
Figure 3: 13C CPMAS NMR spectra of beech. Beech not treated black, densified grey.
As previously reported (Tjeerdsma et al. 1998a, Tjeerdsma et al. 2005), cellulose is the wood chemical component
least affected after a thermal treatment. A similar behaviour is observed in our study on the effects of the densification
process of wood. The carbohydrate most intense peaks are present in the range between 65 and 105 ppm and account
for both disordered amorphous and ordered crystalline cellulose species. For all analysed wood species, changes are
observed in this spectral region, probably indicating a rearrangement or degradation of the amorphous cellulose content
in the samples after treatment. This observation is in agreement with reported results on thermal treatment experiments
(Wikberg and Maunu 2004) suggesting that amorphous cellulose is the most subjected to degradation processes, which
are probably catalysed by acetic acid emission during heating.
It should be also mentioned here that even if the samples were subjected to the heat treatment, peaks assigned to
aldehydes around 190 ppm are not present in NMR spectra. In fact it is known that during a thermal treatment the
chemical reactions lead to carbohydrates degradation with production of aldehydes, as furfuraldehyde whose presence
has been assessed in many studies regarding the thermal treatment of wood (Tjeerdsma et al. 1998a). In our case, the
densification treatment probably has been run at temperature lower than that required for sugar degradation.
The hemicellulose carbohydrate units appear slightly affected by the densification treatment, as previously
discussed for cellulose. Moreover, 13C NMR spectra show that the peaks commonly assigned to hemicellulose acetyl
groups at 172 and 21ppm increase in contrast with reported results on hemicellulose degradation. Indeed, the
carboxylic groups cleavage is known to occur at high temperature and humidity conditions (Tjeerdsma and Militz
2005). On the contrary, an increase of acetyl signals was already reported and attributed not to hemicellulose but to
esterification reactions leading to new lignin complexes since carboxylic groups were detected in lignin fraction (Diouf
et al. 2011, Tjeerdsma and Militz 2005 ). Another characteristic assignment related to hemicelluloses C1 is the small
shoulder at 103 ppm on cellulose signal at 105 ppm. No change of the peak profile was observed in all spectra,
confirming the limited chemical modification of the carbohydrate component.
.
135
In the aromatic region of the 13C spectra, a general intensity increase was detected after densification probably caused
by lignin rearrangement and condensation reactions. These changes are stronger for ash and almost negligible for
poplar.
The spectra of all selected hardwood show signal at 153 ppm assigned to C3,5 syringyl units etherified and peak at
148 ppm of the same carbon not etherified. After densification an increase of both peaks is observed and at the same
time a reduction of relative intensity between signals at 153 ppm and 148 ppm in all species is present. The ratio
between the signals at 153 and 148 ppm in treated and untreated wood spectra has been used to measure the degree of
cleavage of -O-4 syringyl linkages (Wikberg and Maunu 2004). The depolymerization of lignin due of ether bond
creates new reactive sites as free hydroxyl phenolic and carbonyl groups in the propane chain of lignin C or C. This
consideration is in agreement with the observed increase of carbonyl assigned peak at 172 ppm and with signal at 21
assigned to methyl of acetyl group. Moreover the increase of peak at 153 ppm could be assigned to condensation of
guaiacyl units by the formation of carbon-carbon bond. This bond is characteristic of guaiacyl units since syringyl has
C3,5 already occupied by methoxyl group and has been reported in the literature for hardwood (Wikberg and Maunu
2004, Diouf et al. 2011, Tjeerdsma and Militz 2005, (Liitia et al 2002).
4. CONCLUSION
Densification is an important process to increase particular mechanical properties. During densification process,
modification of wood components take place and could be related to mechanical properties. Wood chemistry
modifications of treated beech, ash and poplar wood were observed with support of 13C CPMAS NMR spectroscopy.
Carbohydrate changes were observed in terms of increase in crystalline component with respect to amorphous
cellulose. Hemicellulose seems not to be affected by densification since the shoulder at 103 ppm doesnt decrease.
Increase of carbonyl assigned peaks at 172 ppm and 21 ppm occured due to lignin bond cleavage. Lignin beside
carbonyl formation undergoes the process of cleavage of -O-4 bond of syringyl units and at the same time
condensation reaction of guaiacyl units. All these processes were similar to a mild thermal treatment since increase of
carbonyl peak was found at the initial stage of heat treatment, whereat aldehyde peak were not present. Beech seems to
be the most resistant to chemical changes due to densification then poplar and ash however in all species densification
may be attributable to a mild heat treatment and to a desirable few loss of mechanical properties.
ACKNOWLEDGMENTS
Part of this work has been conducted within the framework of project SWORFISH (team 2009 incoming (CALL 2)
and TrentinoPCOFUND-GA-2008-226070) co-financed by Autonomous Province of Trento and sub-project B2:
Connectors - wooden nails of project SOFIE 2: Research for sustainable construction, manufacturing and wood
characteristics of Trentino region, funded by Provincia Autonoma di Trento.
REFERENCES
Bardet M., G. Gerbaud, M. Giffard, C. Doan, S. Hediger and L. Le Pape. 2009. 13C highresolution solid-state NMR for structural
elucidation of archaeological woods. Prog. Nucl. Magn. Res. Spectrosc. 55:199214.
Bardet M., L. Emsley, M. Vincendon.1997. Two-dimensional spin-exchange solid-state NMR studies of C-13-enriched wood. Solid
State Nucl. Magn. Reson.8:2532.
Bourgois J. and R Guyonnet .1988. Characterization and analysis of torrified wood. Wood Sci Technol 22:143155.
Carrasco F. and Roy C..1992. Kinetic study of dilute-acid prehydrolysis of xylan-containing biomass. Wood. Sci. Technol.
26:189208.
Dietrichs H.H., H. Sinner, J. Puls. 1978. Potential of steaming hardwoods and straw for feed and food production. Holzforschung
32:193199.
Diouf P. N., T. Stevanovicb, A. Cloutierb, C. Fangb, P. Blanchetc, A. Koubaad, N. Mariotti. 2011. Effects of
thermo-hygro-mechanical densification on the surface characteristics of trembling aspen and hybrid poplar wood veneerset.
Applied Surface Science 257: 35583564
Ellis S. and L. Paszner. 1994. Activated selfbonding of wood and agricultural residues. Holzforschung 48:8290
Fengel, D., and G. Wegener. 1989. Wood chemistry, ultrastucture, reactions. New York: De Gruyter.
136
Kaar W. E., L. G. Cool, M. M. Merriman, D. L. Brink. 1991. The complete analysis of wood polysaccharides using HPLC. J Wood
Chem Technol 11:447463
Liitia, T., S. L. Maunu, J. Sipila and B. Hortling. 2002. Application of solid-state 13C NMR spectroscopy and dipolar dephasing
technique to determine the extent of condensation in technical lignins. Solid State Nuclear Magnetic Resonance, 21:
171186.
Mannu SL.. 2002. NMR studies of wood and wood products. Prog. Nucl. Magn. Reson. Spectrosc. 40:15174.
Melkior T., S. Jacob, G. Gerbaud, S. Hediger, L. Le Pape, L. Bonnefois, M. Bardet.2012. NMR analysis of the transformation of
wood constituents by torrefaction. Fuel 92: 271280.
Newman, R. H. and J. A. Hemmingson. 1995. Carbon-13 NMR distinction between categories of molecular order and disorder in
cellulose. Cellulose, 2: 95110.
Navi P. and Sandberg D. 2012. Thermo-hygro-mechanical processing of wood. M. Rappaz (ed.). EFPL Press, Lausanne,
Switzerland. 360p.
Sivonen, H., S. L. Maunu, F. Sundholm, S. Jamsa and P. Viitaniemi.2002. Magnetic resonance studies of thermally modified wood.
Holzforschung, 56: 648654.
Tjeerdsma B.F.and H. Militz. 2005. Chemical changes in hydrothermal treated wood: FTIR analysis of combined hydrothermal
and dry heat-treated wood, Holz Roh Werkst. 63: 102111.
Tjeerdsma, B. F., M. Boonstra, A. Pizzi, P. Tekely, and H. Militz. 1998. Characterisation of thermally modified wood: Molecular
reasons for wood performance improvement. Holz als Roh- und Werkst, 56: 149153
Wikberg H. and S. L. Maunu. 2004. Characterisation of thermally modified hard- and softwoods by 13C CPMAS NMR.
Carbohydrate Polymers 58: 461466
137
Session II
Hardwood processing
and optimization
139
ABSTRACT
One of the important problems in hardwood industry is the understanding of the wood structure prior to any
mechanical processing. Computer tomography (CT) has the potential to see inside a tree by providing
volumetric information of the internal structure without its damage.
The mission of the Hardwood Scanning Center at Purdue University is to increase the global competitiveness of the
United States hardwood industry and to conserve the hardwood resource by development of technologies which
will enable hardwood industry to see inside a tree and use this information to make better processing decisions.
The Hardwood Scanning Center at Purdue University partnered with MiCROTEC, GmbH of Italy in the
development of an industrial grade log CT scanner. First three industrial CT log scanners have been installed
in 2012 in mills around the world and this paper discusses their applications. The Hardwood Scanning Center
developed visualization and optimization software for the hardwood veneer and sawmill operations. This paper
provides an overview of the solution for computer tomography log scanning.
140
Figure 1. Photo of CT.Log installed in a veneer mill log yard (Image courtesy of MiCROTEC, GmbH)
141
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
Figure 2. a) Gantry of CT.Log, b) Image of a log scanned with CT.Log and software-recognized and classified interior features,
and c) CT.Log system with enclosure, conveyer and shielding (Images courtesy of MiCROTEC, GmbH).
142
a)
b)
Figure 3. a) A 3-D realistic visualization of a CT scanned log and b) results of automatic knot detection.
3. INDUSTRIAL IMPLEMENTATION
HARDWOOD VENEER MILL
Traditional log analysis is performed at the inspection yard (Figure 4), where the human inspector carefully
analyzes all external visual characteristics of each log. Based on these observations and the experience of the inspector,
the suggested orientation for cutting the log is marked. The log is then cut into two flitches and each flitch is sliced into
veneer. If the initial cutting orientation was not optimal, the overall value of the resulting veneer can be significantly
decreased.
143
HARDWOOD SAWMILL
The current state-of-the-art grade sawing process in US hardwood sawmill employs 3-D laser scanning of log
surface. This process was adapted from softwood sawmills, where knowing the exact volumetric shape of a log enables
selection of a sawing pattern and positioning of a log that achieves the maximum volume of resulting lumber. In
hardwood sawmill, such volume scanning coupled with computerized log carriage enables a sawyer to laterally position
the log on a carriage while at the same time advancing the log to the saw by a distance that will result in best opening
face. Positioning the log on the carriage in the appropriate orientation, however, as well as all subsequent cuts are left
to the sawyers experience.
144
145
Year
1962
1967
1969
1975
1980
1989
1994
'94-'99
Author
Peter
Peter
Tsolakides
Wagner and Taylor
Richards et al.
Steele et al.
Steele et al.
Chang and Guddanti
# of Logs
10
50
6
10
320
24
6
10
Species
Southern Pine
Yellow Poplar
Red Oak
Southern Pine
Red Oak
Red Oak
Red Oak
Red Oak
%
Improvement
3%
9%
21%
8%
11%
12%
10%
18% -28%
More recently, Gazo and Chang (2010) conducted a proof of concept study based on 60 logs of three log grades and
species of yellow poplar, black cherry, hard maple, white and red oak, scanned on a medical CT scanner and processed
in an actual sawmill. This study (the largest and the most comprehensive to date) shows a potential value improvement
of 46% overall, across all log grades and species.
Now the industrial-grade technology exists to bring the concept of CT scanning of logs and processing optimization
that accounts for internal log defects from the laboratory to the actual practice. The CT.Log scanner and the
LogView software have only been available commercially for less than a year. Two of the first three installations are
in softwood mills and the data from the third system are not publically available at this time. As more systems are
installed over time and data will become available, we will be in better position to analyze and report actual value
improvements in various applications. The proposed technology has the potential to enable significant breakthrough in
many tree and log breakdown operations.
REFERENCES
Benson-Cooper, D.M., R.L. Knowles, F.J. Thompson, and D.J. Cown. 1982. Computed tomographic scanning for the detection of
defects within logs. Forest Research Institute, New Zealand Forest Service, Rotorua, NZ., Bull. No. 8. 9p.
Bhandarkar, S., T.D. Faust, and M. Tang. 2002. Design and prototype development of a computer vision-based limber
production planning system. Image and Vision Computing 20(3):167-189.
Birkeland, R. and S. Holoyen. 1987. Industrial methods for internal scanning of log defects: a progress report on an ongoing
project in Norway. In: R. Szymani (Ed) 2nd International Conference on Scanning Technology in Sawmilling, October 1-2,
Oakland/Berkeley Hills, CA. Forest Industries/ World Wood, San Francisco, CA.
Burgess, A.E. 1985. Potential applications of medical imaging techniques to wood products. In: R. Szymani (Ed) 1st
International Conference on Scanning Technology in Sawmilling, October 5-6, San Francisco, CA. Forest Industries/World
Wood. San Francisco CA.
Chang, S.J. 1992. External and internal defect detection to optimize cutting of hardwood logs and lumber. Transferring
Technologies for Industry No. 3, USDA National Agricultural Library. Beltsville, MD. 24p.
Chang, S.J., P.C. Wang, and J.R. Olson. 1987. Nuclear magnetic resonance imaging of hardwood logs. In: Proceeding of the
2nd International Conference on Scanning Technology for Sawmilling. Miller-Freeman Publ. San Francisco, CA. pp. IX 1-8.
146
147
148
149
International
Scientific Conference on Hardwood Processing (ISCHP 2013)
150
151
152
153
154
155
156
Holzforschung
Technische Universitt Mnchen
Mnchen, 80797, Germany
ABSTRACT
Economically relevant tropical wood species from Africa such as Azobe, Bilinga, Tali, Okan and from South
America Cumaru, Purpleheart, Sapupira contain irregularly but sometimes considerable amounts of brittleheart.
Brittleheart is a well-known phenomenon in many tropical wood species and reduces the strength properties
considerably. According to national strength grading standards such as BS 5756 and NEN 5493 timber with
brittleheart has to be rejected, however in many cases brittleheart is not visible on the outer surface and cross
section of sawn timber. This investigation has been focused on qualitative characterisation of brittleheart in
tropical hardwood in conjunction with non-destructive detection methods. For the investigation specimens only
from full size structural members were used. To determine the morphological characteristics of brittleheart on the
timber, a light microscope and an environmental scanning electron microscope (ESEM) were employed. Five
grades of brittleheart have been characterised by deformation and displacement of wood tissue. Typically
brittleheart covers an area/volume in the wood. In order to follow the propagation of the branched brittleheart
lines and possible fissures in that area, 3D tomographic scanning has been performed. Results show that
CT-scanning is a promising technique for detecting the defined grades. However, limitation of detection is due to
the resolution of the scanners and contrast ability of imaging which has also been effected by the fibre structure of
the wood species. Quantitative studies of brittleheart on mechanical properties are continuing.
1. INTRODUCTION
Timber intended for structural applications should be free of brittleheart. The term brittleheart is mainly used in
wood industry and timber trade, and is related to defects in the wood structure of tropical hardwood caused by
compression action. Brittleheart can be recognized as microscopic failure lines up to macroscopic failure areas of
several millimetres in width which may affect a great part of the stems cross section. It is important to detect
brittleheart when timber is strength graded and assigned to a strength class. In Europe, the assignment to strength
classes is based on EN 14081 and is determined by visual and/or machine strength grading methods. Certain national
standards reject the presence of brittleheart when assigning timber to a strength class, such as the Dutch standard NEN
5493 and the British standard BS 5756. Brittleheart reduces the timber strength and often shows brittle failure when
loaded in bending resulting in fails without any prior warning. On the rough sawn surfaces of tropical hardwood
brittleheart sometimes are very difficult if not impossible to detect with the naked eye. In the most unfavourable
situation it cannot be detected anyway, even after the final processing step.
Non-destructive methods to detect brittleheart do not exist yet on industrial level. Until now, the only way to detect
it is by means of visual inspection. However, this manner is limited to detection of clearly visible brittleheart, in most
cases on planed surfaces. A reliable detection method and methodology, for both less visible and not visible brittleheart
to the naked eye, are required in order to fulfil the requirements of the strength grading rules.
The purpose of this research is to investigate morphological characteristics of different levels of brittleheart (LB)
157
and identify non-destructive detection techniques to be used for different LB. Finally these characteristics will be
related to the bending strength properties of selected tropical hardwoods containing different LB. This paper presents a
part of the on-going investigation. It describes the morphological characteristics of the LB in tropical hardwood with
the aid of non-destructive detection methods based on X-rays. In order to identify the different LB, first the test
specimens of different hardwood species are characterised by visual inspection, CT-scanning (3D), and microscopy.
2. STATE OF THE ART
Research on brittleheart in wood has been reported for more than seventy years (Dadswell and Langlands 1938,
Arnold and Steiger 2006). Many studies regarding brittleheart or compression failures and their influence on timber
strength properties, accompanied by quantitative analysis, have been conducted on softwood, but less on tropical
hardwood. Different terms were used to describe the phenomena of brittleheart. Researchers who investigated
softwood in this respect, mainly have used the term compression failure (Dinwoodie 1968, Hoffmeyer 2003, Wilkins
1968, U.S. Department of Agriculture 1961). For hardwood species only a limited number of researches on the
morphology of LB have been conducted (Dadswell and Langlands 1938, Yang 2001, Kuera and Bariska 1982,
Scurfield et al 1972). These authors prefer the term brittleheart.
Arnold and Steiger (2006) define compression failures in conifers as defects in the wood structure in the form of
buckled cell walls of the wood fibres. They state these compression failures as a well-known natural phenomenon
and as complex three-dimensional geometric structures with more or less fuzzy boundaries which appear in a broad
range of intensities. The appearance of brittleheart affecting the stems cross section of hardwood species, is also
mentioned by Hillis et al. (1973). In their investigation on young and mature karri, Eucalyptus diversicolor, it was
found that brittleheart can extend to the full length of the merchantable trunk. They found that 20% of the mature trees
were affected by brittleheart.
Different experts have characterized visible LB on board surfaces of softwood species in two ways. One way is by
measuring the width of the failure lines (Behzad 2007), and the other way is by visually assigning LB (Hoffmeyer
2003). Ander and Nyholm (2000) present a nomenclature of some of the different types of deformations and their
characteristics according to different authors. In addition, Hoffmeyer (1990) mentions definitions and characteristics
for compression failures in softwood, which is based on the microscopic examination by Dinwoodie (1968). The terms
provided by Hoffmeyer (1990) give the most logic sequence for the appearance of compression failures (LB). A
simplification of the most important definitions linked to their morphology is as follows:
- slip plane:
a zone of re-orientation of microfibrils;
- microscopic LB:
2 to 5 slip planes in depth covering more than 2 cell walls in width;
- macroscopic LB:
more than 5 slip planes in depth and several microscopic LB are combined; appears
also as a local stability failure and moderately visible to the naked eye;
- gross stability LB:
clearly visible LB to the naked eye.
With regard to visual grading of timber, LB are usually difficult to detect, particularly on rough sawn timber, as
stated by Arnold and Steiger (2006). By planing the surfaces it is possible to detect some LB by visual inspection, but
only the bigger damages can be detected.
Special efforts, such as optimum lighting or light wetting of the surface may be helpful for a better detection of LB
with the naked eye. However, sometimes LB could only be detected with microscopic investigation because the failure
lines are too fine. Potential techniques like the ultrasound method and the radiation method, Computerized
Tomography (CT) using X-rays, were investigated for softwood by several researchers such as Brancheriau and Lanvin
(2008). It appears that the ultrasonic signals and the detection of the small density differences by X-rays were not
distinctive enough to recognize the fine lines (Wallace 1993). Sonderegger and Niemz (2004) concluded that the ability
of CT and scanner technology for LB detection strongly depends on the resolution of the equipment used. They find
that CT could easily detect the larger LB and that higher resolutions are needed to detect the lower LB. However, from
this literature study CT scanning so far has not been used for detection of LB in hardwoods.
3. MATERIAL AND METHODS
3.1. MATERIAL
Four tropical hardwood species were selected for this research. The selection was based on the number of samples
with brittleheart available for each timber species, and on the market share of the species. The collected timber was
158
provided by five Dutch wood companies who import these species from Africa, Asia and South-America. The collected
timber was visually graded according to NEN 5493 and rejected because of the presence of brittleheart.
The four selected tropical hardwood species are presented in Table 1. The results of sapeli and sapupira are reported
in this paper, which are representative for the four species used, in the sense that observed LB in these two species are
comparable in size and magnitude to those observed in the remaining two species.
Table 1: Selected hardwood species used for research (FSC certified)
Trade name
Scientific name
Continent of origin
Azob
Africa
Louro faia
South-America
Sapeli
Africa
Sapupira
Hymenolobium spp
South-America
The dimensions of the samples prepared for the CT-scanner Nanotom (further described in section 3.2) were 20 x
20 x 100mm3. These samples were also used for the morphological characterisation of the LB. The samples for the
other CT-scanner Somatom (see section 3.2) had dimensions of 72 x 122 x 600mm3 for both sapeli and sapupira. The
specimens used for the Somatom were conditioned at a temperature of 20C and a relative humidity of 65%, whereas
the samples for the Nanotom were kept in the room with a temperature of around 20C and a relative humidity of
approximately 40%.
3.2. METHODS
Visual characterisation
Since most of the LB of the four tropical hardwood species appeared as branched failure lines on the timber
surfaces, the method of visually assigning grades to LB was preferred and applied. The visual assessment of the visible
LB was based on dry planed timber surfaces having moisture contents less than 15%.
Six grades are used for the visual assessment of the LB. These grades are based on Hoffmeyers investigation on
LB in Norway spruce (Hoffmeyer 2003). The specimens used for his research were conditioned at a temperature of
20C and a relative humidity of 65%. The grades formulated by Hoffmeyer for dry timber surfaces are as follows:
-
CT- scanning
159
Before scanning the samples with the CT-scanners, samples of each timber species with LB are visually
characterized from grade 0 up to grade 5. The samples for the Nanotom were scanned with a resolution of 30m x 30m
x 30m over a length of 30mm, and the samples of the Somatom were scanned with a lower resolution of 0,3mm x
0,3mm x 1,0mm, but over a total length of 200mm. A beam hardening correction of 7,0 was used for the Nanotom
samples, and the timing used by the detector to process a single frame was 0,5 seconds. The samples for the Somatom
are scanned with the spiral scan mode and with a slice distance of 1,0mm. Settings for voltage, current and filters
determine the contrast of the scan images obtained. In our research a low density is represented by a black colour,
whereas a high density is represented by a white colour.
Microscopic characterisation
To determine the morphological characteristics of the LB on the timber, a light microscope and an environmental
scanning electron microscope (ESEM) are used. The magnifications used for the light microscope were in the range
from 6,3x up to 50x and for the ESEM from 100x up to 2000x. The ESEM is used for the morphological characteristics
of the LB on wood cell level.
4. RESULTS
4.1. MORPHOLOGICAL CHARACTERISTICS
Visual characteristics
Different levels of brittleheart called grades on a dry and wet surface are illustrated in Figure 2 and 3 (identical
samples) for sapeli, and in Figure 4 and 5 for sapupira (identical samples). These figures also show the development of
LB within one damaged area. If the surface of the sample is slightly moistened with water (Figure 3), the development
and branching of the LB become more expressed, compared to dry surfaces (Figure 2). As shown in Figure 3 and Figure
5 (at the arrows), even grade 1 become visible. The appearance of the LB in Figure 2 to Figure 5 are representative for
all wood species examined. The development of grades within one damaged zone was an important characteristic for all
four species. The occurrence of only one grade on the timber surface was hardly noticed.
Microscopic characteristics
The LB are also microscopically investigated. The anatomical structure for different grades on the longitudinal
sections is presented in Figure 4 and 5. These pictures are obtained by the light microscope with a magnification of
50x. Figure 4 shows grade 1 (image A) and grade 4 (image B) on a tangential section of a sapeli sample, and Figure 5
for grade 4 and 5 on a tangential section of a sapupira sample. These figures illustrate localized buckling of the wood
cells in the tangential direction for the large LB (grade 4 in Figure 4 and 5).
160
The buckling shape is either a horizontal displacement or a combination of horizontal and vertical displacements.
The morphological structure of the fine lines grade 1 are barely visible and showing no distinct fibre buckling (Figure
4 (A)). These low LB appear as fine lines on the wood surface. Grade 5 is characterised by broken fibres (Figure 5).
Often the brittleheart on the radial sections appear as straight lines or compressed areas perpendicular to the fibre
direction. This orientation of the brittleheart applies to both the high and low LB.
The morphological characteristics of the LB can be observed and analysed in more detail using the ESEM. In
Figure 6 large buckling and compression of the wood fibres is clearly visible on the tangential plane section (see
arrow). The picture presented applies to grade 4 on a sapeli tangential plane.
A horizontal displacement of fibre and ray cells can also be observed on the sapupira tangential plane, see arrows in
Figure 7. As before, grade 5 is characterised by broken fibres,
as shown
in Figure
8 (see
arrows).Processing (ISCHP2013)
International
Scientific
Conference
on Hardwood
Figure 9 shows a typical example of grade 1 (see white arrow), and even smaller (see black arrow). Here also fibre
buckling is seen for these failures, but in a less pronounced manner.
161
Somatom
Figure 13 presents sectional images (A, B, and C) and a 3D image (image D) of a sapeli beam scanned with the
Somatom. The white zones representing the LB grade 4, show a clear contrast with the surrounding wood tissue. Image
162
D illustrates a 3D view of the beam for only the highest densities. This 3D view shows the geometry of the LB existing
in the beam. The white zones are pointed out by the red cross in the four images (A to D).
Figure 13: Sectional (A,B,C) and 3D image (D) of a sapeli beam scanned with the Somatom red cross positioned
in white zones represents brittleheart lines
5. DISCUSSION
5.1. MORPHOLOGICAL CHARACTERISTICS
Mechanical properties such as bending and tensile strength are about forty times higher in the direction parallel to
the fibres, than perpendicular to the fibre direction. Therefore, the orientation of the fibres in the brittelheart zones may
influence the strength properties considerably. The orientation and the condition of the fibres have been chosen as a
microscopic characteristic for the different LB (grades).
On the light microscope a maximum magnification of 50x has been used. On this level different types of appearance
for the microscopic characteristics of the LB have been distinguished on the tangential and radial planes. Grade 1 still
appears as a line on the tangential and radial planes, while the grade 2 to 4 show buckling of the wood cells. On the
radial plane dark lines or areas of highly compressed wood tissue has been observed for the grades 2 to 4. Grade 5 has
been characterised as broken fibres on both planes.
The fine failures grade 1 show slightly horizontal displacements of the fibres when higher magnifications are used,
such as 125x obtained by the ESEM (Figure 9). Larger deformations and buckling of the wood tissue has been observed
for the grades 2 to 4. The difference between the grades depends on the volume of deformed and displaced fibres, and
the obtained strain of the horizontal displacement. Due to the complexity of fibre orientation and buckling, it is not
clear which stress component leads to the rupture of fibres (grade 5). It is possible that the increased longitudinal
compressive stresses lead to tensile stress components on the changed fibre direction, or to tensile stresses
perpendicular to the original fibre axis.
5.2. CT-SCANNING DETECTIONS
3D CT-scanning should have two advantages. Firstly, to be able to follow the propagation of the LB inside the
beam. Secondly, to be able to use this technique for the non-destructive detection of LB on an industrial level.
The most interesting finding is the limit of detection of the Nanotom, which was able to detect all grades, whereas
the Somatom was sensitive to detect some grades depending on the wood species. The differences of capability of LB
detection of both Nanotom and Somatom are caused by the voxel resolution and the contrast of the scan images. The
contrast can also be influenced by the diffuse density distribution within the specimen and distribution of high density
extractives, such as resins or crystals. The higher density counts for LB having a visually assigned grade up to grade 4.
Since a crack is detected as open space by the CT-scanners, a low density is assigned to this area represented by black
zones.
Besides the limits of detection for LB grades, also detection limits exist with regard to the sizes of the samples. For
the examination on the Nanotom, the sample sizes are restricted to 100 x 100 x 290 mm3, while the Somatom can scan
cross sections up to 300 x 400 mm2 without length limitation.
163
6. CONCLUSIONS
The purpose of this paper is to present the qualitative characterisation and non-destructive detection of LB in
tropical hardwoods. The visual and microscopic characteristics show two classes which can be distinguished. Firstly
class 1, where buckling of the wood cells has been observed. This grade contains LB grade 1 (very fine lines, only
visible to the naked eye on planed wet wood surfaces) up to grade 4 (clearly visible brittleheart to the naked eye). The
difference between the grades depends on the volume of deformed and displaced fibres, and the obtained strain level of
the horizontal displacement. Secondly class 2, where class 1 is extended by fibre rupture. This class contains grade 5
(clearly visible brittleheart, accompanied by wood fibre rupture).
It is concluded that the CT-scanners can detect LB successfully. The Nanotom (high resolution micro CT-scanner) can
detect all LB grades. The Somatom (medical CT-scanner) can detect LB starting at grade 2 for sapeli and grade 3 for
sapupira. Both types of 3D CT-scanner were able to detect the propagation of different LB grades throughout the
specimens. Hereby the branched brittleheart and the changeover of diverse LB grades inside the wood could be
detected, which could not be linked to the visual grades on the surface of the wood. In this respect 3D CT-scans are able
to detect the internal expansion of the brittelheart. For the non-destructive detection, the 3D CT-scanning technique is
a promising method to detect brittleheart in tropical hardwood on an industrial level.
Further research is underway in order to quantify the characteristics of the different LB grades. Special attention
should be paid to the volume of the compressed damaged areas, and the level of the buckling of the wood tissue. The
relation between the different LB grades and the decrease of the strength properties will be studied next.
REFERENCES
Ander, P. and Nyholm, K. 2000. Deformations in wood and spruce pulp fibres: Their importance for wood and pulp properties. In:
Proceedings of the international symposium on wood machining, 3-19.
Arnold, M. and Steiger, R. 2006. The influence of wind-induced compression failures on the mechanical properties of spruce
structural timber. Materials and Structures 40:57-68.
Behzad, K. 2007. Effects of compression failure on the mechanical properties of Pinus wood in an experimental planting in north
of Iran. In: ICOMOS IWC XVI international symposium, Florence, Venice and Vicenza.
Brancheriau, L. and Lanvin, J-D. 2008. Research into non-destructive compression crack detection methods for wood structures.
DGAC study n 06/084/MQ/FIN, Final report revised version.
Dadswell, H. E. and Langlands, I. 1938. Brittle heart and its relation to compression failures. Emp. For. J. 17:58-65.
Dinwoodie, J. M. 1968. Failure in timber. Part 1. Microscopic changes in cell-wall structure associated with compression failure.
J. Inst. Wood Sci. 21:37-53.
Hillis, W.E., Hardie, A.D.K. and Ilic, Y. 1973. The occurrence of brittle heart in Eucalyptus grandis grown in Zambia. In:
Proceedings of IUFRO Division 5 meeting, South Africa, 2:485-493.
Hoffmeyer,P. 1990. Failure of wood as influenced by moisture and duration of load. PhD Thesis. State University of New York.
Hoffmeyer, P. 2003. Mechanical properties of timber from wind damaged Norway spruce. In: L. Salmn (Ed) Proceedings of the
second international conference of the European Society for Wood Mechanics, Stockholm, Sweden, 25-28.
Kuera, L.J. and Bariska, M. 1982. On the fracture morphology in wood. Part 1. A SEM-study of deformations in wood od spruce
and aspen upon ultimate axial compression load. Wood Sci. Technol. 16:241-259.
Scurfield, G., Silva, S.R. and Wold, M.B. 1972. Failure of wood under load applied parallel to grain: A study using scanning
electron microscopy. Micron 3:160-184.
Sonderegger, W. and Niemz, P. 2004. The influence of compression failure on the bending, impact bending and tensile strength of
spruce wood and the evaluation of non-destructive methods for early detection. Holz Roh Werkst. 62:335-342.
U.S. Department of Agriculture 1961. Detection of compression failures in wood. Number 1588. Forest Products Laboratory,
Madison, Wisconsin.
Wallace, G. 1993. Defect core detection in radiata pine logs. Report 93/46. Institute of Geological & Nuclear Sciences Limited
Lower Hutt, New Zealand.
Wilkins, A.P. 1968. The nomenclature of cell wall deformations. Wood Sci Technol. 20:97-109.
Yang, J.L. 2001. Bending strength properties of regrowth eucalypt brittleheart. Short Note. Holzforschung 55:183-184.
164
Density of wood can be increased to a certain extent. If compression is done for wet wood in increased
temperature, wood can be densified without severe damages to the cellular structure. Deformations take place
mainly in thin-walled earlywood cells. Light weighted wood species are easier to compress than denser ones, and
increase in density is reached with lower pressure. In case of even distribution of moisture and temperature
through the specimen, also the compression is relatively homogeneous. New ways of using originally light
weighted wood species can presumably be found by compression treatment. However, the compressed cells tend to
recover to their original shape induced by external stimuli, such as temperature or humidity changes. The recovery
may be reduced by proper compression parameters and thermal modification after the compression. In this study,
European aspen (Populus tremula) and silver birch (Betula pendula) boards were modified using different
combinations of compression and thermal modification in order to study the resulting changes in their technical
properties. Compression of 10% and 30% from the green thickness at high temperature and at different initial
moisture content levels, and thermal modification of 3 hours at 190C were used. The effect of compression and
thermal modification with their combinations on degree of compression, density profile and checking of sawn
timber are reported, and their importance in selected end-products is discussed.
1. INTRODUCTION
The average basic density of wood of common Finnish wood species is only between 376 and 480 kg/m3
(Krkkinen 2007). Many mechanical properties of wood are highly dependent on wood density and they set the
boundaries for utilisation of wood in different products. Low density of some common species restricts their use in
products that require high mechanical performance and wear resistance. For example, the use of wood as parquet or
furniture material requires hardness and wear resistance, as well as sufficient bending properties. Therefore, the initial
mechanical properties or durability of wood are often tried to improve with different modification methods or finishing.
Since wood is a porous material, its density can also be increased by impregnating with some chemical (e.g., Lande et
al. 2008), or by compression (e.g., Rautkari et al. 2010). Densification of wood with compression works out especially
for heat-softened moist wood. Large, thin-walled cells have large lumens. Thus, the compression of light weighted
wood species, such as aspen and pine, results in higher relative benefit of densification than in case of denser wood
species. Compression of wood has not become a commercial treatment method since there have not been industrial
systems available, and problems related to the set recovery have not been solved.
In a laboratory scale, compression of wood has been studied by a number of researchers (e.g., Blomberg et al. 2005,
Lamason and Gong 2007, Navi and Girardet 2000, Rautkari and Hughes 2009, Rautkari et al. 2010). In a majority of
the studies, the densification of wood has been targeted to the surface of wood, as in the case of semi-isostatic
compression (Blomberg et al. 2005). The biggest problem in compressed wood has been the set recovery in the course
of time after the compression has been released, the set recovery being accelerated by the high relative humidity and its
* Corresponding author: Tel.: +358 40 801 5053; Fax: +358 29 532 3113; E-mail: veikko.mottonen@metla.fi
165
variation (Blomberg and Persson 2007, Laine et al. 2013b). Being a mechanical treatment, however, compression may
cause damages to the cellular structure, which may appear as checking and deformation of sawn products. However,
according to the results of Laine et al. (2013b), considerable fixation of compressive deformation can be obtained by a
hydrothermal post-treatment. On the other hand, hydrothermal treatment degrades the wood by hydrolysis and
acidification which are in relation to increase in the moisture content (MC), oxygen level and acidity of wood (Aln et
al. 2002).
The objective of this study was to investigate an industrial application of wood treatment, where drying and thermal
modification are carried out in compressed state in order to densify the boards in the direction of their thickness, and to
enhance the product characteristics, e.g., dimensional stability, strength and hardness. The effect of compression
treatment in different stages of drying on checking, degree of compression, and development of density profile in the
direction of thickness were studied. European aspen (Populus tremula L.) and silver birch (Betula pendula Roth) were
selected in the study to represent a relatively light (basic density approximately 390 kg/m3) and dense (500 kg/m3)
Finnish hardwood species.
2. MATERIALS AND METHODS
A total of 112 European aspen and 112 silver birch boards were sawn from freshly harvested logs to nominal
cross-cut dimensions of 40 mm 100 mm. Both flat sawn and radially sawn boards were produced. The modification
processes were carried out in a pilot modification kiln patented by Korwensuun Konetehdas Ltd, Finland. Specimens
for gravimetric moisture content (MC) and basic density measurements were cut from the ends of the boards before
modification treatments. The length of the boards was then adjusted to 2,700 mm, which is the effective length of the
modification kiln. In addition to the modified boards, two reference boards were oven-dried without compression. They
were cut to pieces length of 1 m and dried in a laboratory oven using temperature of 40 C above the wood fiber
saturation point (FSP) and 60 C below the FSP. The approach of FSP was detected by calculating the average MC of
pieces based on the weighing data of the pieces and on initial MC of boards analyzed from small gravimetric samples.
The boards were modified using four different combinations of compression and thermal modification (Table 1,
Figure 1). The system allows drying, compressing and thermal modification in a single process, and different
combinations of process and modification parameters can be used. In the kiln, the boards are stacked between
aluminum plates structured of hollow pipes and the compression is carried out using hydraulic press. The kiln air
circulates through the hollows of aluminum plates which are also perforated in order to enable the evaporation of
moisture from the wood surface. Air and wood temperature, MC of wood, compression force and degree of
compression are measured, each at two different locations in the kiln. In this study, compression was tested both when
the wood material was green and when it was predried down to 20% MC within the process. The drying temperature
was first increased gradually to 90 C for aspen and to 120 C for birch during 3 hours, where it was maintained until the
moisture content of wood was decreased to 30%. Below 30% the drying temperature was increased to 130 C for the
rest of the drying phase. The final moisture content was set to 8%. The target degree of compression for birch and aspen
were set to 10% and 30%, respectively. The different degree of compression was based on differences in the basic
density of the wood species and practical experience of ability of the pilot kiln to compress different wood species.
Further on, half of the material was thermally modified at 190C after the drying and compression phases.
Dimensions of boards were measured both before and after the modification processes from five locations at 50 cm
distances along the boards starting from 25 cm from the other end. Degree of compression was calculated as the
decrease in board thickness in relation to the initial thickness of green boards. Poissons ratio as a relation of the change
in width to the change in thickness was calculated from the dimensional changes of boards. Drying shrinkage was
included in the degree of compression. Length of both surface and end checks was measured for each board both before
and after the modification treatments. Also the length of internal checks was measured. Moisture content of the
modified boards was determined using the gravimetric method. Air-dry density of modified boards was determined
from small specimens by weighing the mass and measuring the stereometric volume of the specimens.
The density profile of untreated and compressed specimens in the direction of board thickness was measured using
X-ray microdensitometry. A cross-sectional specimen of 1 cm was cut from each board, and these specimens were
further cut into rectangular specimens in the direction of board thickness using a twin blade circular saw. The cross
dimensions of the specimens in the microdensitometry tests was 5 mm 5 mm, the length varying according to the
actual thickness of each board. The specimens were not conditioned in order to avoid the possible set-recovery of
densification before the density analysis. A voltage of 30 kV and current of 25 mA was used for X-ray intensity with an
exposure time of 20 ms, and the movement steps between measuring points being 25 m.
166
Table 1: Combinations of compression and thermal modifications used in the processes. G=compression started with green wood,
MC20=compression started at 20% MC, TM=thermal modification.
Process
Birch G
Birch G + TM
Compression
Thermal modification
Starting of compression
Green
10%
Green
10%
3 hours at 190C
Birch MC20
At 20% MC
10%
Birch MC20 + TM
3 hours at 190C
At 20% MC
10%
Aspen G
Green
30%
Aspen G + TM*
Green
30%
3 hours at 190C
Aspen MC20
At 20% MC
30%
At 20% MC
30%
3 hours at 190C
* Modification processes are still going on and data will be available later
Figure 1: Examples of boards before the modification process (reference) and after different combinations of compression and
thermal modification (G=compression started with green wood, MC20=compressed started at 20% MC, TM=thermally modified).
167
This may be caused by slower drying of the centre parts of the boards resulting in postponed change from elastic to
plastic deformation in those locations. Earlier, Laine et al. (2013b) have found an increasing spring-back effect with
increase in wood thickness.
Figure 2: Changes in relative thickness and width (initial value proportioned to 100) of birch (on the left) and aspen (on the right)
boards before and after the modification treatments. G=compression started with green wood, MC20=compressed started at 20%
MC, TM=thermally modified.
Table 2: Degree of compression (%) and average and maximum (in parenthesis) length (mm) of checks on board surface and in the
board ends after modification processes
Process
Degree of
compression
Poissons
ratio
Mean
Surface checks
Max
Mean
End checks
Max
Mean
Internal checks
Max
Birch G
9.6
-1.17
1.01
11.0
1.01
6.0
1.22
3.0
Birch G + TM
12.5
-0.92
0.57
5.0
0.32
3.0
0.13
2.0
Birch MC20
7.7
-1.53
Birch MC20 + TM
10.7
-1.23
0.35
2.0
Aspen G
27.3
0.22
30.75
200.0
8.96
21.0
0.89
5.7
8.0
-0.7
15.70
84.0
10.94
23.0
1.01
5.3
Aspen G + TM
Aspen MC20
Aspen MC20 + TM
The densification of wood caused by the compression covered the possible softening of wood caused by the thermal
modification (Table 3). At short treatment times with hydrothermal treatment, softening is not observed or mechanical
properties may even be enhanced (Sundqvist et al. 2006). In comparison to the untreated specimens, the compression
increased the density of birch and aspen at its most by 21.3% and 51.6%, respectively. The starting of compression at
the moisture content under 20% decreased the degree of compression markedly both in the processes with and without
thermal modification (Table 2) despite the maximum press load of 686 kPa was used. Furthermore, the density of wood
was lower when treated with processes where the compression started under 20% MC, the decrease being higher for
aspen than for birch. As a light weighted wood species, aspen contains more cell lumen volume than birch and can
168
therefore be compressed more easily at high temperature and moisture content. Decrease of moisture content below
20% stiffens the cell walls effectively in both species.
The effect of compression and thermal modification on the density profile in the direction of board thickness was
similar for all modification processes (Figure 4). Mainly, the density profile was even through the entire specimen.
Only in few specimens analysed by X-ray microdensitometry a slightly more densified layer was found near to the
surface of board. Specimen number 178 is an example of surface densified birch board which had the highest density at
the depth of 1 mm from both surfaces (Figure 4). The other compression modified specimens did not show any
exceptionally densified layer through the thickness of board. This result differs from the studies by Rautkari et al.
(2011), Kutnar and Kamke (2012) and Laine et al. (2013a), who observed a clearly denser layer near to the surface, the
location and depth of the peak being dependent on the press time, degree of compression and temperature of the press.
On the other hand, Kutnar et al. (2009) also observed a uniform density profile for the material with 6 mm thickness
when compressed under saturated steam and transient conditions. The main reason for the difference in density profile
formation between this study and the previous ones is the difference in heat softening of wood material; in this study
and in the study of Kutnar et al. (2009) the wood material was heat softened throughout combined with isostatic
pressing. In the studies of Rautkari et al. (2011) and Laine et al. (2013a) the pressing method was more like
semi-isostatic, i.e., heating and pressing on one side mostly. Also the difference in moisture content of wood between
this study and the earlier ones (wood conditioned to 12% MC) most probably had an effect on the difference in the
distribution of density. High MC intensifies the softening of the wood matrix, which is closely related to the glass
transition temperature of lignin (Ibach 2010). Due to the moisture of green wood the lignin softens already at 50100C
(Jskelinen and Sundqvist 2007).
Figure 3: Change of cross-cut dimensions of boards at different locations along board after the modification treatments for birch (on
the left) and aspen (on the right). The measuring points (T1-5, W1-5) were located at 500 mm distances along board starting at 250
mm from the other end of boards. G=compression started when wood green, MC20=compressed started at 20% MC, TM=thermally
modified.
In most cases, drying accompanied with compression in the direction of thickness resulted in negative Poisson
ratios, i.e., reduction in the width of the boards (Table 2). This peculiarity was derived from the fact that drying
shrinkage in tangential direction was larger than the widening of the board as a result of compression. The only case
with positive Poisson ratios was seen in aspen boards, when they were compressed in green condition.
169
Table 3: Density of wood (kg/m3) of reference material and after different modification processes analysed using x-ray (A) and
gravimetric (B) methods. Mean standard deviation.
Birch
Aspen
Reference
A
B
525
50230
38438
A
53945
B
58333
58249
G + TM
A
B
55032 60229
MC20
A
B
51721 59434
43235
MC20 + TM
A
B
53231 60943
Figure 4: Density profile in the direction of thickness of compression and thermally modified birch specimens and reference
specimens of same species measured using x-ray microdensitometry. Distance: location through the specimen thickness, 0 = outer
surface of the board.
The starting of the compression of boards at green state increased checking as compared to the compression at 20%
MC (Table 2). Regarding birch, no checking was found for boards which were started to compress at 20% MC. Both
the surface checking and end checking were much more intensive in aspen boards than in birch boards, which is due to
the higher compressive strains, lower mechanical strength and expectedly uneven moisture distribution in aspen
boards.
6. CONCLUSIONS
The degree of compression was found to differ between the processes which differed from each other by the phases
when compression was started. The targeted degree of compression was achieved only when compression started with
green wood. Also the densification was more intensive with compression of green wood. However, any remarkable
differences in density profiles were not observed between the processes with different compression starting times. The
further interest of research lies in minimizing the recovery of the densification in changing environmental conditions.
Thermal treatment during the final stage of the thermo-mechanical modification process was found to intensify the
degree of compression and densification, and reduce the checking in sawn timber. In this respect, the thermal treatment
completes the thermo-mechanical modification process bringing in also other beneficial effects on wood properties.
The original purpose of using thermal treatment as a wood modification method is changing wood colour and
170
improving its dimensional stability. Including thermal modification to the compression treatment process has potential
to diminish the problems related to the set-recovery of mechanically compressed wood.
7. ACKNOWLEDGEMENTS
Korwensuun Konetehdas Ltd. co-funded the study, and performed the modifications. The Vocational Training
Association of Woodworking Men co-funded the study. Mr. Juha Metros and Mr. Hannu Koivunen collected the wood
material and made the gravimetric, stereometric and visual analysis of boards, and Mr. Kari Sauvala performed the
X-ray analysis. The authors wish to present sincere thanks to all these institutions and individuals.
REFERENCES
Aln, R.; Kotilainen, R.; Zaman, A. Thermochemical behavior of Norway spruce (Picea abies) at 180225C. Wood Science and
technology 2002, 36, 163-171.
Blomberg, J., Persson, B. and Blomberg, A. 2005.Effects of semi-isostatic densification of wood on the variation in strength
properties with density. Wood Science and Technology 39: 339350.
Blomberg, J., Persson, B. 2007. Swelling pressure of semi-isostatically densified wood under different mechanical restraints.
Wood Science and Technology 41:401415.
Ibach, R.E. 2010. Specialty Treatments. In: Forest Products Laboratory. Wood handbookWood as an engineering material.
General Technical Report FPL-GTR-190. Madison, WI: U.S. Department of Agriculture, Forest Service, Forest Products
Laboratory. 508 p.
Jskelinen , A.-S. and Sundqvist, H. 2007. Puun rakenne ja kemia. University Press Finland Ltd, Helsinki. 142 p. ISBN
978-951-672-351-1
Kutnar, A. and Kamke, F.A. 2012. Compression of wood under saturated steam, superheated steam, and transient conditions at
150C, 160C, and 170C. Wood Sci Technol 46: 7388.
Kutnar, A., Kamke, F.A. and Sernek, M. 2009. Density profile and morphology of viscoelastic thermal compressed wood. Wood
Science and Technology 43: 5768.
Krkkinen, M. 2007. Puun rakenne ja ominaisuudet.
Laine, K., Rautkari, L. and Hughes, M. 2013a. The effect of process parameters on the hardness of surface densified Scots pine
solid wood. European Journal of Wood and Wood Products 71: 1316.
Laine, K., Rautkari, L., Hughes, M. and Kutnar, A. 2013b. Reducing the set-recovery of surface densified solid Scots pine wood by
hydrothermal post-treatment. European Journal of Wood and Wood Products 71: 1723.
Lamason C, Gong M (2007) Optimization of pressing parameters for mechanically surface densified aspen. Forest Products
Journal 57: 6468.
Lande, S., Eikenes, M., Westin, M. and Schneider, M. H. 2008. Furfurylation of wood. Chemistry, properties and
commerialisation. In: T. Schultz, H. Militz, M. H. Freeman, B. Goodell and D. D. Nicholas (Eds.), Development of
commercial wood preservatives. ACS Symposium Series 982.
Navi, P. and Girardet, F. 2000. Effects of thermo-hydro-mechanical treatment on the structure and properties of wood.
Holzforschung 54: 287-293.
Rautkari. L. and Hughes, M. 2009. Eliminating set-recovery in densified wood using a steam heat-treatment process. In:
Proceedings of the 4th European conference on wood modification, Stockholm, Sweden
Rautkari, L., Properzi, M., Pichelin, F. and Hughes, M. 2010. Properties and set-recovery of surface densified Norway spruce and
European beech. Wood Science and Technology 44: 679691.
Rautkari, L., Laine, K., Laflin, N. and Hughes, M. 2011. Surface modification of Scots pine: the effect of process parameters on the
through thickness density profile. Journal of Material Science 46: 47804786.
Sundqvist, B., Karlsson, O. and Westermark, U. 2006. Determination of formic-acid and acetic acid concentrations formed during
hydrothermal treatment of birch wood and its relation to colour, strength and hardness. Wood Sci. Technology. 40:
549-561.
171
ABSTRACT
Improvement of visual attraction of wood surface is one of the most important reasons for the thermal modification
of sawn timber in wood products industry. Therefore, thermal modification of light-coloured wood species is an
increasing modification method of wood in Northern Europe. The impressive dark brown colour obtained by the
Thermowood thermal modification process resembles the colour of tropical wood species, which raises the value of
wood markedly. In addition, dimensional stability of wood material increases due to the decrease in
hygroscopicity. Besides the positive effect on colour and hygroscopicity, thermal modification lowers the
mechanical strength of wood mainly because of the degradation of hemicellulose chains. One of the greatest
challenges is to produce thermally modified wood which is intensively coloured but, at the same time, has the
original strength and durability. In this study, the reasons for the variation of colour and mechanical properties
found in the industrial production of thermally modified sawn timber are considered from the point of view of
pre-drying of wood material. Freshly sawn Silver birch (Betula pendula) boards were pre-dried using
conventional low temperature drying at 60C to different moisture content levels (10%, 20%, 30% and 40%) before
thermal modification process. The thermal modification process consisted of high temperature drying at 140C
and thermal modification at three different temperatures (170C, 185C and 200C). The effect of pre-drying on
colour, Brinell hardness, modulus of elasticity and modulus of rupture are reported.
1. INTRODUCTION
The colour of most utilized hardwood species (beech, birch, aspen) in Northern Europe is light compared to the
colour of imported wood species from tropical and subtropical regions. Therefore, thermal modification for visual
improvement of wood surface of light-coloured wood species is becoming more and more a general method that is used
in furniture, paneling and parquet industries for reproducing the aesthetic of tropical woods. The impressive dark
brown colour obtained by the thermal modification raises the value of wood markedly. Thermal modification also
increases dimensional stability of wood by lowering the equilibrium moisture content (Gndz et al. 2008). Besides the
positive effects on colour and stability, thermal modification also weakens the mechanical strength of wood mainly
because of the degradation of the hemicelluloce chains (Aln et al. 2002, Kamdem et al. 2002, Tjeerdsma et al. 1998).
One of the greatest challenges is to produce thermally modified wood which is intensively coloured but, at the same
time, has the original strength and durability. There would be many new applications for thermally modified wood if the
mechanical properties could be remained at the level of untreated wood or even improved during the process. In this
study, the variation of colour and mechanical properties found in the industrial production of thermally modified sawn
timber are considered.
* Corresponding author: Tel.: +358 40 801 5053; Fax: +358 29 532 3113; E-mail: veikko.mottonen@metla.fi
172
The discolouration and the depression of mechanical strength of wood are specific for each wood species and
affected by the modification temperature and the moisture in wood. In birch wood, the changes in lightness occur after
a short period of time and the colour saturation continues as the concentration of acids increases in wood (Sundqvist et
al. 2006). The changes in wood colour are connected to change of number and type of chromophores, which are the side
chains of pigment molecules. The depression of mechanical strength in turn is connected to the degradation of cellulose
chains as a result of hydrolysis (Aln et al. 2002). Thus, there may be different chemical reactions behind the colour
change and depression of the mechanical strength which may also differ by time. The change in colour is not permanent
and the dark colour turns to grayish even as soon as after 6 months if the uncoated wood is exposed to UV-light (Jms
et al. 2000, Miklei et al.2011). The colour changes the more the higher is the thermal modification temperature.
Compared to Scots pine wood birch wood needs clearly lower modification temperature to obtain the considerable
colour change.
The Finnish ThermoWood process for hardwood species is based on classification where the dark colour and
biological durability are mentioned as main advantages of the higher modification temperature (200C) and durability
the main advantage of the lower modification temperature (185C) (Thermowood handbook 2003). The difference in
the modification temperature is so large that the average colour differs between the two classes markedly. In the
industrial production of thermally modified wood, there exists clear variation in the wood colour although the process
parameters are tried to be standardized and sometimes the sufficient colour change is not achieved. One possible reason
for colour variation could be the variation in the initial moisture content of wood. In this respect, the two most common
thermal modification processes, Thermowood and Plato, differ from each other as undried or pre-dried wood is used in
the first phase of thermal modification in Plato process (Militz 2002). The variation in the initial moisture content is
eliminated in Thermowood process by pre-drying of the wood material to the absolute dry condition before the thermal
modification process. However, the equal drying of wood is problematic because of the large variation in the initial
moisture content of wood. The effect of the changes in wood during the drying phase on the properties of thermally
modified wood is unknown. In addition, the moisture in wood would probably contribute the adequate colour change
(c.f., Sundqvist et al. 2006).
The objective of this study was to investigate the effect of the initial moisture content and pre-drying of the wood
raw material on the technical and mechanical properties of thermally modified wood intended for flooring or furniture
products. These properties include development of evenly darkened colour during the thermal modification processes,
the hardness of wood surface, and the bending stiffness and strength. Green wood material and wood material pre-dried
to the moisture content of 10%, 20%, and 35% was thermally modified at 170C, 190C, and 212C. Differences in
wood colour, Brinell hardness and bending properties were analysed between the initial moisture content levels of
wood and between the thermal modification temperatures.
2. MATERIALS AND METHODS
The experiments of this study were made using 120 freshly sawn Silver birch (Betula pendula Roth.) boards,
dimensions of 25 mm 100 mm 1000 mm. The initial moisture content of the green boards varied from 55.3% to
86.8%. The boards were divided to four groups. Three of the groups were pre-dried to the moisture content levels of
10-15% (MC10), 20-25% (MC20) and 30-35% (MC30), and the group of green boards (MCg) was not dried before the
thermal modification process. In each MC level, the drying was carried out at 60C and the decrease in MC during the
pre-drying was monitored from five sample boards with a resistance moisture content meter. The duration of the
pre-drying phases within the groups MC10, MC20 and MC30 was 380, 240, and 170 hours, respectively. The boards
were divided to groups which were thermally modified at 170C, 190C and 212C according to the ThermoWood
process in the thermal modification kiln of the Lappeenranta University of Technology.
A drying phase of about 12 hours preceded the thermal modification process and consisted of three different steps.
During the first step the kiln temperature was gradually increased within 4 hours to 100C. In the second step the
constant temperature of 100C was maintained until the temperature of wood reached also the same temperature. In the
last step the drying temperature was increased to 130C, and kept at that level for 6 hours. The effect of the high
temperature drying phase of 12 hours (preceding the thermal modification) on the wood moisture content was tested
with the green samples. The moisture content of wood of the test samples was 35.4%. Between the stages of high
temperature drying and thermal modification, the moisture content of the other pre-dried groups could not be analyzed.
The thermal modification process was carried out at saturated steam pressure, in order to intensify the heat transfer into
the wood and to prevent the ignition of the wood material. The duration of the thermal modification was three hours at
each temperature. Equalizing phase was not included in the end of the thermal modification process in order to avoid
the fading of the differences between study variables.
173
Effect of pre-drying on colour and mechanical properties of thermally modified Silver birch timber
From each board, two defect-free samples for static modulus of rupture (MOR) and static modulus of elasticity
(MOE) measurements were prepared to the size of 17 mm 35 mm 350 mm. The same samples were used also for the
colour and Brinell hardness measurements. A total of 240 samples were tested. The MOR was calculated in accordance
with standard EN 310 as follows:
max
3 Fmax L
2 b h2
(1)
where Fmax = maximum load [N],), L = distance between the lower supports [mm], b = specimen width [mm], h =
specimen height [mm]. The MOE was calculated in accordance with standard EN 310 as follows:
Fp L3
(2)
4 f p b h3
where Fp = load at the proportional limit [MPa], L = distance between the lower supports [mm], fp = strain at the
proportional limit [mm], b = specimen width [mm], h = specimen height [mm].
The Brinell hardness was tested in accordance with standard EN 1534 as follows:
HB
2 F
Db Db Db 2 d 2 2
(3)
where HB is the Brinell hardness (MPa), F is the nominal force (N), is the pi factor ( 3.14), Db is the diameter of the
steel ball (mm) (10 mm in this study), and d is the mean diameter (measured with calliper both parallel and
perpendicular to the grain) of the residual indentation (mm).
Samples for spectral measurements of the wood were prepared and the reflectance spectra were measured from the
face which was revealed in the core of the split wood samples. Spectral measurements of wood were made with a
portable spectrophotometer (Minolta CM-2500d) using 2 standard observer and D65 standard illuminant (Precise
color 1994). The spectrum of reflected light in the visible range (360 740 nm) was measured. From each sample
board, three measurements, avoiding knots or other extraneous discolourations, were made and averaged to one
observation. The three-dimensional L*a*b* colour scale (recommended in 1976 by the International Commission on
Illumination, CIE), where +a* stands for red, a* stands for green, +b* stands for yellow, and b* stands for blue, was
used to quantify changes in the tint of the wood (see e.g. Hunt 1998). The L* axis represents nonchromatic changes in
lightness from an L* value of 0 (black) to an L* value of 100 (white).
3. RESULTS AND DISCUSSION
The average value of MOR decreased with the increase in thermal modification temperature by 9.7% from 170C
to185C and by 8.5% from 185C to 200C, respectively (Figure 1). Hence, the total decrease of MOR was 17.4%
when the thermal modification temperature increased from 170C to 200C. Within the thermal modification
temperature, the MOR did not differ between the initial MC levels. Increase in moisture content, temperature and
treatment time in thermal modification has been found to increase the formation of formic and acetic acids by
hydrolysis (Sundqvist et al. 2006). The acidification degrades the wood which is shown as lower mechanical properties.
However, the degradation of wood due to the increased moisture content was not observed in this study. Probably, the
degradation of wood by acidification ceased rapidly with the decrease of moisture content during the relatively fast
drying and thermal modification process.
The average value of MOE increased with the increase in thermal modification temperature both between 170C
and 185C and between 185C and 200C (Figure 2). An exception was the initially green sawn timber (MCg) in which
the MOE decreased by 6.7% when the thermal modification temperature increased from 185C to 200C. The
174
dependence between MOR on MOE (Figure 3) was strong (Pearson correlation 0.78) in the group of thermally
modified sawn timber at 170C, but regarding the entire material, the dependence weakened with the increase in
thermal modification temperature to 185C (0.50) and 200C (0.47). The weakening of the dependence between MOE
and MOR indicates the formation of structural defects during thermal modification, probably the increasing
degradation of cellulose chains by hydrolysis (c.f., Tjeerdsma et al. 1998). However, in the group of initially green
sawn timber, the dependence between MOR and MOE remained strong at each thermal modification temperature,
indicating smaller changes in the structure of the main components, probably mainly the hemicelluloses. Thus, the
presence of moisture at high content in wood in the beginning of the thermal modification process seems to delay the
beginning of the degradation process of wood caused by hydrolysis.
Figure 1. Bending strength (MOR) by treatment temperature and initial moisture content.
Figure 2. Modulus of elasticity (MOE) by treatment temperature and initial moisture content.
175
Effect of pre-drying on colour and mechanical properties of thermally modified Silver birch timber
Figure 3: Dependence of static MOR on static MOE and the linear regression lines by initial moisture content level.
176
Similar to the MOR, also the Brinell hardness decreased with the increase in thermal modification temperature
(Figure 4). However, the moisture content level had an effect on the Brinell hardness, the values being the highest for
initial moisture content levels of MC20 and MC30. This difference in the moisture related behavior between the MOR
and the Brinell hardness is probably caused by the fact that the determination of Brinell hardness represents a
combination of measurement of friction, shear strength and compression strength (Kollman and Ct 1968). Thus, the
differences in the initial moisture content level may be seen differently in the Brinell hardness values than in the MOR
values.
Figure 4: Brinell hardness of wood at different initial moisture content levels by thermal modification temperature.
Regarding each initial MC level, the CIEL*a*b* colour values showed increase of redness and yellowness and
decrease of lightness with the increase of thermal modification temperature (Figure 5). The decreased lightness of birch
wood with the increase in modification temperature was connected more like with the increase in redness than in
yellowness. The colour change of wood induced by the thermal modification is very obvious for most wood species
(Bekhta and Niemz 2003, Dagbro et al. 2010, Johansson 2008, Unsal et al. 2003). The effect of initial moisture content
level on colour was observed only in the redness values which increased with the increase in the initial moisture
content. This result indicates that the chemical changes involved in mechanical and spectral properties are different.
Therefore, the change in the wood colour by thermal modification would be possible to achieve with minimum effect
on mechanical strength and wear resistance.
177
Effect of pre-drying on colour and mechanical properties of thermally modified Silver birch timber
Figure 5. Variation of the CIELab colour values by the thermal modification temperature and initial moisture content level.
178
The results of this study indicate that wood discoloration and deterioration of the strength of wood can occur at
various stages of heat treatment. The further interest of research lies in minimizing the deterioration of the strength with
adequate change in colour. Initial moisture content of wood at the drying stage of the thermal modification had a clear
influence on the colour formation. Therefore, the control initial moisture has the potential of improving the quality of
heat-treated wood.
REFERENCES
Bekhta, P. and P. Niemz. 2003. Effect of high temperature on the change in color, dimensional stability and mechanical properties
of spruce wood. Holzforschung 57:539546.
Gndz, G., Niemz, P. and Aydemir, D. 2008. Changes in Specific Gravity and Equilibrium Moisture Content in Heat-Treated Fir
(Abied nordmanniana subsp. bornmlleriana Mattf.) Wood. Drying Technology, 26, 1135-1139
Dagbro, O., Torniainen, P., Karlsson, O. and Morn, T. 2010. Colour responses from wood, thermally modified in superheated
steam and pressurized steam atmospheres. Wood Material Science and Engineering 5: 211-219.
Hunt, R.W.G. 1998. Measuring Colour. Third edition. Fountain Press, Kingston-upon-Thames, England. pp. 53-72.
Johansson, D. 2008. Heat Treatment of Solid Wood: Effects on Absorption, Strength and Colour. Doctoral Thesis. Lule
University of Technology
Jms, S., Ahola, P. and Viitaniemi, P. 2000. Long-term natural weathering of coated ThermoWood. Pigm. Resin Technol. 29:
68-74.
Kamdem, D.P., Pizzi, A. and Jermannaud, A. 2002.Durability of heat-treated wood. Holz als Roh- und Werkstoff 60: 1-6.
Kollman, F. and Ct, W.A. 1968. Principles of wood science and technology. 1. Solid wood. Springer Verlag. Berlin
Heidelberg-New York. 592 p.
Miklei, J., Jirou-Rajkovi, V., Antonivi, A. and pani, N. 2011. Discolouration of thermally modified wood during simulated
indoor sunlight exposure. Bioresources 6: 434-446.
Militz, H. 2002. Heat Treatment Technologies in Europe: Scientific Background and Technological State-of-Art. In: Proceedings
of Conference on Enhancing the durability of lumber and engineered wood products February 11-13, 2002, Kissimmee,
Orlando. Forest Products Society, Madison, US.
Precise Color Communication. 1994. Color control from feeling to instrumentation. Minolta Co., Ltd. Osaka, Japan. 49 p.
Sundqvist, B., Karlsson, O. and Westermark, U. 2006. Determination of formic-acid and acetic acid concentrations formed during
hydrothermal treatment of birch wood and its relation to colour, strength and hardness. Wood Sci. Technology. 40:
549-561.
Tjeerdsma, B. F., Boonstra, M., Pizzi, A., Tekely, P. and Militz, H. 1998. Characterisation of thermally modified wood: molecular
reasons for wood performance improvement Holz Roh- Werkstoff 56:149-153.
ThermoWood handbook. 2003. Finnish ThermoWood Association. www.thermowood.
Unsal, O., Korkut, S, and Atik, C. 2003. The effect of heat treatment on some properties and colour in eucalyptus (Eucalyptus
camaldulensis Dehn.) wood. Maderas. Ciencia y tecnologa. 5(2): 145-152.
179
In order to make a prediction about the optical properties of high-quality walnut wood, various wood anatomical
parameters and color were measured before, during and after drying. It should be investigated whether easy
recorded parameters, such as the annual ring structure correlate with the sorting class of walnut. The results
showed that color measurements can at no time in the production chain predict the final quality of the timber. Also
the measured density of the boards is not correlated with the quality of the walnut samples. Further results show
that the annual ring structure and the distance of the board to the centre of the tree have a significant impact on the
quality class of the walnut boards. Due to the two measured variables, a prediction can be made about the quality
by using a discriminant function.
1. INTRODUCTION
Wood as a versatile renewable material is often found in high quality furniture, not only because of its mechanical
abilities, but also because of its unique appearance. Hence, not every timber can be used for furniture, the aesthetics and
the visual appearance is of great importance. For this reason, wood is undergoing a sorting. State of the art provides
several rules that allow an assessment of the quality (NORM EN 13017-2 (2001)). The general requirements as well
as the appearance-classes for single-and multi-layer solid wood panels produced from hardwoods are determined in this
standard.
However, it is known that high quality products would like to outperform the specified requirements. Thus, the
sorting is done mostly by corporate color cards and specific quality criteria. The sorting itself is thus performed by
employees who decide individually for each presented board. But chromatic aberrations of wood are often recognized
late in the process chain. The timber must often be dried, planed and sanded previously, before the errors can be seen.
To minimize mistakes from employees and safe process costs a pre-sorting before drying of the sawn timber could be
helpful. By pre-sorting, substandard timber could be rejected and the proportion of high-quality boards could be
increased.
So far, the pre-sorting of wood was investigated using acoustic measurements (Wang and Simpson 2006). A
significant correlation between the acoustic properties and the grade loss caused by exceeding warp limits could be
determined. Also by using stress wave and transverse vibration techniques a pre-sorting of lumber were shown (Halabe
et al. 1996), but only mechanical properties could be predicted.
This work deals with a prediction of the optical properties of high quality walnut (Juglans nigra L.) timber prior to
drying. Due to this issue, several wood parameters were measured in the process chain, which may contribute to an
early sorting. Specifically, the anatomical characteristics and color measurements of wood were thereby taken into
consideration.
* Corresponding author: Tel.: +43 (01) 47654-4228 ; Fax: +43 (01) 47654-4295 ; E-mail: m.mueller@kplus-wood.at
180
Figure 1: The color chart used in the sorting of the walnut boards
2.1. DENSITY
The influence of the density on the color development was reviewed in the context of this investigation. Of each
board, two dice samples (3 cm x 3 cm x 3 cm) were cut from the heartwood. After conditioning of the samples, at 20 C
and 65 % relative humidity, the density was determined according to (DIN 52182 1976). Two Density measurements
were made per board - one from the edge area of the board and one selected from the centre of the board. The total
density resulted from the averaged value.
n =
mn
in g/cm
Vn
(1)
where, n stands for the density, mn stands for the mass and Vn stands for the volume of each sample.
2.2. ANNUAL RING ANALYSIS
An annual tree ring analysis was carried out, in which the average annual ring width and the widest annual rings of
each board was recorded (4074-5 2008)). For each measurement, a 2 cm wide strip was cut and afterwards sanded (600
grain size). Due to the very low proportion of juvenile wood and the generally high level of homogeneity, all anatomical
wood structures were included in the analysis.The calculation of the average annual ring width resulted from the
following formula:
181
J br =
l
[mm]
n
(2)
where, Jbr stands for the annual ring width, l stands for the length of the sample and n stands for the amount of annual
rings.
In addition to the mean annual ring width values, the widest annual ring of each board was detected and measured
with a sliding gauge. Due to wider annual rings of possible juvenile wood near the centre of the tree, additionally the
distance of the board to the centre of the tree was measured. For these measurements, a template for each sample was
prepared. The template has been adapted to the sample cross-section until the annual rings of the template matched the
course of annual ring of the board. Then, the distance to the centre of the tree was read from a scale.
4. RESULTS AND DISCUSSION
Figure 2 shows the result of the sorting according to the industrial standard, after planing and sanding the 150
walnut samples. It can be seen that most of the boards were placed in class A. In the class C the least boards were
assigned, substantially less than in B.
Figure 2: Quality distribution of the analyzed walnut boards, sorted in classes A, B and C
For the determination of appropriate parameters for an early sorting a discriminant analysis was performed. In the
model, the color measurements before, during and after the drying, as well as the density and the annual ring analysis
was included. By the mean values of the independent variable, presented in groups A-, B- and C-Ware, the explanatory
power of each variable can be calculated (Table 1).
Table 1: The Group statistic of the annual ring measurements
Wood Quality
A Class
B and C Class
Parameter
Avr. annual ring width
Widest annual ring
Distance to centre of tree
Avr. annual ring width
Widest annual ring
Distance to centre of tree
Average
2.21
5.16
53.11
2.92
6.43
61.45
Standard deviation
1.01
2.10
43..25
1.26
2.63
38.47
[n]
85
85
85
65
65
65
The more the average values within each quality group (A; B/C) differ from each other, the easier to differentiate
the groups. A good model is characterized by the fact that the average function values of the individual groups differ
significantly from each other.
182
(3)
where, x1 stands for the variable distance to the center of tree and x2 stands for the variable widest annual ring.
The function indicates that the distance to the center of tree and the widest annual ring is positively connected.
It can therefore be concluded, that the quality is decreasing the further the board is removed from the center of the tree
and the broader the widest annual ring. An auto-correlation of these two factors could be excluded (p = 0.1). These
results are not in contrast to the knowledge that juvenile wood, which is most likely to be found near the log center, is
generally of lower quality (Bendsten 1978). Because the proportion of juvenile wood of the measured boards is
negligibly, this contradictory factor can be excluded. Why nevertheless the boards closer to the middle are of higher
quality, can specifically for the walnut not be answered and could not be found in the literature. One explanation would
be that walnut trees were growing at a young age in a more dominated forest structure and thus have rather thin annual
rings. The period of this state as well as the further development of the forest structure is therefore crucial to the quality
of the wood.
The calculated classification results show the prediction result compared to the actual group membership. Of 85
A-grade boards, 64 will be correctly predicted and 21 incorrectly assigned. This represents a prediction rate of 75.3%.
Of 62 boards B-/C-Ware will be 37 correctly assigned and 25 incorrectly classified. This represents a prediction rate of
59.7%. It can thus be stated that an presorting before processing with the variables distance to the center" and "widest
annual rings", could lead to an increased proportion of" A quality " timber in production.
5. CONCLUSION
The investigations carried out for a preliminary sorting of walnut wood showed that color measurements before
during and after drying, cannot predict the future quality of the resulting board. Rather, due to an annual ring structure
analysis of walnut boards, a prediction for the future quality sorting can be made.
REFERENCES
BENDTSEN, B.A. 1978. PROPERTIES OF WOOD FROM IMPROVED AND INTENSIVELY MANAGED STANDS. FOR. PROD. J. 28(10): P61-72.
BRUNNER, C., G. SHAW, D. BUTLER, AND J. FUNCK. 1990. USING COLOR IN MACHINE VISION SYSTEMS FOR WOOD PROCESSING. WOOD
DIN 4074-5. 2008. SORTIERUNG VON HOLZ NACH DER TRAGFHIGKEIT - TEIL 5: LAUBSCHNITTHOLZ.
DIN 52182. 1976. PRFUNG VON HOLZ; BESTIMMUNG DER ROHDICHTE.
HILLER, C., F. FREESE, AND D. SMITH. 1972. RELATIONSHIPS IN BLACK WALNUT HEARTWOOD BETWEEN COLOR AND OTHER PHYSICAL
HALABE, U., G. BIDIGALU, H. S. GANGARAO, AND R. ROSS. 1996. NONDESTRUCTIVE EVALUATION OF GREEN DIMENSION LUMBER
USING STRESS WAVE AND TRANSVERSE VIBRATION TECHNIQUES. PAGES 1891-1894 IN D. THOMPSON AND D. CHIMENTI, EDITORS.
183
184
CNR-IVALSA
Sesto Fiorentino (FI), I-50019, Italy
MiCROTEC Srl
Bressanone (BZ), I-39042, Italy
ABSTRACT
Since some years now, strength grading is a mandatory step for the structural use of timber. To satisfy this
requirement there are two generally recognized methods: visual strength grading bases the assignment of each
structural element to a grade on the timber characteristics that are visually measurable (i.e. knots, slope of grain,
annual ring width); while machine grading uses the non-destructive measurement of physical and mechanical
properties of wood, most frequently the dynamic modulus of elasticity. The efficiency of strength grading is closely
linked to the value of correlation between these characteristics and the true strength of the timber elements. At the
same time, chestnut, largely used for structural purposes in Italy, is known to be a wood species difficult to deal
with, and the visual grading rule of the Italian standard provides only one level of performance. Furthermore, the
use of non ordinary assortment characterized by irregular cross section (squares with wane) is very common, but
it is not included in the current reference standards for the structural use of solid wood. As concerns machine
grading, at the moment there is no certified machine available for chestnut. Therefore an improvement was tested
to grade chestnut timber with wane by means of machine measurements. 4 provenances (2 from Italy and 2 from
France) were sampled. The dynamic modulus of elasticity of 580 timber elements was measured by means of
ViSCAN grading system; afterwards bending tests were performed on each piece and bending strength, as well as
modulus of elasticity and density, were determined. Here, the results of the machine grading of chestnut timber
with wane are presented.
1. INTRODUCTION
Chestnut (Castanea sativa Mill.) is one of the most common hardwoods used in construction in the
Mediterranean countries. It can be found in ancient and prestigious buildings as well as in new structures, so to be an
integral part of the Italian architectural culture. Both in the past and present, chestnut is broadly used as squared with
wanes. This is a particular assortment called in Italy Uso Fiume and it is described by an irregular cross section,
that maintains natural wanes along the length of the log, so to adapt sawing to the trunk shape and reduce material
losses, very important for chestnut typically characterized by irregular stem form.
For some years now, however, to use timber elements in construction it is mandatory to strength grade them. It
means that each piece of timber has to be evaluated by means of non destructive measurements and assigned to a
strength class, the characteristic strength properties of which are defined by the European standards (EN 338 2009).
The assignment can be achieved by a visual inspection (visual grading) of the main strength reducing characteristics,
i.e. knots, grain deflection, annual ring width, or by an instrumental assessment (machine grading). The
effectiveness of the grading is linked to the prediction capability of the estimating properties in respect to the real
strength, stiffness and density of timber.
* Corresponding author: Tel.: (+39) 055-5225754; Fax: (+39) 055-5225643; E-mail: nocetti@ivalsa.cnr.it
185
IT-1
82
77
159
Sub-sample
(N of pieces)
IT-2
FR-1
56
61
117
69
59
59
187
FR-2
60
59
119
Section
(mm)
Length
(mm)
120X120
150x150
100x100
150x150
150x150
190x190
180x180
190x190
180x180
3000
3000
2500
3500
3000
4500
4500
4300
4500
The specimens were squared edge logs with wane, i.e. logs edged on the four sides, with boxed heart and
maintaining the wane along the length of the timber piece, but with constant cross section along the entire length
(Figure 1). Whether not further specified, for cross section is intended the nominal cross section, defined by the
rectangle circumscribing the timber element. The pieces were sampled, examined and tested in green condition
(moisture content > 30%).
Each specimen was measured by means of ViSCAN grading system (by MiCROTEC). It determines the
dynamic modulus of elasticity of a timber piece from information on its natural frequency of vibration in
longitudinal direction and wood density. The excitation necessary to make the board vibrate is provided by an
impact and the natural frequency of vibration is measured by a non-contact laser interferometer. The weight and
dimensions of each specimen were measured, so that timber density could be determined by equation 1 and the
dynamic modulus of elasticity by equation 2.
186
w
b!h!l
(1)
Edyn = 4 ! l 2 ! f 2 ! dens
(2)
dens =
where dens = density used in the calculation of the dynamic modulus of elasticity; w = weight; b = base; h = height
of the timber piece; l = length of the timber element; Edyn = dynamic modulus of elasticity, f = natural frequency
Once finished the machine measurement, four point bending tests were carried out in accordance with EN 408
(2012). The critical section of each specimen was defined by the visual examination and placed in the mid-span. The
local deformation was measured in the neutral axis on both sides of the beam and the mean of the two was used to
calculate the local modulus of elasticity (El). Simultaneously, the total deformation was measured in the central
point on the tension edge of the element and used to calculate the global modulus of elasticity (Eg). Then the test
went on until failure and the maximum load was recorded to calculate the bending strength (fm). Both for machine
measurements, strength and stiffness calculation, the reference cross section was the rectangle circumscribing the
timber element (nominal cross section). Detail on test configuration and the equation used to calculate mechanical
properties are reported in the standard (EN 408 2012).
After failure, wood density (!) was measured on a sample cut from each specimen with dimensions equal to the
full cross section, free from defects and as close as possible to the fracture, as recommended by the EN 408 (2012).
The slice was cut, conditioned to the reference conditions of 20C and 65% RH and then measured to determined
the wood density.
The moisture content (MC) at the time of bending test was determined on 5 sections of the full cross section
taken close to the supports, the loading points and the mid-span. The oven dry method was used (EN 13183-1 2003).
The moisture content of each specimen was the average of the 5 sections.
Finally, modulus of elasticity and density were adjusted for moisture content: the values of modulus were
increased of 1% for each percentage point difference in moisture content to the reference value of 12%, but only
from 18% and not from their actual value of moisture content; density was decreased by 0.5% for every percentage
point difference (EN 384 2010). Finally, bending strength was adjusted for timber size to the reference value of
width = 150 mm, as recommended by EN 384 (2010).
2.2 DATA ANALYSIS
Firstly, data were analysed by means of linear regression to quantify the amount of variation in the mechanical
properties explained by the machine measurements.
187
MC
(%)
fm
(MPa)
El
(MPa)
Eg
(MPa)
!
(kg/m3)
IT-1
159
44.2
(10.2)
42.7
(7.8)
12 700
(1 740)
11 400
(1 240)
578
(51)
IT-2
117
65.4
(15.5)
40.1
(7.1)
11 500
(1 780)
10 500
(1 390)
563
(34)
FR-1
187
64.4
(14.8)
39.1
(6.2)
11 000
(1 500)
10 200
(1 310)
562
(40)
FR-2
119
59.5
(11.3)
40.3
(6.3)
10 900
(1 610)
10 100
(1 350)
552
(45)
ALL
582
58.1
(15.8)
40.5
(7.0)
11 500
(1810)
10 600
(1400)
565
(44)
The results of the regression analysis are presented in Table 3 and in Figure 2. The independent variable of the
linear regression was the Indicating Property (machine dynamic modulus of elasticity), while the dependent
variables were the strength and stiffness properties measured in laboratory, as well as timber density alternatively. In
Table 3 the coefficients of determination of the linear regressions are shown for each sub-sample and for the whole
sample. As expected, the highest coefficients were noticed for the regression with static modulus of elasticity
(slightly higher with global modulus than local), whereas they were lower with timber density firstly and with
bending strength secondly.
The linear regression calculated to predict bending strength from modulus of elasticity and density measured in
laboratory illustrated the difficulties in predicting the strength of chestnut timber: with static modulus of elasticity as
188
El (MPa)
60
40
50
IT-1
IT-2
FR-1
FR-2
Eg
0.77
0.69
0.62
0.66
0.71
!
0.17
0.27
0.28
0.36
0.28
IT-1
IT-2
FR-1
FR-2
20
8000
30
fm (MPa)
El
0.60
0.62
0.69
0.62
0.67
fm
0.35
0.31
0.23
0.28
0.31
Sub-sample
IT-1
IT-2
FR-1
FR-2
All
8000
10000
12000
14000
16000
8000
10000
IP_MOE (MPa)
16000
(b)
IT-1
IT-2
FR-1
FR-2
450
8000
500
550
!(kg/m3)
10000
600
650
700
IT-1
IT-2
FR-1
FR-2
12000
14000
14000
IP_MOE (MPa)
(a)
Eg (MPa)
12000
8000
10000
12000
IP_MOE (MPa)
(c)
14000
16000
8000
10000
12000
14000
16000
IP_MOE (MPa)
(d)
Figure 2: Indicating Property (dynamic modulus measured by the machine: IP_MOE) vs (a) bending strength fm, (b) local
modulus of elasticity El, (c) global modulus Eg and (d) density !.
189
C18
18
9 000
320
C27
27
11 500
370
Strength class
C30
D24
30
24
12 000
10 000
380
485
D27*
27
10 500
510
D30
30
11 000
530
*D27 strength class will be included in the next version of the standard
Table 5: Number of pieces and yield as a results of maching grading (assigned grade) and a perfect grading based on the
laboratory measurements (optimum grading) for three D-classes combinations. REJ = reject.
Grade
combination
D24
REJ
D27
REJ
D30
D24
REJ
Assigned grade
N
(%)
577
99.1
5
0.9
478
82.1
104
17.9
180
30.9
397
68.2
5
0.9
Optimum grade
N
(%)
582
100
0
0
555
95.4
27
4.6
483
83.0
79
13.6
20
3.4
Table 6: Number of pieces and yield as a results of maching grading (assigned grade) and a perfect grading based on the
laboratory measurements (optimum grading) for two C-classes combinations. REJ = reject.
Grade
combination
C27
REJ
C30
C18
REJ
190
Assigned grade
N
(%)
577
99.1
5
0.9
511
87.8
66
11.3
5
0.9
Optimum grade
N
(%)
582
100
0
0
575
98.8
0
0
7
1.2
191
192
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)2013)
INTRODUCTION
Wood can be used for many applications because of its excellent material
properties. Thermal conductivity of wood is an important physical property for drying of
the lumber, heating logs in veneer mills, hot pressing composites, and insulating
applications like doors, windows, cladding, saunas and etc.
1
193
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
Ors and Senel (1999) have studied the thermal conductivity coefficients of the woods of
Cedrus, Fir, Populus nigra, and Beech. Results showed that the average thermal
conductivity coefficients in dry air conditions are 0.1128 for Fir, 0.1146 for Populus
nigra, 0.1253 kcal/mhC of Cedrus, 0.1580 kcal/mhC for Beech, 0.1783 kcal/mhC for
flat-pressed particleboards, 0.1998 kcal/mhC for MDF's and 0.2263 kcal/mhC for Hardfibre boards. Dubey et al. (1999) determined thermal conductivity of 11 Indian wood
species and a relation between thermal conductivity and both specific gravity and
temperature has been established. The results indicate that thermal conductivity of
hardwoods increases linearly with the increase in specific gravity as well as mean
temperature within the range from 18C to 48C. Ruiying et al. (2005) analyzed
similarity of some properties between thermal conductivity and electrical conductivity.
The results showed that the theoretical values of thermal conductivity was fairly identical
to their experimental values, in which the maximum error was 14.1% and mean error was
less than 7%. Their conclusion provided a feasible method to study thermal properties of
wood theoretically. Hunt and Gu (2006) obtained a new regression equation for radial
and tangential effective thermal conductivities as a function of density (porosity). The
results have applied to a two-dimensional soft wood board model for transient thermal
analysis. Stavros and Lazaros (2005) have drawn an artificial neural network model that
estimates wood thermal conductivity under a wide range of conditions of moisture
content, temperature, and apparent density. Wang et al. (2008) have studied heat
conductivity and tissue structure of 3-and 30-year-old populus tomentosa Carr. Their
results show that the temperature on leaf surfaces among the various varieties of trees
was almost the same. Leaf heat conductivity, temperature, and water content of leaves
are positively correlated. The thicker the leaf tissue structures the larger heat resistance.
In the study described in this manuscript, thermal behaviour of Platanus
Orientalis, Fagus Orientalis and Pinus Elderica, have been studied experimentally. A
number of experiments have been conducted to obtain empirical correlations for
prediction of thermal conductivity of the species in different ranges of temperature at the
steady-state condition using an ASTM C177 device.
EXPERIMENTAL
Material and Methods
The samples of each species have a size of 2020cm2 and 1cm thickness. The
samples were sawn radially and all of them were tested in the same condition with
EMC=12%. The density of Platanus Orientalis, Fagus Orientalis and Pinus Elderica,
was found to be 0.681025, 0.656675 and 0.561325 gr/cm3, respectively.
In this experiment a flexible heater with 200 mm width, 200 mm length and thickness of
2mm was utilized as heat source and two cold plates with width of 200 mm, length of 200
mm with thickness of 30 mm employed for cooling another side of the samples. The
heater, cold plates, and wooden samples, were located in a wood frame and fixed to each
other with a spring loaded nut. They were insulated by polyurethane to prevent heat loss.
Various amount of heat loads was provided by a power supply. Also, a Lutron DW-6060
wattmeter was used to measure the exact value of the input power. Two PT-100
2
194
T
x
(1)
Where Q is heat load, k is the thermal conductivity of samples; A is the area of samples
and T is temperature difference between both sides of the samples in the steady state.
Rearranging Eq. (1) to obtain k gives:
k
Q. x
2 A T
(2)
For the cases studied here, x 0.005 m and A 0.04 m 2 . Therefore, the above equation
gives:
k 0.0625
Q
T
(3)
195
RESULTS
Thermal behavior of the samples has been studied at various temperatures.
Figures (2) (4) depict the temperature difference (TD) between the hot and the cold side
of the samples, and the average temperature (AT) versus time in different values of heat
flux. Figures show that TD and AT increases over time and heat flux. Also, it can be
seen that increasing the heat flux leads to decreasing the time of reaching steady state
conditions. The results indicate that Pinus Elderica has the higher thermal resistance
rather than Platanus and Fagus Orientalis. After considering the thermal behavior of the
three samples in Figures 2, 3, and 4, we consider the thermal conductivity of the samples
in the steady-state condition in different heat fluxes and AT.
Fig. 2. a) Variation of average temperature and b) temperature difference between hot and cold
sides of Platanus, versus time
196
Fig. 3. a) Variation of average temperature and b) temperature difference between hot and cold
sides of Fagus Orientalis, versus time
Fig. 4. a) Variation of average temperature and b) temperature difference between hot and cold
sides of Pinus Elderica, versus time
Table.1 shows the results for the Platanus Orientalis samples. It can be seen from the
table that thermal conductivity of Platanus Orientalis varies from 0.209 W/m. oC at 35.95
o
C to 0.216 W/m. oC at 52.70 oC. Tables 2 and 3 display the results of experiments for
5
197
Fagus Orientalis (Table 2) and Pinus Elderica (Table 2), respectively. The tables show
that thermal conductivity of Fagus Orientalis, changes from 0.203 W/m. oC at 38.45 oC
to 0.211 W/m. oC at 57.90 oC and for Pinus Elderica varies m 0.139 W/m. oC at 40.70
o
C to 0.190 W/m. oC at 57.30 oC, respectively.
Table 1. Measurements for Platanus at steady-state condition
Experiment No.
1
2
3
4
Average
Temp.
(C)
Q
(Watt)
(C)
35.95
42.80
47.70
52.70
46
66
81
95
13.7
19.2
23.4
27.4
Ave. Thermal
conductivity, k
(W/m.K)
0.209854
0.214844
0.216346
0.216697
Average
Temp.
(C)
Q
(Watt)
(C)
1
2
3
4
38.45
46.15
51.7
57.9
46
64
78
94
14.1
19.5
23.2
27.8
Ave. Thermal
conductivity, k
(W/m.K)
0.203901
0.205128
0.210129
0.211331
Average
Temp.
(C)
Q
(Watt)
(C)
1
2
3
4
40.7
47.35
51.9
57.3
45
64
77
90
20.2
22.3
25.6
29.6
Ave. Thermal
conductivity, k
(W/m.K)
0.139233
0.179372
0.187988
0.190034
The curve fitting technique was used to fit a line to the experimental data using the
software MAPLE 11 between 35-60 oC. It is obvious from the figure 5 that Platanus
Orientalis has the highest thermal conductivity in all temperatures in compare with the
other two species. Another point is that thermal conductivity of Fagus Orientalis and
Platanus Orientalis increases linearly with AT while it is parabolic for Pinus Elderica.
198
Fig.5. Variation of Platanus, Fagus Orientalis and Pinus Elderica thermal conductivity versus
average temperature.
Accurate correlations for thermal conductivity of each species can be seen in Table 4
which were obtained with the most accuracy using mathematical software MAPLE 11.
Table 5. Empirical correlations for prediction of thermal conductivity
Empirical correlation for Thermal conductivity (W/m.K)
species
Platanus
Fagus Orientalis
Pinus Elderica
DISCUSSION
Thermal behavior of three wood species has been studied in both transient and steadystate conditions and three useful correlations are presented for predicting the thermal
conductivity of the species in the applicable temperature range. As expected average
temperature and temperature difference between hot and cold sides of increased with time
and this is due to the fact that increasing the heat the molecular motion of the wood
7
199
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
elements is increased and leads to releasing more energy and subsequently increases the
temperature. Totally, the advantages of the work can be drawn as below:
1- Investigation of temperature variant through Platanus Orientalis, Fagus
Orientalis and Pinus Elderica in different heat fluxes for the first time.
2- Finding thermal conductivity of the species as an important feature for insulation
or heat transfer applications. In future one can find this parameter without
conducting any experiment which need spending time and equipments in that it is
easy to use the empirical correlations.
3- It was found that which species between theses three industrial and important
woods are more suitable for any heat transfer process or industrial usage.
All in all the experiments lead to finding accurate and useful data for researchers in the
field and are also useful for industries and are easy to understand.
CONCLUSIONS
Thermal behaviour of Platanus Orientalis, Fagus Orientalis and Pinus Elderica is
obtained experimentally using ASTM C177. The results show that thermal
conductivity of these woods increases with temperature. It was found that Platanus
Orientalis has the highest thermal conductivity due to its highest density in respect
to Fagus Orientalis and Pinus Elderica. Hence, Pinus Elderica is more suitable for
insulation applications in compare with the other rwo species.
The study of thermal conductivity of woods is always of interests for researchers due
to the wide application of wood materials in industries and results of this paper can
be useful for experts in the field to have more efficient systems.
ACKNOWLEDGMENT
The author would like to thank the grant that supported the work on behalf of
University of Tehran and Alborz wood research center.
REFERENCES CITED
Bean, W., (1981), Trees and Shrubs Hardy in Great Britain, Vol. 1 4,.
Ors, Y., senel, 1999, A., Thermal Conductivity Coefficients of Wood and Wood-Based
Materials, Turkish Journal of Agriculture and Forestry,23: 239-246.
Dubey, Y.M, Jain, JD., Gupta, S., 1999, Studies on thermal conductivity of some Indian
hardwoods. Journal of the Timber Development Association of India, 45(1-2), pp, 33-38.
8
200
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
International
Scientific
Conference
on Hardwood
Processing
(ISCHP
2013)
Chen, R. Y., Xie, Y. Q., Yang, Q. X., Liao, Y. Q., Lin JG. Study on wood thermal
conductivity in transverse direction by analogism, Fujian Agriculture and Forestry
University, Fuzhou 350002, China, Chinese Academy of Forestry.
Hunt JF, Gu H, Two-dimensional finite element heat transfer model of soft wood, Part I ,
Effective thermal conductivity, wood and fiber science, 2006, 38(4), pp, 592-598.
Stavros A, Lazaros I, Predicting wood thermal conductivity using artificial neural
networks. Wood and fiber science, vol. 37, No.4, 2005, pp, 682-690.
Wang M, Zhang WJ, Xiao J, Zhang ZY, Liu J Forestry Studies in China. Vol. 10, No. 3,
2008, pp, 173-178.
Zarr RR, Dalton GR, Fiorovante SM, Development of a NIST Standard Reference
Database for Thermal Conductivity of Building Materials, National institute of standards
and technology.
201
Department of
Architecture and Design.
Politecnico di Torino
Turin, Italy.
Department of
Agricultural, Forestry and Food
Sciences. University of Turin
Turin, Italy.
ABSTRACT
The paper illustrates the results of an interdisciplinary work flow concerning the design and construction of a
prefabricated structure using poplar plywood. The study, promoted by Politecnico and University of Torino, is
related to the development of a local wood resource focussing on the most traditional and important panel in Italy
which, from a consolidated position in the furniture market, is now searching potential and innovative applications
especially in the building sector, where the use of modern wooden materials is growing quickly. The Mona Lisa
Wood Pavillon investigates the wood design issue from three main points of view. First, the structure is part of the
promotion path of poplar plywood, conceptualizing its manufacturing process in an architectural project; then it
highlights the role of technology by using an engineered veneer-based product with interesting properties; finally
it explores the role of digital design in exploiting the potentialities of the material, in particular through a
parametric approach which permits the optimization of design and construction. The project was carried out
thanks to the joined efforts of complementary sectors and to the coordinated work of architects, engineers and
wood experts. The resulting architecture was built on the occasion of the 2012 Made Expo event, with the
patronage of the Italian woodworking associations Federlegno, Assopannelli and Conlegno, together with the
technical contribution of several industrial companies, which have sustained this research with the aim of
developing a manifesto for a better knowledge of poplar plywood.
1. INTRODUCTION
The Mona Lisa Wood Pavilion is the outcome of a research project developed by Politecnico di Torino, University
of Torino and some institution and enterprises of the Italian wood and plywood sector. Starting from the Mona Lisa
workshop, we want to foster an experimental study, referring to a number of international teaching models, that exam
the potential of computational technologies in architectural projects producing three-dimensional structures with flat
elements in plywood. The initiative is part of the more general framework of experiences, part didactic and part
research, that the Department of Architecture promoted with particular reference to the process of innovation in
construction technologies using wood.
202
The Mona Lisa workshop took the name of Leonardo da Vincis famous painting, which was produced on a poplar
board between 1503-1514, as its icon and cultural reference, in order to highlight the theme of valorization of poplar
plywood in the area of technological and design research, this being the aim pointed out by the industrial partners in the
initiative (IBL, Panguaneta, Invernizzi, Cima, Vigolungo, Toro Compensati).
The merger in Leonardos work between art (ability) and science (knowledge, theory) was the cultural basis of the
research project developed during the workshop, which focused on the creation of a manifesto on the value and
potential of poplar in architecture. The enigmatic expression of Lisa Gherardini, the Gioconda also known as Mona
Lisa reveals how Leonardo's approach to scientific knowledge was visual, considering the eye as his main instrument
and drawing as the resource to formulate his conceptual models. Through his activity, Leonardo da Vinci made it clear
for the first time that the field of science and the field of applications were inseparable. His work reveals a concept of
design and procedure that is extremely modern, where the inventor is someone who creates a product or a work of art
by combining the various elements into a new configuration never to be seen in nature, through continuity between the
process of design i.e. the abstract configuration of various elements and their material incorporation (As made clear
by Fritjof Capra in his text The Universal Science: art and nature in Leonardos genius, Leonardo da Vinci was more
interested in the process than in the physical realization of his projects. Most of the machines and mechanical devices
he described in his drawings and technological sketches were never built. Few of his plans of military and civil
engineering ever became concrete. Even as a painter he often seemed more interested in the solution of problems of
composition than in the effective completion of the work).
For Leonardo, painting is art, but also science, a science of natural shapes based on the study and observation of
living shapes, continuously molded and altered by innate processes: rocks and sediments of the earth shaped by water,
the anatomy of animal bodies in movement, the development of plants, molded by their metabolism and botanical
diversity.
This dimension of investigation formed the basis of the work described hereafter, and turned out to be fundamental
for a structured workshop, even more than for the goal of developing an architectural project on the study and analysis
of the specific characteristics of a wood species like poplar with the features and peculiarities that determined the nature
of the Mona Lisa Wood Pavilion.
Like other similar international experiences taken as reference, the design aspect was developed starting from an
in-depth analysis of the poplar products for construction prepared by the industrial partners of the initiative, which
emphasized and focused on the development of panels combining functional and innovative features, capable of
satisfying the increasing demand for specific technical and performance characteristics.
The activity was structured in the form of a project workshop, through the coordination between the courses of
architectural technology (prof. G. Callegari with arch. A. Spinelli, Politecnico di Torino) and construction methods
(Prof. M. Sassone, Politecnico of Torino), and involved students at the Department Architecture and Design of
Politecnico di Torino in the three levels of training (Bachelor of Science course in Architecture, Master Degree in
Urban Architecture Construction, Ph.D in Architecture and Building Design), organized in the sphere of the Woodlab
project serving for the cultural promotion of the initiative ( financed by funds for students project at Politecnico of
Torino, aims to create an interest and network of people and institutions around the theme of innovative use of wood in
construction).
203
From the standpoint of method, after a first stage of study of poplar, preparatory to the workshop, the activities
focused on the most significant international experiences of three-dimensional architectural design in plywood.
Computational design enables architects to integrate ever more multifaceted and complex design information, while the
industrial logics of conventional building construction are eroding rapidly in a context of increasingly ubiquitous
computer-controlled manufacturing and fabrication. A novel convergence of computation and materialization is about
to emerge, bringing the virtual process of design and the physical realization of architecture much closer together, more
so than ever before. (Menges, 2012)
This permits to optimize all the potentialities of material features and design techniques, allowing the freeform
research in the field of plywood. The particular need to create a structure that would exalt the peculiarities of poplar
wood (light weight, pale color, extreme uniformity with barely perceptible veining, ease of processing, gluing and
finishing) centered the initial part of the workshop activities on the production and processing of plywood, and
characterized the end product.
Poplar has by now achieved a certain maturity in the field of plywood, in terms of development (for the high levels
of productivity attained) and market. It is therefore possible to identify certain margins for improvement with reference
to the potential for innovation determined by its diversification, through alternative uses in non-traditional sectors and
the development of products with higher added value, capable of optimizing its specific features in respect of the needs
of conformity to qualitative and performance standards, reliability and eco-compatibility (Zanuttini, 2003; Castro,
Zanuttini 2008). It is in this concrete perspective that the experience of the workshop acquires relevance through the
realization of a showpiece product the Mona Lisa Wood Pavilion serving to create a manifesto on poplar and
explore the uses of this wood species in architecture
2 POPLAR FEATURES
Poplar is a fast-growing tree, of many species/clones, easy to propagate vegetatively, and highly adaptable to
different climatic and soil conditions. These attitudes, combined with the wide range of wood, fiber, fuel and other
products and services provided, have led to the widespread cultivation of poplars, mostly in plantations, at present in
Europe the total amount of Poplar cultivated area is about 1 million hectares (Coaloa 2009). Considering that
plantations, which can be certified in terms of sustainable management, can significantly contribute to the reduction of
the pressure on natural forests, it is estimated that in the near future most of the timber for the wood industry will derive
from dedicated stands grown in an arboricultural system.
In Italy poplar is still the only renewable domestic source of wooden raw material which can be planned, capable of
meeting the needs of industry over the (relatively) short term, since it has a growth cycle of 10-12 years. Poplar wood
products have been confirmed as highly competitive and hard to replace both in the furnishing sector, where they are
used in various forms (mainly plywood and particle-board), and for some types of packaging for fruits and vegetables.
The CRA-PLF (Research Unit for Intensive Wood Production) deals with studies on various fast growing species
(such as willow, eucalyptus and black locust), suitable for intensive wood production for industrial uses, but its main
activity is focused on poplar (Castro and Zanuttini, 2008). Here, research work is chiefly aimed at developing new
poplar clones with particular qualities such as quick growth, disease resistance and high-quality wood that is adaptable
to varying soil environments. A few of the poplar clones obtained, after having successfully undergone experimentation
204
in Italy and France, have been recently protected by Community plant breeders rights, and, being registered in the
National Register of Forest Clones (RNCF), they are now available for commercialization in Europe.
The last part of the breeding activity, which can only be carried out on 9-10 years old plants, involves the assessment
of the wood suitability for industrial applications. This is partly done through laboratory tests (to assess technological
characteristics) and partly in cooperation with plywood industries (to evaluate the peeling attitude, as this is by far the
most important transformation for poplar wood in Italy), following the method in use at the CRA-PLF (Castro and
Miegge, 1999; Alga et al., 2008).
In fact, among several possible uses of the Italian poplar, plywood is the most important; being by far the most
remunerative, it also represents the main target for poplar growers. Poplar plywood is nearly always destined to the
furniture sector (including motor-home applications) because its low density and light colour are very appreciated and
practically impossible to achieve with alternative, more economic or easier to work materials.
However, plywood is already a mature product for which further developments and applications are hard to
imagine, it suffers from the ever increasing use of new and competitive wood-based products (e.g. OSB) and the
number of technical applications whose strict requirements are hardly met by other panels are rather limited
However it must be pointed out that the reason why the range of potential applications of poplar wood is still quite
small, lies not in its weak points (like low natural durability and quite poor mechanical properties) but mostly in the
fact that its use as solid wood or in wood-based panels for structural purposes is still quite limited in many Countries,
like Italy, where poplar is traditionally cultivated in specialized plantations. In many other geographical areas,
however, - Canada, United States, France, Germany and some North-European countries among others - the use of
poplar wood for structural applications is much more frequent, and solid wood (typically beams) is often substituted
with the so called engineered products, specially designed to meet specific performance requirements.
The main reasons of their success are the following:
Optimization of physical-mechanical behaviour and reduction of wood variability, obtained through selection of
the raw material and a controlled re-composition, in order to obtain more isotropic and homogeneous final products;
Improvement of the durability, reached by chemical or physical means (i.e. treatments or choice of adequate
adhesives).
Thanks to these advantages, these products - plywood included - have been able to gain important sectors of the market.
In this way a product considered a commodity (semi-finished product for general purposes) can be changed into a
specialty (i.e. a service-product aimed at satisfying a well defined/specific use) (Castro and Fragnelli, 2006).
As an example, the following diagram shows the mechanical performance of plywood made with the new clones
and with I-214; it can be easily seen that some of the new clones show significantly higher bending strength, while at
the moment none has a comparable low density (and light colour).
Figure 4_Resistance of plywood specimens made by different clones from recent selection in Italy (average values) and
comparison with I-214. The basic density refers to solid wood.
In this context, poplar wood is much appreciated for its visual aspect and for the high-quality finishing achievable.
Due to precise breeding choices and to the intensive cultivation techniques adopted, it generally shows low density and,
205
accordingly, poor mechanical properties (even if compared with aspen, yellow poplar or other similar wood species
which grow in a forestry environment). However, thanks to its high performance-to-weight ratio, its structural
efficiency is very interesting and, should the market require poplar wood with improved strength properties, farmers
would be ready to plant new clones, other than the widely known I-214. These, in addition to growing faster, present
a denser and more resistant wood (20 30 % stronger than I-214, and even more) but at present they are not cultivated
on a large scale because their wood is less appreciated by the plywood manufacturers for the traditional use in the
furniture sector. The mechanical properties of the plywood obtained from the above clones, according to the results of
preliminary tests, proved to be significantly higher than those of analogous panels made with I-214; together with the
greater amount of raw material obtainable, this makes the new clones particularly suitable for innovative uses, in
particular in the building sector, with or without structural functions, in which the performance aspects are determining.
2. MONA LISA PAVILION WORKSHOP: A FRACTAL FOREST
Contemporary architecture of the 21st century had to face the trend towards complex forms. The recent
development of cutting-edge computational parametric design, digital fabrication and rapid prototyping systems
developed this taste in present architectural styles enormously and rapidly. Mona Lisa Pavilion is an outcome of the
WoodLab workshop which was started with the intention to produce an architectural showpiece that can use advanced
digital-aided design and fabrication technologies and become contextual to the contemporary trend of architectural
styles. The aim of the workshop was to produce an architectural showpiece that will serve to create a manifesto on
poplar and to explore the uses of this wood in architecture.
2.1. DESIGN CONCEPT: RECREATING A POPLAR FOREST
The main objective of the workshop was to promote poplar plywood and to exhibit its versatility in an innovative
way for creative purposes in the industrial sector. The intention was to showcase the unique quality of poplar plywood,
which is smooth in color, lightweight, flexible and easy to cut. To achieve this challenging task, design ideas were
communicated through sketches. The concept development exercise started with a thread, cannot an architectural
piece be a visual narration of a story of the production of construction material itself, if its intention is to emphasize the
materials usefulness and versatility? Poplar sprouts from its seed, and then gradually grows into a young plant and
finally a full-grown woody tree. Poplar trees altogether exist in a plantation like a family, provide shading to others and
finally they become useful for making plywood needed for manufacturing furniture, interiors and architectural objects.
The growth and age of a poplar tree is recognized by the increasing of its height, trunk radius and number of its
Figure 4_ Replication of poplar tree growth. Left: poplar growth annually and its representation by design elements. Right: front view of
Fractal Pavillon.
branches. All these elements, which mark poplars growth, were symbolically taken as design vocabularies to narrate
the design concept. Therefore, the idea was to recreate a poplar plantation as a human designed pavilion that can shelter
us and provide furniture for human needs. However, this whole conceptual idea had to be transformed into a design
model.
2.2. DIGITAL DESIGN & PARAMETRIC MODELING
206
branches. All these elements, which mark poplars growth, were symbolically taken as design vocabularies to narrate
the design concept. Therefore, the idea was to recreate a poplar plantation as a human designed pavilion that can shelter
us and provide furniture for human needs. However, this whole conceptual idea had to be transformed into a design
model. Scientific Conference on Hardwood Processing (ISCHP 2013)
International
2.2. DIGITAL DESIGN & PARAMETRIC MODELING
The first step of transforming the idea into a design model was done through hand sketch drawings. For this
workshop project, sketching played the foremost and most significant role, helping students to externalize their idea
into a design model. They dedicated the longest hours of this workshop to this brainstorming phase, because thinking
and simultaneously drawing sketches is a tool to solve problems, create new ideas and assist communications (Hank
and Belliston, 1992). Although sketching eases the process of transforming the idea onto a sheet at early concept level,
an additional device is necessary to take the design to a more detailed level. Recent development of digital design
makes it convenient to develop a detailed design from sketches. The first conceptual model was generated in SketchUp
for quick visualization and understanding of the design concept transformed by hand sketches (Figure 5). However,
advanced development of computer aided design using parametric modeling and computational techniques as powerful
tools take this concept externalization stage a step further for form generation and optimization based on defined
parameters and rule-based constraints.
Figure 6_Parametric modeling in Rhinoceros, (a) A tree unit with branches, (b) Parametric modeling of pavilion
profile, (c) Parametric modeling of sitting benches, (d) Unrolled components on XY plane ready for CNC cutter.
For generating a parametric model of the pavilion, Grasshopper scripting was used in Rhinoceros. This was done in
two different parts. First part dealt with generating tree unit with branches (Figure 6a), whereas second part for
modeling two long benches that defined the pavilion margins (Figure 6c). In the first part, branches were produced by
fractal logic of iterative process, bifurcating a trunk into two branches; each branch was then again split into two
branches and so on until the third iterative steps. The aim of the branch production process for making a tree unit was
to follow the growth phenomena of poplar tree, thus imitating nature. However, this part was a challenging task for the
students, because the tree units had to be designed such a way as to make it possible to realize them using real poplar
plywood, showing its versatility, but taking into consideration its strength and weakness points. Moreover, the
branching pattern needed to be parametric so that, if necessary, it was possible to achieve the desired configurations of
tree units by changing the parameters, taking into considerations the low cost, less material and simple
constructability in mind. In the second part, making benches was an easy but a tricky task. Conceptually, each piece of
bench had to represent a poplar seed and its successive growth until its final use as furniture, i.e., a complete sitting
bench. By using simple Grasshopper scripting, students could easily produce the benches in the desired, although
comfortable, sitting. Parametric tool allowed students to optimize the number, sizes and thickness of the bench pieces
easily keeping costs low and using the minimum amount of material. However, the tricky task in modeling the benches
was to merge them with the base of each tree unit, which was a part of the whole sitting bench, with the additional
backrest. After completing the digital model of the complex shaped Mona Lisa Pavilion (Figure 6b), each design
component was unrolled on XY plane to make it ready for cutting in a CNC cutter (Figure 6d).
207
2.3. REALIZATION
As soon as the parametric modeling of the pavilion was finally completed on the computer, a small scale (1:20)
physical model was made (Figure 7a) to assess its physical behavior: every piece was cut in a CNC cutter in FabLab
Torino, and all pieces were assembled to get the final physical outcome. No serious problem appeared because of the
smaller scale, except the heavy exposure of too many metal bolts used for jointing pieces. But, in making real scale
physical prototype, we faced two main challenges. The first challenge was to create one long but complete strip by
jointing two different pieces of poplar plywood boards. This was done through loose tongued joints (Figure 7b). Four
such strips were jointed together by screwing tightly with brass screws, so that the colors of poplar and brass screws
could merge to avoid the heavy exposure of metals (Figure 7c) However, the joints had to be made strong enough to
carry the load of the upper part of the strips, i.e., the group of branches which acted as cantilevers.
Figure 7_ (a) Small scale (1:20) physical model, (b) Loos tounged joint, (c) Connection of two ply pieces.
For physical testing of internal stresses and bending behavior, students jointed two larger curve-shaped strips in the
University workshop building and then pulled the tails of each strip manually for getting large deflection bending in
order to obtain the curved branching configuration. At certain level of stress a crack appeared at the bent area due to
torsion effect. Therefore, jointing plywood pieces was not the only critical aspect since ensuring the crack-less large
deflection of bending was the another main challenge. Accordingly, in real scale production of one tree unit, screws
were placed diagonally at turning position but not perpendicular to the tangent of bending curve so that it would not
only ensure crack-less bending but also offer a relaxation space in torsion. One of the most important elements was the
wedges for generating branches. A triangular wooden wedge was produced on the basis of the dimensions provided by
the Rhinoceros model of a branch (Figure 6a) and then it was inserted between two strips, thus the wedge forced the
strips to deform outwards in opposite directions, forming the branch. The sides of the wedges were gently curvilinear so
as to give branches a natural shape (Figure 8).
Figure 8_ Creating branches by inserting curved triangular wedges for making physical model and protypes.
The next step of the workshop event was to start realizing all individual components for making a real scale
prototype of two tree units. After cutting all pieces in a CNC cutter, all the components were assembled and connected
to form two different tree unit prototypes. Two units were then placed facing each other to observe their structural
stability as well as their aesthetical appearances (Figure 9). The final and last challenge was yet to be faced: realization
and installation of the complete structure in real scale. After a long period of designing, modeling, structural testing and
making prototype units, final production of all units was done in an industrial workshop building by the industrial
participants and supervised by WoodLab team from Politecnico di Torino. Then all assembled units were transported
to the exhibition center in Milan two days before the event of MadeExpo2012 exhibition. It took one full day to
fabricate and install the final real scale pavilion structure on the event site, and on the first day of exhibition it was
finally ready for inauguration.
208
Figure 9_ Production progress of real scale tree unit in the university workshop.
The final and successful outcome of the Mona Lisa Pavilion and the overwhelming appreciations by visitors at
exhibition centre rewarded the WoodLab team of all the work done: right from the evolution of their design concept
through conceptual sketching to parametric modeling, and from small-scale prototype testing to final real scale
fabrication and installation. In a way, it was the success of WoodLab workshop that offered students multi-dimensional
lessons and the occasion to learn some parametric and fabrication skills.
Poplar plywood could be a key leverage point in the field of sustainable wood structure, enhancing great potential in
architectural design and engineering concepts. Computational design could achieve the performance of the material
and translate in input for the project, permitting a close relation between functionality and sustainability.
ACKNOWLEDGEMENTS
The Mona Lisa project has been possible thanks to the contributions of many individuals, institutions and private
organizations such as the Politecnico di Torino, Conlegno, Federlegno, Assopanelli, Pro-Populus, FabLab Torino, the
C.R.A.-PLF Institute in Casale Monferrato and the University of Turin. An important role in the fate of the project was
played by the companies who have supported the initiative, not only through sponsoring, but also with technical support
for the works, identifying certain key aspects. For a complete list of companies see the official website of the project
(www.woodlab.polito.it).
REFERENCES
Aichim Menges (2013) Material Computation Higher Integration in Morphogenetic Design, March-April 2013
Alga R., Castro G., Zanuttini R. (2008). Otto nuovi cloni sperimentali di pioppo: prime valutazioni sulle caratteristiche
tecnologiche. Sherwood, Foreste e Alberi Oggi n 41, marzo 2008: 5-10.
Callegari G., Sassone M., Spinelli A. (2013). Wood digital works, Milano, Edizioni Ambiente.
Callegari G., Sassone M.,Spinelli A., Zanuttini R., Mona Lisa Wood Pavilion: poplar pl(a)ywood, (2013) ICSA2103
(International Conference on Structures and Architecture) Proceedings, Portugal, 24-26 luglio, 2013
Castro G. Zanuttini R. 2008 Poplar cultivation in Italy: history, state of the art, perspectives, in 23rd session of the
International Poplar Commission Beijing
Coaloa D. (2009) La pioppicoltura in Italia e nel mondo, stato attuale e prospettive future. In Dendronatura 30: (1)
21-26
Gasparini P. Tabacchi G. (2011). LInventario Nazionale delle Foreste e dei serbatoi forestali di Carbonio INFC 2005.
Secondo inventario forestale nazionale italiano. Metodi e risultati. Ministero delle Politiche Agricole, Alimentari e
Forestali; Corpo Forestale dello Stato. Consiglio per la Ricerca e la Sperimentazione in Agricoltura, Unit di ricerca
per il Monitoraggio e la Pianificazione Forestale. Edagricole-Il Sole 24 ore, Bologna. 653 pp.
Castro G., Miegge D. (1999). Caratteristiche fisiche e idoneit alla sfogliatura di alcuni cloni di pioppo. Sherwood,
Foreste e Alberi Oggi n 5 (4): 41-44.
Castro G., Fragnelli G. (2006). New technologies and alternative uses for poplar wood. Boletin del CIDEU n. 2: 27-36.
Castro G., Zanuttini R. (2008). Poplar cultivation in Italy: history, state of the art, perspectives. Cost Action E44
Wood Processing Strategy. Final Conference Milan, ITALY May 30th - June 3rd 2008: 141-154.
Zanuttini, R. 2003 I moderni pannelli a base di legno. I.F.M.. LVIII(1), 34-53
Hanks & Belliston. (1992). DRAWA visual approach to thinking, Learning and Communicating
209
Holzforschung Mnchen
Technische Universitt Mnchen
Winzererstrasse 45, 80797 Mnchen, Germany
ABSTRACT
Within the last years research focused on the investigation of beech lumber as a material for load bearing
structures. As a result technical approval for beech glulam and beech hybrid glulam could be achieved in Germany.
However, despite several advantages, the price of beech glulam in comparison to spruce glulam, is still too high,
creating a barrier to market acceptability. This high price is due to the high cost of lamellas production because the
yield of sawn timber is low. This paper analyses the production of beech lamellas for glued laminated timber.
Roundwood of different dimensions and qualities was cut using three different sawing technologies. Lamellas with
different cross sections were kiln-dried and planed, then visually and mechanically strength-graded. The entire
manufacturing process was evaluated with regard to yield, quality and product cost of the lamellas. The results
show that the yield was low in general and depending both on sawing technology and roundwood quality. A value
of 17 % was observed for a profiling unit. A value of 22 % was achieved virtually by cutting all around the log with
a band saw, leaving a heart plank which contains the pith. The low yield increases the product cost to more than
700 /m.*
1. INTRODUCTION
1.1. BEECH GLULAM
Glued laminated timber (glulam) was developed in the beginning of the 20th century (Rug 2006). Due to the good
workability and availability of roundwood more than 90 % of glulam produced in Europe is made of spruce and pine
(Mack 2006). The production of glulam with softwoods can be considered as a technically mature process, based on
know-how with regard to stress grading, finger jointing an bonding of the lamellas.
To enhance strength and stiffness of glulam the use of hardwoods has been discussed already in the 1960s (Egner
and Kolb 1966). First systematic investigations were carried out in Switzerland. A test series with finger jointed
lamellas and glulam made of beech was performed, giving first impressions concerning the behavior of lamellas and
glulam under load (Gehri 1980, 1985). Later Frhwald et al. (2003) performed bending tests on small glulam beams
made of beech and glulam hybrid beams made of spruce in the inner and beech in the outer lamellas. The results
confirmed the high strength values obtained by Gehri. However, predicting the strength of beech glulam and beech
hybrid glulam was not possible, because no information about strength grading and strength of finger joints of
lamellas was available. This changed when Blass et al. (2005) developed models that allow for predicting the behavior
of beech glulam beams under load.
One of the most important requirements for the use of hardwood for structural elements are adhesive systems that
provide a durable bonding during service life. The extensive research work of Schmidt et al. (2010) result in the
development of an adjusted bonding technology for beech that meets the requirements given in adhesive standards
(EN 301).
210
211
Diameter in cm
25 d 49
25 d 49
25 d 69
Length in m
4
3
3-5
2.2. PROCESSING
SAWING, DRYING AND PLANING
For all three studies the cross sections of the lamellas where chosen to meet the requirements of the technical
approval for beech glulam (thickness 20 mm 40 mm, width 80 mm 150 mm). The boards were cut with appropriate
oversize considering shrinkage and waste due to drying and planing. The cross sections of both boards and lamellas
are given in Table 2.
Table 2: Cross sections of boards and lamellas
Study
I
II
III
In study I (Figure 1a) the logs were shaped to an angled model by a profiling unit. Then, side boards of various
thicknesses (side product) were cut from the model, making it square in cross section. Finally the Model was cut
by a circular saw into square edged boards (main product). The cross sections of the boards varied depending on the
log diameter. Boards with a width of 215 mm and 255 mm, respectively were cut lengthwise into narrower boards
before planing.
In study II (Figure 1b) a vertical band saw was used. To increase the yield of sawn timber boards of 30 mm
thickness were cut of the sides first for other purposes where possible. Thereafter unedged boards were cut from two
sides of the log and sharp-edged by a circular saw until the remaining model had a width of 140 mm. The model
was then cut into boards with the band saw.
In study III the logs were cut with a vertical band saw using two sawing patterns. By cutting all around, the log was
rotated every time at 90 after cutting a board (Figure 1c) until a heart plank with a cross section of approximately
80 x 80 mm and the pith inside remained. By quarter cutting boards were cut from one side up to the heart plank,
then the log was rotated at 90 . Afterwards the boards were cut lengthwise and square edged by a circular saw,
212
respectively. The cross sections of the boards varied depending on the log diameter. Both boards for lamellas with a
width of 85 mm and 105 mm were produced from logs with a diameter smaller than 40 cm while boards for wider
lamellas were produced from logs with a diameter higher than 40 cm.
All square edged boards were kiln-dried to a moisture content of approximately 12 % in conventional dryers and a
vacuum dryer for half of the material in study III, respectively. To achieve a rectangular shape and to prepare the
surface for strength grading, the boards were planed on all four sides to raw lamellas.
Figure 1: Example of sawing patterns of study I (a), II (b) and III (c) for log diameter = 40 cm
STRENGTH GRADING
The technical approval for beech glulam allows for the use of visual strength grades LS10 and LS13 according to
DIN 4074-5. These grades are further broken down by knot area ratio and dynamic MOE for particular glulam
strength classes (Table 3).
Table 3: Grading requirements for lamellas for beech glulam
Class
Bending strength fm,k in N/mm
GL 28h
28
Grade
Edyn in N/mm
Knot area ratio (DEB)
LS10
Grade
Edyn
LS10
GL 32c
GL 36c
32
36
Outer lamellas (h/6 per side)
LS13
LS13
0.04
Inner lamellas
LS10
LS10
GL 40c
40
GL 44c
44
GL 48c
48
LS13
14000
LS13
15000
LS13
16000
0.04
LS10
13000
LS10
14000
LS10
14000
The lamellas were graded according to DIN 4074-5. The dynamic MOE was derived from natural frequency from
longitudinal oscillation measured by accelerometer according to equation (1):
Edyn 4 l 2 f 2106
in N/mm
(1)
where, Edyn = dynamic MOE, l = length in m, = density in kg/m and f = natural frequency in Hz.
2.3. YIELD AND ECONOMIC ANALYSIS
-
The volume of wood of the several production stages was derived as follows:
Roundwood: By mid diameter under bark and length.
Wet boards: By setting size of the saw (thickness and width), measured length and quantity of the boards.
213
Dry boards: By measured values for thickness and width and quantity of the boards. As length is considered to
remain constant (Knigge and Schulz 1966) no measurements were taken.
Planed raw lamellas: By setting size of the planing machine (thickness and width), measured length and quantity of
the lamellas.
Missing data were derived as follows:
The yield of lamellas produced from the side product by using the value for the main product (study I).
The waste (%) of production stage (4) of study III by using the value of study I. For that purpose the amount of raw
lamellas was reduced by the amount of lamellas containing pith.
The volumetric waste (%) of production stage (5) for study III by the value of study I.
A economic analysis was only carried out for studies II and III using the following prices and product cost as basis
for the calculation:
Roundwood 53.50 /m, sawing 45 /m, kiln-drying 50 /m, planing 25 /m and 44 /m proceeds from sale of
the waste for study II.
Roundwood 80 /m, sawing 65 /m, kiln-drying 55 /m, planing 60 /m and 44 /m proceeds from sale of the
waste for study III.
A
25
B
42
57
C
33
100
43
214
II
III
Product
Roundwood
Boards (wet)
Boards (dried)
Raw lamellas
Lamellas DIN 4074-5
Lamellas glulam
Roundwood
Boards (wet)
Boards (dried)
Raw lamellas
Lamellas DIN 4074-5
Lamellas glulam
Roundwood
Boards (wet)
Boards (dried)
Raw lamellas
Lamellas DIN 4074-5
Lamellas glulam
Production stage
(PS)
(1) Sawing
(2) Drying
(3) Presorting + planing
(4) Grading I
(5) Grading II
(1) Sawing
(2) Drying
(3) Presorting + planing
(4) Grading I
(5) Grading II
(1) Sawing
(2) Drying
(3) Presorting + planing
(4) Grading I
(5) Grading II
Main product
Main product + side product
Yield in
Waste in
Yield in
Waste in
m
%
m
% % PS m
%
m
% % PS
23.6 100 12.5 53
53
23.6 100 9.7 41
41
11.1 47
1.7
7
15
14.0 59 2.1
9
15
9.4
40
3.5
15
37
11.9 50 4.4 18
37
5.9
25
2.6
11
44
7.5 32 2.8 12
37
3.3
14
0.5
2
15
4.7 20 0.7
3
15
2.8
12
4.0 17
4.8 100 1.9
40
40
2.9
60
0.5
10
17
2.4
50
1.1
21
46
1.3
29
0.8
19
62
0.5
10
0.1
2
20
0.4
8
66.3 100 33.8 51
51
32.1 49 3.4
5
11
28.7 43 8.3 13
29
20.4 31 3.1
5
15
17.3 26 2.6
4
15
14.7 22
(1) Due to the production of the heart plank, which is not used for lamellas, study III shows the lowest yield of wet
boards (49 %). Furthermore it is noticeable that both study I and study II show the same yield (59 % resp. 60 %)
due to the same dimension of roundwood and similar sawing patterns (model) in spite of different technologies.
For cutting of beech roundwood with diameter between 18 cm and 30 cm Ehlebracht (1999) obtained a yield of
59 % for a profiling unit and 58 % for a horizontal frame saw combined with a circular saw, which basically
emulates the procedure of study II. For spruce a yield of approximately 70 % is reached, if processed by a profiling
unit (45 % main product, 25 % side product). This is a widely-used technology for the production of glulam
lamellas.
(2) After drying, the boards study I showed an averaged loss of 8 % for both thickness (3 mm) and width (14,5 mm on
average and 10 20 mm depending on primary width, respectively). The boards of study II showed a slightly
higher average loss of 10 % for thickness (4 mm) and 8 % for width (12 mm on average and 10 20 mm
depending on primary width, respectively). However, lower values were detected for study III for thickness (4 %
resp. 1.7 mm) and width (8 % resp. 10.4 mm with a variation of 6.2 9.4 mm). This agrees the values given in
standards (DIN 68100) for radial and tangential shrinkage fairly well due to annual rings with tangential grain
orientation within the boards. The absolute yield amounts to 50 % (I + II) and 43 % (III).
(3) Within this production stage boards of the studies I and II with high warp were cut to a length of at least 2 m.
Boards with too much warp were rejected in all the studies and the remaining boards were planed to raw lamellas.
Both rejection and planing caused high percentages of waste, especially in studies I (37 %) and II (46 %). As the
rejection in study III was very low, most of the waste (29 %) is produced by planing. This may be due to the better
roundwood quality, better drying quality or the effect of boards without pith. The absolute yield amounts to
32 % (I), 29 % (II) and 31 % (III).
(4) In general the waste within this production stage is very high: For study I it amounts to 37 % and for study II to
62 %. The influence of the pith becomes more apparent when raw lamellas containing pith are evaluated
separately. In study I 71 % and in study II 48 % of the waste was formed by lamellas containing pith while the rest
was rejected due to wane or warp for the most part. For study III no data for waste are available yet. But it was
observed that only a few lamellas still contain pith an adjusted calculative value of 15 % waste was derived from
study I as described above (amount of lamellas containing pith: 2 m).
The absolute yield of lamellas amounts to 20 % (I), 10 % (II) and 26 % (III).
215
(5) The absolute yield of lamellas for glulam according to the technical approval amounts to 17 % (I), 8 % (II) and
22 % (III). The value for study III was estimated by using the percentage of study I. Overall these values are low
compared to spruce, for which the yield of a profiling unit reaches approximately 40 %. The quality of the lamellas
was very high: 73 % of study I can be used for glulam class GL48c and 60 % of study II can be used for GL44c.
Furthermore 12 % of study I can be used for glulam class GL40c minimum, while the rest of both studies (15 %
and 40 %, respectively) is usable for class GL28h.
Study III showed the highest yield. Furthermore it remains to be seen if either the cutting method or the roundwood
dimensions and quality influence the result. Study II showed the lowest yield. The comparison of study I and study II
shows several similarities. Dimensions of roundwood are the same, the sawing patterns (model) are similar in spite
of different technologies and the yields of both wet and dried boards amount to the same. Hence it can be concluded
that the lower roundwood quality (100 % C) of study II leads to a lower yield of glulam lamellas.
The yield of lamellas would increase by 3 % (I), 2 % (II) and 4 % (III), respectively if lamellas of grade LS7 could
also be used for glulam, e. g. GL28. However, this proportion is not that high when compared to the amount of waste
due to reject and pith. Since the grading rules of DIN 4074-5 are very strict a revision as suggested by Leeb and
Hbner (2006) seems reasonable to increase the yield.
The yield can also be increased if lamellas with variable cross sections are produced from logs with the same
diameter at once. This is shown by study I, where the amount of glulam lamellas increased by 5 % if thinner lamellas
were produced from the sides of the logs. For study III it is observed that the yield would have been higher if the
unedged boards were cut more rationally by the circular saw using individual adjustments depending on the width of
the unedged board.
Frhwald et al. (2003) give a yield of 72 % for the manufacturing of spruce glulam from kiln-dried lamellas,
whereby it is not clear if the material is pre-graded. In this research work, the yield varies between 17 % (I) and 51 %
(III). Additionally the yield of kiln-dried lumber in relation to roundwood is normally higher for spruce due to both
longer and straighter logs and lower shrinkage of boards at drying. Hence the difference of the yield will become even
more evident.
3.3. ECONOMIC ANALYSIS
The calculation of product cost for the lamellas shows average values of about 1200 /m for study II and 740 /m
for study III. Compared to a price of currently 250 /m for spruce lamellas this is three times as much for beech at
best.
This research work is limited on lamellas for glulam. Any other side products are not considered and would
increase the overall yield, naturally resulting in a reduction of the price for beech lamellas. However, side products of
hardwoods in general including beech generate a low added value compared to spruce. Furthermore the studies
showed a higher yield with increase of log diameter and quality. In spite of higher prices the use of bigger hardwoods
is more efficient as the cost for processing decrease proportionally (Ehlebracht 1999).
4. CONCLUSIONS
Overall the investigations on the supply of beech lamellas for glulam showed a low yield, but an impact of sawing
technology was found. A relatively high yield of lamellas can be achieved by cutting all around the log and leaving a
heart plank containing the pith. As lower roundwood quality leads to a lower yield the roundwood quality should not
be too bad. However, it remains to be analysed if the amount of lamellas can be increased with a more efficient sawing
technology and better adapted rules for strength grading. The product cost for beech lamellas were found to be very
high in comparison to spruce but would decrease with higher yields. Furthermore it remains to be analysed if an
adapted processing speed for instance by using several band resaws in series and therefore, a higher throughput of
roundwood reduces product cost.
Acknowledgements The authors thank Caroline Fischer who collected the data for study I within her masters
thesis and Jonas Hnnebeck who collected the data for study II within his diploma thesis.
216
REFERENCES
Blass, H. J., J. K. Denzler, M. Frese, P. Glos and P. Linsenmann. 2005. Biegefestigkeit von Brettschichtholz aus Buche. Karlsruher
Berichte zum Ingenieurholzbau. Vol. 1. Universittsverlag, Karlsruhe. 134 pp.
Deutsches Institut fr Bautechnik (DIBT). 2009. Allgemeine bauaufsichtliche Zulassung. BS-Holz aus Buche und
BS-Holz-Buche-Hybridtrger. Zulassungsnummer Z 9.1 679
DIN 4074-5. 2008. Strength grading of wood Part 5: Sawn hard wood. DIN Deutsches Institut fr Normung e. V.
DIN 68100. 2010. Tolerance system for wood working and wood processing Concepts, series of tolerances, shrinkage and
swelling. Normenausschuss Holzwirtschaft und Mbel (NHM) im DIN Deutsches Institut fr Normung e. V.
Egner, K. and H. Kolb. 1966. Geleimte Trger und Binder aus Buchenholz. Bauen mit Holz 68:147 - 154
Ehlebracht, V. 1999. Untersuchung zur verbesserten Wertschpfung bei der Schnittholzerzeugung aus schwachem
Buchenstammholz (Fagus sylvatica L.). Dissertation. Schriftenreihe Agrarwissenschaftliche Forschungsergebnisse. Vol. 19.
Dr. Kovac, Hamburg. 276 pp.
EN 1316-1. 1997. Hardwood round timber. Qualitative classification Part 1: Oak and beech. Normenausschuss Holzwirtschaft
und Mbel (NHM) im DIN Deutsches Institut fr Normung e. V.
EN 1912. 2012. Structural timber Strength classes Assignment of visual grades and species. Normenausschuss Bauwesen
(NABau) im DIN Deutsches Institut fr Normung e. V.
EN 301. 2006. Adhesives, phenolic and aminoplastic, for load-bearing timber structures Classification and performance
requirements. Normenausschuss Holzwirtschaft und Mbel (NHM) im DIN Deutsches Institut fr Normung e. V.
Frese, M. 2006. Die Biegefestigkeit von Brettschichtholz aus Buche. Experimentelle und numerische Untersuchungen zum
Laminierungseffekt. Berichte zum Ingenieurholzbau. Vol. 5. Universittsverlag, Karlsruhe. 152 pp.
Frese, M. and T. Riedler. 2010. Untersuchung von Buchenschnittholz (Fagus sylvatica L.) hinsichtlich der Eignung fr
Brettschichtholz. Eur. J. Wood Prod. 68:445 - 453
Frhwald, A, J. B. Ressel and A. Bernasconi. 2003. Hochwertiges Brettschichtholz aus Buchenholz. Abschlussbericht. Institut fr
Holzphysik und mechanische Technologie des Holzes der Bundesforschungsanstalt fr Forst- und Holzwirtschaft, Hamburg
Frhwald, K. and G. Schickhofer. 2005. Strength Grading of Hardwoods. In: Proceedings of the 14th. International Symposium on
Nondestructive Testing of Wood, May 2 - 4 University of Applied Science, Eberswalde, Germany. pp. 199 - 210
Gehri, E. 1980. Mglichkeiten des Einsatzes von Buchenholz fr Tragkonstruktionen. Schweiz. Bauwirtsch. 56: 17 - 21
Gehri, E. 1985. Verbindungstechniken mit hoher Leistungsfhigkeit Stand und Entwicklung. Holz Roh Werkst 4:83 88
Glos, P. and B. Lederer. 2000. Sortierung von Buchen- und Eichenschnittholz nach der Tragfhigkeit und Bestimmung der
zugehrigen Festigkeits- und Steifigkeitskennwerte. Abschlussbericht Nr. 98508, Holzforschung Mnchen, Technische
Universitt Mnchen
Glos, P. and T. Nher. 2005. Aufnahme der einheimischen Holzarten Buche (Fagus sylvatica), Eiche (Quercus petraea, Quercus
robur) und Douglasie (Pseudotsuga menziesii) in die europische Norm EN 1912. Abschlussbericht Nr. 05510.
Holzforschung Mnchen, Technische Universitt Mnchen
Glos, P., J. K. Denzler and P. Linsenmann. 2004. Strength and Stiffness Behaviour of Beech Laminations for High Strength
Glulam. CIB-W18/37-6-3. In: Proceedings of Meeting Thirty-Seven. International Council for Research and Innovation in
Building and Construction. Working Commission W18 Timber Structures. August 31 - September 3, 2004, Edinburgh,
Scotland
Kollmann, F. 1951. Technologie des Holzes und der Holzwerkstoffe. Erster Band, zweite Auflage. Springer, Berlin Gttingen
Heidelberg. 1050 pp.
Leeb, W. and U. Hbner. 2006. Merkmale der Festigkeitssortierung von Buche und Esche. Schlussbericht. holz.bau forschungs
gmbh, Graz. 154 pp.
Leicht, T. and T. Lundt. 2007. Mrkte bei Brettschichtholz in Bewegung. Holz-Zentralblatt 15/2007:392
Mack, H. 2006. Der europische Markt fr Brettschichtholz (BSH). Ergebnisse einer Umfrage innerhalb der europischen
Leimholzindustrie. Wiener Leimholz Symposium 2006, Wien. pp. 35 50
Ranta-Maunus A., J. K. Denzler and P. Stapel. 2011. Strength of European Timber. Part 2. Properties of spruce and pine tested in
Gradewood project. VTT Working Papers 179. VTT Technical Research Centre of Finland. 113 pp.
Rug, W. 2006. 100 Jahre Hetzer-Patent. Bautechnik 83:533 - 540
Schmidt, M., P. Glos and G. Wegener. 2010. Gluing of European beech wood for load bearing timber structures. Verklebung von
Buchenholz fr tragende Holzbauteile. Eur. J. Wood Prod. 68:43 - 57
Schrder, S. 2007. Analyse der Strukturen und Innovationen der deutschen Laubholzsgeindustrie. Diploma thesis.
Albert-Ludwigs-Universitt Freiburg. Fakultt fr Forst- und Umweltwissenschaften
Knigge, W. and H. Schulz. 1966. Grundriss der Forstbenutzung. Parey, Hamburg Berlin. 584 pp.
217
Session III
Hardwood product
development
219
1. INTRODUCTION
Studies on the changes in wood at relatively low treatment temperatures, with the practical aim mainly to improve
the form stability, were started in the 1940s and were intensified in the 1960s 1970s, embracing physical properties
(Burmester 1973; Kollmann and Schneider 1963), the changes in the chemical composition (Sandermann and Augustin
1964) and biological durability (Stamm and Baechler 1964). The interest in thermal modification was resumed in the
1990s and resulted in the comparatively fast implementation of the technologies in industry. The crucial driving force
of the development of wood modification methods was the necessity of replacing, for ecological reasons, traditional
wood preservatives components (arsenic, chromium). One of the most important aims of all wood modification
treatments is the enhancement of the durability of wood, replacing biocides.
On an industrial scale, the ThermoWood process is used most often (annual volumes of production exceed 100 000
m3). The second most important process is the Plato process (The Netherlands). Both the above-mentioned methods
embrace wood hydrothermolysis thermal treatment of wood in the water vapour medium. In the rectification process
developed in France, the heating of wood proceeds in a neutral (nitrogen) medium. It is known that the
depolymerisation and hydrolysis of wood components is favoured by the presence of an abundant content of water
(Garrote et al. 1999). This principle forms the basis of the comparatively new WTT (Wood Treatment Technology)
method developed by a Danish company, which, in contrast to the above-mentioned technologies, is hydrothermal
one-stage treatment in water vapour medium at elevated pressure, which intensifies the chemical transformations that
*
220
occur in wood. Hence, the dynamics of the material properties changes are different from those in other hydrothermal
treatment processes.
In the present study, modification of deciduous wood in the WTT process was investigated to determine the
relationship between the treatment parameters and the properties of the obtained material. The potential application
fields of modified wood are wood products in elevated moisture and outdoor conditions (without the contact with soil),
respectively, in the medium favourable for the development of wood fungi.
2. EXPERIMENTAL
Hydrothermal modification (HTM) was carried out in a WTT experimental multifunctional pilot device in a water
vapour medium at temperatures of 160-170oC for 1 h or 3 h under elevated pressure (0.62-0.84 MPa) conditions. The
heat treatment was performed in one treatment stage with three sub-stages: (1) heating up to the modification
temperature; (2) holding at this temperature, and (3) cooling of the reactor. The time to reach the maximum pressure
and temperature was 12-14 h, and the total process time from the samples placing into the autoclave till taking them out
from it was 24-28 h. To ensure the necessary amount of saturated steam in the reactor, additionally, water was filled in
the autoclave. The amount of the filled water was calculated by taking into account the initial moisture content of wood
and the volume of the autoclave.
For modification, birch (Betula spp.), grey alder (Alnus incana) and aspen (Populus tremula) wood boards without
any visible defects were chosen. For each timber species, at least 16 boards were treated for a set of process conditions.
In contrast to the studies on semi-dry (W 20-30%) wood, carried out by Burmester (Burmester 1973), in our
experiments, the moisture content of wood before the treatment was 10-12%; for modified wood after the treatment, it
was within a range of 4-6%. The boards average sizes were: length 1000 mm, width 100-105 mm, and thickness 28-32
mm. Before and after modification, the mass and density of wood samples were investigated.
To determine the physical and mechanical properties of wood, the samples, before the tests, were conditioned till
the equilibrium moisture at (20 2)oC/(65 5)% RH. Bending strength properties of wood were determined in a
3-point bending test according to DIN 52186 (specimen with the size 360 x 20 x 20 mm, conditioned according to the
DIN 50014 requirements at (20 2)oC/(65 5)% RH). Samples without visible wood defects, with their fibres parallel
to the samples longitudinal axis, were chosen. 20 samples from each treatment and different boards were taken for the
tests. The difference of densities for parallel samples was in the range of 10%. Bending strength was determined,
using a material strength testing device ZWICK Z100. Speeds of loading were appropriate for each group individually,
so that to achieve the destruction maximum within 90 10 sec. The results were compared to the bending properties of
the untreated samples of the corresponding species.
Wood surface hardness according to Brinell for untreated and thermally treated wood was determined by the EN
1534 test method. Using the material hardness testing device, the ball semisphere (ball diameter 10 mm) was pressed
into the wood with a constant strength of 1000 N. The semisphere impress hollow diameter was determined with a
reflecting light binocular microscope. Samples with the sizes 25 x 25 x 50 mm without visible defects were conditioned
for 4 weeks. The difference of densities for parallel samples was in the range of 10%. For each series, 20 parallel
samples were chosen.
The equilibrium moisture content (EMC) discussed in this paper was determined at the air humidity 98% RH. The
studies were carried out also at the relative humidity of air RH 65%, but the results obtained at a higher RH (98%)
demonstrated more clearly the water vapour sorption properties of HTM wood in conditions of higher humidity. 10 test
specimens were placed in a climate chamber at 20 2C in a desiccator over a saturated salt solution of copper (II)
sulphate pentahydrate. Constant mass for all specimens was reached in 60 days.
The change in the hygroscopic properties of modified wood of different species was investigated by the method of
cyclic sorption-desorption of water vapours. The two-cycle sorption of water vapours was measured on a vacuum
balance with quartz spirals as the sensible element (sensitivity 1.5-2.0 mm/mg) at the temperature (22 0.1)C. The
residual pressure upon vacuum treatment was 1 P. The time of reaching the equilibrium in each point of the isotherm
was 20-24 h. The isotherms were analysed by the comparative method in combination with the BET method (Gregg and
Sing 1982).
The contact angle the angle between the solid surface base line and a tangent to the drop contour, passing through
the intersection point solid-liquid-air, was measured with a Dataphysics OCA20 device, equipped with a video camera
and software for the analysis of the drop shape, contact angle and solid surface free energy computation. From the drop
contour, two values of the contact angle can be obtained on the left and right points; the result is the mean value of the
two contact angles. Before determination of the contact angle, all samples used were stored in a climate chamber.
Before any measurement, each sample had been sanded to obtain a surface as smooth as possible; the contact angle was
measured immediately after the sanding procedure. The liquid used was water; the device was equipped with an
221
automated dispenser to have always the same volume of liquid (10 l of water).
The durability of modified and non-modified wood against the destructive action of brown (Coniophora puteana,
Poria placenta) and white (Trametes versicolor) rot fungi was tested in compliance with the modified Standard method
EN 113 and, after leaching, according to EN 84. The test duration was 6 weeks. The samples were dried before and
after fungi tests at (103 2oC).
Carrying out the hydrothermal modification of soft deciduous wood in a temperature range of 140-180oC, it has
been found that the main changes in the chemical composition of wood begin at a temperature of 160oC; in its turn, the
treatment at 180oC decreases the bending strength by more than 50% for birch and aspen, thereby considerably limiting
the applicability of this material under load conditions (Andersons et al. 2010).
To reach as high as possible biological durability at lower strength losses and to choose the optimal parameters for
different deciduous tree species, in the present work, two treatment regimes were compared, namely, holding for 3 h at
160oC and for 1 h at 170oC. The data for the regime 160oC/1 h demonstrate the effect of the increase in temperature at
10oC on the properties of the modified material. It has been found that soft deciduous wood has no essential distinctions
between the mass and density losses in both the treatments (Table 1), namely, the difference is in the range of ~2%.
Table 1: Mass and density loss of hydrothermally modified wood
Treatment, oC/h
Aspen
160/1
160/3
170/1
5
9.6
11.4
160/1
160/3
170/1
5.2
7.3
8.9
Alder
Mass loss, %
6.3
11.3
13.6
Density loss, %
8.1
11.2
12.8
Birch
5.2
10
11.8
4.7
7.1
8.6
With decreasing wood density, bending strength also decreases (Table 2). The greatest decrease (by ~20%) at
170oC/1 h is for alder; while for other species, it is only within several percent. For birch (to a lesser extent, also for
aspen), the thermal treatment in the water vapour medium at moderate temperature (140-160oC/1 h) acts plasticising
and enhances the bending strength and modulus of elasticity, compared to the case of unmodified wood.
Table 2: Bending strength of hydrothermally modified wood
Treatment,
o
C/h
Control
160/1
160/3
170/1
Aspen
Alder
Birch
Bending strength, N.mm-2
79
82
124
67
56
136
59
60
97
60
48
94
With increasing modification time, the wood surface hardness according to Brinell also decreases (Table 3).
Table 3: Wood surface hardness according to Brinell
Treatment,
o
C/h
Control
160/1
160/3
170/1
222
Aspen
Alder
Birch
Hardness according to Brinell
2.26
2.37
2.81
1.72
1.78
2.23
1.50
1.56
1.91
1.51
1.53
1.97
It has been shown that, for the soft deciduous wood under study, the decrease of bending and hardness is similar for
both the regimes, namely, with elevating temperature by 10oC or extending the exposure time at 160oC from 1 h to 3 h.
Probably, this is the peculiarity of the investigated species, because, for thermomodified pine wood with a similar
moisture content (7%), temperature has a greater effect on the strength properties than the duration (Ates et al. 2009).
As a result of the thermal modification, the wood hygroscopicity decreases (Rowell et al. 2009) and the cell wall
microstructure changes (Andersons et al. 2007). HTM wood has a lower equilibrium moisture content (EMC) 7-8%
at the relative moisture of air of 98% (19-20% for unmodified wood), and it is similar for all woods. The difference of
EMC between the regimes 160oC/3 h and 170oC /1 h is less than 1%.
Summarising and making calculations from the data of the HTM wood water vapour sorption-desorption
experiments, it can be concluded that, as a result of the thermal treatment, the surface accessible for water vapours and
mass hygroscopicity decreases, although the changes in the hydrophilic centres concentration have an accidental
character. This is explained by the varied changes in wood during the hydrothermal treatment, which embraces the
components chemical changes and the transformations in the cell wall structure, with the formation of interstructural
bonds. It should be mentioned that, for the thermomodified wood, in the second cycle of sorption, irrespective of the
treatment method (water or another inert medium), the hydrophobisation effect decreases 1.5-2 times, then the structure
and properties of the samples get stabilised (Chirkova et al. 2005). The results testify a partial reversibility of the
hygroscopic properties of thermally modified wood. The water sorption process is almost identical for the soft
deciduous wood, treated in the regimes 160oC/3 h and 170oC /1 h (for example, see hygroscopicity changes of
unmodified and modified birch in two sorption cycles, Table 4).
Table 4: Hygroscopicity changes for birch wood as a result of hydrothermal modification
Treatment,
o
C/h
A*, in the
I cycle, m2/g
A/Auntreat. in
the I cycle, %
Control
330
160/1 h
135
45.0
160/3 h
144
43.6
170/1 h
114
34.6
*A - accessible specific surface, m2/g (water surface)
Birch
A, in the II
cycle, m2/g
A/Auntreat.. in
the II cycle, %
Change of
A/Auntreat. in the
II cycle, %
300
215
208
211
71.7
69.3
70.3
26.7
25.7
35.7
For the wood modified at both the regimes under study, the values of the contact angle of the wood surface also
differ little (Table 5).
Table 5: Contact angle of water of thermomodified wood
Treatment,
o
C/h
Control
160/1
160/3
170/1
Aspen
62.06
81.15
87.39
85.11
Alder
Birch
Contact angle, o
53.58
61.67
91.02
72.65
87.01
70.77
89.42
70.27
The hydrophobicity properties of HTM wood explain the lower capillary water uptake (CWU) and the smaller
changes in the samples sizes, cyclically impregnating with water and drying (higher antiswelling efficacy ASE)
(Biziks et al. 2012).
Our results show that, in the case of modifying in the regimes 160 oC/3 h and 170oC/1 h, there are no essential
differences between the hygroscopic properties of wood. The bending strength of the species under study (except alder)
and surface hardness differ little also for both the regimes.
Wherewith, of crucial importance is the durability against biological decay for wood, hydrothermally modified at
both the regimes. It is known that durability is one of the most important aspects of wood modification. The
hydrothermal modification of wood in the regime 160oC/1 h, although improve, but does not fully ensure the durability
against wood fungi. The extension of the treatment time from 1 h to 3 h improves the biological durability, although the
best results are obtained, modifying the wood in the HTM regime 170 oC/1 h (Tables 6-8).
223
Various fungi degrade to a different extent the wood of different species, hydrothermally modified under the same
conditions. The results of the mass losses after the effect of the fungi are analysed according to the Standard EN 350-1
(Table 8), where the X-value is the mean mass loss of the tested wood (in %) / mean mass loss of the control after fungi
tests (in %). The X-values apply to one of the five durability classes.
Table 6: Wood degradation with the brown rot fungus Coniophora puteana according to EN113 and after leaching according to
EN84
Treatment, oC/h
Control
160/1
160/3
170/1
113
52
7.1
0.3
1.0
Aspen
84
43.4
17.2
1.7
0.2
Table 7: Wood degradation with the white rot fungus Trametes versicolor according to EN113 and after leaching according to
EN84
Treatment,oC/h
Control
160/1
160/3
170/1
113
30.1
25.3
8.1
6.5
Aspen
Table 8: Wood degradation with the brown rot fungus Poria placenta according to EN113 and after leaching according to EN84
Treatment,oC/h
Control
160/1
160/3
170/1
Aspen
113
30.6
26.4
12.8
1.8
C. puteana is a brown rot fungus, which degrades most intensively both coniferous wood and our investigated
unmodified deciduous wood (except alder, which is degraded more by the white rot fungus C.versicolor). On the
contrary, it degrades least the hydrothermally modified soft deciduous wood, namely, the treatment at 160 oC/3 h
ensures the durability class 1 (very durable, x 0.15). However, in the case of the action of T.versicolor and
P.placenta, the wood modified in this regime has only the durability class 2 (durable, x > 0.15 0.30) or even class 3
(moderately durable, x > 0.3 0 0.80). For alder and aspen, the durability class 1 is ensured by the regime 170oC/1 h.
On the contrary, HTM birch is only moderately durable against the test fungi. The hydrothermally modified wood is
degraded most intensively by P.placenta this has been found also in other thermowood studies (Welzbacher and
Rapp 2002).
After leaching with water (EN 84), thermowood of all tested soft deciduous species becomes durable against
biodegradation (mass losses decrease), also against the action of P.placenta. This can be explained at least partially by
the leaching of the products of thermodestruction of wood, easily utilisable by fungi.
224
Treatment, oC/h
160/3
170/1
Treatment, oC/h
160/3
170/1
Treatment, oC/h
160/3
170/1
113
0.01
0.02
113
0.27
0.22
113
0.43
0.06
Aspen
Aspen
Aspen
84
0.04
0
84
0.05
0.01
113
0.54
0.29
84
0.25
0.03
113
0.61
0.44
Birch
Birch
84
0.09
0.08
84
0.38
0.04
In the case of the modification at 170oC, partial release of the easily metabolised degradation products probably
occurs, and the wood becomes more durable against biodegradation. Different explanations for the durability
mechanisms of thermally modified wood have been hypothesised, including the formation of new free molecules with
fungicide properties in wood, degradation of hemicelluloses as the most important nutrient element for fungi, and
components transformations, owing to which the fungus fails to recognise the wood structure (Weiland and Guyonnet
2003). Some authors have concluded that the reason for biodurability against brown rot fungi is the lower
hydrophobicity of wood (Tjerdsma et al. 2000). Our results of the HTM wood fungi tests do not verify this viewpoint,
because the wood moisture after the action of fungi is within the range of 30-70%, necessary for their development. The
durability of thermowood is determined by the totality of different conditions, including the presence of the products of
wood components degradation, metabolised by the fungi. We have found that the reason can also be the different
production of fungal ligninolytic and cellulolytic enzymes in contact to unmodified and thermally modified wood
(unpublished data).
6. SUMMARY
1. In the one-stage modification, soft deciduous wood with lower equilibrium moisture, lower water vapour sorption
and higher surface hydrophobicity, which improve the materials form stability, is obtained.
2. There are no essential differences between the obtained physical properties, mass, density and strength (statistical
bending, hardness) properties decrease, carrying out the treatment in the regimes 160oC/3 h and 170oC/1 h.
3. Different durability against the brown and white rot fungi (standard EN 113) is obtained for wood of soft deciduous
tree species, modified in both the regimes under study. The comparatively lower durability properties are
demonstrated by hydrothermally modified birch wood, probably owing to the polysaccharides component
composition.
4. After leaching (Standard EN 84), the durability of modified wood increases. Probably, biodegradation is favoured
by the wood components degradation products, which are water-soluble and easily metabolisable by fungi. After
their removal, the wood durability is enhanced.
ACKNOWLEDGEMENT
The authors gratefully acknowledge the financial support by the ERAF (Grant No. 2010/0324/2DP/2.1.1.1.0 /10/
APIA/VIAA/057) and by the Latvian State Research Program NatRes.
225
REFERENCES
Andersons B., I. Andersone, J. Chirkova, E. Suttie and D. Jones. 2007. Changes in the wood cell wall microstructure as a result of
modification processes. In: Proceedings of the 3rd European Conference on Wood Modification [ECWM3], October 15-16, 2007,
Cardiff, UK, pp. 393-400.
Andersons, B., I. Andersone, V. Biziks, I. Irbe, J. Chirkova, E. Sansonetti, J. Grinins and H. Militz. 2010. Hydrothermal
modification for upgrading the durability properties of soft deciduous wood. International Research Group Wood Preservation,
Document IRG/WP 10-40494.
Ates, S., M.H. Akyildiz and H.Ozdemir. 2009. Effects of heat treatment on Calabrian pine (Pinus brutia Ten.) wood. BioResources,
4(3): 1032-1043.
Biziks, V., B. Andersons, I. Andersone and J. Grinins. 2012. Effect of one stage heat treatment on the dimensional stability of
hardwood. In: Proceedings of the 7. Europaeischer TMT-Workshop, 26-27 April 2012, Dresden Germany, pp. 60-70.
Burmester, V.A. 1973. Einfluss einer Wrme-Druck-Behandlung haldtrockenen Holzes auf seine Formbestndigkeit. Holz als
Roh-un Werkstoff, 31: 237-243.
Chirkova, J., B. Andersons and I. Andersone. 2005. Water sorption properties of thermo-modified wood. In: Proceedings of the
2nd European Conference on Wood Modification [ECWM2], October 6-7, 2005, Gttingen, Germany, pp. 65-69.
Garrote, G., H. Dominguez, J.C. Parajo. 1999. Hydrothermal processing of lignocellulosic materials. Holz als Roh- und
Werkstoff, 57: 191-202.
Gregg, S.J. and K.S.W. Sing. 1982. Adsorption. Surface and Porosity. Academic Press, London/New York, 303 pp.
Kollmann, F. and A. Schneider. 1963. On the sorption-behaviour of heat stabilized wood. Holz als Roh- und Werkstoff, 21 (3):
77-85.
Rowell, R.M., R.E. Ibach, J. Mcsweeny and T. Nilsson. 2009. Understanding decay resistance, dimensional stability and strength
changes in heat-treated and acetylated wood. Wood Materials Science and Engineering, 1-2:14-22.
Sandermann, W. and H. Augustin. 1964. Chemical investigation on the thermal decomposition of wood. Part III: Chemical
investigation on the course of decomposition. Holz als Roh- und Werkstoff, 22 (10): 377-386.
Stamm, A.J. and R.H. Baechler. 1964. Decay resistance and dimensional stability of five modified woods. Forest Prod. J. 10:
22-26.
Tjerdsma, B.F., M. Stevens and H. Militz. 2000. Durability aspects of (hydro)thermal treated wood. International Research Group
on Wood Preservation. Document no. IRG/WP 00-40160.
Weiland, J.J. and R. Guyonnet. 2003. Study of chemical modification and fungi degradation of thermally modified wood using
DRIFT spectroscopy. Holz als Roh- und Werkstoff 61: 216-220.
Welzbacher, C.R. and A.O. Rapp. 2002. Comparison of thermally modified wood originating from industrial scale processes
durability. International Research Group Wood Preservation, Document IRG/WP 02-40229.
226
GESAAF University
of Florence Firenze,
50145, Italy
DiSAFA
University of Turin
Grugliasco (TO), 10095, Italy
ABSTRACT
The present paper covers the preliminary results of an experimental study carried out in order to assess the effects
of heat treatment on poplar plywood intended for motorhome application. Tests were performed on poplar (I-214
clone) plywood glued with melamine-urea-formaldehyde (MUF) resin subjected to heat treatment before or after its
composition. The treatment was than performed on the single veneers or on the final panels. Veneers and panels
were treated at 180-190 C, respectively, until a dry mass loss of about 5%. Different properties were
assessed before and after treatment according to the reference standards: moisture content, density, Modulus of
Elasticity (MOE), bending strength and bonding quality. As expected, both moisture content and density at
standard environmental conditions decreased after treatment. MOE did not show significant variations, while
strength and bonding quality suffered serious reductions. Even if some issues require further investigation,
the above experimental analysis showed that heat treatment could be successfully applied to poplar plywood.
1. INTRODUCTION
Heat Treatment (HT) is applied to wood in order to decrease its hygroscopic behavior and to improve both its
dimensional stability and fungi decay resistance. On the other side, various mechanical properties such as stiffness
and strength can be lowered as a result of HT. HT also causes a darkening of wood.
Significant reviews on the effect of HT of wood are available: (Sandberg, Haller, & Navi, 2013), (Navi &
Sandberg,
2012), (Finnish ThermoWood Association, 2003), (Kamdem, Pizzi, & Jermannaud, 2002) while literature regarding
the effect of HT on poplar wood and especially on plywood is limited. Several tests on the mechanical characteristics of
HT poplar LVL were performed by (Nazerian & Ghalehno 2011) and other studies on the mechanical resistance of
silver nanoparticles impregnated poplar were conducted by (Taghiyari 2010).
HT poplar plywood panels were considered by the authors as a possible interesting solution for the motorhome
sector. In fact, after HT the rough material becomes lighter, with an improved dimensional stability and more durable
than conventional poplar plywood. Such products also could be suited for external use where the dimensional stability,
durability and water repellent surfaces are essential requirements. The experiments concerned several topics regarding
plywood glued before and after HT and in particular:
effects of HT on the physical properties of the veneers and of the panels;
* Corresponding author: Tel.: (+39) 0553290656674; Fax: (+39) 055319179; E-mail: giacomo.goli@unifi.it
227
International
Scientific
Conference
International Scientific Conference on Hardwood
Processing
(ISCHP
2013) on Hardwood Processing (ISCHP 2013)
effects of the HT on stiffness (MOE);
effects of the HT on bending strength (MOR);
effects of the HT on bonding quality.
veneers;
plywood panels assembled with MUF resins before heat treatment;
plywood panels assembled with MUF resins after heat treatment of the veneers.
On the whole, the main objective was to determine if HT can be considered consistent with the glues currently used
for plywood production and if HT should be applied on the veneers or could be directly applied on the assembled
panels.
2. M ATERIAL
AND METHODS
Legend:
In the present paper the following codes and acronyms have been used:
#: number of samples;
SEC: standard environmental conditions - 20 C and 65 % RH;
ODS: oven dry state - 1033 C;
M0: untreated oven dry mass [kg];
M12: mass at SEC;
ML0: mass loss due to thermal treatment at ODS, referred to M0 [%];
ML12: mass loss due to thermal treatment at SEC, referred to M12 [%];
12: untreated density at SEC [kg m-3];
L12: density loss due to thermal treatment at SEC, referred to 12 [%];
fv: shear strength;
ACWF: apparent cohesive wood failure.
Veneers or panels with dimensions 400x400 mm were heat treated at 180 C until a 5% of ML0 (T1). They were
first conditioned to standard environmental conditions than measured and weighted. The material used in the tests is
reported in Table 1.
Table 1: Heat treated poplar base material.
Material
# Layers
Veneer
Plywood with MUF resin
1
5
Nominal thickness
[mm]
2.1 #15
9.3 #06
In order to start the treatment exactly from the same stage, samples were oven dried and the mass measured (M0) as
well as the dry mass loss (ML0) was computed. Both veneers and panels were heat treated. After HT, veneers were
assembled into panels with MUF resins and tested. The moisture content and the density were determined according to
EN 322 and EN 323 standards.
Modulus of elasticity (MOE) and bending strength (MOR) were determined for longitudinal (L) and transversal (T)
directions according to the EN 310 standard. The bonding quality was determined according to the EN 314 standard
using pre-treatment 5.1.2 (6 hours in boiling water followed by cooling in water at 20 C for at least 1 hour). As regards
the bonding quality, fv was determined. As regards acwf values, it must be underlined that after HT the failure of several
specimens did not occur in the shear area of 25x25 mm, but perpendicular to it. In this case, according to EN 314, the
test should be considered as non-valid and thus repeated. Owing to this, it was then impossible to determine acwf.
The complete set of panels with their dimensions, number of layers and test performed is reported in Table 2 (panels
glued before HT present the suffix _pre, panels glued after HT using treated veneers present the suffix _post).
228
Material
# Layers
Plywood panel_Test
Plywood panel_T1_pre
Plywood panel_Test
Plywood panel_T1_post
5
5
5
5
Panel thickness
[mm]
9.3#06
9.2#02
8.8#10
7.6#10
Test
EN 310
MOE-MOR
MOE-MOR
MOE-MOR
MOE-MOR
Test
EN 314-1
(L-T)
(L-T)
(L-T)
(L-T)
5.1.2, fv
5.1.2, fv
5.1.2, fv
5.1.2, fv
The significance of the results was analyzed by a non-parametric Kruskal-Wallis statistic test. A non-parametric test
was chosen because of the variable number of samples and because of the limited number of samples for some case. A
Wilcoxon Rank Sum Test for paired or for independent values was used as post hoc test according to the types of
data.
3. RESULTS
3.1. PHYSICAL
CHARACTERIZATION
Density at standard environmental conditions (12), density loss at standard environmental conditions (L12) and dry
mass loss (ML0) of veneers and panels glued before treatment (pre) are presented in Table 3. The results (except 12)
refer to the same specimens before treatment (paired data). For the panels glued after treatment (post), being impossible
to make the measurement on the same material, density loss at standard environmental conditions (L12) was
computed from control samples (independent data). The data of post panels are reported in Table 4.
Table 3: Rough material subjected to thermal treatment and main physical properties before and after treatment. pre= glued before
heat treatment. #: number of samples. SD, where available, in brackets.
Materia
l
12 pre
treatment
[kg m-3]
Veneers
/
panels
V
Material heat treated directly in the form of panels
Plywood panel_Test
454#03 (17)
P
Plywood panel_T1_pre
459#04 (18)
P
Treat.
temp.
[C]
Treat.
Time
[hh]
L12
[%]
ML0
[%]
ML12
[%]
Thickness
[mm]
/
180
/
23
/
NA
/
5.7
/
8.3
/
/
/
190
/
8
/
7.0
/
4.9
/
9.8
9.4 (0.1)
9.0 (0.0)
Table 4: Main physical properties of panels composed from thermally treated veneers. post= glued after thermal treatment. #:
number of samples. SD, where available, in brackets.
Material
12 post
treatment
[kg m-3]
Veneers
/
Panels
L12
[%]
Thickness
[mm]
/
-1.2
8.8 (0.1)
7.6 (0.4)
Concerning panels assembled post-treatment using HT veneers as shown in Table 4, density loss at standard
environmental conditions (L12) present an opposite behavior if compared to pre-treatment panels. In fact Plywood
panel_T1 post presents an increase of density. This depends on permanent deformation during pressing (a
consequence of the reduction of the strength of the material after HT). In fact as from Table 4, even if after treatment a
permanent reduction was expected, the thickness of the plywood panel_T1_post resulted to be 7.6 mm against the
8.8 mm of the test, showing that this variation can be ascribed to plastic deformations consequent to panel pressing.
This highlights the problem of pressing parameters for treated veneers when assembled into panels after HT process.
229
International
Scientific
Conference
International Scientific Conference on Hardwood
Processing
(ISCHP
2013) on Hardwood Processing (ISCHP 2013)
3.2. MECHANICAL
CHARACTERIZATION
Both longitudinal (L) and transverse (T) modulus of elasticity (MOE) and bending strength (MOR) were
determined. The test results are summarized in Table 5. For every material the values of the treated samples were
compared with the control sample by means of a Wilcoxon Rank Sum Test for independent values. The statistical
analysis did not show significant differences between MOE both for L and T cases. For MOR a significant difference
was highlighted between the control and T1 treatment (Figure 1).
Table 5: Mechanical performances of treated and test panels according to EN 310. post= glued after heat treatment, pre= glued
before heat treatment.
Material
Plywood panel_Test
Plywood panel_T1_pre
Plywood panel_Test
Plywood panel_T1_post
MOE L
[N mm-2]
5448#30
4924#18
6507 #6
6496 #12
Group
a
a
a
a
MOR L
[N mm-2]
48#30
30#18
51 #6
38 #12
Group
a
b
a
b
MOE T
[N mm-2]
3123#30
3020#18
2304 #6
2143 #12
Group
a
a
a
a
MOR T
[N mm-2]
33#30
23#12
30 #6
19 #12
Group
a
b
a
b
40
35
30
25
20
MOR L
MOR T
15
10
5
0
Plywood panel_T1_pre Plywood panel_T1_post
Figure 1: MOR reductions for transverse and longitudinal specimens referred to test samples.
3.3. BONDING
QUALITY
The main results are summarized in Table 6. As can be observed, treated panels present a lower shear strength
(fv) compared to control panels. These differences were verified by a Wilcoxon Rank Sum Test.
230
Material
Test
Plywood panel_Test
Plywood panel_T1_pre
Plywood panel_Test
Plywood panel_T1_post
5.1.2
5.1.2
5.1.2
5.1.2
fv
[N mm-1]
1.3#45
0.4#45
1.4 #18
0.5 #30
Group
a
b
a
b
Figure 2 shows the reduction of fv after T1 treatment referred to the control sample. This is likely due to the
decreased wood mechanical properties, to the thermal degradation of the glues during HT for pre-treatment panels,
and to the lower adhesion properties of the glues to the heat treated wood for post-treatment panels.
70
fv reduction [%]
69
68
67
66
65
64
63
62
61
Plywood panel_pre
Plywood panel_post
As already said it was not possible to determine the acwf for HT panels because of the rupture in a direction
perpendicular to the glue line. Therefore further investigations are required in order to understand the effects of HT on
both glue and wood, this behavior seem to depend mainly on the reduction of strength of HT wood.
4. CONCLUSIONS
All panels showed a dry mass reduction after treatment. A reduction was observed for mass and density at standard
environmental conditions as well. As regards the mechanical performances of the panels, no statistically significant
variations were observed for MOE for either the longitudinal or transverse samples. Statistically significant differences
were observed for MOR both for longitudinal and transverse samples. The bonding quality suffered an important
reduction of the shear strength after heat treatment with a higher reduction for the panel glued before heat treatment,
showing that the effect of the treatment on the glue has a not negligible influence on the final properties of the panel.
5. ACKNOWLEDGMENTS
The authors acknowledge the financial support of Toscana Regional Administration with the POR CReO projects
line as well as the FESR.
231
International
Scientific
Conference
International Scientific Conference on Hardwood
Processing
(ISCHP
2013) on Hardwood Processing (ISCHP 2013)
REFERENCES
CEN. (1994a). EN 322: Wood-based panels - Determination of moisture content.
CEN. (1994b). EN 323: Wood bassed panels - Determination of density.
CEN. (1994c). EN 310 - Wood-based panels - Determination of modulus of elasticity in bending and of bending strength.
CEN. (2005). EN 314-1: Plywood - Bonding quality - Part 1: Test methods.
Finnish ThermoWood Association. (2003). ThermoWood Handbook. Changes. Helsinki: Finnish ThermoWood Association.
Retrieved from www.thermowood.fi
Kamdem, D. P., Pizzi, a., & Jermannaud, a. (2002). Durability of heat-treated wood. Holz als Roh- und Werkstoff, 60(1), 16.
doi:10.1007/s00107-001-0261-1
Navi, P., & Sandberg, D. (2012). Thermo-Hydro-Mechanical Processing of Wood (p. 376). EPFL Press.
Nazerian, M., & Ghalehno, M. D. (2011). Physical and Mechanical Properties of Laminated Veneer Lumber Manufactured by
Poplar Veneer, 1, 10401045.
Paul, W., Ohlmeyer, M., & Leithoff, H. (2006). Thermal modification of OSB-strands by a one-step heat pre-treatment Influence
of temperature on weight loss, hygroscopicity and improved fungal resistance. Holz als Roh- und Werkstoff, 65(1), 5763.
doi:10.1007/s00107-006-0146-4
Poncsk, S., Shi, S. Q., Kocaefe, D., & Miller, G. (2007). Effect of thermal treatment of wood lumbers on their adhesive bond
strength and durability. Journal of Adhesion Science and Technology, 21(8), 745754. doi:10.1163/156856107781362653
Sandberg, D., Haller, P., & Navi, P. (2013). Thermo-hydro and thermo-hydro-mechanical wood processing: An opportunity for
future environmentally friendly wood products. Wood Material Science and Engineering, 8(1), 6488.
doi:10.1080/17480272.2012.751935
Sernek, M., Boonstra, M., Pizzi, A., Despres, A., & Grardin, P. (2007). Bonding performance of heat treated wood with structural
adhesives. Holz als Roh- und Werkstoff, 66(3), 173180. doi:10.1007/s00107-007-0218-0
Taghiyari, H. R. (2010). Study on the effect of nano-silver impregnation on mechanical properties of heat-treated Populus nigra.
Wood Science and Technology. doi:10.1007/s00226-010-0343-5
232
InternationalScientific
Scientific Conference
Conference on
Processing
(ISCHP
2013)2013)
International
onHardwood
Hardwood
Processing
(ISCHP
Property Variation in Black ash (Fraxinus nigra M.), a case for its utilization in thermally
modified timber (TMT) products
Mathew Leitch1*, Scott Miller2, Brent Forbes3, and Jim Riffel4
1,2,3,4
Abstract
Identification of underutilized species in Northwestern Ontario, Canada, with potential for new manufacturing opportunities
through marketing could help strengthen the regional forest sector. For this to occur improved knowledge of the physical and
mechanical properties of these species and how this can be applied to end use attributes in value added forest products, is
required. Black ash is an abundant hardwood grown in Northwestern Ontario that is currently underutilized and not a
commercial species. Intensive physical and mechanical inherent property maps in the axial and radial directions have been
created for Black ash grown in Northwestern Ontario. Property maps include MOE, MOR, compression parallel to grain and
relative density. Results display Black ash is very consistent in its wood properties in the axial and radial directions within a site,
while significant differences were only found between sites. In a second study Black ash was thermally modified in a TMT kiln at
two kiln temperatures (160C 180C; 180C 200C) and tested for mechanical properties and also color properties with a
Colorimeter. It was found that due to the consistent nature of Black ash properties along the trees main stem it is an ideal species
for thermal modification producing a consistent product with consistent color. Black ash with its consistent wood properties could
be a species used in an emerging TMT manufacturing industry for high value products such as furniture, instruments and
hardwood flooring in the region.
1. INTRODUCTION
The forest industry in Canada has been in decline for much of the past decade (CCFM, 2008). The Ontario forest
sector contributes over $12 billion annually to the Canadian economy (Ontario Auditor General, 2011). In addition it
is estimated that approximately 170,000 people are directly employed by the industry in Ontario, many of which are
in small one-industry communities (Ontario Auditor General, 2011). Many of these jobs and communities are
located in Northwestern Ontario where there are large forest resources and many mill operations. However,
pressures on the industry including appreciation of the Canadian dollar recently, economic downturn in the United
States, rising input costs, increased global competition and new lost cost competitors has shut many mills down
recently (CCFM, 2008; Auditor General of Ontario, 2011). In the last decade many saw mills, pulp mills and panel
mills have shut down (Kelly, 2013). The CCFM (2008) has stated that in order for the Canadian industry to remain
competitive and relevant on a global scale the industry needs to focus on innovation and research and development
of existing and new products.
One of the areas highlighted, as a new opportunity is value-chain optimization to improve our competitiveness
with a focus on under-utilized forest resources (CCFM, 2008). The Ontario Ministry of Natural Resources (OMNR,
2007) recognizes 12 species as common or abundant in Northwestern Ontario. Much of the research and attention in
Northwestern Ontario has focused on softwood species such as Black and White spruce and Jack pine, however little
attention has been focused on the hardwood species available (Cutter et al., 2004; Evans et al., 2000; Harrington and
DeBell, 1980; OKeefe, 1982; OMNR, 2007). The need for research on under-utilized species is mentioned
extensively throughout the literature (Adamopoulos et al., 2007; Carmean and Boyce, 1973; Kellogg, 1982; Koga
and Zhang, 2004; Maeglin, 1976; Springer and Olson, 1987; Zhang et al., 1994; Zobel, 1964), as is the need for
mechanical property testing of these species in order to identify suitable utilization (Evans et al., 2000).
One species found across Ontario and the Lake States is Black ash (Fraxinus nigra Marsh) (OMNR, 2007;
Wright and Raucher, 1990), which is considered an under-utilized species with little information on its properties
and characteristics (Alden, 1994; Cassens, 2007; Wright and Raucher, 1990). The OMNR (2007) estimates that
233
International
ScientificConference
Conference on
(ISCHP
2013)
International
Scientific
onHardwood
HardwoodProcessing
Processing
(ISCHP
2013)
Black ash has approximately 20 million cubic meters of growing stock in Ontario. Even though the species in underutilized Alden (1994) states that the wood of Black ash ranks favorably when compared to other native hardwoods.
With this said there is still a large void in the literature for the properties of Black ash and how they may vary within
a single tree, between trees and across sites. The literature reports that properties vary within a tree in three distinct
ways; within a growth ring, radially from pith to bark, and longitudinally from stump to crown (Koga and Zhang,
2004; Zobel and van Buijtenen, 1989). Black ash displays an average density of approximately 500 kg/m3 (Jessome,
1977; Almedag, 1984). This density is slightly less than species that Black ash would be equivalent to from a
product point of view, such as White ash, Green ash, Oak, Maple and White Birch. Black ash is utilized in secondary
products (Cassens, 2007), such as tool handles, furniture, mill-work, trim, cabinetwork, flooring, and novelty
products (Perem et al., 1981; Wiemann, 2010).
One method of increasing value and enhancing a woods marketability is by utilizing the thermal modification of
timber (TMT) technology. This is a clean technology that does not require chemical inputs to increase durability,
and as stated by Richards (1978) several decades ago society is looking for alternative wood preservation strategies.
Heat treatment has been used for centuries to improve wood stability and durability (Militz, 2002; Shi and Zhang,
2007). TMT research has been a core research subject in Europe for over a decade (Finnish Thermowood
Association, 2003; Hill, 2006; Rapp and Sailor, 2000) and has shown that the process increases stability and
durability (Gonzalez-Pena and Hale, 2009; Hill, 2006; Shi and Zhang, 2007). Another benefit of TMT systems is the
enhancement to the aesthetic appearance of the wood making it very appealing in the market place (Militz, 2002).
Wood color has been described as a very valuable characteristic for its use (Gonzalez-Pena and Hale, 2009). Color
space (CEI L*a*b*) is one of the most common methods of measuring color, with each color being expressed as a
point within the space defined by three coordinates (Gonzalez-Pena and Hale, 2009). The vertical coordinate L*
represents the position on the black-white axis, a* represents the position on the green-red axis, and b* represents the
position on the blue-yellow axis (Gonzalez-Pena and Hale, 2009). A difference in color referred to as Delta E*,
represents the distance between the control and treatment colors (Tappi, 1994).
This paper presents results from, 1) a study looking at the variation in physical and mechanical properties in the
radial and longitudinal direction of Black ash growing in Northwestern Ontario, 2) a study on Black ash properties
utilizing TMT technology to improve and enhance product potential and marketability, and 3) a study on the color of
TMT Black ash as a possible means of in-line color grading in a TMT processing facility.
2. MATERIALS AND METHODS
2.1 VARIATION IN PHYSICAL AND MECHANICAL PROPERTIES
Three sites were selected in the Thunder Bay Seed Zone that represented the eco-sites (Racey et al., 1996) that
the species is found growing on. Within each site three mature healthy trees with no visible signs of defect were
selected for a total of 9 mature trees across three sites. These trees were felled after which 1 m bolts from 0%, 25%,
50% and 75% of total tree height were removed, labeled and returned to the lab for processing. Tree characteristics
and measurements were recorded for each tree according to the American Society for Testing and Materials (ASTM)
Standard Practice for Sampling Forest Trees for Determination of Clear Wood Properties (D5536-94) (2010) as well
as the Ontario Forest Growth and Yield Program Field Manual for Establishing and Measuring Permanent Sample
Plots (Hayden et al., 1955). Bolt samples were divided radially into three positions, juvenile wood, transition wood,
and mature wood. From each bolt and position the following tests were performed: MOE, MOR, compression
parallel to the grain, and relative density. The tests MOE, MOR, and compression parallel to the grain were
performed on Tinius Olsen Universal Wood Testing Machines (H10KT, H50KT) following ASTM Standard D553694 (2010) and ASTM Standard (D143 09) (2009). Density was measured according to ASTM Standard D 4442
07 (2007).
An analysis of variance (ANOVA) test was carried out with a general linear model at 95 percent confidence using
the Statistical Program for Social Sciences (SPSS) 19.0 software. Interactions were pooled when no significance was
determined and Duncans post hoc test was utilized to identify statistically similar values. A linear model was
developed, which includes the components radial position, longitudinal position, site and tree, which is nested in site.
Factors tree, radial position and longitudinal position were fixed while site was random. The relationship between
relative density12 and measured mechanical properties; MOE, MOR and compression parallel to the grain were
investigated using linear, logarithmic and exponential equations developed in SPSS 19.0 software. Regression
analysis was compared using coefficients of determination to establish if the variation in mechanical properties due
to relative density12 could be explained via the above regression equations.
2.2 PROPERTIES OF TMT BLACK ASH
234
Measured Property
Density12 (Kg/m3)
Relative Density12
Relative DensityOD
MOE (MPa)
MOR (MPa)
Compression Parallel (MPa)
1
625
557
591
6855
77
36
Site
2
660
591
630
8736
91
46
3
670
582
623
9691
97
39
All
Sites
652
577
614
8437
88
41
Jessome
(1977)
494
539
13500
84
41
and OD where slightly higher in Site 2. Site 1, the lowland swamp, displayed the lowest values for all properties
measured. Relative Density and Density were found to be significantly different between sites, with a Duncans Post
Hoc Test identifying Site 1 as a subset distinctly different than Site 2 and 3, the second subset. However, variations
between longitudinal and radial positions were not found to be significantly different at the 95% confidence level. A
general trend of decreasing values from pith to bark was seen in the radial direction, which has been reported in
previous studies (Burdon et al., 2004; Fukazawa, 1984; Hamilton, 1961; Paul, 1930; Phelps and Chen, 1989;
Wheeler, 1987; Woodcock and Shier, 2002; Zhang et al., 1994; Zobel and Sprague, 1998). Montes et al. (2007)
explain a lack of significant radial variation by noting that radial density variation is less for denser hardwoods, such
as black ash, than for less dense species. Woodcock and Shier (2002) and Pliura et al., (2006) noted that young trees
have wide variations in density, whereas older, mature trees utilized within this study tend to converge upon a sort of
optimal density. Most importantly, Zobel and Sprague (1998) attribute the absence of radial variation in hardwoods
to the lack of significant difference between juvenile and mature wood. The trends in density variation found in ring
porous woods can be further attributed to the decreasing amounts of latewood within the growth ring as the distance
from the pith increases (Fukazawa, 1984; Wheeler, 1987; Woodcock and Shier, 2002). Longitudinal density values
showed a trend of increased values in the base followed by a decrease near the center of the tree and then an increase
towards the top of the tree. This pattern has been noted in a number of hardwood species (Panshin and deZeeuw,
1980; Zobel and van Buijtenen, 1989; Zobel and Sprague, 1998). However, the most common trend in longitudinal
235
InternationalScientific
ScientificConference
Conference on
Processing
(ISCHP
2013)
International
onHardwood
Hardwood
Processing
(ISCHP
2013)
density is to display little variation and is reflected in the lack of variance between samples (Stringer and Olsen,
1987; Taylor 1979; Zobel and Sprague, 1998).
For MOE, again the sites are all significantly different from one another as is the lowest longitudinal position
different from the 3 positions above it. All radial positions for MOE were not significantly different at the 95%
confidence level. The general trends displayed MOE decreasing from pith to bark, which is consistent with findings
provided by Hamilton (1961) and Clarke (1935). Longitudinally MOE increased from the base of the tree to the top
of the tree. These results are contradictory to the available literature, which suggests that butt sections are often
superior in selected properties (Carmean and Boyce, 1974; Clarke, 1935; Kraemer, 1956; Myer, 1930). However,
Burdon et al., (2004) suggests variation in properties is negligible past 3.5 metres. This trend is supported by post
hoc test results from longitudinal variation of MOE, which demonstrates two subsets of positions; the first
longitudinal position and longitudinal positions two through four.
MOR displayed the same result trend as Relative Density with all sites being significantly different from one
another and both longitudinal and radial values were not found to be significantly different. This is contrary to
literature provided by Hamilton (1961) and Clarke (1935) who found a radial decrease in properties. However,
Koehler (1939) found no relationship between toughness, a measure related to MOR, and radial position. As with
MOE, MOR longitudinal results are contradictory to research within the literature, which suggests that the butt logs
are often superior in select properties (Carmean and Boyce, 1974; Clarke, 1935; Kraemer, 1956; Myer, 1930). It
should be noted that Zobel and van Buijtenen (1989) mentioned that many studies prior to 1980 only measured the
mature wood properties so radial variation did not include the core juvenile wood, meaning some early trends are
misleading.
Compression Parallel to the Grain displayed the same result trend as MOE with all sites being significantly
different from one another while the lower longitudinal position 1 was significantly different than position 3 and 4 with
longitudinal position 2 being similar to position 1 as well as 3 and 4. Longitudinal variation of compression parallel to
the grain generally demonstrated a pattern of increasing compression values from stump to crown on all sites. The
results are consistent with Clarke (1935) who suggested that the highest values in compression parallel to the grain
occur in the upper sections of the tree. Larger compression values in the upper sections of the tree suggest that the
increased percentages of juvenile wood result in the increased values (Hamilton, 1961; Park et al., 2009; Zobel and van
Buijtenen, 1989; Zobel and Sprague, 1998).
Results from all tests consistently demonstrated that properties displayed larger mean values on the second site,
the upland site, or the third site, the well-drained site compared to the first site, the lowland swamp, which
consistently displayed inferior properties. These results are consistent with findings in a number of studies which
suggest that growth and resulting properties of Black ash are improved on drier sites as opposed to the wetland sites
to which it is more commonly found (Benedict and Frelich, 2008; Carmean, 1979; Erdmann et al., 1987;
Hildebrandt, 1960; Howe, 1970; Keeland et al., 1987; Stewart and Krajicek, 1978; Tardiff and Bergeron, 1999).
Values for MOE and Compression Parallel to the Grain in the juvenile wood core display significant differences
between sites but not between longitudinal positions in the tree. Again Site 1 is different than Sites 2 and 3, which
are similar. When linear, logarithmic and exponential equations were developed to test relative density12 as a
function of the measured properties MOE, MOR and Compression Parallel to the Grain a positive correlation
between relative density12 and the measure properties was found at the 95% confidence level. A number of
researchers have noted this trend on a per species basis (Bendtsen and Senft, 1986; Kretschmann, 2010; Zhang,
1995; Zobel and van Buijtenen, 1989). The Forest Product Laboratory exponential equations (USDA Hardwood
Model) predicted the properties to within 4% for MOR, 12% for compression parallel to the grain and 12% for side
hardness compared to the measured properties while for MOE the FPL equations were 25% off of the measured
values.
Zhang (1995) concluded that a site-specific approach is required when predicting relationships, as
silviculture and environmental factors can greatly affect growth rate, which in turn affects relative density and
mechanical properties (Zhang, 1995).
3.2 PROPERTIES OF TMT BLACK ASH
Table 2 presents the mean values from the tested materials. Relative density did not display any significant
differences between treatments at the 99% confidence level with an average across all three treatments of 610 kg/m3.
MOE was found not to be significantly different between treatments at the 99% confidence level and displayed an
average across all three treatments of 9132 MPa, however there was a trend of increasing values from Control to
Durability treatment. This is in contrast to the literature that states MOE and MOR values decrease as the process
temperature increases (Finnish ThermoWood Association 2003; Shi and Zhang, 2007).
236
Measured Property
Relative DensityOD
MOE (MPa)
MOR (MPa)
Compression Parallel (MPa)
1
614
9162
97
63
Treatment
2
3
597
623
9904
10030
103
77
69
67
All
610
9132
89
60
Jessome
(1977)
539
13500
84
41
However, Shi and Zhang (2007) did find an increase in MOE values for fir, aspen and birch. MOR on the other
hand, did show a significant difference at the 99% Confidence level with the Durability treatment (77 MPa) lower
than the Stability (103 MPa) and Control (97 MPa) treatments. These subsets were confirmed with a Tukeys Post
Hoc Test. This does agree with the literature for thermally modified wood MOR values (Finnish ThermoWood
Association 2003; Shi and Zhang, 2007). Shi and Zhang (2007) did report that white birch displayed a 6% increase in
MOR values rather than a decrease found in other species.
The color analysis displayed interesting results where the original data in the form of L*a*b* color values
measured were transformed into X,Y,Z values to accommodate comparison between dependent variables and
between species, according to ASTM Standard D2244-11. Box Plots displaying all treatments in L*a*b* coordinates
and X, Y, Z Values confirmed the transformation had no effect on the relationships or variability of Treatment
means. It was found that color within a treatment was consistent while between treatments there was a significant
difference between the Durability (0.566 X coordinate) treatment and the Control (0.714 x-coordinate) and Stability
(0.698 x-coordinate) treatments. Gonzalez-Pena and Hale (2009) reported a darkening of all TMT species tested
according to the L*a*b* color space. Thompson et al. (2005) reported a consistent tan color of TMT wood of red
alder. A linear regression suggests that color could be a predictor of MOR and Hardness with adjusted R2 values of
0.4468 and 0.3888, respectively. Using a MANOVA 6 random new samples were selected and tested for color
combined with known species with specific color signatures, these results were transformed into X, Y, Z values to
allow comparison. The results display a significant difference in color and following a Tukeys Post Hoc Test 3
homogeneous subsets were found, which places each of the Control, Stability and Durability treatments in their own
subset. The Durability subset included the species African mahogany and Black walnut, the Stability treatment
included the species Black cherry and White oak while the Control treatment included Yellow birch and Hard maple.
Multiple regression was performed on the data which displayed color change was predictable between treatments.
Linear regression gave a predicte model with an adjusted R2 of 0.846; however a second order polynomial regression
better explained the change in color for the durability with an adjsuted R2 value of 0.9586 (Figure 1).
Color Change
Between Treatments
Light to Dark
0,800
0,700
0,600
0,500
0,400
0,300
0,200
0,100
0,000
Ab (G)
Ab (C)
Ab (S)
Thermal Treatments
Ab (D)
237
InternationalScientific
Scientific Conference
Processing
(ISCHP
2013)2013)
International
ConferenceononHardwood
Hardwood
Processing
(ISCHP
A comparison between commercial values of the select species was completed using retail prices in North
Western Ontario. Cost analysis shows the potential for as much as 16 times increase in value in black ash after
themal modification when considered as a repalcement for more valueable scarce hardwood species. The results
suggest that the thermowood process is in fact ideal for this species where the natural properties of the wood appear
to be consitent enough to allow consistent products from this treatment. It has been reported that density is very
important to the thermal process (Finnish ThermoWood Association 2003; Shi and Zhang, 2007; Gonzalez-Pena and
Hale, 2010) and even though there was differences in growth rate within treatments the results clearly display that
color analysis separates product based on color characteristics rather than density values. It was also noted that the
Durability treatment was the only treatment to lower property values, this was mainly due to increased brittleness of
the material after treatment, another result that has been reported in the literature (Hill 2006; Shi and Zhang, 2007;
Finnish ThermoWood Association 2003).
4. CONCLUSIONS
It can be seen that Black ash has reasonably consistent properties from a manufacturing perspective and that
the species is a good choice for TMT modification in order to increase the value of the species. This preliminary study,
although limited by a small sample size, warrants a larger more comprehensive study in this area. In addition the TMT
treatment opens many opportunities in the value-added sector to be used in flooring, fine furniture and other high value
products based on the natural beauty of the grain and the aesthetic improvements produced through the TMT treatment.
It has also been shown that Black ash displays a very consistent product following TMT treatment and has properties
that are ideal for many products. The color study displayed that this type of technology could serve as an easy in-line
system in a production facility to grade the product as it leaves the planer mill, ensuring a consistent quality product to
value-added manufacturers. This also brings an under-utilized species into the commercial realm, which would mean
better management of this resource. A final note to the potential benefit of TMT treatment is that this could be a means
of thermal sanitation of ash wood affected by the emerald ash borer, which could then be exported making this process
a cost effective means for utilizing infested wood that displays marketable properties.
5. REFERENCES
Adamopoulos, S., Passialis, C., and E. Voulgaridis. 2007. Strength properties of juvenile and mature wood in Black locust
(Robina pseudoacacia L.). Wood and Fibre Science 39(2): 241-249.
Alden, H.A. 1994. Wood Technology Transfer Fact Sheet: Fraxinus spp. Centre for Wood Anatomy Research. USDA Forest
Service, Forest Products Laboratory. Madison. 3pp.
Alemdag, I.S.1984. Wood Density Variation of 28 Tree Species from Ontario. Canadian Forest Service. Petawawa National
Forestry Institute. Information Report PI-X-45. 12pp.
ASTM. 2007. Standard test methods for specific gravity of wood and wood-based materials; Standard test methods for direct
moisture content measurement of wood and wood-base materials. D4442-07. in ASTM. 2011. Annual Book of ASTM
Standards. Section 4: Construction, Volume 4.10; Wood. ASTM International, West Conshohocken, Pa. 854 pp.
ASTM. 2009. Standard test methods for small clear specimens of timber. D143-09. in ASTM. 2011. Annual Book of ASTM
Standards. Section 4: Construction, Volume 4.10; Wood. ASTM International, West Conshohocken, Pa. 854 pp.
ASTM. 2010. Standard practice for sampling forest trees for determination of clear wood properties. D553694. in ASTM. 2011.
Annual Book of ASTM Standards. Section 4: Construction, Volume 4.10; Wood. ASTM International, West
Conshohocken, Pa. 854 pp.
ASTM. 2011. Standard practice for Calculation of color tolerances and color differences from instrumentally measured color
coordinates. D2244-11. in ASTM. 2011. Annual Book of ASTM Standards. Section 4: Construction, Volume 4.10; Wood.
ASTM International, West Conshohocken, Pa. 854 pp.
ASTM. 2012. Standard practice for Identification of instrumental methods of color or color=difference measurements of
materials. E805-12a. in ASTM. 2012. Annual Book of ASTM Standards. Section 4: Construction, Volume 4.10; Wood.
ASTM International, West Conshohocken, Pa. 854 pp.
Bendtsen, A.B. and J. Senft. 1986. Mechanical and anatomical properties in individual growth rings of plantation-grown Eastern
cottonwood and Lobolly pine. Wood and Fibre Science 18(1): 23-38.
Benedict, M.A. and L.A. Frelich. 2008. Site factors affecting black ash ring growth in Northern Minnesota. Forest Ecology and
Management 255: 3489-3493.
Burdon, R.D., Kibblewhite, P.R., Walker, J.C.F., Megraw, R.A., Evans, R and D.J. Cown. 2004. Juvenile versus mature wood: a
new concept, orthogonal to corewood versus outerwood, with special reference to Pinus radiate and Pinus taeda. Forest
Science 50(4): 399-415.
Canadian Council of Forest Ministers (CCFM). 2008. Forest Sector Innovation Framework. Queens Printer. Ottawa, ON. 98 pp.
Carmean, W.H. 1979. Site index comparisons among Northern hardwoods in Northern Wisconsin and Upper Michigan. USDA
Forest Service, North Central Experiment Station, Research Paper NC-169. 17 pp.
Carmean, W.H. and S.G. Boyce. 1973. Hardwood log quality in relation to site quality. USDA, Forest Service, North Central
Forest Experiment Station.7 pp.
238
InternationalScientific
Scientific Conference
Conference on
Processing
(ISCHP
2013)2013)
International
onHardwood
Hardwood
Processing
(ISCHP
Cassens, D.L. 2007. Hardwood lumber and veneer series: Ash. Purdue University Extension Series. Purdue University
Department of Agriculture. West Lafayette. 4 pp.
Clarke, S.H. 1935. Recent work on the relation between anatomical structure and mechanical strength in English ash. Forestry 9:
132-138.
Cutter, B.E., Coggeshall, M.V., Phelps, J.E. and D.D. Stokke. 2004. Impacts of forest management activities on selected
hardwood quality attributes: a review. Wood and Fibre Science 36(1): 84-97.
Erdmann, G.G., Crow, T.R., Peterson, R.M. and C.D. Wilson. 1987. Managing black ash in the lake states. USDA Forest Service,
North Central Forest Experiment Station. Technical Report NC-115.
Evans, J.W., Senft, J.F. and D.W. Green. 2000. Juvenile wood effect in Red Alder: analysis of physical and mechanical data to
delineate juvenile and mature wood zones. Forest Products Journal 50 (7/8): 75-87.
Finnish ThermoWood Association. 2003. Thermowood Handbook. Wood Focus, Helsinki, Finland. http://www.thermowood.fi
Accessed October 21, 2009.
Fukazawa, K. 1984. Juvenile wood of hardwoods judged by density variation. IAWA Bulletin 5(1): 65-73.
Gonzalez-Pena, M. and M. Hale. 2009. Colour in thermally modified wood of beech, Norway spruce and Scots pine. Part 1: Colour
evolution and colour changes. Holzforschung, Vol. 63, pp. 385393.
Gonzalez-Pena, M. and H. Hale. 2010. Rapid Assessment of Physical Properties and Chemical Composition of Thermally
Modified Wood by Mid-Infrared Spectroscopy. Wood Science and Technology. DOI 10.1007
Hamilton, J.R. 1961. Variation of wood properties in southern red oak. Forest Products Journal 11: 267-271.
Harrington, C.A. and D.S. DeBell. 1980. Variation in specific gravity of red alder. Canadian Journal of Forest Research 10: 293299.
Hayden, J. J., Kerley, J., Carr, D., Kenedi, T. and J. Hallam. 1995. Ontario Forest Growth and Yield Program Field Manual for
Establishing and Measuring Permanent Sample Plots. Ontario Ministry of Natural Resources, Ontario Forest Research
Institute, Sault Ste. Marie, Ontario. 145 pp.
Hilderbrandt, G. 1960. The effect of growth conditions on the structure and properties of wood. Proceedings of the Fifth World
Forestry Conference. The World Forestry Conference. Seattle, WA. pp. 1348 1352.
Hill, C. 2006. Wood Modification. Chemical, Thermal and Other Processes. Wiley, Chichester. 99, 239pp.
Howe, J.P. 1970. Effect of irrigation on wood formed at various heights in ponderosa pine. Wood & Fibre Science 3(1):126-128.
Jessome, A.P. 1977. Strength and related properties of woods grown in Canada. Eastern Forest Products Laboratory, Ottawa,
Ontario. Forestry Technical Report 21. Ministry of Supply and Services Canada, ISBN: 0-662-00571-6.
Keeland, B.D., Conner, W.H. and R.R. Sharitz. 1987. A comparison of wetland tree growth response to hydrologic regime in
Louisiana and South Carolina. Forest Ecology and Management 90: 237-250.
Kellogg, R.M. 1982. Coming to grips with wood quality. The Forestry Chronicle 58(6): 254-257.
Kelly, C. 2013. Industry Innovation: Emerging bioeconomy in Northwestern Ontario. Bioenergy Now: Spring edition:29-30.
Koehler, A. 1939. Heredity versus environment in improving wood in forest trees. Journal of Forestry 37(9): 683-687.
Koga, S. and S.Y. Zhang. 2004. Inter-tree and intra-tree variations in ring width and wood density components in balsam fir
(Abies balsamea). Wood Science and Technology 38: 149-162.
Kraemer, J.H. 1956. Growth Strength relations of White ash: a study of the effects of three influencing factors upon the
mechanical properties of the wood of White ash, (Fraxinus americana, L.) Purdue University Agricultural Experiment
Station, Bulletin Number 640. 27 pp.
Kretschmann, D,E. 2010. Mechanical properties of wood pp. 1-46 in Forest Products Laboratory. 2010. Wood handbook: Wood
as an Engineering Material. Centennial Edition. General Technical Report. FPL-GTR-113. USDA, Forest Service,
Washington, DC. 463 pp.
Maeglin, R.R. 1976. Natural variation of tissue proportions and vessel and fibre length in mature northern red oak. Silvae
Genetica 25(3-4): 122-126.
Militz, H. 2002. Heat Treatment Technologies in Europe: Scientific Background and Technological State-of-Art Proceedings of
Conference: Enhancing the durability of lumber and engineered wood products. February 11-13, 2002, Kissimmee,
Orlando. Forest Products Society, Madison, US.
Montes, C.S., Hernandez, R.E. and J. Beaulieu. 2007. Radial variation in wood density and correlations with growth of
calycophyllum spruceanum at an early age in the Peruvian Amazon. Wood and Fibre Science 39(3): 377-387.
Myer, J.E. 1930. The structure and strength of four North American woods as influenced by range, habitat and position in the
tree. New York State College of Forestry Technical Bulletin Number 31. 8 pp.
OKeefe, T.G. 1982. Hardwood utilization: a down-east view pp. 11-13 in Lees, J.C. 1982. Fuel or Fibre; Hardwood Utilization
and Marketing Opportunities. Canadian Forestry Service, Maritimes Forest Research Centre, Fredericton, NB.
Information Report M-X-138. 39 pp.
Ontario Auditor General. 2011. Annual Report: Office of the Auditor General of Ontario. Queens Printer of Ontario. Toronto,
ON. 457 pp.
Ontario Ministry of Natural Resources. 2007. Ontario tree species distribution. State of the forest report. Queens Printer of
Ontario. Toronto, ON. 24 pp.
Panshin, A.J. and C. de Zeeuw. 1980. Textbook of Wood Technology: Structure, Identification, Properties, and Uses of the
Commercial Woods of the United States and Canada. Fourth Edition. McGraw-Hill Inc., New York, NY. 722 pp.
Park, D.Y., Koubaa, A., Brais, S., and M.J. Mazerolle. 2009. Effects of cambial age and stem height on wood density and growth
of jack pine grown in boreal stands. Wood and Fibre Science 41(4): 346-358.
Paul, B.H. 1930. The application of silviculture in controlling the specific gravity of wood. USDA Forest Service. Technical
Bulletin Number 1228. 97 pp.
239
International
ScientificConference
Conference on
(ISCHP
2013)
International
Scientific
on Hardwood
HardwoodProcessing
Processing
(ISCHP
2013)
Perem, E., McBride, C.F. and C.T. Kieth. 1981. Commercial woods pp. 9-40 in Mullins, E.J. and McKnight, T.S. Canadian
Woods: Their Properties and Uses. Third Edition. University of Toronto Press. Toronto, ON. 389 pp.
Phelps, J.E. and P.Y.S. Chen. 1989. Lumber and wood properties of plantation-grown and naturally grown black walnut. Forest
Products Journal 39(2): 58-60.
Pliura, A., Zhang, S.Y., Bousquet, J. and J.MacKay. 2006. Age trends in genotypic variation of wood density and its intra-ring
components in young poplar hybrid crosses. Annals of Forest Science 63: 673-685.
Rapp, A.O. and M. Sailor, 2000. Heat treatment in Germany. Proceedings of Seminar Production and development of heat
treated wood in Europe, Nov. 2000, Helsinki, Stockholm, Oslo.
Richards, B. 1978. Wood Preservative. The Construction Press Ltd. Lancaster England.10 pp.
Shi, J.L., and J. Zhang. 2007. Mechanical behaviour of Quebec wood Species Heat Treated using ThermoWood Process.
Department of Forest Products, Mississippi State University. 5 pp.
Springer, J.W. and J.R. Olsen. 1987. Radial and vertical variation in stem properties of juvenile Black locust (Robina
pseudoacacia). Wood and Fibre Science 19(1): 59-67.
Stewart, H.A. and J.E Krajicek. 1978. Ash: an American wood. USDA Forest Service, North Central Forest Experiment Station.
Technical Report FS-216. 8 pp.
Tappi standard (1994) T 527. Om-94. Color of paper and paper board. In: Tappi Test Methods. Tappi Press, Atlanta, GA. 5 pp.
Tardiff, J. and Y. Bergeron. 1999. Population dynamics of Fraxinus nigra in response to flood-level variations in Northwestern
Quebec. Ecological Monographs 69(1): 107-125.
Taylor, F.W. 1979. Property variation within stems of selected hardwoods grown in the Mid-South. Wood Science 11(3):193-199.
Thompson, D., R. Kozak and P. Evans 2005. Thermal modification of colour in red alder veneer. 1. Effects of Temperature,
Heating Time and Wood Type. Centre for Advanced Wood Processing, Department of Wood Science. University of British
Columbia. Wood and Fibre Science 37(4): 653-661.
Van Buijtenen, J.P. 1969. Controlling wood properties by forest management. TAPPI Journal 52(2): 257-259.
Wheeler, E.A. 1987. Anatomical and biological properties of juvenile wood in conifers and hardwoods. 41 st Annual Meeting,
Forest Products Research Society. Louisville, Kentucky. 75 pp.
Wiemann, M.C. 2010. Characteristics and Availability of Commercially Important woods in Forest Products Laboratory. 2010.
Wood handbook: Wood as an Engineering Material. Centennial Edition. General Technical Report. FPL-GTR-113.
USDA, Forest Service, Washington, DC. 463 pp.
Woodcock, D.W. and A.D. Shier. 2002. Wood specific gravity and its radial variations; the many ways to make a tree. Trees;
Structure and Function 16: 437-443.
Wright, J.W. and M.H. Raucher. 1990. Fraxinus nigra Marsh. in Burns, R.M and Barbara, H.H. 1990. Silvics of North America:
Volume 2, hardwoods. USDA Forest Service, Agriculture Handbook 654. 154pp.
Zhang, S.Y. 1995. Effect of growth rate on wood specific gravity and selected mechanical properties in individual species from
distinct wood categories. Wood Science and Technology 29: 451-465.
Zhang, S.Y., Nepveu, G. and R.E. Owoundi. 1994. Intratree and intertree variation in selected wood quality characteristics of
European oak (Quercus petraea and Quercus robur). Canadian Journal of Forest Research 24: 1818-1823.
Zobel, B. 1964. Breeding for wood properties in forest trees. Unasylva 18(2-3) (73-74): 89-103.
Zobel, B.J. and Sprague, J.R. 1998. Juvenile Wood in Forest Trees. Springer-Verlag. Berlin, Germany. 300 pp.
Zobel, B.J. and van Buijtenen, J.P. 1989. Wood Variation: Its Causes and Controls. Springer-Verlag. Berlin, Germany. 363 pp.
240
R&D Department
Vinavil S.p.A.
via Toce, 7. Villadossola (VB), Italy
CNR-IVALSA
National Research Council of Italy
via Madonna del Piano, 10. I-50019,
Sesto Fiorentino (FI), Italy
ABSTRACT
The chromatic alteration of the poly(vinyl acetate) (PVAc) bondline in wood assemblies represents an aesthetic
defect appearing over time, and it is mainly important for wood-based panels such as those used in furniture
industry. This occurrence is almost peculiar of vinyl-acetate dispersions, and it is mainly evident in selected wood
species. Among the several factors affecting this phenomenon, it is known in the practice that light exposure and
temperature are the most important ones. The paper reports on the investigations carried out to give some
clarifications on the mechanisms causing the discoloration in PVAc adhesives. A commercial PVAc dispersion
was spread on specimens of selected hardwood species (cherry, oak, maple and beech), which were then exposed
to UV irradiation. These tests evidenced appreciable differences in the discoloration degree among the considered
species. Moreover, GC-MS analysis showed that specific flavonoids are responsible for the wood-alone surface
discoloration after UV irradiation. Considering the complexity in evaluating all the substances present into the
wood surfaces, attention was paid to this specific class of compounds, in order to replicate the discoloration
process by using selected model-flavonoids. It is shown that the phenomenon can be replicated in laboratory, and
that discoloration is related to the presence of post-transition metal ions in the dispersion. A new water based
poly(vinyl acetate) dispersion specifically formulated without the presence of these ions showed a very limited
chromatic alteration of the bondline after irradiation.
1. INTRODUCTION
The types of adhesives commonly used in wood industry for furniture are both thermosetting and thermoplastic.
The choice of the adhesive depends on different criteria: final use, required performances, costs, application process,
environmental issues etc. With reference to thermoplastic polymers, water based poly(vinyl acetate) (PVAc)
dispersions are widely used. Required performances in the specific application mostly address to products
complying with D3 class according to EN 204. These adhesives give to the final assembly excellent tolerance to
water and to humidity exposure, and very good resistance to both heat and constant load. However, a peculiar
behavior of this class of adhesives is the chromatic alteration of the bondline. This inconvenience appears over time,
and represents an aesthetic defect for wood-based panels such as those used in furniture industry. Among the several
factors affecting this phenomenon, it is known in the practice that light exposure and temperature are the most
important ones. Moreover, this occurrence is mainly evident in selected wood species.
Several works in the literature have given evidence of chromatic alteration of the PVAc glue line exposed to
temperature and light during time (a good list of references has been reported in Pucci, 2007). In fact, these external
conditions both give to the dispersion film the needed energy for developing chemical reactions able to produce
* Corresponding author: Tel.: (+39) 0324-503-277; Fax: (+39) 0324-366; E-mail: f.chiozza@vinavil.it
241
Product A
Product C
Solid content, %
52 1
50 2
Viscosity, mPa!s
15000 3000
8000 2000
pH
3.0 0.2
5.3 0.2
" 8.0
" 8.5
"2
"2
> 21
> 21
"7
"7
Measurements were also carried out both on four different wood species commonly used in the woodworking
industry and on polymeric films applied to wood surfaces. The used wood species were: oak (Quercus sp.p.), beech
(Fagus sylvatica L.), cherry (Prunus sp.p.), and maple (Acer sp.p.). Wood samples for all tests were taken from
boards stored in standard conditions (20C and 65% relative humidity), corresponding to approximately 12% of
moisture content in wood.
Furthermore, selected analyses were performed on solutions prepared by adding naringenin (chosen as a
model-flavonoid) to hydro-alcoholic solutions. In some cases, aluminum was also used together with naringenin (see
the following text for further details). In the case of spectrophotometric UV/VIS measurements, 1:1
water/isopropanol solutions were used as the solvent (naringenin concentration 5!10-4 % w/w); as for TLC, 1:1
butanol/water solutions were tested (naringenin concentration 5!10-2 % w/w).
2.1. UV EXPOSURE
Four exposure processes were carried out on: a) polymeric films alone (with and without the addition of
naringenin); b) wood surfaces alone; c) films applied on wood surfaces; d) solutions of naringenin with and without
the addition of aluminum. In the case (a), the two polymeric films were spread on polyethylene plates and then
conditioned for one week at 23C/50% r.h. before exposure. In the case (c), both adhesives were spread on the four
242
243
Figure 1: Aspect of both polymeric films (A and C) exposed to UV (upper parts in the Figure). The portion of each film covered
by an aluminum foil during UV exposure is also shown (lower parts in the Figure).
Figure 2: Comparison of the discoloration effect for the two adhesives (A and C) when applied to different wood species: from
left, maple, beech, oak and cherry. In the Figure, the upper parts of each sample were those exposed to UV irradiation, whereas
the lower ones were those covered by aluminum foils.
Table 2: Colorimetric measurements obtained for the two adhesives (A and C) applied to wood and for the wood itself after 95 h
*
*
UV irradiation. In the Table, !E is the total color difference, !C * is the chroma difference, and !L the lightness difference.
!C *
!L*
!E *
maple
!C *
!L*
!E *
oak
!C *
!L*
!E *
cherry
!C *
!L*
Product
!E *
beech
Wood as such
10.2
-9.3
-3.9
15.5
-13.8
-7.0
10.3
-8.5
-5.8
10.2
-6.9
-7.5
17.0
-14.1
-9.6
22.5
-19.2
-11.7
9.2
-4.2
-7.9
15.0
-1.7
-12.9
9.2
-8.4
-3.4
16.0
-14.0
-7.8
7.5
-4.6
-5.9
5.5
-3.5
-4.3
Colorimetric tests proved an influence of the type of wood on the wet adhesive, and forced a closer investigation
on the wood changes after exposure, in the attempt of identifying the compounds responsible for discoloration in
wood, and verifying if they were also responsible for the alteration of color in the adhesives.
3.2. EXTRACTIVES ANALYSIS
The comparison of chromatograms of cherry wood extractives between exposed and fresh surfaces put in
evidence a very strong increase of two flavonoids (namely the peaks at 14.05 min assigned to luteolin, and at 16.95
min assigned to naringenin) after exposure (Figure 3). Luteolin belongs to natural dyes and it is commonly extracted
from weld (Reseda luteola L.); it is one of flavonoids that are the main chromophores in the most common used
yellow dyes, and occurs in plants as sugar derivates (in the case, present as 7-glucosideluteolin). Luteolin gives one
244
Figure 3: Cherry wood extractives. The peaks corresponding to luteolin (14.05 min) and naringenin (16.96 min) are evidenced.
In maple the analysis put in evidence the increase of some chemical compounds, such as isofraxidin (retention
time of 14.43 min, data not shown), also belonging to flavonoids. This molecule is similar to scopoletine that is
considered responsible of chromatic alteration of maple (Yoshimoto, 1989). Fraxidin, a molecule very similar to
isofraxidin, has been already detected in maple (Manville and Levitin, 1974).
The comparison of chromatograms in the other considered species did not allow evidencing so clear differences
between exposed and fresh surfaces related to possible interactions with the adhesives. More in detail, in oakwood
the strongest difference in extractives observed after UV irradiation was imputable to the appreciable increase of
pyrogallol (retention time 9.55 min, data not shown). This product originates from exposure of gallic acid to oxygen
at high temperature or in weak acid condition, and its brown color is apparently responsible of the chromatic
alteration in oakwood. Potential interaction between pyrogallol and adhesive (pH, oxidant environment, chelations)
could be hypothesized. However, on the basis of results reported in Table 2, the effect of this eventual interaction
should be limited. In beech, the comparison between exposed and fresh surfaces did not evidence any relevant
compound that could be considered crucial for chromatic alteration. This allows supposing that the observed
increase in !E * shown by the wood substrate (Table 2) is mainly related to light effects associated to surface
changes in lignin after UV irradiation (Fengel and Wegener, 1989).
3.3. MEASUREMENTS CARRIED OUT ON FLAVONOID-METAL COMPLEXES PREPARED IN LABORATORY
Considering that flavonoids were individuated as the main responsible of discoloration in wood, some attempts
were carried out to reproduce in laboratory the observed effects, making use of a selected model-flavonoid to which
a metal ion was added in order to simulate the conditions in the commercial PVAc dispersion. Attention was mainly
paid to cherry wood, which evidenced the strongest chromatic alteration among the considered bondlines (Figure 2
and Table 2). Therefore, naringenin was selected as model-flavonoid, and aluminum as metal ion. In fact, it is
known that aluminum is commonly added to commercial PVAc latexes to improve their resistance to water
(Chiozza et al., 2012). Prepared hydro-alcoholic solutions of naringenin evidenced a strong absorption band at 289
nm (Figure 4). However, when the naringenin-aluminum complex was considered, the shape of the curve changed:
245
Figure 4: Absoprtion curves of naringenin (red curve) and of the naringenin/aluminum complex (blue curve). In the same Figure
it is also shown the absoprtion curve of the complex when the naringenin solution was considered as the blank (green curve). In
such a way, the modification of the curve imputable to the complex could be better evidenced.
Naringenin was also directly added to polymeric films, in order to evaluate its effect on color before and after
UV irradiation. These effects are shown in Figure 5a (that can be compared to Figure 1): it was evident how the
addition of naringenin had a more marked effect on the commercial dispersion (A), although this effect was very
limited before exposure, whereas the variations were less evident in the newly synthesized latex (C). Moreover,
considering that spectrophotometric analysis showed that the complex absorbs in the UV range, we decided to
observe the same surfaces under a low-pressure mercury-vapor fluorescent UV lamp (Philips TL-D/08) (Woods
lamp). Observation evidenced how unexposed (covered) films where naringenin was added were fluorescent for
product A, whereas they were not for C (Figure 5b, lower parts). Therefore, these observations confirmed the
presence of a naringenin/metal complex in the commercial product (A), and they also attested the suitability of the
used approach to verify spectrophotometric analysis (which showed that this complex absorbs in the UV range).
a)
b)
Figure 5: Aspect of both polymeric films (A and C) additivated with naringenin and then exposed to UV (upper parts in the
Figure): a) films observed under normal light; b) films observed under the Woods lamp. In both (a) and (b), the portions of the
film covered by an aluminum foil during UV exposure is also shown (lower parts in the Figure).
Observations also evidenced that UV exposure induced a marked decrease, or even the disappearance, of
fluorescence in adhesive A (Figure 5b, upper parts). On the opposite, in C the presence of a slight fluorescence
could be observed after exposure. It is worth to notice that the same results obtained for the mixtures prepared in
laboratory were also found when the two products were spread on the cherry wood surfaces (already shown in
Figure 1) and observed under the Woods lamp (Figure 6). In fact, unlike C, product A showed fluorescence in
protected (covered) surfaces, thus evidencing the presence of a complex between a metal ion and flavonoids with
chemical structure similar to naringenin. Moreover, in product A fluorescence was noticeably reduced after UV
exposure, whereas in C it slightly appeared so that both products showed similar fluorescence, Figure 6).
This decrease in fluorescence after UV exposure denoted a modification of the complex after irradiation. This
occurrence was verified using thin layer chromatography (TLC). TLC experiments evidenced that naringenin
followed the solvent front in the mixture butanol/water, whereas the naringenin/aluminum complex (molar ratio 1:1)
distributed in three spots (Figure 7): a transparent spot (in the lower part of the picture, corresponding to the area of
246
Figure 6: Products A and C spread on the cherry wood surface: Comparison between observation under normal light (left) and
under Woods lamp (right). In the Figure, the upper part of the sample has been exposed to UV irradiation, the lower one was
covered by an aluminum foil during irradiation.
The naringenin/aluminum complex prepared in laboratory and exposed to UV irradiation for 95 h was
appreciably more colored than the unexposed complex (Figure 7b). When the UV irradiated complex was put in the
TLC plate, four different spots were observed (drop CE in Figure 7c): the transparent, highly fluorescent one in the
drop deposition area (imputable to the complex, which did not migrate with the solvent), the transparent, slightly
fluorescent one in the middle (both were already observed also for the unexposed complex, CNE), and two colored
spots near the solvent front, both not fluorescent. Considering that the middle spot was found both in the unexposed
(CNE) and the UV irradiated (CE) complex, it can be suggested that such a spot is related to some intermediate form
(e.g., differently coordinated with the metal) of the complex. Instead, the two spots near the solvent front were one
slightly colored and one darkly colored. The former spot was attributable to not-complexed naringenin, as already
observed for exposed naringenin solutions (NE in Figure 7a), whereas, by exclusion, the darker spot was attributable
to a reaction product of the naringenin/aluminum complex after UV irradiation (in fact, the irradiated complex
became darker than the un-exposed one, as shown in Figure 7b).
Figure 7: Pictures (a) and (c) show the TLC plates with naringenin as such (NAS in the Figure), naringenin as such exposed to
UV irradiation (NE), naringenin complexed with aluminum (CNE), and naringenin complexed with aluminum and exposed to
UV irradiation (CE). Picture (b) shows a drops of CE and CNE put on a paper sheet, evidencing the different coloration between
the two solutions. In both (a) and (c) the drop was put in the lower parts of the TLC plates. In (c) the fluorescence observed under
the Woods lamp was evidenced by drawing a line with a pencil.
247
a)
b)
c)
Figure 8: Comparison between the observation under normal light (left in each photo) and under UV lamp (right in each photo)
for products A and C applied on: oakwood (a), maple (b), and beech (c). In the Figure, the upper part of each sample has been
exposed to UV irradiation, the lower one was covered by an aluminum foil during exposure.
4. SUMMARY
The chromatic alteration of the PVAc bondline in assemblies with selected hardwood species represents an issue
in wood-based panels. Some investigations have been carried out to give some clarifications on the mechanisms
causing this discoloration, through tests performed both on polymeric films alone and on the dispersions spread on
wood surfaces. Measurements evidenced that the substances causing discoloration in wood surfaces alone (most of
them belonging to flavonoids) are also responsible for the chromatic alteration of commercial dispersions containing
metal ions inside, owing to the formation of a metal-flavonoid complex. The presence of this complex has been
evidenced on adhesives spread on cherry wood after observations under UV light (spectrophotometric measurements
evidenced that the complex as such absorbed in the UV range rather than in the visible one). However, such an
occurrence depends on the absorption characteristics of the specific compound involved in the complex, which also
explains the appreciable differences in the discoloration degree among the considered species. In any case, UV
irradiation in cherry wood was able to produce new compounds, more colored and no more UV-fluorescent (all
these occurrences were evidenced after TLC tests). Once again, for the other wood species the effective mechanism
depended on the specific complex generated on the surfaces. The importance of the presence of the metal ions (most
probably of a post-transition metal, like the aluminum, which was considered in present work) on the bondline
discoloration was also confirmed by the fact that chromatic alteration was not evident (or it was evident only
248
249
Holzforschung Mnchen
Technische Universitt Mnchen
Winzererstrasse 45, 80797 Mnchen, Germany
ABSTRACT
Bonding strength and its durability are basic requirements for the utilisation of wood in structural products such
as glued laminated timber (glulam). Due to the gaining importance of hardwood species in Central Europe the
bonding quality of ash for use in glued laminated timber was determined. The bonding quality was examined in
shear tests (EN 302-1) and delamination tests (EN 302-2). Three surfacing methods (peripheral planing, sanding
(80 grit) and face milling) as well as three adhesives (PUR, PRF and MUF) were included in the survey. The
bonding performance could be improved by using sanding or face milling as surfacing method. This generally
applied both for shear strength and wood failure as well as for resistance to delamination for all three adhesives.
The best overall performance in combination with ash was found for the PRF adhesive.
1. INTRODUCTION
It becomes more and more apparent that the world wide climate change will have an impact on our environment. On
the one hand the growth conditions in Central European forests are expected to alter (Pretzsch and Dursky 2002,
Koelling et al. 2007) and at the same time more extreme meteorological events are expected to occur (IPPC 2007). In
this context, hardwoods such as ash (Fraxinus excelsior L.) play an important role to establish stable forests. However,
the increasing importance of hardwoods is in contradiction to the limited economic usability. Therefore, the
development of new markets is required. In this connection, the utilisation of ash in load-bearing timber structures can
be an important area of application that so far has been given little attention.
Ash is a wood species that is well known for its good mechanical properties (Kollmann 1941, 1951). However, the
utilization of ash requires the determination of characteristic properties in dimensions that are relevant for load-bearing
structures. Therefore, research was performed by Frhwald and Schickhofer (2004) and by Glos and Torno (2008). The
latter survey resulted in the assignment of design values of the strength class D40 according to EN 338 for visually
graded sawn ash timber by means of DIN 4074-5, Grade S10 and better. This assignment confirmed the high potential
of ash wood for use in load-bearing constructions. However, the production of sawn timber from hardwoods with large
cross-sections and great lengths is problematic and therefore, ash is only regarded valuable as a material for glued
laminated timber (glulam).
A basic requirement for the utilisation of ash wood in glulam is the realization of reliable and durable bonds
between lamellas as well as in finger joints. Bonding quality depends on a number of factors though and todays
adhesives are mainly designed for the production of engineered wood products made of spruce. However, ash shows
important differences for example in wood anatomy and wood chemistry in comparison to spruce. Indeed, the higher
* Corresponding author: Tel.: +49(89) 2180-6458; Fax: +49(89) 2180-6464; E-mail: knorz@hfm.tum.de
250
density of ash involves a lower porosity, in particular in the late wood areas. As a consequence, the penetration
behavior of adhesives is assumingly distinctly lower in ash wood. This leads to the conclusion that the surface topology
of ash wood may have a distinct influence on the bonding quality.
The most common surfacing method in glulam production is peripheral planing. Planed adherends often show
crushed cells and damages in the surface and the subsurface (Seltman 1995). This may lead to limited glue penetration
(de Moura and Hernandez 2005) and therefore, to lower bond quality. In addition, damages in the wood surface may
cause a weak boundary layer and hence, influence bond strength (Dunky and Niemz 2002). Sanding of adherends
results in a different surface topology available for bonding. According to Murmanis et al. (1983) sanded surfaces show
damaged cells that partially prevented the penetration of adhesives. The influence of the surfacing method on the
topology was confirmed by Cool and Hernandez (2011) who found the most severe damages on sanded wood surfaces
when comparing them to planed and face milled surfaces. With face milling a distinct difference to planing and sanding
is the machining direction. While the two latter methods typically produce surfaces in grain direction, the machining
direction with face milling is perpendicular to the grain. This leads to a fibrillation of the surfaces (Hernandez und Cool
2008) which causes an increase of the surface area that is available for the formation of adhesive bonds. According to
Cool and Hernandez (2011) open cell lumens and fibrillation as generated with face milling facilitate the penetration of
adhesives. In addition, a positive impact of face milling as surfacing method was found by Cool and Hernandez (2011)
with Black Spruce. In shear tests after accelerated aging the authors obtained higher strength values with face milled
surfaces than with sanded and planed surfaces.
In this research work, the bonding quality of ash was examined for the three surface manufacturing methods
peripheral planing, sanding and face milling in combination with three different adhesive systems. The bonding quality
of ash was determined by means of two well-established test methods - the delamination test (EN 302-2) and the shear
test (EN 302-1). Delamination tests and shear tests are used in different forms both for adhesive approval tests and for
quality control in the glulam industry.
2. MATERIALS AND METHODS
2.1. WOOD AND SURFACING METHODS
For the making of the specimens, ash wood originating in Southern Germany was used. Ash boards with approx.
42 mm thickness were cut into lamellas with 550 mm length and 170 mm width. For conditioning reasons, the lamellas
were stored at 20 C and 65 % relative humidity (20/65) for several weeks. For determination of density and moisture
content (MC) before bonding, small sections with a length of 20 mm were cut from the lamellas. The mean and the
standard deviation of the wood density were 644 ( 87) kg/m and 10.4 ( 1.8) % for MC.
The adherend surfaces were prepared with three different surfacing methods:
-
Peripheral planing: The ash wood was planed with a conventional planing machine (Martin, Germany) with a
feed speed of 6 m/min. The planing knives were freshly sharpened.
Sanding: The sanded surfaces were made by means of a cylinder sanding machine (Kndig, Switzerland). A
feed speed of 7 m/min was used. The sanding belt had 80 grit and has not been used before. Subsequent to the
sanding process the surfaces were cleaned with compressed air.
Face milling: Face milling was carried out with a conventional Ledinek Rotoles machine (Slovenia). The
knives were freshly sharpened and the cutting depth was 2.5 mm. A feed speed of 10 m/min was used.
251
Parameters that were used for ash bonding in this survey are displayed in Table 1. The bonding of the test specimens
was performed within a time period of 6 h after producing the surface. The bonding and the pressing procedure were
carried out in a climate room at 20/65.
Table 1: Bonding parameters
Adhesive
MUF
PUR
PRF
100/50
n/a
100/20
400
140
450
<5
<5
<5
30
20
60
Pressure in N/mm
1.2
1.2
1.2
Pressing time in h
2.5
The shear test specimens were made according to EN 302-1. Therefore, panels with 300 mm length, 130 mm width
and a thickness of 5 mm were prepared with the three surfacing methods. The angle between annual growth ring and
surface to be bonded was between 30 and 90. After bonding of the panels the bonded assemblies were conditioned at
20/65 for seven days. Then the test specimens as shown in Figure 1 were cut from the bonded assemblies. For each
surfacing method and adhesive, a total of 80 test specimens were produced.
150
20
10
The test specimens for determination of resistance to delamination were made according to EN 302-2. For the
bonded members six lamellas with 500 mm length, 160 mm width and 30 mm thickness were assembled. All lamellas
showed flat sawn annual growth rings. After bonding and pressing the bonded members were conditioned at 20/65 for
seven days. Then the width of the bonded members was reduced to 150 mm and test specimens as shown in Figure 2
were cut from the bonded members. For each surfacing method and adhesive, two bonded members were produced and
from each bonded member two test specimens were obtained.
252
180
150
75
fV
Fmax
A
in N/mm
(1)
where, Fmax = applied load at failure in N and A = bonded test surface in mm.
In addition to the shear strength the specimens were assessed with regard to their failure pattern. The wood failure
percentage was determined for each specimen with 10 % accuracy. Some test specimens were excluded from the
analysis, as sometimes the specimens were distorted after the water immersion and consequently suffered damage when
clamping them in the testing machine..
DELAMINATION TEST (EN 302-2)
The delamination test was performed according to EN 302-2. The high temperature delamination procedure was
applied for type I adhesives according to EN 301 and EN 15425, respectively. The delamination test includes both the
soaking of the test specimens and a subsequent drying procedure. Before test the mass of the specimens was
documented. The soaking-drying cycle is then repeated for three times. For the soaking procedure the test specimens
are submerged in water in a pressure vessel. Alternating pressure levels (25 5 kPa for 15 min and 600 25 kPa for
253
60 min) guarantee for high water absorption and the maximum swelling of the test specimens. After the soaking
procedure the specimens are dried at 65 3 C temperature, 12.5 2.5 % relative humidity and 2.25 0.25 m/s air
speed. The drying procedure is finished when the mass of the specimens is between 1 and 1.1 times the initial mass
before soaking.
After each drying cycle the delaminations were marked on the end-grain surfaces. After the third cycle the
delamination D was calculated according to equation (2):
l1
100
l2
in %
(2)
where, l1 = the total delamination length on both end-grain surfaces in mm and l2 = the total nominal length of the glue
lines on both end-grain surfaces in mm.
3. RESULTS AND DISCUSSION
TENSILE SHEAR TEST (EN 302-1)
The results of the shear test according to EN 302-1 after treatment A1 are displayed in Figure 3. The mean values
for shear strength range from 11.3 N/mm (MUF, sanded surface) to 13.6 N/mm (PRF, face milled surface). The
highest strength values were found for the face milled surfaces for all three adhesives. While PUR- and PRF-bondings
showed the lowest mean strength with planed surfaces, the lowest mean value for MUF was found for the sanded
surface. The approval of adhesives according to EN 301 (MUF, PRF) and EN 15425 (PUR) requires an average shear
strength of 10 N/mm with specimens made of beech and a thin glue line of 0.1 mm. This requirement could also be
fulfilled with ash bondings.
Figure 3: Shear strength (left) and wood failure percentage (right) for ash bondings after treatment A1.
The results found for shear strength are reflected by the wood failure percentages obtained in this survey. The
lowest wood failure percentages were found for the planed surface for all three adhesives. While the highest values in
wood failure could be determined with the face milled surfaces for PUR- and PRF- bondings, the MUF-adhesive
showed its best performance with regard to wood failure with sanded surfaces.
When examining the shear strength and wood failure in combination, the results show a clear pattern. With
increasing shear strength in the order (1) peripheral planning, (2) sanding, and (3) face milling, also the wood failure
percentage rises. This picture can be found for PUR- and PRF-bondings and for MUF-bondings with planed and face
milled surfaces. The only exceptions to this are the MUF-bondings with sanded surfaces where a lower shear strength
254
is accompanied by a very high wood failure percentage. However, due to the high wood failure the lower strength value
is assumingly caused by a lower wood quality.
Figure 4 shows the shear performance of ash bondings according to EN 302-1 after treatment A4. With wet
condition of the specimens the obtained mean shear strengths range between 5.6 N/mm (PUR, planed surface) and
7.7 N/mm (PRF, planed surface). The highest strength values were found for PRF-bondings with means > 7.0 N/mm
for all surface treatments. While for MUF-bondings distinctly higher strengths could be found with sanded and face
milled surfaces in comparison to planed surfaces, the differences in shear strength were small for PUR-bondings here,
the means range between 5.6 N/mm (planed surface) and 6.0 N/mm (sanded surface). When comparing the results
with the requirement of 6 N/mm (average strength value) given in EN 301 and EN 15425, respectively, it can be
concluded that PRF-bondings fulfill this specification after water immersion independently from the surfacing method.
In contrast, while MUF-bondings meet the requirement with sanded and face milled surfaces, PUR-bondings reach the
boundary value only with sanded surfaces.
The anaylsis of the wood failure percentages shows high values 70 % for PRF with all surfaces. The high wood
failure percentage correlates well with the highest bond strengths for PRF for all three surfaces. In contrast, the wood
failure percentage for PUR bondings does not exceed a median of 40 %. The highest wood failure is obtained for
sanded surfaces which is reflected by the highest mean shear strength for PUR. At the same time, both the lowest
strength value and the lowest wood failure could be found for PUR with planed surfaces. This is confirmed for the MUF
adhesive with the weakest performance for peripheral planing and with increasing values in shear strength and wood
failure for sanded surfaces. The best performance for MUF-bondings was found for the face milled surfaces.
Figure 4: Shear strength (left) and wood failure percentage (right) of ash bondings after treatment A4.
255
As stated above, low mean values for resistance to delamination were found for ash bondings with PRF and MUF.
Nevertheless, the requirements for adhesive type I according to EN 301 and EN 15425, respectively, could not be
fulfilled. This is due to the fact that the requirement of maximum 5 % delamination applies to every individual test
specimen. As mentioned before, the low mean delamination values were found for the PRF-bondings with sanded
(2.5 %) and face milled surface (3.7 %). However, the maximum delamination for individual specimens reached 6.8 %
for the sanded and 6.1 % for the face milled surface and therefore, exceeded the 5 % boundary value.
4. CONCLUSIONS
The influence of the three surfacing methods, peripheral planing, sanding and face milling, has been determined for
ash bondings made with three adhesives (PUR, PRF and MUF). Both in shear test according to EN 302-1 and in
resistance to delamination (EN 302-2) a significant impact of the surfacing method on the bonding quality was found.
The surprising result was that peripheral planing as the most common surfacing method in glulam industry showed
significantly lower shear performance and resistance to delamination than both alternative methods. The comparison of
the sanded and face milled surfaces showed heterogeneous results. While the resistance to delamination was higher
with sanded surfaces for the PRF and MUF adhesives, the delamination performance was better with the face milled
surface in combination with PUR. In general, the PRF showed the best performance both in shear and in delamination
tests. Due to the fact that the requirements for adhesive type I according to EN 301 and EN 15425, respectively, could
not be fulfilled the utilization of ash in glued laminated timber is currently not possible. Therefore, more research work
is required to optimize the bonding quality of ash wood.
REFERENCES
Cool, J. and R. E. Hernandez. 2011. Evaluation of four surfacing methods on black spruce wood in relation to Poly(Vinyl Acetate)
gluing performance. Wood and Fiber Science 43(2):194-205.
de Moura, L. F. and R. E. Hernandez. 2005. Evaluation of varnish coating performance for two surfacing methods on sugar maple
wood. Wood and Fiber Science 37(2):355-366.
DIN 4074-5. 2008. Strength grading of wood Part 5: Sawn hard wood. Deutsches Institut fr Normung DIN, Berlin.
Dunky, M. and P. Niemz. 2002. Holzwerkstoffe und Leime. Technologie und Einflussfaktoren. Springer, Berlin/Heidelberg.
954 pp.
EN 301. 2006. Adhesives, phenolic and aminoplastic, for load-bearing timber structures Classification and performance
requirements. Comit Europen de Normalisation CEN, Brussels.
EN 302-1. 2004. Adhesives for load-bearing timber structures. Test methods. Part 1: Determination of bond strength in
256
257
CNR-IVALSA
via Madonna del Piano, 10. I-50019, Sesto
Fiorentino (FI), Italy
ABSTRACT
The adhesion between two solids glued together depends on the intimate contact at the atomic level promoting the
attractive force between the surfaces. Both wood surface anatomy and machining process may affect adhesion.
Thus, for an optimal interaction and to limit eventual problems, a good adhesive should mold, when still liquid, to
the wood rough surface and penetrate inside cell cavities to increase the contact area. In this work the effect of
wood anatomy on the adhesion properties of three commercially valuable hardwood species: beech (Fagus
sylvatica L.), chestnut (Castanea sativa Mill.), and birch (Betula alba L.), together with Scots pine (Pinus sylvestris
L.) as a representative conifer, was evaluated by using a commercial polyvinyl acetate (PVAc) glue, normally used
in woodworking industry. Microscopy observations (carried out by Environmental Scanning Electron Microscopy,
ESEM, on glued joints) were correlated to the adhesive strength of the same joints, evaluated according to
standards EN 204-205, suitable for non-structural adhesives. Results showed that anatomical features of wood
greatly influenced the interface quality. The presence of vessels had a dual effect: on one hand, vessels favored
mechanical interlocking, especially when the adhesive penetrated inside them (as in the case of birch); on the
other hand, vessel-rich zones generated weak links areas. Moreover, penetration of adhesive inside the wood
vessels reduced the glueline thickness, thus leading to a more starved joint. In addition to anatomical issues, also
the applied pressure was able to modify the adherend surface morphology, leading to interface damage. This was
mainly observed in the case of chestnut, birch and pine, probably because cell walls in those species are thinner
than in beech. Moreover, conifer joints also suffered of the presence of the resins contained in the parenchymatic
cells: Scots pine was the only species among those considered in which resin leakage was observed, which
appreciably weakened the mechanical strength of joints.
1. INTRODUCTION
The use of wood is closely connected to the use of adhesives: about 80% of the finished products of wood are
glued, and around 70% of the entire volume of adhesive products in the world are destined to this area (Pizzi 1998).
In fact, wood adhesives are used in a lot of various fields, from the do it yourself and hobby to the sector of
furniture and construction, and wood species to be glued vary greatly depending on the final use of the products.
The mechanisms of adhesion have been studied for a long time in order to optimize the adhesion process (Pizzi
1994, McBain and Hopkins 1925, Good 1972, Bikerman 1961, Collett 1972). The various mechanisms proposed are
distinguishable according to the type and distance of the wood-adhesive interaction. However, they can be grouped
in two main types, one involving mechanical aspects, and one comprising the attraction between charges. In general,
the more reliable theories that underlie the adhesion mechanisms in wood include mechanical interlocking,
interdiffusion, weak layers (weak boundary layers, WBL), chemical bonding, and specific adhesion.
Mechanical interlocking was proposed for the first time by McBain e Hopkins (1925). In that case, it is
postulated that the adhesive adheres to surface irregularities through the penetration in cavities and pores when still
*
Corresponding author: Tel.: (+39) 0461-660232; Fax: (+39) 0461-650045; E-mail: santoni@ivalsa.cnr.it
258
Wood samples of four species were used in present work. They were: beech (Fagus sylvatica L.), chestnut
(Castanea sativa Mill.), and birch (Betula alba L.), together with Scots pine (Pinus sylvestris L.) as a representative
conifer. A commercial water resistant Poly(vinyl acetate) (PVAc) adhesive belonging to the D3 class as defined by
EN 204:2001, normally utilized in the woodworking industry, was used as the glue, and its main characteristics are
reported in Table 1.
Table 1: Main chemical, physical and mechanical properties of the PVAc-based adhesive used in present work. The properties
related to wood were measured on beech, as specified into the standard EN 204:2001.
Parameter
PVAc adhesive
Solid content, %
52 1
Viscosity, mPa!s
15000 3000
pH
3.0 0.2
3
" 8.0
"2
> 21
"7
Wood assemblies were prepared by gluing two different thin boards (640 x 130 x 5 mm3 each), according to EN
205 (suitable for non-structural adhesives): these thin boards were sanded with grit 150 and, after spreading the
adhesive (150 g/m2) on both sides, they were overlapped and put under a press (pressure applied: 0.8 MPa, the same
259
a)
b)
Figure 1: Secondary electron image of transversal sections of the bonded beech (Fagus sylvatica L.) joints. The surface of the
wood appeared compact and not deformed by neither the pressure nor the machining process (in the specific case surfaces were
sanded). On the right (b) a higher magnification (600X) of the same image shown on the left (a) was reported. Only a few vessels
filled with adhesive were visible at the sample interface (b).
260
a)
b)
Figure 2: Transversal sections of the bonded birch (Betula alba L.) joints. The Figure shows two different samples: in (a)
scalariform perforations are visible, and the glueline is quite straight; in (b) crushed vessels and heavy cell damages at the
interface are evident. Both images show a wood surface deformed by the bonding pressure. Adhesive penetration into the vessels
is mainly visible in (b).
In the transversal section of chestnut, the porous ring in the earlywood (a typical characteristic of this species)
was clearly visible (Figure 3). It was observed that some large early vessels (up to 500 m) were often occluded by
tyloses. Because of this phenomenon, the chestnut heartwood is not easily penetrable in deep by liquids such as
adhesives. This occurrence was apparently confirmed in present case, where the adhesive did not seem to have
penetrated deep inside the vessels (although they were very large). However, in one case we noted the presence of
glue in a large vessel (Figure 3a), and moreover the measured thickness of glueline (about 60 m) suggested that
part of the glue is penetrated into vessels and fibers, even if not in depth because of the tyloses presence. Rather,
both vessels and wood fibers appeared as partially crushed by the pressure applied during bonding (Figure 3b): this
created regions where there was a severe split in wood near the interface (Figure 3a). As reported by Hernndez and
Naderi (2001), cells with large lumina and thin cell walls undergoes to damage during machining and gluing
process, and the same authors also observed that where the crushed cells were big the glueline became thicker. In
our case, the glued interphase was somewhat irregular, and the first rows (1-5) of cells were deformed, thus
constituting a zone of weakness for the mechanical strength. Consequently, the glueline was rather irregular as well.
Although it was 60 m thick on average, a broadening was sometimes observed, mainly where the vessels were
compressed (Figure 3).
Analyzed specimens of Scots pine showed (as expected) tracheids with variable wall thickness and diameters of
the lumens (10-30 m), depending on the vegetative period in which they were formed (Figure 4a). On the
interphase, latewood cells (smaller and with thicker cell walls compared to the earlywood tracheids) were not
damaged. In these areas, the adhesive completely wetted the cell surfaces, and there was no discontinuity in the
261
a)
b)
Figure 3: Transversal sections of the bonded chestnut (Castanea sativa Mill.) joints. Wood surface was deformed during the
gluing process. The Figure shows two different parts of a same sample: in (a) both the tyloses filling the vessels and some
adhesive into a vessel are visible. A split between bondline and wood surface is also visible; in (b) both crushed vessels and
heavy cell damage at the interface are evident.
on the same side of the observed detachments (Figure 4c). It is known that in general resins cause a considerable
worsening of adhesion (Christiansen 1991). As already noticed in the case of chestnut, cell deformations and
detachments induced the not-uniformity of the glueline, which was not very regular: it was thinner (approximately
100 m) close to latewood, and more enlarged (240 m) close to the earlywood cells, where wood density is lower
and wood cell walls thinner. A similar phenomenon was already observed previously in softwoods such as Douglas
fir and Southern pine (Murmanis et al. 1986, Jokerst and Stewart 1976, Hernndez and Naderi 2001).
a)
b)
c)
Figure 4: Transversal sections of the Scots pine (Pinus sylvestris L.) bonded joints. Picture of different samples, (a) an overview
of the interface; (b) wood surface in part deformed by the conditions of the gluing process; (c) drops of resin close to the contact
surface.
262
a)
b)
Figure 5: Values of tensile shear strength (a) and of wood failure percentage (b) for the various species considered. Boxes extend
from the lower quartile to the upper quartile values, the medians are given by a thicker line within the boxes, whereas the
whiskers represent data from the 5th- to 95th- percentile. The notch in boxplots represents the extension of the 95% confidence
interval for the medians.
Beech wood represents the reference species for the European standards on wood adhesives, because it can be
considered as an easy to glue wood. Tests reported in present work confirmed this occurrence, and in fact beech
evidenced the highest shear strengths compared to the other species. It is worth to note that micromorphological
observations showed for beech the regularity of the glueline, without any crushing of vessels or weakness regions. It
is also worth to note that the wood failure percentage in beech joints was as high as 100% (Figure 5b). This
occurrence is often associated to lower joint strengths, for instance because of anatomical defects of wood (usually a
slight slope of grain). However, in present case the grain of tested specimens was straight (not shown), and hence we
can conclude that the reason of such a behavior was different (see the following text).
Chestnut did not reach the limit threshold of shear strength provided in EN 204:2001 (which is 10 N/mm2,
although this value is referred to beech assemblies) (Figure 5a). In this case, the fact that the measured wood failure
percentage was 100% in all tested specimens (Figure 5b) was imputable to a wood-side weakness of the assemblies.
Micromorphological observations allowed evidencing that the reason of such a behavior was related to the local
deformations and presence of crushed cells at the wood-adhesive interphase that give rise to an extremely irregular
glueline; therefore, the anatomical properties of this species negatively affected the adhesion.
A similar occurrence was also found on birch, whose shear strength median value was however higher than for
chestnut (around 10 N/mm2 compared to 7.5 N/mm2, Figure 5a). In the case of birch, the distribution of wood failure
percentages (which ranged from approx. 90% to 100%, Figure 5b) evidenced an interphase also weak, but slightly
stronger than for chestnut. This fact was confirmed both by the measured higher mechanical values and by the
micromorphological analysis, which did not allow finding also in birch the wood splits that were instead observed in
chestnut.
The mechanical behavior of pine was similar to that of chestnut (Figure 5a), whereas the wood failure
percentages fell in the range 90-100% (Figure 5b). In that species the interphase showed regular zones adjacent to
263
264
mechanisms overpassed (in terms of importance) the opposite process by which a larger and rougher contact area (as
that one generated by a distorted interphase) would increase the positive effect of the mechanical interlocking and of
van der Waals-type interactions. Moreover, conifer joints suffered of the presence of the resins contained in the
parenchymatic cells: Scots pine was the only species among those considered in which resin leakage was observed,
International Scientific Conference on Hardwood Processing (ISCHP 2013)
which appreciably weakened the mechanical strength of joints.
REFERENCES
Bikerman, J.J. 1961. The Science of Adhesives Joints. Academic Press, New York.
Christiansen, A.W. 1991. How overdrying wood reduces its bonding to phenol-formaldehyde adhesives: A critical review of the
literature. Part II. Chemical reactions. Wood Fiber Sci. 23, 6984.
Collett, B.M. 1972. A review of surface and interfacial adhesion in wood science and related fields. Wood Science and Technology
6:142.
Good, R.J. 1972. Theory of cohesive vs. adhesive separation in an adhering system. J. Adhesion 4:133154.
Hernndez, R. E and N. Naderi. 2001. Effect of knife jointing on the gluing properties of wood. Wood Fiber Sci. 33(2):292301.
Jokerst, R. W., and H. A. Stewart. 1976. Knife-versus abrasive-planed wood: quality of adhesive bonds. Wood Fiber Sci 8(2):107
113.
Kamke and Lee .2007. Adhesive penetration in wooda review. Wood Fiber Sci, 39(2): 205 220
Marra 1992. Technology of wood bonding: Principles in practice, 1 Ed., Van Nostrand Reinhold, New York. 7680 pp.
McBain, J.W. and D.G. Hopkins, 1925. J. Phys. Chem. 29: 188.
Murmanis, L., B. River, and H. Stewart. 1986. Surface and subsurface characteristics related to abrasive-planing conditions.
Wood Fiber Sci. 18(1):107117.
Pizzi, A. 1994. Advanced wood adhesives technology. Marcel Dekker, Inc. New York, NY. 289 pp.
Pizzi, A. 1998. Special section: Wood adhesives. Foreword. Int. J. Adhesion & Adhesives 18: 67.
River, B.H. and V.P. Miniutti,. 1975. Surface damage before gluing weak joints. Wood Wood Prod. 80(7):3536.
List of cited standards
E N 204. 2001. Classification of thermoplastic wood adhesives for non-structural Applications
EN 205. 2003. Adhesives-Wood adhesives for non-structural applictinos. Determiantion of tensile shear strengh of lap joints.
265
Nowadays hardwood is increasingly used in timber construction, sometimes as a complete substitution for
softwood members and sometimes in combination with softwood. In principle, hardwood shows a similar
orthotropic behavior to softwood; however the ratios of the mechanical parameters between the three anatomical
directions and the magnitudes are different and depend strongly on the individual microstructure of the species.
The aim of our study is to characterize the three-dimensional elastic behavior of common ash (Fraxinus excelsior)
by tensile and compression tests on dog-bone-shaped specimens in the three anatomical directions and at stepwise
angles in between, using a universal testing machine in combination with a digital image correlation technique. As
a result, we determined Youngs moduli and Poissons ratios for the different load directions using the same
sample. Furthermore, several important mechanical properties (tensile, bending, compression, and shear strength
as well as fracture toughness) were determined for ash within this study. From prior studies it is known that in the
radial-tangential plane of wood the elastic moduli in the principal directions is clearly higher than the off-axis
elastic moduli. This angular dependency has not previously been experimentally proven for hardwoods commonly
used in timber engineering. Furthermore, the dependency of mechanical parameters on the fiber-load angle is
experimentally determined. The measurements for principal and off-axis load directions provide a sound basis for
modeling of hardwood structures.
1. INTRODUCTION
In the forests of central Europe a semi-natural forestry tendency is observable. This means that, in the long term, a
higher percentage of hardwoods will be present in the structure of forest stands. In the last decade, a reduction of 40%
in hardwood production was recorded which, for example, has contributed to an under-use of about 50% for this
species group in Switzerland (Schmitt, 2012). One of the intrinsic problems is the assortment distribution. Substitution
materials such as synthetic materials, metal, and concrete as well as other wood products, reduce the demand for
hardwood products so that about 60% of the harvest is used directly for energy purposes without flowing up the value
chain. Thus it is worth considering a more holistic use of the hardwood natural resource. It is plausible to assume that
hardwood would be more highly demanded if more detailed species specific data about construction relevant
properties was available.
Elasto-mechanical material data is needed for nearly every material model to describe the mechanical properties of
wood. The determination of the parameters used in the model is an intrinsic part of the model development and the
accurateness of the determined data is decisive for the quality of the model. In the case of wood, the mechanical
properties differ significantly within the three anatomical directions: longitudinal, radial and tangential. In general, the
material behavior of wood has been widely investigated. Based on solid state physics and 3D elastic material behavior,
it can be described using the compliance matrix (eq. 1) with the engineering elastic parameters, where ii are the
normal strains, ij the shear strains, ii the normal stresses and ij the shear stresses, Eii the moduli of elasticity, Gij the
shear moduli and ij the Poissons ratios. Thereby, the first index signifies the strain direction and the second index the
stress direction.
* Sebastian Clau: Tel.: (+41) 44-63-36866; Fax: (+41) 44-63-21174; E-mail: sclauss@ethz.ch
266
1
11 E
11
12
22 E
11
13
33 E
11
23 0
13 0
12 0
21
1
E33
1
G23
1
G13
E22
1
E22
31
23
E22
E33
32
E33
G12
11
22
33
23
13
12
(1)
Selected property parameters of hardwood have been published by Kollmann (1941), Kollmann (1951), Bodig and
Jayne (1993), Szalai (1994), Pozgaj et al. (1997), Sell (1997), Wagenfhr (2007), and Sliker (1993), amongst others.
However complete data sets for the three main wood directions that are sufficient for static calculations and modeling
in wood construction and also for calculations on multi-layered boards, parquet or musical instruments with the finite
element method rarely exist. In the majority, investigations are carried out only parallel to the grain under tension,
compression or bending load. Parameters for the other main directions transverse to the grain, which are required for
finite element calculations, as well as the influences of fiber and growth ring angles, are limited for hardwood or quite
simply not available. Equally, the influences of the load type (tension, compression, bending) and of the moisture
content on the elastic constants are scarcely investigated. Table 1 shows a summary of different mechanical and hygric
parameters of common ash presented in Bonoli et.al. (2005). In most of the cases only general values are given without
differentiating between the anatomical directions.
Since the usage of ash and also other hardwood species is becoming an issue of increasing interest in timberwork,
we aimed to provide a comprehensive dataset of mechanical and hygric properties of ash for practical utilization,
where the following material properties were determined within this framework:
Swelling ratios
Thermal conductivity
Key mechanical properties were tested at 50, 65, 80, and 95% relative humidity and 20C in different anatomical
directions or in the planes that arise from them, respectively:
Impact bending
Fracture toughness
In addition, the stiffness of the material was determined in and between the anatomical directions with an
increment of 15 in the fiber or growth ring angle, respectively. This part of the investigation will be presented within
this paper. The other parts of this comprehensive study will be presented in upcoming publications.
267
Unit
kg/m3
kg/m3
MPa
Value
450690860
700
680760
450..650880
510705888
680760
410650820
640700
0,2
4,55,0
5,3
5,0
8,08,4
8,7
8,0
44001250018100
13000
43001340018100
11900..13900
58001420020900
71001260019500
1190013900
235280
Origin
Germany
Switzerland
Belgium
Germany
Germany
Germany
MPa
MPa
MPa
MPa
Impact bending
268
kJ/m2
MPa
20..4468
4359
375977
315172
70165293
130
75176293
130160
7...11,2
58120210
105
49102178
100127
65130194
71119174
106880
1070250
6788
2090250
0,7
0,650,750,87
Wagenfhr (1996)
DIN 63364
Kollmann (1941)
Sell (1997)
Khne (1951)
Leclerco (1975)
Kollmann (1941)
Sell (1997)
Wagenfhr (1996)
Wagenfhr (1996)
Kollmann (1941)
Germany
Wagenfhr (1996)
DIN 63364
Germany
Switzerland
Belgium
Wagenfhr (1996)
DIN 63364
Kollmann (1941)
Sell (1997)
Khne (1951)
Leclerco (1975)
Germany
Wagenfhr (1996)
Germany
50
Compression strength (L)
Source
DIN 63364
Germany
Switzerland
Belgium
Germany
Germany
Germany
Germany
Germany
Switzerland
Belgium
Germany
Germany
Switzerland
Germany
Kollmann (1941)
Sell (1997)
Khne (1951)
Leclerco (1975)
Wagenfhr (1996)
DIN 63364
Kollmann (1941)
Sell (1997)
Wagenfhr (1996)
Wagenfhr (1996)
DIN 63364
Kollmann (1941)
Sell (1997)
Khne (1951)
Leclerco (1975)
Wagenfhr (1996)
Kollmann (1941)
Sell (1997)
Khne (1951)
Wagenfhr (1996)
Kollmann (1941)
The material used was common ash (Fraxinus excelsior) located near Zurich, Switzerland with an average density
of 0.60 g/cm and an average moisture content of about 12%. All test specimens for the determination of the physical
and mechanical properties were cut from the same trunk, however only sapwood was used for the investigation. All
knots or defects in the wood structure were excluded from testing in advance (i.e. only clear specimens were used).
Between the anatomical directions, specimens were cut at increasing angles of 15 along the fiber direction. In
total, 21 different configurations resulted that were tested at normal climate (20/65) only. Figure 1 shows how the
specimens were cut out of the boards for the individual angles. Since the subsequent angles were not always
comparable to the desired ones in each case, all fiber and growth ring angles of the individual specimens were
measured optically using image analysis. The test was conducted using 95 mm long dog-bone-shaped specimens (Fig.
2a). This shape allows transferring the forces at the clamps to the specimen without deformation in the radial and
tangential directions under longitudinal tensile load. At least 10 specimens were tested per individual direction. Since,
especially in the tangential direction, it is hardly possible to gather specimens with straight alignment of the growth
rings over the whole length of the sample, supports were glued on the boards made out of beech wood to obtain the
desired specimen dimensions.
a) LR plane specimens
L
R
b) TR plane specimens
Fig 1: Position and pre-cut of the specimens (arrows show load transmission direction during the test)
269
The quasi-static determination of the MOE was performed under tension and compression load using the same
specimen with a Zwick 100 (ZWICK/ROELL) universal testing machine equipped with a 100 KN load cell. The test
270
was performed with displacement control using a testing speed of about 1 mm/min. In the compression test, a simply
supported pressure plate was used to avoid bending effects on the specimen. In the tension test, conical clamp jaws
were used to achieve a central load transmission in both configurations.
For the strain measurement in this study we used an optical video correlation system (VIC 2D, Correlated
Solutions Inc., Columbia USA) that allows evaluation of the strain distribution over the complete specimen surface
using an optimized correlation algorithm to provide full-field displacement and strain data for mechanical testing on
planar specimens. The in-plane movement was determined as a mean strain in the load direction and across this
direction for the selected measurement area of interest, with a size of about 11 11mm (Fig. 2b) in between the
parallel section of the specimen. The system can measure in-plane displacements and strains from 50 microstrain to
2000 % strain and above. In the actual measurements, lateral strains in the longitudinal direction under tangential load
reach values of about 0.015 %. The specimen preparation only requires the application of a random speckle pattern.
For the tension and compression measurements, three cameras with a resolution of 2048 2048 and one with
1628 1236 pixel resolution, respectively (models AVT Pike F421B and Dolphin F201b, Allied Vision
Technologies), were used to determine the displacement data from each side of the specimen individually in the same
run (Fig. 2c). The image capturing was synchronized with the load signal via an A/D converter. After testing all
samples, the calculation of the strain data was performed in the linear range of 12.5 to 50% of the actual force at 0.2%
plastic deformation (F0.2) (Fig. 2d). These orientation dependent maximum forces were determined in a preliminary
test for the main directions with F0.2(L) = 60.00 MPa, F0.2(R) = 18.84 MPa and F0.2(T) = 13.22 MPa. The forces in the LR
and LT planes for the angles in between the main directions were calculated using Hankinsons formula (eq. 2) with
an exponent n=2.5 for the LR and the LT planes.
=
.2().2(/)
.2() +.2(/)
(2)
The strength in the TR plane varied as a sinusoidal function (eq. 3) described in Bodig and Jayne (1993). This
equation was used for the calculation of the maximum forces in the TR plane with the empirical constant K=0.2.
2
(3)
The MOEs for the individual specimens were calculated from the stress-strain data for each specimen surface
individually. From these values an outlier test was performed excluding individual specimens with values lying
outside the one and a half interquartile range of the complete sample. In several cases, bending effects led to differing
values of the opposite cameras, however by averaging the values, the errors cancelled each other out. From the
reduced sample the mean value and variance was calculated.
3. RESULTS AND DISCUSSION
The results of the tension and compression tests measurements are summarized for the anatomical directions in
steps of 15 (Table 2). For the three anatomical directions, twice as many samples were tested (see Fig. 1: LT 0 = LR
0 = longitudinal, TR 90 = LR 90 = radial, TR 0 = LT 90 = tangential), however, the groups belong to samples from
different parts of the trunk. The individual mean values do not always correspond exactly to each other, with the
greatest variation occurring in the radial direction, which indicates that the position in the trunk is of particular
importance. The number of samples varies slightly in the different orientations caused by individual outliers and
available samples.
The mean values in the fiber direction reach MOEs around 9.0 GPa in compression load and about 8.3 GPa in
tension load. The radial direction shows values of about 1.0 GPa in compression and tension loads. The tangential
direction reveals the lowest values with 0.6 GPa in compression and tension loads. According to the results, no
significant variation between the types of load was obtained.
As shown before in a study by Baumann (1992), the values in tension and compression load do not significantly
differ. He also obtained slightly higher values in the longitudinal direction for the compression load. On comparison
with other studies on ash (Table 2), the currently obtained values are significantly lower, depending on individual
wood specific parameters. As is well known, the density in particular plays a major role in regard to the mechanical
properties. The results of the displayed values in Table 3 obviously show an increasing MOE with increasing density
in the longitudinal direction. The directions perpendicular to the grain do not follow this correlation. On the whole, one
can see a variation in the values of about 25% in the literature, which is not unusual in the case of wood. The ratios
271
between the individual directions, however, do not differentiate to such a magnitude. The mean ratio between
longitudinal : radial : tangential is about 13.4 : 1.6 : 1.0, which is in accordance with these measurements.
In Figure 1 the MOE from the tension and compression tests are displayed in and between the anatomical
directions. In general, the dependency of the MOE on the orientation is particularly pronounced. Up to 30 deflection
from the longitudinal direction results in the modulus decreasing to about one third of the maximum value. In principle
the gradient can be described with Hankinsons formula (eq. 2) with an exponent n=2.5 for the description of the
obtained data in the LT plane and 2.25 in the LR plane, for tension as well as for compression. The data for the TR
plane are in good accordance with the fitted curve using Equation 3 with the empirical factor K=0.4. The values of
MOE in the main directions for calculation of the fit were taken from the results obtained in the individual plane.
In the TR plane a minimum of the off-axis modulus is revealed at about 45 between the radial and tangential
directions, as shown in Grimsel (1999) for Norway spruce and beech by a theoretical transformation of stress-strain
equations to a 3D coordinate system. Garab et al. (2010) compared experimental results of softwoods with distinctly
different anatomic structures with those calculated by a tensor transformation to determine whether it is admissible to
treat the investigated wood species as orthotropic materials. The assessment of common ash and a comparison of it
against the wood specific features of ring porous hardwood will be the next step for our investigation. Since, in the
current investigation, data in and between all three anatomical directions were obtained, a 3D comparison between
calculated and experimentally obtained data will be possible.
Table 2: MOE determined in tension and compression tests with loading in and between
the three anatomical directions L, R, T ( = 121%)
x
0 n
x
15 n
x
30 n
x
45 n
x
60 n
x
75 n
x
90 n
ELT
9196
7
0.601
7004
8
0.617
2403
9
0.615
1488
9
0.622
984
10
0.652
704
10
0.608
696
9
0.62
11.6
5
17
2.6
4
2.3
10
1.9
14.4
8.1
7.2
2.3
4.9
1.8
compression
ELR
8792
10.4
10
0.606
3
7023
32.9
10
0.617
1.1
2970
7.8
10
0.624
2.1
1684
7.7
10
0.611
3.8
1185
4.8
10
0.617
3.2
880
3.6
10
0.597
5.9
838
2.9
10
0.589
0.7
ETR
578
9
0.577
602
10
0.617
559
10
0.592
463
10
0.554
663
10
0.599
1002
10
0.603
1143
9
0.563
29.6
4.2
12.5
1.8
7
1.4
2.2
0.5
4.4
0.5
6.2
0.8
2.3
1.4
ELT
8446
9
0.601
6106
10
0.612
2966
9
0.615
1620
9
0.623
1052
10
0.652
740
10
0.608
637
10
0.62
9.9
5
12.2
2.8
11.8
2.3
11.5
1.9
18.7
8.1
8.4
2.3
6.9
1.8
tension
ELR
8235
12
9
0.608
3
6305
18
9
0.617
1.1
2901
7.9
10
0.624
2.1
1694
8.3
9
0.614
3.7
1155
6
10
0.617
3.2
849
3.1
10
0.597
5.9
866
3.1
10
0.589
0.7
ETR
625
9
0.573
574
10
0.617
523
10
0.592
433
9
0.553
655
10
0.599
995
10
0.603
1193
9
0.565
2.7
3.8
14.5
1.8
5.5
1.4
3.2
0.4
2.9
0.5
8.6
0.8
3.9
1.4
Eij MOE for each orientation in MPa in the ij-plane where the i-orientation has an angle of 0 and the j-orientation has
an angle of 90; x = mean value in MPa; = variation coefficient in %; = raw density in g/cm3, n = number of
samples
272
Publication
Own measurements
Pozgaj et al. (1993/1997)
Niemz (1993)
Hearmon (1948) in Szalai (1995)
Stamer (1935)
Baumann (1922)
Load type
compression
tension
compression
compression
compression
tension
12
12
10-12
12
9
14
12
-
8994
8341
15798
13000
15800
15000
8994
10336
8947
991
1030
1875
1500
1510
1640
991
1669
1590
637
631
1268
820
800
970
637
1062
1033
273
274
Compression and tension tests on the same specimens from common ash (Fraxinus excelsior) of various
orientations, with respect to the fiber and growth ring orientation, were conducted. The off-axis MOE was obtained
and predicted by Hankinsons formula. The dependency on the fiber and growth ring direction obtained by the
experiment and regression coincided well with each other in the off-axis fiber and growth ring angles. The empirical
parameters obtained for Hankinsons formula amounted to n=2.5 in the LT and n=2.25 in the LR planes. The empirical
factor for the calculation of the off-axis modulus in the TR plane amounted to K=0.4. The data obtained for common
ash provides a sound basis for material modeling purposes.
5. LITERATURE
Baumann, R. 1922. Die bisherigen Ergebnisse der Holzprfungen in der Materialprfungsanstalt an der TH Stuttgart. Verlag des
Vereins Deutscher Ingenieure, Berlin.
Bodig, J. and Jayne. 1993. Mechanics of Wood and Wood Composites. Krieger Publishing Company.
Bonoli, C., P. Niemz, P. and D. Mannes. 2005. Investigation into certain mechanical properties of ash. Schweizer Zeitschrift fr
Forstwesen 156, 432-437.
DIN 68364 2003. Properties of wood species - Density, modulus of elasticity and strength
Hearmon, R. and G. Britain. 1948. The elasticity of wood and plywood. HM Stationery Office London.
Garab, J., Keunecke, D., Hering, S., Szalai, J. and P. Niemz. 2010. Measurement of standard and off-axis elastic moduli and
Poissons ratios of spruce and yew wood in the transverse plane. Wood Sci. Technol. 44, 451-464.
Grimsel, M. 1999 Mechanisches Verhalten von Holz: Struktur- und Parameteridentifikation eines anisotropen Werkstoffes.
Dissertation, Dresden, 89p.
Kollmann, F. 1941. Die Esche und ihr Holz. Verlag Julius Springer, Berlin. 147 pp.
Kollmann, F., 1951. Technologie des Holzes und der Holzwerkstoffe. Springer, Berlin.
Khne, H. 1951. Untersuchung ber einige Eigenschaften des Eschen- und Robinienholzes im Hinblick auf dessen Verwendbarkeit
fr Werkzeugstiele. Forschungsbericht Empa
Leclerco, A. 1975. La qualit du bois de frne. Bull. rech. Agron. Gembloux 10(4): 497526.
Niemz, P. 1993. Physik des Holzes und der Holzwerkstoffe. DRW-Verlag.
Ozyhar, T., Hering S. and P. Niemz. 2013. Moisture-dependent orthotropic tension compression asymmetry of wood.
Holzforschung. 67(4): 395404
Pozgaj, A., Chovanec, D., Kurjatko, S. and M. Babiak, M. 1997. Struktura a vlastnosti drevan. Priroda.
Sell, J. 1997. Eigenschaften und Kenngrssen von Holzarten. Baufachverlag, Dietikon. 87 pp.
Schmitt, J. 2012. Analyse Laubholzmarkt (Inputpapier Laubholz-Beirat)
Sliker, A., and Y. Yu. 1993. Elastic constants for hardwoods measured from plate and tension tests. Wood and Fiber Science 25,
8-22.
Stamer, J. 1935. Elastizittsuntersuchungen an Hlzern. Archive of Applied Mechanics 1, 1-8.
Szalai, J. 1995. A faanyag s faalap anyagok anizotrp rugalmassg- s szilrdsgtana. Hillebrand Nyomda Kft., 1-304. cessed
November 4, 2008.
Wagenfhr, R. 1996. Holzatlas. Fachbuchverlag Leipzig. 688 pp.
275
276
277
278
279
280
281
282
ABSTRACT
Chestnut was one of the timber species commonly used for structures in Portugal, namely for roof and floor
structures. Despite this, the design was mostly based on empirical experience, and for this reason hardly any
information is available regarding its mechanical properties. This paper describes a study undertaken to assess the
most relevant mechanical properties, from an engineering point of view, for Portuguese chestnut structural beams.
In the study, non-destructive techniques were used in combination with destructive approaches. A sample of new
timber specimens with structural size was selected. Each of the specimens was tested to determine the dynamic
modulus of elasticity. The technique used was axial vibration. Next, all beams were tested in accordance with EN
408 to determine the global modulus of elasticity and the bending strength. After failure, in bending, the test
samples were collected, near the supports where no damage was observed, to determine the material
compression strength. The results obtained are presented and discussed in the paper.
1. INTRODUCTION
Timber and stone were the main construction materials in Portugal until the end of XIX Century. This situation
was completely changed in the first half of the XX century due to the fast development of concrete technology (Dias,
1999). Traditionally, the species used in Portugal were: Maritime pine (Pinus pinaster Ait.), Pitch pine (Pinus
palustris Mill.), Pine Redwood (Pinus silvestris L.), Oak (Querqus robur L.) and Chestnut (Castanea sativa Mill.)
(Machado et al., 2009). The specie used in each particular situation was closely connected to the availability of the
material and for this reason oak and chestnut were predominant in the North of Portugal while the pine was the
predominant specie in the South (Feio, 2005). Chestnut was a popular wood due to its stability and durability
properties, combined with adequate mechanical properties. This was considered a noble specie, used in building
structures as well as in Palaces and Castles (Feio, 2005). At that time structural design was not based on technical
information and knowledge but on empirical experience obtained from previous constructions. For these reasons
hardly any information was available about this wood type.
In this century and t h e last decade of the XX century, the interest for timber as a structural material has
increased. This renewed interest was inspired by a number of emblematic structures built in Portugal (e.g., Utupia
Pavillon). Despite this, the use of chestnut timber is mostly limited to rehabilitation/renovation applications. As a
consequence, several research studies have been undertaken to obtain the technical information that is required for the
design of structures made with chestnut. Loureno et al. (2007) undertook a study to assess the compression
perpendicular to the grain required in the design of the traditional Portuguese roof systems. The same authors also
assessed the possibility of using non-destructive methods (ultrasounds, needle penetration and micro-drilling) in the
evaluation of timber structures (Feio et al., 2007). Dias et al. ( 2 0 0 7 ) tested chestnut specimens to obtain the
embedding strength of this timber. The use of chestnut timber in new products, such as glue- laminated timber, has
also been tested (Ribeiro et al., 2009).
In spite of these research studies the information available is still limited and this poses one of the major
difficulties that results in the unavailability of a grading standard for this specie. To overcome it some authors
(Machado et al., 2009) suggest the use of the Italian standard UNI 11035-2 (UNI, 2010).
* Alfredo Dias: Tel.: (00351) 966634821; Fax: (00351) 239797233; E-mail: alfgdias@dec.uc.pt
283
EA
2 u x, t m 2u x, t
0
L t 2
x 2
(1)
Where, = axial stiffness, (, ) = longitudinal displacement, = mass of the specimen, = length of the specimen
f
n
2L
(2)
Where, f = natural frequency, = vibration mode, = length of the specimen, E = modulus of elasticity and
= density
2
2
E dyn 4L f
Where,
(3)
= axial dynamic modulus of elasticity, = length of the specimen, f = natural frequency and = density
In these tests the vibration was initiated by a longitudinal impact at one end of the beam and the acceleration was
measured by a piezoelectric accelerometer (Figure 1). The results of natural frequencies and the axial dynamic
modulus of elasticity are presented at Table 1.
284
E m, g
3al 2 4a 3
w w1
6a
2bh 3 2 2
5Gbh
F
F
2
1
(4)
Where, , = global modulus of elasticity, = distance between loading head and the nearest support, = with of
cross section, = depth of cross section, 2 1 = increment deformation and 2 1 = increment of load
The bending strength was determined according to the Equation (5), from EN 408 (CEN, 2010) and the load was
applied to reach the maximum load at 300 120 seconds. Figure 2 illustrates the test arrangement and a typical
failure.
fm
3Fa
bh 2
(5)
285
Where,
F
bh
(6)
Figure 3: Typical visible gross shear band of compression strength parallel to grain tests.
286
b (mm) h (mm)
L (mm)
(kg/m3)
w (%)
f (Hz)
Edyn (MPa)
Em,g (MPa)
fm (MPa)
fc,0 (MPa)
73.1
96.4
1976
544.5
15.1
1211
12460.5
9256.3
44.2
41.9
74.0
100.9
1975
685.9
16.4
972
10108.0
9087.6
54.1
30.4
74.1
97.2
1972
511.2
15.5
1168
10854.6
9202.2
51.5
31.2
74.4
102.1
1974
677.7
16.0
1071
12112.4
8041.0
35.8
31.0
74.3
98.2
1979
551.4
17.7
1268
13896.2
11740.6
58.4
42.0
72.0
97.4
1973
536.8
15.0
1280
13684.2
8636.3
60.2
39.0
75.7
101.0
1980
530.2
15.9
1073
9580.2
8520.6
43.1
29.2
72.0
101.3
1977
556.2
17.0
1209
12707.8
10713.6
48.9
47.6
72.5
97.6
1978
697.7
17.4
995
10808.4
8105.8
43.2
29.3
10
74.1
96.9
1985
532.9
15.7
1063
9491.7
8592.5
51.3
32.3
11
74.8
94.3
1970
696.9
23.4
1075
12503.0
9622.6
48.8
33.5
12
72.9
101.1
1978
558.1
15.9
1015
8999.0
7697.7
45.3
28.5
b Thickness; h Width; L Length; Density; w Moisture content; f Frequency of vibration; Edyn Dynamic
modulus of elasticity; Em,g Global modulus of elasticity; f m Bending strength; fc,0 Compression strength parallel to
grain
According to EN 384 (CEN, 2009b), if the specimens are not tested at the reference conditions (12% of moisture
content) but with a mean moisture content ranging from 10% to 18% the modulus of elasticity, density and
compression strength parallel to grain should be adjusted according to Equations (6), (7) and (8), respectively.
12 w 1 0.005w 12
(6)
Where, 12 = density at 12% of moisture content, w = density at the moisture content of the test, w = moisture content
E 0 E0, w 1 0.02w 12
(7)
Where, E0 = modulus of elasticity at 12% of moisture content, E0,w = modulus of elasticity at the moisture content of
the test
f c,0 f c,0, w 1 0.03w 12
(8)
Where, fc,0 = compression strength parallel to grain at 12% of moisture content, fc,0 = compression parallel to grain at
the moisture content of the test
The values of modulus of elasticity and density adjusted to 12% moisture content are presented at Table 2.
287
Table 2: Values of density (12), global modulus of elasticity (Em,g) and compression strength parallel to grain (fc,0)
adjusted to a moisture content of 12%
N
12 (Kg/m3)
Em,g (MPa)
fc,0 (MPa)
536.1
9824.1
45.7
670.9
9881.3
34.4
502.2
9852.5
34.5
664.1
8684.3
34.7
535.5
13086.9
49.2
528.7
9160.2
42.5
519.8
9190.9
32.6
542.4
11777.8
54.7
678.9
8975.8
34.0
10
523.0
9234.1
35.9
11
657.3
11810.1
45.0
12
547.1
8303.3
31.8
In spite of the small size of the sample considered in this study, there is a set of results that allow useful conclusions
from an engineering point of view.
The mean value of the density obtained was 575.5Kg/m3, ranging between 519.8 Kg/m3 and 678.9 kg/m3. These
values, although slightly lower, are in line with what was obtained by Feio et al. (2007) either for new or old timber.
The mean value obtained for the dynamic elasticity modulus was 11433.8MPa which is 14% higher than what was
obtained afterwards for the static modulus of elasticity. Feio et al. (2007) obtained 15000MPa and 18000MPa for the
dynamic elasticity modulus (new and old timber, respectively) using the ultrasound technique. For the static modulus
of elasticity the mean value was 9981.8MPa, ranging between 8303.3MPa and 13086.9MPa.
Regarding the bending strength the mean value was 48.7MPa, ranging between 35.8MPa and 60.2MPa. For
compression strength parallel to grain the mean value was 39.6MPa, ranging between 31.8MPa and 54.7MPa. Feio et
al. (2007) obtained mean values of 42.9MPa and 47.6MPa for new and old chestnut wood respectively. The values
obtained by Feio are 24% and 37% higher than the results obtained in this study.
Taking into account all of the properties obtained in these tests, the properties of the chestnut are close to the
strength class D30 as defined in EN 338 (CEN, 2009a).
In spite of the small sample, the correlations between different properties can be useful since they give some insight
on the feasibility of the method to predict other properties that usually requires additional work for its determination or
involves sample failure such as the bending strength or compressive strength parallel to grain tests used in this
study. The correlations between modulus of elasticity (dynamic vs global) (Figure 3) and between modulus of
elasticity (dynamic and global, separately) and bending strength were established (Figure 4 and 5, respectively). The
correlations were made considering the results adjusted to 12% moisture content. A summary of correlation
coefficients between the properties determined in this study are presented in Table 3.
y = 0,5427x + 3776,3
R = 0,3867
9500,0
8500,0
7500,0
8000,0 9000,0 10000,0 11000,0 12000,0 13000,0 14000,0
Dynamic modulus of elasticity (MPa)
288
65,0
60,0
y = 0,0014x + 32,169
R = 0,1228
55,0
50,0
45,0
40,0
35,0
30,0
8000,0
9000,0
10000,0
11000,0
12000,0
13000,0
14000,0
Figure 4: Correlation between bending strength and axial dynamic modulus of elasticity.
65,0
y = 0,0022x + 27,004
R = 0,2112
60,0
55,0
50,0
45,0
40,0
35,0
30,0
7500,0
8500,0
9500,0
10500,0
11500,0
12500,0
13500,0
Edyn
Em,g
fm
fc,0
-0.62
-0.01
-0.23
-0.35
-0.36
-0.62
0.79
0.49
0.49
0.81
Edyn
-0.01
0.79
0.62
0.35
0.79
Emg
-0.23
0.49
0.62
0.46
0.80
fm
-0.35
0.49
0.35
0.46
0.38
fc,0
-0.36
0.81
0.79
0.80
0.38
Statistical analysis of the correlation coefficients presented in Table 3 shows that these vary from -0.01 (no
significance) up to 0.81 (high significance). The higher correlation coefficient was determined between frequency and
compression parallel to grain and the lower was between density and axial dynamic modulus of elasticity. Also, the
correlations between modulus of elasticity and compression strength parallel to grain resulted in high correlation
coefficients (0.79 and 0.80). Considering the correlations with bending strength, 0.49 was the highest correlation
coefficient (frequency from non-destructive tests) followed by static modulus of elasticity (0.46). The correlations
which combined density with another property gave poor correlation coefficients probably due to the dispersion of the
density results obtained.
Although there were some high correlation coefficients, most of them are lower than those found for other
Portuguese species namely Maritime pine that were obtained by Morgado (Morgado, 2013), with the exception
being the correlation between static modulus of elasticity and compression strength parallel to grain.
289
4. CONCLUSIONS
This study was undertaken to help remedy the lack of knowledge about the mechanical properties of Portuguese
chestnut wood, namely through the experimental assessment of the most important engineering properties. Two
different approaches were used: non-destructive and destructive tests. Longitudinal vibration was the method of nondestructive testing used to determine the axial dynamic modulus of elasticity. The assessment of the other
mechanical properties (modulus of elasticity, bending strength and compression strength parallel to grain) was done
in accordance with the static tests prescribed in the EN 408 (CEN, 2010). The compression strength parallel to grain
was determined with one sample collected from each specimen tested to bending strength.
The tests on a sample of twelve beams with structural size (ranging from 72 to 75.7 mm in thickness, 94.3 to
102.1 mm in width and 1970 to 1985 mm in length) yielded a density mean value of 575.5Kg/m3, an axial modulus
of elasticity mean value of 11433.8MPa, and a static modulus of elasticity mean value of 9981.8MPa. Related to
bending strength and compression strength parallel to grain, the mean values were 48.7MPa and 39.6MPa,
respectively. According to the EN338 (CEN, 2009a) and taking into account the properties obtained in this study, the
properties of the chestnut are close to the strength class D30. When compared with data available in the bibliography
for other species the values obtained here are slightly lower Feio et al. (2007).
Correlations were established between all the properties determined in this study to find the best predictor of
bending strength and compression strength parallel to the grain. Despite the small size of the sample there are a number
of conclusions that can be drawn. The highest correlation coefficient was obtained between frequency and
compression parallel to grain (0.81) followed by static modulus of elasticity (0.80) and axial dyna mic modulus of
elasticity (0.79). In the correlations including bending strength, the highest correlation coefficient was determined
with frequency (0.49) followed by static modulus of elasticity (0.46). The worst correlation coefficients were
determined for those correlations that included the density variable.
The values obtained for some correlations were high, nevertheless a sample with more specimens is required to
improve the reliability of the correlations before the validity and usefulness of employing the longitudinal vibration
method as a predictor of stiffness and strength properties can be established.
REFERENCES
CEN 2009a. EN 338 - Structural timber - Strength classes. Brussels: CEN.
CEN 2009b. EN 384 - Structural timber - Determination of characteristic values of mechanical properties and density. CEN.
CEN 2010. EN 408 - Timber structures - Structural timber and glued laminated timber - Determination of some physical and
mechanical properties. Brussels: CEN.
CLOUGH, R. W. & PENZIEN, J. 1993. Dynamics of structures, McGraw Hill International Editions.
DIAS, A. M. P. G. 1999. Estruturas mistas madeira-beto. Msc.
DIAS, A. M. P. G., LOPES, S. M. R., VAN DE KUILEN, J. W. G. & CRUZ, H. M. P. 2007. Load-carrying capacity of
timber-concrete joints with Dowel-type fasteners. Journal of Strutural Engineering -ASCE, 133, 720-727.
FEIO, A. 2005. Inspection and Diagnosis of Historical Timber Structures: NDT Correlations and Structural Behaviour. PhD,
University of Minho.
FEIO, A. O., LOURENCO, P. B. & MACHADO, J. S. 2007. Non-Destructive Evaluation of the Mechanical Behavior of Chestnut
Wood in Tension and Compression Parallel to Grain. International Journal of Architectural Heritage, 1, 272-292.
LOURENCO, P. B., FEIO, A. O. & MACHADO, J. S. 2007. Chestnut wood in compression perpendicular to the grain:
Non-destructive correlations for test results in new and old wood. Construction and Building Materials, 21, 1617-1627.
MACHADO, J. M. S. R. S., DIAS, A. M. P. G., CRUZ, H. M. P., CUSTDIO, J. E. P. & PALMA, P. M. C. 2009. Avaliao e
Reforo de Estruturas de Madeira, Verlag Dashofer.
MORGADO, T. F. M. 2013. Classificao e Aplicao Estrutural da madeira de Seco Circular de Pinheiro bravo. PhD Thesis,
University of Coimbra.
RIBEIRO, A. S., SALAVESSA, E., LIMA, A. M. & LOUSADA, J. L. 2009. Estudo do Comportamento Mecnico de Vigas Macias e
Laminadas de Madeira de Castanho. In: FLORESTAIS, S. P. D. C. (ed.) 6 Congresso Florestal Nacional. Ponta Delgada
- Aores.
UNI 2010. UNI 11035-2 Legno strutturale - Classificazione a vista dei legnami secondo la resistenza maccanica. Part 2: Regole
per la classificazione a vista secondo la resistenza meccanica e valori caratteristici per tipi legname strutturale. In: UNI
(ed.). UNI.
VRIES, d. P. A. 1998. Strength grading of small-diameter Larch. Report C4-98-02. Delft Univeristy of Technology.
290
ABSTRACT
The use of high density hardwood species for furniture and interior purposes can be limited because the weight.
Keeping the weight low is important both from a user perspective and for logistic reasons in the manufacturing and
distribution process. This work describes the construction and mechanical characteristics of a new type of
light-weight panel in wood. The panel is a sandwich construction in three layers with hardwood as the surface layers.
The surface layers are made of 6 mm thick solid beech and the core consists of solid pine wood in thicknesses of 24 or
96 mm cross-laminated to the surfaces. The total panel density was then 373 and 294 kg/m3 given a beech surface
layer with a density of 725 kg/m3. The presented light-weight panel resulted in a 50-60 % decrease in the use of wood
compared to a traditional edge-glued panel. Tests of the mechanical properties showed a bending stiffness in the
longitudinal direction (of the core) of 2.9 kNm2 for a 36 mm thick panel and 221 kNm2 for a 108 mm thick panel. In
the transverse direction, the corresponding values were 11 kNm2 and 88 kNm2. The overall results indicate that the
light-weight panel presented shows promising technical and environmental properties and can thereby contribute to
an increased use of hardwood species.
1. INTRODUCTION
Sawn hardwood is an important raw material for the furniture and joinery industries. Compared to the common softwood species
in northern Europe, the high density hardwoods (e.g. beech, birch and oak) that are now becoming more common in forestry have
several advantageous properties such as greater strength, greater hardness and greater abrasion resistance. Their higher weight and
in many cases greater distortion are, however, properties that limit the use of these species and the search for ways to overcome these
disadvantages is an important issue for the wood material industry. The development of light-weight materials can be one way
towards a better future.
In northern Europe and especially in the Scandinavian countries and Russia, softwood species are much more common than
broad-leaved trees in the forests, with most of the hardwoods being used in the pulp and paper industries. This has been the trend
since the 1950s, but it has changed drastically during the last decade. The reasons for this change are: the increased awareness of the
need for a sustainable environment and especially biological diversity; the restructuring of the pulp and paper industry towards new
products, e.g. textiles, which change the raw material needs; and an increase in a silviculture that promotes the qualities of the forest
as an environment for everyday recreation. This has led to greater volumes of hardwood on the market and a need to develop new
products that take into account their species-specific properties.
Light-weight materials have been used in the aircraft and space industries for a long time, where low weight is needed in
combination with high stiffness. The common core materials are made of aluminium, aramid-paper-reinforced plastic or impregnated
paper (Blitzer 1997). The main benefits of the light-weight materials (panels) for furniture applications are their high
strength-to-density ratio and low weight. A low weight is advantageous in the transport and handling of products and it also lowers
the transportation costs (Michanickl 2006). In terms of recycling and resource-efficient material management, light-weight materials
made of wood, paper or recycled cellulosic materials have been developed in recent years mainly in the furniture industry. Feifel et
al. (2013) have shown that an increased used of light-weight wood-based materials for furniture purposes in general can greatly lower
the environmental impact of greenhouse gas emissions, non-methane volatile organic compounds (NMVOC) and formaldehyde from
that industry. Light-weight structural materials based on wood and paper layers are being increasingly considered as substitutes for
291
commonly used materials such as fibre- or particle-boards. In terms of recycling and energy consumption, foam and composite cores
based on non-renewable resources are, however, critical.
Light-weight materials can be divided according to their function and structure into three groups: 1) light-weight materials
which combine materials with a low weight-to-strength/stiffness ratio, 2) structural light-weight materials which are structures
mainly for building purposes that, with a minimum of weight, can distribute applied loads, and 3) light-weight systems the
superposition of functions of light-weight materials, providing not only supporting functions but also thermal insulation, etc.
In design, there are several ways of reducing the weight of a product, e.g., a piece of furniture, without losing function and
without any deterioration in the appearance of the final product. Different types of core materials for use as components in
light-weight materials have been on the market since the 1940s (Zenkert 1997). Pflug et al. (2004) distinguish two ways of supporting
the top layers of a sandwich construction: homogeneous support of the top layers (e.g., foam cores) and structured support of the
top layers (e.g., textile/wire cores, air-cushion cores, wave cores, and honey-comb cores).
Top layers can be produced from a wide range of panel-shaped materials like fibreboard, chipboard, veneers or high-pressure
laminates (HPL). The typical thickness of the top layer is 2 to 8 mm. Such a top layer has to have several important properties
to strengthen the whole sandwich construction: a) high bending strength, b) high impact strength, c) high internal bond strength
(chip/fibreboards), and d) high tensile strength and compression strength. The lower these properties are, the greater is the effect of
the core material (Riepertinger 2004).
To benefit from the properties of the outer layer and the core, both parts have to be glued to create a rigid composite with good
shear stiffness and tensile strength. The typical mode of failure of the composite is delamination between the core and the top layer
under a shear load or tensile load. In a typical paper-based honeycomb core the adhesive must therefore form a concave fillet weld,
otherwise the connection is limited to the small surface of the honeycomb ligament.
With a lower weight, new problems with joint and edge stability and small face coating arise. For this reason, the first panels were
made with wooden frames to guarantee overall stability and make joining possible. This technology, however, restricts continuous
panel production and variable formatting in individual sizes. For this reason, frameless panels were developed (Britzke 2003) with
new joining technologies and machine tools to set the joints (Herold et al. 2006). A great disadvantage of light-weight panels with
paper cores, and also with other types of core material, is their low resistance to moisture and free water, which results in low
durability in moist environments. One way to overcome this problem is by impregnation of the core material to prevent moisture
uptake or to use core materials that interact with moisture without losing their mechanical properties, i.e., solid wood.
During recent years, several ideas for light-weight panels of exclusively solid wood that combine lower weight with high strength
have been presented. These panels are mainly for buildings and other structural systems due to their wide range of advantages, i.e.,
high strength-to-weight ratio, reduction in the cost of framework and good thermal and sound insulation properties (Pokharel, 2003).
Dendrolight is a panel that is available on the market that consists of a core layer of cross-aligned boards with longitudinal kerfs,
which are glued to the surface layer at an angle of 45 degrees (Berger 2006). Skuratov (2010) has presented a three-layer-panel with
a core of low grade sawn timber that has a pattern of hollow cells which reduce its weight.
Wu and Vlosky (2000) studied particle board, medium density fibreboard, and plywood from the perspective of furniture and
cabinet manufacturers. They state that fundamental characteristics that are important to consider in choosing a panel product include
customer objection or approval, economics, weight, finishing properties, shape stability (mainly warp), fastening issues, and, mainly
in the case of plywood, the treatment of the edges. Following Wu and Vlosky, the present study has focused on weight and shape
stability of a new type of light-weight solid-wood panel for furniture and interior uses. The main objective has been to create a
wood-based material with low weight while at the same time maintaining high bending stiffness and bending strength, where the
homogeneous fibre orientation in the core layer supports a form-stable product. Finishing properties are equal to those of natural
wood. The intent is to create a material that may complement traditional solid wood characteristics and uses in the aforementioned
applications.
2. MATERIALS AND METHODS
MATERIALS AND SAMPLE PRODUCTION
A sandwich construction in three layers was used for the study. The sandwich design gives a light-weight structure, as shown in
Figure 1. The surface consists of solid wood and the core consists of webs of solid wood. The design of the webs is described in Table
1. In the panel construction, the surfaces (top and bottom layers) are oriented with their fibre direction in direction L1, while the fibres
in the webs (core layer) are oriented in the W1-direction. The surface layers of the panels were made from beech wood (Fagus
sylvatica L.) and the webs were made of pine (Pinus sylvestris L.). The annual rings in the webs (the tangential direction) were
oriented parallel to the plane of the surface layers. The webs were glue-bonded (Casco system 1974/1993 and PVAC D-3) to the
surfaces. Before gluing, the raw material for the panels was conditioned to equilibrium (EMC) at 20C and 65 % RH. The pressure
during gluing was 0.8 MPa (calculated with respect to the area of the webs), the pressing time was 0.5 hours, and the pressing
temperature was 20C. The panels were constructed in thicknesses of 36 mm and 108 mm, with a constant thickness of the top and
bottom layer of 6 mm. The panels were 1500x1500 mm in original size. After manufacturing, all samples were conditioned for 14
292
days to EMC at 20C and RH 65 % before the tests. The test groups are described in Table 2. 36 mm and 108 mm thick panels are
separated into two groups according to the direction of the bending test in L1 and W1 directions. The densities of the samples in the
two groups were 294 and 373 kg/m.
(a)
Figure 1:
(b)
Table 1: Description of the webs used for the panels. Abbreviations in parenthesis [ ], see Fig. 1a.
Panel thickness [T3]
Distance [L2]
(mm)
(mm)
(mm)
(mm)
(mm)
36
24
25
42
96
108
96
25
42
96
Table 2: Description of the panel samples used in the study. Abbreviations in parenthesis [ ], see Fig. 1a.
Group
No. of
samples
Thickness [T3]
(mm)
Test
direction
D e ns it y
(kg/m)
36
[W1]
373
36
[ L1 ]
373
108
[W1]
294
108
[ L1 ]
294
293
International
International Scientific Conference on Hardwood Processing
(ISCHP Scientific
2013) Conference on Hardwood Processing (ISCHP 2013)
fm =
3 Fmax l2
(Eq. 1)
bt 2
where
fm - bending strength (MPa)
Fmax - maximum load (N)
l2 - distance between an inner load point and the nearest support (mm)
b - width of the specimen (mm)
t - thickness of the specimen (mm)
The bending stiffness value was determined according to the equation:
=
where:
(2 )2 2
(Eq. 2)
(2 )
Figure 2: Test set-up for the bending test showing a 36 mm thick panel loaded in the L1-direction according to Figure 1a.
RESULTS AND DISCUSSION
In this study, a light-weight panel intended for furniture and interior use was manufactured and tested. The results of the bending
test are described in Table 3.
294
Table 3: Results of bending tests on panels. Values for variables marked with * showing min. mean max. values.
Abbreviations in parenthesis [ ], see Fig. 1.
Figure 2: Test set-up for the bending test showing a 36 mm thick panel loaded in the L1-direction according to Figure 1a.
36
[W1]
61 77 91
2.5
36
[ L1 ]
9.6
108
[W1]
22 23 26
70 71 72
190
108
[ L1 ]
10 11 12
84 88 92
75.6
The density was 373 kg/m3 and 294 kg/m3 respective for the panels of thickness 36 and 108 mm y. This means a reduction in
density of about 50 % for the 36 mm panel and o f about 60 % for the 108 mm panel, compared to a massive beech panel with
a density of 725 kg/m3.
The samples bending properties were tested in two directions, see Figure 1 and Table 1. All the samples showed one strong and
one weak direction. The 36 mm panel had the highest bending stiffness and highest E-modulus in the L1-direction, while the 108
mm panel had the highest bending stiffness and E-modulus in the W1-direction.The thickness of the panel is important in
determining the failure load in the W1-direction. The failure load in the L1-direction is about the same for the two panels of
different thicknesses. It would be possible to make a panel that has the same failure load in both directions, but the bending strength
and E-modulus would then be different in the main directions of the panel. The very low bending-strength values in test group four
can be related to failure in the area between the webs and the surface layers.
Comparison of strength values with those of e.g. the DendroLigth panel shows that the tested 36 mm panel has bending
strength values comparable with DendroLight 3-layer panel with 4 mm HDF as surface layers and a total thickness of 25 mm.
The design of the webs in the manufactured panel is important for both the perspective of material characteristics and production
and resource effectiveness. The design of the webs in the panel design studied here was selected in order to give the panels equal
glue area, and the width of the webs was independent of panel thickness (36 mm and 108 mm). In a future study, the design of the
webs may be optimized toward a fixed strength requirement. The choice of pine as the species used for the webs is also something
that may be varied in relation to availability and the required characteristics of the panel. The homogeneous annual ring orientation
in the webs (horizontal annual ring direction in relation to the surface of the panel) is selected to give the panels a homogeneous
movement in relation to surrounding moisture conditions.
The thin surface layers in a three-layer sandwich construction are sensitive to shock or impact loads. The surface bears tensile
and compressive loads, while the core bears shear loads. This is a problem in both the use and the design of fastening solutions.
Products within the automotive and ski manufacturing and construction industries employ various solutions to the problem of
fastening to sandwich constructions. At least two main types of fastenings are in use. It is possible to reinforce the panel in relation
to the fastenings when the panel is manufactured. This decreases the flexibility of the panel, but the fastenings can be made stronger.
If flexibility is important, it is possible to use retrofitted fastenings. Examples of retrofitted fastenings include different sorts of
plugs, tape, screws, or rivets.
In further studies, factors not included in the study should be analysed: economics, fastening issues and edge treatment.
Economics must be investigated in relation to potential production layout. Shape stability must also be tested, as well as long-term
creep resistance. Another important aspect to consider is the acoustic behaviour of the material. The material will probably be
compared to a solid wood panel, and the hollow structure will produce a different sound that will probably affect consumer
experience. From an industrial perspective, customer approval is a key issue.
CONCLUSIONS
This study presents a new three-layered light-weight sandwich construction based on solid wood components. In the use of the
material, the orientation of the core layer must be considered in relation to both the assumed load conditions and the thickness
295
of the panels. The panels promising technical properties and natural wood surface may complement traditional solid wood
characteristics and uses in furniture and interior applications. The panel may therefore contribute to an increased use of hardwoods.
REFERENCES
Berger, J. 2006. Bauplatte oder der gleichen, deren herstellung und werwendung. (Building panel or the like, and production and
use thereof.) Patent No. PCT/AT2006/000141.
Blitzer, T. 1997. Honeycomb technology Materials, design, manufacturing, applications and testing. Chapman & Hall, London.
233 pp.
Britzke, M. 2003. Diploma thesis TU Dresden.
EN 789. 2004. Timber structures test methods. Determination of mechanical properties of wood-based panels. European
Committee for Standardization, Brussels, Belgium.
Feifel, S., W-R. Poganietz, and L. Schebek. 2013. The utilization of light weight boards for reducing air emissions by the German
wood industry a perspective? Environmental Sciences Europe 25(5), doi: 10.1186/2190-4715-25-5.
Herold, J., K. Rehm, U. Krppelin, and M. Britzke. 2006. Neues Verfahren zum Erzeugen von Krafteinleitungspunkten in leichten
Sandwichplatten mit Papierwabenkern. (New procedure for generating loading points in lightweight sandwich panels with
honeycomb paper core.) Holztechnologie, 47(3):43-44.
Michanickl, A. 2006. Development of a new light wood-based panel. 5th European Wood-Based Panel Symposium. Hannover,
Germany.
Pflug, J., B. Vangrimde, I. Verpoest, D. Vandepitte, M. Britzke, & A. Wagenfhr, 2004. Continuously produced paper honeycomb
sandwich panels for furniture applications. In: Proceedings of the 5th Global Wood and Natural Fibre Composite
Symposium, Kassel, April 27-28.
Pokharel, N. 2003. Behaviour and design of sandwich panels subject to load buckling and flexural wrinkling effects. PhD thesis,
School of Civil Engineering, Queensland University of Technology
Riepertinger, M. 2004. Deckschichten als wichtigstes Qualittskriterium fr Leichtbauplatten.(Layers as the most important
quality criterion for lightweight panels.) Wabenplatten-Kolloquium, Schopfloch.
Skuratov, N. 2010. New lightweight solid wood panels for green building. In: Proceedings of the International Convention of
Society of Wood Science and Technology and United Nations Economic Commission for Europe Timber Committee
October 11-14, 2010, Geneva, Switzerland. 7 pp.
Wu, Q. and R.P. Vlosky, 2000. Panel products a perspective from furniture and cabinet manufacturers in the Southern United
States. Forest Products Journal, 50(9):45-50.
Zenkert, D. 1997. The handbook of sandwich construction. Engineering Materials Advisory Services Ltd., Cradley Heath, UK.
442 pp.
296
Session IV
Hardwood market
and sustainability
297
Department of Sustainable
Biomaterials
Virginia Tech
Blacksburg, VA 24061
ABSTRACT
Forest certification systems are voluntary, market-based initiatives to promote the sustainable use of forests. These
standards assume that consumers prefer products made from materials grown in an environmentally sustainable
fashion, and this in turn creates incentives for companies to adopt responsible environmental practices. One of the
major reasons for the creation of forest certification systems was to prevent deforestation in tropical forests. However,
after 20 years of certification, only 10% of the global forest area is certified, mostly in countries that, arguably, already
practiced responsible forest management before certification. Only 2% of tropical forests have been certified (81
million hectares), while in the same period 283 million hectares were lost to deforestation. Africa and Latin America
are the only regions with a net loss of forest area in the 2000-2010 decade. In this paper, the status of forest
certification is analyzed, and challenges and opportunities are evaluated using the case of Bolivia. In a short period of
time, Bolivia became the leader in tropical forest certification. However, in recent years, the area of Bolivian forest
under certification has fallen sharply; and deforestation has actually increased in the 2000-2010 period, compared with
the previous decade. This research tries to uncover the reasons for the rapid initial growth of certification in Bolivia
and its subsequent decline. The information generated from this research is valuable input for policymakers and
support organizations to better assist countries with their efforts to sustainably manage their forests.
1. INTRODUCTION
The environmental, or green, movement is one of the most important developments in modern society. The
environmental movement started in the 1960s out of concerns about environmental sustainability (Friedman, A.
2007, Kibert, C.J. 2005, United Nations 1987). These concerns have encouraged governments and other policysetting bodies from around the world to issue strategies aimed at ensuring the sustainable use of the planets natural
resources and the prevention of waste. For this purpose, international organizations such as the World Wildlife Fund
(WWF), Greenpeace (Greenpeace 2011), or the United Nations Environment Programme (UNEP) have been created
to promote the responsible utilization of our resources (UNEP 2011, WWF 2011c). Also, businesses are making
efforts to include environmental issues in their corporate culture, motivated not only by regulation but also by
consumer demand (Bowyer, J. 2008).
In the first part of this manuscript, details are provided about the evolving forest certification systems. Then, a
presentation and discussion follows about the results from a qualitative research about the status of forest
certification in Bolivia, considered by many a success story of certification (FSC 2011).
1.1 FOREST AREA, TYPE OF FORESTS, GEOGRAPHICAL DISTRIBUTION, AND OWNERSHIP
Forests cover about 4 billion hectares, a third of the total global land area. More than half of that area is
concentrated in five countries: the Russian Federation, Brazil, Canada, United States, and China (FAO 2010).
Primary forests, defined as forests not significantly disturbed by human activity (FAO 2010), constitute 36 percent
of the total forest area and 57 percent are classified as regenerated forests (where humans have intervened, chiefly
*
298
Corresponding author: Tel.: (651) 624-0770; Fax (651) 624-; E-mail: espinoza@umn.edu
with selective logging); and 7 percent is planted forests (e.g., plantations). More than a quarter of the total forest area
is designated primarily for production (Table 1). Ownership of worlds forests is largely public (80% according to
FAO 2010). However, communities, individuals, and private enterprises have been playing an increasingly
important role in forest ownership and management. Examples are the increasing private ownership of forests in
Russia and Africa, or more community involvement in forest management in South America and Southeast Asia
(FAO 2010).
1.2. DEFORESTATION AND RESPONSES TO DEFORESTATION
Forests, in addition to representing a major resource for human sustenance, provide carbon sequestration,
temperature regulation, protection against erosion and flood control, water storage, and habitat to half of all known
plant and animal species (Bswald, K. 1996, &KOHEHN$DQG-DDEi0, Lenz, H. 1967, WWF 2011b). A
quarter of the worlds population directly relies on forests for its subsistence, while all humans are in one way or
another dependent on products from our forests (FAO 2011). Forests offer shelter for most of our terrestrial
biological diversity, and play a vital role in conserving water and providing wood and non-wood products for
communities around the world. Also, recently, the importance of forests for climate change has been recognized
more prominently. While forests are major carbon sinks, global forest degradation or clearing also contributes a
sixth to global carbon emissions. However, if forests were managed sustainably on a global scale, they would have
the potential to absorb one tenth of the projected carbon emissions by 2050 (FAO 2011). The trunk of the trees is an
extremely environmentally-friendly material. In fact, wood is considered the most environmentally friendly material
of all materials used by humans in large quantities as long as timber harvesting is done in a sustainable way.
As discussed above, forests are of critical importance for the future of humanity and the survival of species
(Bswald, K. 1996, &KOHEHN$DQG-DDEi0, FAO 2011, Lenz, H. 1967, WWF 2011b). However, forests
have been and continue to be subject to intensive utilization and logging activities. As a result, large areas of forest
lands have been lost mostly to agricultural use but also to urbanization. The world lost 13 million hectares forestland
per year during the last decade as compared to 16 million hectares per year during 1990e2000 (FAO 2010). While
most boreal and temperate forest areas have been stabilized lately, the annual loss of 13 million hectares per year
during the last decade occurred entirely in tropical forests, one of the most valuable areas in respect to biodiversity
and natural regeneration. In fact, the deforestation of tropical forests is estimated to contribute about 15 percent of
global total carbon dioxide emissions, a greenhouse gas (Achard, F., Stibig, H.J., et al. 2010). However, there are
vast regional differences in respect to deforestation. South America and Africa account for most of the deforestation
(7.4 million hectares of net change per year), while Asias net change in forest area was positive with more than 2.2
million hectares added per year. In fact, China alone added a net 2.0 million hectares annually (FAO 2010), as China
has increased planted forests more than all other countries combined during the last two decades (FAO 2010). These
additional forest areas are needed. To maintain the current level of global wood consumption per capita (0.6m3 per
person per year), 2.5 billion m3 wood harvest will have to be added to the current 3.5 m3 over the next decade to
account for population growth and increasing standards of living (Sutton, W.R.J. 1999).
As one response to these challenges, governmental regulation and non-governmental certification systems of
forests have been established. The former include laws, restrictions, and taxes, among others; while private
initiatives include, prominently, forest certification (McGinley, K. and Cubbage, F.W. 2011).
1.3. FOREST CERTIFICATION
Forest certification systems are initiatives with the purpose of promoting the sustainable utilization of forests
from an environmental, economic, and social point of view (Bowyer, J. 2008, Hansen, E., Fletcher, R., et al. 2006).
Differently from regulatory approaches to forest sustainability, the stated driving force for the adoption for forest
certification standards was market demand (McGinley, K. and Cubbage, F.W. 2011). One main driver for the
creation of forest certification systems was to address the need of communicating the environmental quality and
sustainability of forest products to consumers. The underlying assumption for the establishment of forest
certification systems is that consumers prefer products manufactured by companies responsible with the
299
environment. The hope is that such consumer preference creates a pull effect by which companies are incentivized
to adopt environmentally friendly practices (Vidal, N., Kozak, R., et al. 2005).
The first forest certification system with a global scope was the system developed by the Forest Stewardship
Council (FSC 2013). The FSC system, formed in 1992 by several major environmental non-government
organizations (NGOs) and global retailers (Nussbaum, R. and Simula, M. 2005, Van Kooten, G.C., Nelson, H.W., et
al. 2005, WWF 2011a), today maintains an important place in shaping forest certification globally as it is
responsible for the certification of the largest forest area of any stand-alone certification system. Other certification
systems include the Sustainable Forest Initiative in the United States, launched in 1994 (SFI 2011), the American
Tree Farm System incepted in 1941 (ATFS 2010), or Canadas Standards Association system, founded in 1996
(CSA 2011). In 1999, European small forest owners founded the Program for the Endorsement of Forest
Certification systems (PFEC 2013). The PFEC is an umbrella organization that endorses 29 national forest
certification systems in 28 countries, including SFI, ATFS, and the CSA system (PFEC 2013). Other certification
systems are the Indonesian Ecolabelling Institute (LEI 2011), the Brazilian Programa Brasileiro de Certificao
Florestal (CERTFOR 2011), the Malaysian Timber Certification Council (MTCC, 2006), and Chiles Sistema
Chileno de Certificacin de Manejo Forestal Sustentable (CERTFOR 2011). These national systems were developed
based on other standards and guidelines; for example the Malaysian MTCC is based on the International Tropical
Timber Organization (ITTO 2011) criteria, as is the Brazilian CERFLOR, while the Indonesian LEI system
incorporates criteria from the ITTO and FSC(Perera, P. and Vlosky, R.P. 2006).
1.4 CERTIFICATION IN TROPICAL COUNTRIES
After 20 years of forest certification, 416 million hectares of forests have received certification from one of the
two global forest certification organizations (174 million ha by FSC and 242 by PEFC; FSC 2013, PFEC 2013).
This represents about 10% of the global forest area of 4.03 billion hectares (FAO 2010). However, an overwhelming
majority of the total certified forest area is located in the northern hemisphere (UNECE/FAO 2012), and only 2% of
the tropical forests have been certified. In fact, a net loss of forest area occurs only in Africa, Latin America, and the
Caribbean. Arguably, the countries where most of the certified area is located already had well managed forests
before certification (Howe, J. and Fernholz, K. 2012). The U.S., Canada, and the countries of the European Union
contain more than three quarters of the total certified area (FAO 2012). The forestland area in the U.S., for example,
has not changed in almost 100 years (MacCleery, D.W. 2011).
Latin American countries have rich forest resources, covering about half of South American (FAO 2010).
Unfortunately, deforestation is very high in this region, most of it due to conversion of forested areas to croplands
and pasture lands (Bowyer, J.L. 1997, Lindsey, R. 2007). Bolivia, Brazil, and Venezuela are among the ten countries
with the largest loss of forest area during 1990-2010 (FAO 2010). The problem of deforestation is particularly
serious in Bolivia. This countrys forest loss is the highest in the Amazonian basin (Forero, J. 2012) and it is the fifth
highest when its population is considered (deforestation per capita calculations based on population and forest loss
figures found in FAO 2010). Similarly to other countries in the region, much of the deforestation in Bolivia is
caused by land conversion to agriculture, but also due to migration of people to the lowlands, which represent more
than half of the Bolivian territory. According to the Food and Agriculture Organization, the annual rate of
deforestation in Bolivia during the 200-2010 decade has been of almost 0.5%. Another recent study, which used
satellite images and the Normalized Difference Fraction Index (NDFI), estimates a much higher annual forest land
loss figure of 0.56% in the 2000-2005 period and of 0.78% in the 2005-2010 period (Cullar, S., Rodrguez, A., et
al. 2012). This is particularly worrisome considering the fact that Bolivia has once been considered leader in tropical
forest certification, as the next section explains.
1.5 FOREST CERTIFICATION IN BOLIVIA
Total forest area of Bolivia is 53 million hectares, of which 28.8 million are potential productive area (Hjorts,
C.N., Jacobsen, J.B., et al. 2006). With the purpose of ensuring sustainable utilization of its rich forest resources,
Bolivia passed a highly successful New Forestry Law in 1996, introducing sweeping changes to the way forests
300
were used until then (Quevedo, L. 2004). Some of the most important changes were (1) the creation of the Forestry
Superintendence, a government agency charged with enforcing the forestry law; (2) implementation of a system of
forestry concessions, with fee payments based on area, which discouraged the highly selective harvesting; (3)
elaboration and approval of forest management plans (Hjorts, C.N., Jacobsen, J.B., et al. 2006); and (4) established
legal recognition of new actors in sustainable forest management, notably indigenous and local communities
(Quevedo, L. 2004, Urioste, J.L.d., Quevedo, L., et al. 2010). According to Colchester et al. (Colchester, M.,
Boscolo, M., et al. 2006), with the new forestry law, Bolivia had the most progressive legal framework in terms of
its accommodation of rural livelihoods in forest, compared to many countries. Forest certification was implemented
simultaneously with the Forestry Law, in what was considered by many as a success story of certification (Duery, S.
and Vlosky, R. 2005).
International organizations played an important role in the promotion and implementation of forest certification
in Bolivia. The U.S. Agency for Development Aid (USAID) provided financial, marketing, and technical support to
these efforts (Chemonics 2004). The certified area grew at a high pace in the decade from 1996 to 2005, reaching at
one point more than 2 million hectares, a larger area of tropical forest than what Brazil certified at the time (Duery,
S. and Vlosky, R. 2005). However, the certified area fell dramatically since 2008, as illustrated in Figure 1.
2.5
2.0
Million Ha
1.5
1.0
0.5
0.0
Figure 1. FSC-Certified Forest Area in Bolivia (Camara Forestal de Bolivia 2011, FSC 2012).
Some important developments coincided with the period of rapid decline of forest deforestation in Bolivia (2008
to 1010). The great recession officially started in the U.S. in December of 2007 (National Bureau of Economic
Research 2010), and had a great impact on imports during the following years (imports of goods in 2009 and 2010
dropped by 26% and 10% relative to 2008, respectively, Bureau of Economic Analysis 2013). Likewise, the global
recession caused a big decline of the EU 27 countries imports, which 2009 and 2010 monetary value dropped by
22% and 3% from the 2008 level, respectively (Eurostat 2013). This is relevant since United States and the
European Union have been historically the most important markets for Bolivian forest products exports (CFB 2012).
OBJECTIVE
The objective of this research was to understand the reasons for the initial success of forest certification in
Bolivia, and also explain the rapid decline in certified area in recent years. Of interest was also learning the
perceptions of the stakeholders regarding the major benefits obtained from certification.
METHODOLOGY
To accomplish the objectives of this research, both secondary and primary sources of information were used. An
extensive analysis of the literature was conducted to gather facts about forest certification in general and specific to
the Bolivian case. This literature review was followed by semi-structured interviews with major stakeholders of
forest certification in Bolivia. During Summer of 2012 and Spring of 2013, 23 interviews were conducted, including
301
Government support
Regulatory framework
International financial and technical support
Improvement in the standard of living of timber-reliant communities
Improvement in image of the forestry sector
Better, more stable markets for certified products
Land tenure security
Markets (or lack thereof) for certified products
302
Social Control of Forests and Lands, or ABT) interviewed for this study, indicated that the sustainable use of forests
may be better served by developing a national certification system, less costly and of easier access for those who do
not have the financial capability for international certification. This was tried by other countries with different levels
of success (see Introduction).
ROLE OF INTERNATIONAL COOPERATION AND NON-GOVERNMENTAL ORGANIZATIONS
Most of the participants coincided in that international aid and non-governmental organizations (NGOs) played a
significant role in promoting and implementing forest certification in Bolivia. The U.S. government, through its
international development agency (USAID 2013), provided considerable financial support for the promotion of
certification; it financed the creation of BOLFOR, or Bolivian Sustainable Forest Management Project, which was
instrumental in harmonizing the new forestry law and certification. BOLFOR also participated in the creation of
CFV, the Bolivian Council for Voluntary Certification, in 1994. CFV was active in the elaboration of region-specific
standards for certification in agreement with FSC principles (Duery, S. and Vlosky, R. 2005).
NGOs played a significant role in promoting certification by providing technical and financial assistance to
companies interested in certification. NGOs were especially important in the promotion of community-based forest
certification; one participant went as far as to suggest that the decline in this certification category was in great part
due to the removal of NGO support (the first certificate in Bolivia was awarded to an indigenous community;
currently there are no communities holding certificates).
BENEFITS FROM CERTIFICATION
Participant identified the following benefits from certification.
x Improvement in the standard of living and work conditions of timber-reliant communities. Forestry professionals
interviewed for this study agreed in that one of the most important benefits from forest certification is the
improvement in standard of living of communities dependent on forest utilization and the work conditions of
forestry field personnel. This is consistent with principles of certification, which require forest managers to
comply with all laws, regulations, treaties, conventions and agreements, together with all FSC Principles and
Criteria (1st principle), and to maintain or enhance forest workers' and local communities social and economic
well-being (4th principle) (FSC 2013). What is more remarkable, according to representatives of communities
with certification experience interviewed for this study, these benefits tend to stay even if these operations do not
renew their certificates. These participants also indicated that certification has allowed them to raise the standard
of living and educational level of their members.
x Improved image of the forest products sector. Historically, forests have not been well managed in Bolivia. Until
1996, harvest operations were extremely selective, with over 90% of harvested timber focused on three species
(Espinoza, O.A., Bond, B.H., et al. 2007). Despite efforts to reduce it, illegal logging is still a big problem
(Harvey, F. 2013), some estimated that half of the timber consumed domestically is illegally sourced (Pacheco, P.
n.d.). The benefits from the forest reached only a few, and work conditions were poor at best (Pacheco, P., de
Jong, W., et al. 2010). All this contributed to create a negative perception of the industry and has led to conflict
with indigenous people, which in part motivated the state to start promoting communal participation in forest
management in the mid-1990s (de Jong, W., Ruiz, S., et al. 2006) . Participants to this study indicated that the new
forestry regime started by the 1996 law and the implementation of forest certification have contributed to improve
the image and credibility of this important sector of the Bolivian economy, by demonstrating it is capable of
meeting internationally-recognized standards of sustainable forest management, from environmental, social, and
economic perspectives (Nebel, G., Quevedo, L., et al. 2005). Most participants coincided in that forest-certified
products did not command a significant price premium compared to non-certified products in international
markets, but allowed companies access to more stable markets for certified products. A lack of price premiums for
certified products has been confirmed by some studies (Anderson, R.C., Laband, D.N., et al. 2005, Espinoza, O.,
Buehlmann, U., et al. 2012, Hubbard, S.S. and Bowe, S.A. 2005). One interviewee mentioned that, in one recent
303
meeting about forest products business in Bolivia, it was concluded that the drop in demand from European and
US markets did not affect certificate holders as severely as other businesses.
LAND TENURE SECURITY
This was a common theme during the interviews, with most of the participants indicating that forest certification
efforts are hampered by recent developments in land tenure conflicts; particularly, the higher frequency of illegal
settlement on lands under forest management (Condori, I. 2012, Mendoza, L. 2012). The second principle of FSC
certification requires forest managers to to define, document and legally establish long-term tenure and use rights
(FSC 2013); thus this poses a particularly difficult situation for certificate holders or operators in the process of
certification. Interviewees reported at least one company lost its certificate due in part to illegal occupation and
another planning to certify was unable to continue due to illegal occupation of its concession.
MARKETS FOR CERTIFIED FOREST PRODUCTS
Forest certification systems are market-based, non-regulatory (i.e., voluntary) schemes to promote sustainable
utilization of forest resources; therefore, their success depends on the promise of economic advantages to encourage
adoption by industry participants. Such economic advantages can originate from (1) growing market demand for
certified products, (2) the ability of producers to charge price premiums on certified or standard-conforming
products, or (3) benefits from creating access to attractive markets at home or abroad, among other things (Espinoza,
O., Buehlmann, U., et al. 2012). In Bolivia there is very little or no domestic demand for certified products,
according to the companies interviewed for this study, thus certification is sought only by operations that sell, or
intend to sell, to international markets, traditionally Europe and the U.S. During the growth of certification in
Bolivia, the increases in certified area were accompanied by growth in the exports of certified products, with the
latter growing six-fold in the 1999-2002 period (Duery, S. and Vlosky, R. 2005). This included not only
unprocessed lumber, but also value-added products, such as millwork and flooring (Nebel, G., Quevedo, L., et al.
2005). One very large integrated timber company interviewed for this study stated that the main motivation for
maintaining their certification is having a stable market for these products.
However, exports of Bolivian forest products declined significantly since 2010, falling 26% in 2011 and then
again 19% in 2012 (INE 2013). For the first time, the department of Santa Cruz, which represents the majority of the
timber products economic output, imported more timber products than it exported in 2012, mostly from Brazil
(Bolpress 2012). This fall in exports is attributed by most to the U.S. and European economic downturn (Los
Tiempos 2012). However, other reasons indicated by our industry participants are a tighter control by the ABT and
an increased paperwork in the exporting process, which increases the price of timber sold in the domestic market
and also makes exporting less attractive, thus producers prefer to sell internally. One interviewee stated that their
company even sold unlabeled certified product domestically, since this was a more profitable option.
Although overall exports declined during the 2010-2012 period, it is also worth noting that there has been a shift
in international markets for Bolivian timber products. Asia, and particularly China, has become a more important
buyer of Bolivian timber products. Bolivian exports to China have increase more than 4,600% in the 2002-2011
period (Xinhua 2012). China buys mostly low-value added products, such as rough lumber, which it then uses to
manufacture goods for domestic consumption but mostly to export to other markets, such as the Europe (Bolpress
2012). Regional markets have also grown in importance. Latin American has been experiencing a sustained growth
during the last years, with an average growth in economic output of 4.1% per year in the 2005-2012 period (compare
this with average growth in advanced economies, which averaged 1.4% during the same period; IMF 2013). Chile,
Venezuela, Argentina were the most important regional markets for Bolivian timber products in 2012 (Los Tiempos
2012). According to industries interviewed for this study, this is relevant for certificate holders, since there is no
demand for certified material in these markets, thus reducing the motivation for new certifications or for certificate
holders to maintain their status.
304
Another factor affecting exports of Bolivian timber products, and particularly products carrying forest
certification, is the construction boom that is taking place in Bolivia. The economic output from construction
industry has grown at an average annual rate of 9.1% in the 2005-2012 period (INE 2013). This has increased
domestic demand for wood products and consequently prices of timber and timber products (for example, the price
for ocho, a species commonly used in construction, has increased 2.5-fold in the 2006-2012 period; INE 2013).
Companies that otherwise would have targeted exports markets, produce for the domestic market instead, which as
mentioned before, does not require certification. It is expected that, as the construction industry decelerates, exports
markets will become more attractive.
One last market factor which may act as an incentive for certification is the implementation of more stringent
controls for illegally sourced timber, such as the Europes Forest Law Enforcement, Governance and Trade, or
FLEGT Action Plan (FLEGT 2012). The FLEGT program aims at excluding illegal timber from European markets.
The basis for FLEGTs implementation is a voluntary agreement between the European Union (EU) and
governments of exporting countries (known as Voluntary Partnership Agreements, or VPAs), by which a
government agrees in ensuring that timber exported from its country complies with the countrys laws (Carden, C.,
Wijers, R., et al. 2012). Only six countries have signed VPAs to this day, and another six are in negotiations; Bolivia
is not part of either group. Another important component of FLEGT is that it requires importers to apply Due
Diligence, which means providing access to information about the origin of the timber and implement a risk
assessment procedure to evaluate the risk of timber coming from a specific country being illegally sourced; and also
implement risk mitigation measures to keep the risk at a minimum. In the absence of a VPA between the EU and
Bolivia, it is expected that FLEGT will encourage forest certification in the short term, as a certificate will most
likely be accepted as Due Diligence system (Carden, C., Wijers, R., et al. 2012).
SUMMARY
After 20 years of forest certification, only 10% of the total global forest are is under some form of certification,
most of it in temperate latitudes. Only 2% of tropical forests are certified. In Bolivia, a country regarded as a success
story of tropical forest certification, the certified area never exceeded 7% of the total countrys total forest area, and
4% of the potential productive forest. The country at one point held over 2 million hectares of certified forests, but
this area has fallen dramatically during the last five years. As of May, 2013, the certified are was less than 900,000
hectares (FSC 2013). This project intended to uncover the reasons for the initial success of certification in Bolivia,
and its subsequent weakening.
Different stakeholders for forest certification were interviewed for this research, and some major themes were
identified: government support and international cooperation were critical in the implementation of forest
certification in Bolivia. Particularly, a regulatory framework and strong enforcement were conducive to compliance
with forest certification principles. Major benefits identified by interviewees were the improvement in the standard
of living and work conditions of timber-reliant communities, the improvement in the prestige of the forest products
sector, and access to better markets for certified products. Participants identified land tenure insecurity and a lack of
markets for certified products (and the existence of more profitable markets for timber products) as major
impediments for certification.
Forest certification was conceived as a market-based approach to ensure the sustainable use of forests, from a
social, economic, and environmental perspective. However, great part of the support for certification in Bolivia
came from non-private sources, which calls into question the premise that forest certification is a market-based
approach.
Given that deforestation in Bolivia is caused by land use change, namely mechanized agriculture, small-scale
agriculture, and cattle pasture (Mller, R., Pistorius, T., et al. 2013), it is unlikely that certification will have a
significant impact in reducing deforestation. Policy instruments and strong enforcement of land laws are required to
address these issues.
305
Department of Sustainable
Biomaterials
Virginia Tech
Blacksburg, VA, 24061, USA
Fachhochschule Rosenheim
Hochschulstrasse 1
D-83024 Rosenheim, Germany
ABSTRACT
The past decades have seen numerous U.S. secondary wood processing companies shift their production to
overseas locations, mainly in Southeast Asia. The remaining companies have been hit hard by the downturn in
housing markets and the following recession. Thus, many large customers of the U.S. hardwood lumber industry
have reduced or stopped the purchase of products, leaving mostly smaller firms as the drivers of demand for the
U.S. hardwood industry. Therefore, it has become important to understand these small firms needs in the current
business environment.
This study consisted of a mail survey asking participants in six U.S. states questions to help determine the unique
characteristics of small firms in the woodworking industry. Both large and small firms attributed much of their
success to their manufacturing capabilities and leadership prowess. However, small firms used less formal paths to
gather information and planned less investments. Small firms, in general, also purchased more hardwood lumber
using hardwood distributors and requested fewer specialized services from their lumber suppliers.
1. INTRODUCTION
The U.S. manufacturing sector has experienced many challenges during the past decades, with manufacturing
employment recently reaching its lowest level since 1950 (Pierce and Schott 2012, Forbes 2004) and problems being
exacerbated further by the recent economic recession (International Forest Industries 2009). The major driver behind
the decline of U.S. manufacturing competitiveness in many sectors is the ongoing globalization of trade.
Unfortunately, these trends have affected many parts of the wood products industry, with U.S.-based hardwood
lumber-consuming industries such as furniture, flooring, and millwork being especially hard hit by the worldwide
opening of markets for goods and services. More recently, U.S. woodworkers also have been impacted by the housing
and global financial crisis, which has dampened demand for construction-based wood products such as cabinets. The
decline in manufacturing operations producing hardwood-based products has resulted in a corresponding decline in the
demand for hardwood lumber as well. These declines affected the industries and forest landowners upstream from the
final manufacturer (Grushecky et al. 2006).
* Corresponding author: Tel.: (540) 231-9759; Fax: (540) 231-8868; E-mail: buehlmann@gmail.com
306
To counteract these unfavorable trends, efforts are needed to support and grow local, regional and national
industries using U.S. hardwoods. The increasing importance of the customized economy, where customers expect
products and services to be customized to their needs and expectations (Schuler and Buehlmann 2003, Lihra et al.
2008, Buehlmann and Schuler 2009), call for nimble entities that are able to interact with individual customers and
quickly deliver the desired product or service. Small enterprises are well positioned to provide high levels of
customization while also being able to profit from those trends. Their size makes them flexible to tailor products and
services to particular market niches and needs, which helps them overcome some of the other inherent disadvantages of
being small (Penrose 1995). Several studies have highlighted the importance of small enterprises * and their crucial
role in economic and social welfare (Storey 2003, Mulhern 1995). Such enterprises provide local employment, supply
needed products and services to local end-users and enterprises, are flexible and innovative, and have the potential to
grow into larger, national and international enterprises over time.
For example, Bumgardner et al. (2011) showed the potential of small businesses serving a market niche and
being located in a local industry cluster. The Amish wood furniture cluster in Ohio, consisting of nearly 500 companies
with a median size of 4 employees, expanded operations while U.S.-based manufacturing in the broader U.S. furniture
industry declined. Nearly all of these Amish furniture firms are small enterprises, focusing on narrow market segments
and sourcing wood materials and components from local and regional markets.
The objective of the current study was to better understand the importance of small enterprises active in the
secondary U.S. forest products industry in the current business environment. Although small enterprises are becoming
increasingly important to hardwood lumber demand in the United States, they might at the same time present new
challenges to hardwood suppliers, as smaller orders to more customers must be facilitated (Buehlmann et al. 2010). Six
eastern hardwood states were included in the study, including Virginia (VA), West Virginia (WV), Ohio (OH),
Wisconsin (WI), Tennessee (TN), and North Carolina (NC). Questions pertaining to the competitive attributes of small
firms were analyzed. Also of interest were the sourcing practices and informational and training needs of smaller wood
products firms. Such information is important in assisting the secondary hardwood industry and determining research
priorities, especially for the numerous smaller firms that are increasingly important to the U.S. hardwood market.
2. METHODS
2.1 QUESTIONNAIRE
Secondary hardwood forest products manufacturers in the study states were surveyed in 2011 using Dillmans
(2009) Total Design Method. A questionnaire was developed and pre-tested by members of academia and five
secondary hardwood forest products manufacturers, resulting in some questionnaire adjustments. The final
questionnaire consisted of 26 questions and covered aspects related to firm characteristics, perceptions about the
competitive environment, supply chain issues, and information needs. The survey employed questions with categorical
responses, rating responses (5-point scales), and open-ended questions (qualitative responses).
2.2.DATA COLLECTION
An address list was compiled using state directories supplied by researchers or extension professionals in the
study states. Addresses for OH were compiled by Virginia Tech using Mantas online business listings, the
manufacturer index of the Wood Products Manufacturers Association, and the membership list of the Architectural
Woodwork Institute. The final list contained 4,980 firms.
After the initial mailing in the spring of 2011, a reminder postcard, another questionnaire with accompanying
letter, and one last reminder postcard were mailed with a two-week separation between each mailing. At the closing of
the survey, 395 usable questionnaires were obtained. For another 337 entities, responses either were not usable (many
being primary manufacturers not included in the target population) or the company had gone out of business, etc. After
accounting for these firms, the adjusted response rate was 9 percent. Usable responses were obtained from each of the
six states (as well as a few in other states), while some respondents chose to not indicate their state: VA (n=99), NC
(n=83), WI (n=81), OH (n=75), WV (n=21), TN (n=12), other states (n=12), state not indicated (n=12).
*
In the U.S., small businesses are those with less than 500 employees, according to the Small Business Administration.
In the EU, small business have less than 50 employees and medium sized businesses less than 250 employees. For
studies in forest products, small firms have been defined as much smaller, for example with fewer than 20 employees
(Bumgardner et al. 2011).
307
Nonresponse bias was assessed by comparing early and late respondents, assuming there is a continuum from
early respondents to late respondents, and that late respondents can be used as a proxy of nonrespondents (Lahaut et al.
2003). Respondents were categorized as early respondents (n=267) or late respondents (n=128) depending on whether
they returned the questionnaire from the first or second mailing. Four categorical variables were analyzed using
Chi-Square tests of independence, including number of employees, sales in 2010, change in sales from 2010 to 2011
(better, worse, or unchanged), and type of product produced (cabinets, furniture, architectural millwork/fixtures, and
other). None of the tests were significant indicating that early and late respondents were similar in terms of size,
performance, and product types produced, and thus that nonresponse bias was not a major factor in the study.
However, some caution is warranted in interpreting the results given the low response rate.
Responses were obtained primarily from company owners (53 percent), and also from persons in corporate or
operations management (32 percent), production management/engineering (4 percent), marketing, sales, and design (3
percent), and 8 percent either worked in other positions or indicated more than one of the above categories (8 percent).
In terms of products manufactured, kitchen/bath cabinet companies represented 22 percent, followed by architectural
millwork/store fixtures (16 percent), furniture (including household, upholstered, institutional, 14 percent), wood
components (10 percent), flooring (6 percent), wood windows/doors (3 percent), and 29 percent indicated another
product type or produced a combination of the above. Most respondents operated at higher price-points, with 75
percent indicating their price point was either medium or medium-to-high. Most respondents also indicated that they
produced mostly made-to-order (61 percent) or semi-custom (21 percent) products, with only 16 percent producing
stock items and 2 percent producing in multiple categories.
Hardwood lumber was, with an average of 45 percent, the largest component of wood materials costs for
responding companies, followed by composite and engineered products (27 percent), softwood lumber (17 percent),
dimension and components (7 percent), and other (4 percent). Fourteen percent of respondents indicated that their firm
has increased the use of imported finished products over the past 5 years, while 21 percent indicated that their company
had increased the use of imported components or lumber over the same period. Conversely, only 16 percent of
respondents indicated that their firm had increased the export of their products outside of the United States over the last
5 years. Eighty-one percent of respondents worked in a single-facility company and 63 percent of respondents worked
in a company with less than 20 employees. Similarly, 52 percent of respondents reported sales of $1 million or less
and another 25 percent had sales between $1 million and $5 million. Thus, the sample generally reflected the
perceptions and practices of small firms, although further breakdown by small and large firms was used in the following
analyses to help further discern characteristics of small firms.
3. RESULTS AND DISCUSSION
Factors important to business success
Respondents were presented with a list of fourteen factors and asked to choose four that they perceived to be
the most important to the success of their business. Seven of the factors were classified (by the researchers, consistent
with Rogoff et al. (2004)) as being external to individual firms and seven were classified as internal factors, but this was
not indicated on the questionnaire. The results are shown in Table 1. Respondents chose manufacturing capabilities
the most frequently, followed by general economic conditions, and cost of raw material and energy inputs as the top
three factors. While there was a general trend of external factors being rated as less important overall (four of the
bottom five), external factors also represented two of the top three positions. Thus, in general, both types of factors
were seen as impacting the success of secondary wood products manufacturers in the current business environment.
Many internal factors such as marketing activities, human resources management, and owner/manager characteristics
also were rated relatively highly.
The results in Table 1 were then broken down by firm size to make comparisons between small (1-19
employees) firms (n=240) and large (20+ employees) firms (n=142). Both groups ranked manufacturing
capabilities, economic conditions, and costs of raw material/energy as the most important factors. However, for small
firms, marketing activities and owner/manager characteristics were especially important while organizational
efficiency and product characteristics relative to the competition were especially important for large firms; all four of
these factors could be considered as internal to the individual firm.
308
Factor
Percent of
firms
indicating1
External (E) or
internal (I) to the
firm2
63%
57%
43%
Marketing activities (e.g., reaching new customers, good customer service, effective
product promotion)
36%
32%
30%
25%
24%
22%
13%
Regulatory conditions
9%
7%
Competition-driven innovation
7%
6%
1
2
Each respondent was asked to check four factors, thus the total is greater than 100%
As predetermined by the authors
309
310
Lean practices
The survey also inquired about companies practices concerning Lean, a management philosophy pursuing
more effective and efficient operations through the elimination of waste. Fifty-five percent of all respondents indicated
that they are implementing Lean in their operation, while another 9 percent are planning to implement Lean. The
remaining 36 percent do not have any plans to pursue Lean in their operations.
Overall, a sizeable portion of the respondents knew about Lean and recognized terms like lean manufacturing
(68 percent), lean management (53 percent), lean production (51 percent), lean thinking (33 percent), and Toyota
Production System (29 percent). However, there were notable differences by firm size, with large firms being
somewhat more familiar with all of the lean terms investigated (Figure 4). Only in the case of lean manufacturing did
more than 50 percent of small firms indicate that they had heard of the term.
311
Figure 4: Percentages of respondents indicating they have heard of different Lean systems
4. SUMMARY
This research investigated current perceptions and practices of firms in the U.S. secondary hardwood industry.
Focus was given to factors and issues particularly important to small enterprises, defined as firms with less than 20
employees. Such firms ranked their manufacturing capabilities, external economic conditions, and input costs as the
most important factors affecting the success of their businesses in the current environment. Small firms investment
plans were found to focus on marketing communications, finishing, and employee training. Also, these small firms rely
mostly on conversations with customers to gather information regarding industry trends and rely on distributors &
concentration yards to source the majority of their hardwood lumber, even though they requested fewer services from
their suppliers. It was encouraging that many firms expressed familiarity with several Lean terms, and a majority of the
overall sample reported to be implementing Lean practices, However, most small firms were not familiar with several
Lean terms.
As summarized here and elsewhere (see Buehlmann et al. 2013 for more in-depth analysis and related
discussion), small and large firms in the secondary wood products industry can differ in substantive ways. Discerning
these differences is critical given small firms important role in U.S. hardwood markets. Understanding the needs of
smaller manufacturers is becoming increasingly important to analyzing demands for hardwood lumber and the
associated product specifications and distribution channels.
5. ACKNOWLEDGEMENTS
The work upon which this publication is based was funded in part through a grant awarded by the Wood
Education and Resource Center, Northeastern Area State and Private Forestry, U.S. Forest Service. Thanks also go to
university and extension cooperators in the study states, including S. Bowe, S. Grushecky, D. Saloni, and A. Taylor.
Use of trade names in this article does not constitute endorsement of any product or service.
6. REFERENCES
Alreck, P. L. 2004. The Survey Research Handbook, 3rd ed., McGraw-Hill/Irwin, Boston, MA. 463 p.
Buehlmann, U. and A. Schuler. 2009. The U.S. household furniture manufacturing industry in 2008 status and
opportunities. Forest Products Journal. 59(9):20-28.
312
Buehlmann, U., O. Espinoza, M. Bumgardner, and B. Smith. 2010. Trends in the U.S. hardwood lumber distribution
industry: changing products, customers, and services. Forest Products Journal. 60(6):547-553.
Buehlmann, U., M. Bumgardner, and M. Sperber. 2013. How small firms contrast with large firms regarding
perceptions, practices, and needs in the U.S. secondary woodworking industry. BioResources.
8(2):2669-2680.
Bumgardner, M.S., G.W. Graham, P.C. Goebel, and R.L. Romig. 2011. How clustering dynamics influence lumber
utilization patterns in the Amish-based furniture industry in Ohio. Journal of Forestry. 109(2):74-81.
Bumgardner, M., U. Buehlmann, A. Schuler, and J. Crissy. 2011. Competitive actions of small firms in a declining
market. Journal of Small Business Management. 49(4):578-598.
Dillman, D.A. 2009. Internet, Mail, and Mixed-mode Surveys: The Tailored Design Method, 3rd ed., Wiley & Sons,
Hoboken, N.J. 499 p.
Forbes, K.J. 2004. U.S. manufacturing: Challenges and recommendations. NABE 2004 Washington Economic Policy
Conference. March 25, 2004. url:
http://www.whitehouse.gov/cea/forbes_nabe_usmanufacturing_3-26-042.pdf. Accessed on Jan. 6, 2008.
Grushecky, S.T., U. Buehlmann, A. Schuler, W. Luppold, and E. Cesa. 2006. Decline in the U.S. furniture industry: a
case study of the impacts to the hardwood lumber supply chain. Wood and Fiber Science. 38(2):365-376.
International Forest Industries. 2009. Updated softwood lumber market outlook for the US.
http://www.internationalforestindustries.com/2009/11/19/updated-softwood-lumber-market-outlook-for-the-us/
Lahaut, V.M.H.C.J., H.A.M. Jansen, D. van de Mheen, H.F.L. Garretsen, J.E.E. Verdurmen, and A. van Dijk. 2003.
Estimating non-response bias in a survey on alcohol consumption: comparison of response waves. Alcohol
and Alcoholism. 38(2):128-134.
Lihra, T., U. Buehlmann, and R. Beauregard. 2008. Mass customization of wood furniture as a competitive strategy.
International Journal of Mass Customization. 2(3/4):200-215.
Mulhern, A. 1995. The SME sector in Europe: A broad perspective. Journal of Small Business Management.
33(3):83-87.
Penrose, E. 1995. The Theory of the Growth of the Firm. 3rd ed. Oxford, NY: Oxford University Press. 272pp.
Pierce, J. R. and P. K. Schott. 2012. The surprisingly swift decline of U.S. manufacturing employment. National
Bureau of Economic Research (NBER) working paper series. Working paper 18655.
http://www.nber.org/papers/w18655. Accessed July 7, 2013.
Rogoff, E.G., M.-S. Lee, and D.-C. Suh. 2004. Who done it? Attributions by entrepreneurs and experts of the factors
that cause and impede small business success. Journal of Small Business Management. 42(4):364-376.
Schuler, A. and U. Buehlmann. 2003. Benchmarking the Wood Household Furniture Industry: A Basis for Identifying
Competitive Business Strategies for Todays Global Economy. USDA Forest Service General Technical
Report. GTR-NE-304. 18pp.
Storey, D. J. 2003. Enterpreneurship, small and medium sized enterprises and public policies. In: Zoltan, J., B. David
and B. Audretsch (Editors). 2003. Handbook of entrepreneurship research: An interdisciplinary survey and
introduction. Kluwer Academic Publishers.
313
Assessment,
conservation
of non-residential
forest resources:
Ecodesign management
of appearanceand
wood
products for
the FAO
perspective
applications
Aline Cobut1*, Pierre Blanchet2, and Robert Beauregard3
1
3
FPInnovations
Facult de Foresterie, Gographie
2
Qubec,
QC,
G1P
4R4,
Canada
et Gomatique
Forestry Officer, Forest Assessment,
Forestry Officer, Forest
Assessment,
Universit
Laval
Management and Conservation Division,
Management and Conservation
Qubec,
G1V 0A6, Canada
Forestry Department,
Division,
ForestryQC,
Department,
Although sustainable development has been first introduced in the Brundtland report in 1987, major
efforts are still needed from the current governments and institutions for mankind to reach a sustainable level of
development. As environmental awareness grew on, societal demand for more environmentally friendly products
has led to the creation and proliferation of environmental labelling. But competition on the green market opened
ABSTRACT
the way to greenwashing and consequently confusion
among consumers. Thus, it seems important to understand
the facts behind the seals. In parallel, ecodesign is one of the applications of the sustainable development concept.
FAO
is of
a specialized
international
the United
global
efforts(LCA)
to eradicate
A wide
panel
specific tools
is availableagency
to help of
throughout
theNations
process.leading
Life Cycle
Analysis
identifies
hunger, eliminate
povertywith
and apromote
sustainable
management
natural resources.
In aan
environmental
impacts rural
associated
producttheentire
life cycle,
thereforeof helping
in designing
environmentally-friendly
product.
In the end,
products
have
to be for
put food
on the
market
the question
world with an expected
population
of 9 ecodesigned
billion people
by 2050,
demand
and
otherand
products
and
of ecolabelling
needs
to be will
considered.
services from
nature
be greater than ever. This pressure will change both the extent and the quality
of a substantial portion of the worlds forests. Understanding these changes is essential if investments
are to be made to improve forest management to both provide for a growing human population and to
This research explores several concepts linked to the practice of ecodesign in the appearance wood
maintain crucial forest biodiversity. FAO is committed to generating reliable information on forests and
products sector. One concept would be environmental labelling. Ecolabelling is discussed in terms of
forestry practices
the conservation
and sustainable
use ofwood
forest
resources
environmental
relevance to
andpromote
performance
of existent programs
for appearance
products.
The worldwide.
legitimacy of
Clearly,isadditional
neededofinecodesign
research,for
whilst
technical
support,
field activities
still
ecolabelling
addressedinvestment
later in theiscontext
wood
products.
Anotherand
concept
would beare
LCA
as a
tool for
ecodesign.
ecodesign
study based
on aarea,
cradle-to-gate
LCA
hastropics.
permitted to identify and propose
needed
to slowAnthe
loss and case
degradation
of forest
particularly
in the
alternatives for environmental issues related to the life cycle of the most specified interior wood product in
non-residential building projects.
1. INTRODUCTION
1. INTRODUCTION
The Food and Agriculture Organization of the United Nations (FAO) is a specialized international
agency
leading DEVELOPMENT
global efforts to eradicate hunger, eliminate rural poverty and promote the sustainable
1.1
SUSTAINABLE
management of natural resources.
FAO serves its 192 member countries3 as a neutral forum, retaining the
For the first time in the 20th century in 1972, the Club of Rome exposes the idea that the current development
following
core functions:
1) facilitate
andonsupport
countries
in the
development
and implementation
rate
of our society
was not sustainable
based
scientific
predictions
about
natural resources
scarcity. In the of
late
normative
and standard-setting
instruments
such as international
of conduct,
technical
1980s,
Our common
future also known
as the Brundtland
report, issuedagreements,
by the Worldcodes
Commission
on Environment
and
Development
(WCED),
introducedanalyse,
and defined
what
was
going to
be the
challenge
our society
ever
standards
and others;
2) assemble,
monitor
and
improve
access
to biggest
data and
information,
in areas
faced,
sustainable
Sustainable
development
is a policy
development
that
needs of
present
related
to FAOs development.
mandate; 3) facilitate,
promote
and support
dialogue
at meets
global,the
regional
andthe
country
without
the ability
of future
generations
to meet and
theirregional
own needs
1987). It stresses
levels; compromising
4) advise and support
capacity
development
at country
level(Brundtland
to prepare, implement,
monitorthe
inter-related nature of economic, environmental and social activities.
and evaluate evidence-based policies, investments and programmes; 5) advise and support activities that
Even though
sustainability
seems to
an actual
issue and challenge,
it has
seen
earlier in
thatthetheareas
premises
assemble,
disseminate
and improve
thebeuptake
of knowledge,
technologies
andbeen
good
practices
of
are more than twenty to thirty years old. It can be explained by the lack of information communicated to the public
FAOs mandate; 6) facilitate partnerships for food and nutrition security, agriculture and rural development
in relation to industrial activities and their respective environmental impacts. In fact, the environmental awareness
between
governments,
development
partners,
civilmedia
society
and the dedicated
private sector;
7) advocate and
communicate
among
people
has developed
accordingly
to the
coverage
to environmental
disasters
related to
at national,
regionalThe
andactual
globalknowledge
levels in areas
of climate
FAOs mandate.
industrial
activities.
about
change and its short-term and long-term consequences
stresses the need for a sustainable development (IPCC 2007). The increasing number of developing countries in
parallel with world population growth put an even greater stress on natural resources and surrounding ecosystems,
which should also be considered as a driving force for sustainability implementation.
1.2 BUILDING SECTOR ENVIRONMENTAL FOOTPRINT
3
314
FAO has 192 member states along with the European Union, Faroe Islands and Tokelau which are associate members
among people has developed accordingly to the media coverage dedicated to environmental disasters related to
industrial activities. The actual knowledge about climate change and its short-term and long-term consequences
stresses the need for a sustainable development (IPCC 2007). The increasing number of developing countries in
parallel with world population growth put an even greater stress on natural resources and surrounding ecosystems,
which should
also beConference
consideredon
asHardwood
a driving force
for sustainability
implementation.
International
Scientific
Processing
(ISCHP 2013)
1.2 BUILDING SECTOR ENVIRONMENTAL
International FOOTPRINT
Scientific Conference on Hardwood Processing ,ISCHP13, Florence, Italy
The building sector has a large environmental impact when looking at carbon dioxide emissions, energy
consumption and material extraction (Gonzlez and Garca Navarro 2006). According to Bribin et al. (2011)
building construction and civil works use 60% of the raw material extracted from the lithosphere and the building
sector represents 24% of these global extractions. Moreover, Esin (2007) explains as regards to the total impact on
environment, argues that the impact incurred during the production process of building materials has an important
role because of building materials own life cycles. Furthermore, among all the main carbon dioxide emitting
activities, the building sector is one where practice improvement may have significant environment impact
reduction, with minimum change in the western world lifestyle, (Levine et al. 2007; Barker et al. 2007).
Buildings are also confronted to indoor air pollution. The US Environmental Protection Agency (EPA) pointed
out that the concentration of airborne contaminants indoor is two to five times higher than outdoor levels, then
contributing to higher cases of building related illness (Bonda and Sosnowchik 2007).
Green building certification programs have been developed over the last two decades (Fullana et al. 2008). The
oldest, the most globally recognized and the most comprehensive would probably be the English BRE
Environmental Assessment Method (BREEAM) that was first launched in 1990 . In the USA, the Leadership in
Energy and Environmental Design (LEED) is another widely recognized program . American programs like
GreenGlobes and Collaborative High Performance Schools (CHPS) are also listed as sustainable construction
standards. In Japan, green building standards have emerged with the Comprehensive Assessment System for Building
Environmental Efficiency (CASBEE). In France, the Haute Qualit Environnmentale (High Environmental Quality in
French; HQE) was developed .
1.3 WOOD PRODUCTS, AN ENVIRONMENTAL REVIEW
When considering environmentally friendly materials, there is a broad agreement around the virtues of using
wood (Sathre and O'Connor 2010). Wood made products show numerous environmental benefits, but they also face
some environmental issues.
Wood is a renewable resource from the forests. Nevertheless, without proper forest management and harvesting
methods, its usage can lead to soil deterioration, forest degradation and deforestation (Shvidenko 2008). These are
current environmental issues in the tropical rain forests whereas European and Northern American forests are
generally considered to be managed in a sustainable manner (FAO 2010). Wood, a naturally grown material, is
constituted by the action of photosynthesis that uses the sunlight to fix atmospheric carbon dioxide to grow. As a
result, carbon dioxide is stored in the wood structure until the material burns or decays after the death of the tree,
through the action of insects, fungi, and bacteria. Even when wood is processed to become a product, the carbon is
still stocked until the product end of life. However, Buchanan and Bry Levine (1999) consider that carbon storage in
wood products is not a long term solution considering the steady state level of carbon storage in product (finite life)
and the increase in cumulative carbon emissions from fossil fuels utilization in their manufacturing. Within the
forest products industry, the balance in equivalent carbon dioxide emissions is almost neutral, due to the low level of
industrial processing and can be negative, representing a net absorption of emissions, if the products are recycled or
reused instead of landfilled at the end of their life (Bribin et al. 2011).
In wood based appearance products manufacturing (e.g. tables, desks, shelves, etc.) adhesives, glues, coatings or
paints, taints, and varnishes that can contain volatile organic compounds (VOCs) or hazardous substances are used.
Some adhesives are made of chemicals emitting VOCs, the most mentioned being formaldehyde and its derivatives,
which are known to have negative effects on human health (Gminski et al. 2010). VOCs can react with ozone
molecules, even at low concentration, inducing sub-micron particles and by-products, that may provoke harmful
consequences on the health of some sensitive populations (US EPA 2010). Furthermore, additives like
flame-retardant (halogen based), fungicides, pesticides and other biocides improve the material durability against
mold, decay or fire but can be a source of toxic exposure. All these additives are the main reason of the increase of
environmental impacts of wood products.
1.4 ECODESIGN
A relevant application of sustainable development concept in product design and development is ecodesign. It
embraces two of the three constituent parts of sustainability that are economics and environment(Le Pochat 2005).
Ecodesign involves integrating environmental aspects at all stage of the product development process. In this
process, the environment is given the same status as more traditional industrial values such as profit, functionality,
aesthetics, ergonomics, image and overall quality (Brezet and Van Hemel 1997). Ecodesign can also be referred to
as design for environment, life cycle design, and environmentally-conscious design and manufacturing.
315
316
Category
Purpose
Ecolabels
FO
SFI CoC1, FSC CoC, CSA Z809 CoC, PEFC CoC, SGEC CoC
IA
Multi-attributes
LC
A Chain of Custody (CoC) certification is the link between wood products and the origin of wood. Obtaining a CoC certificate
implies to have a forestry management certificate issued by the same organization. For instance a FSC CoC certificate can be
issued if the raw material comes from FSC certified forests and not from other forest management certification schemes.
Sustainable Forestry Initiative (SFI), Canadian Standards Association for sustainable forest management (CSA Z809),
Sustainable Green Ecosystem Council (SGEC), Forest Stewardship Council (FSC), Programme for the Endorsement of Forest
Certification (PEFC)
2
Sustainable Materials Rating Technology (SMaRT)
The FO category includes ecolabels that deal with forestry and forest products. The IA category comprises
ecolabels that address the issue of indoor air quality. The MA category is composed by multi-attributes ecolabels as
opposed to FO or IA categories. The LC category is constituted by ecolabels that includes a life cycle thinking
approach.
It has been perceived that single criteria approach as found in FO ecolabels and IA ecolabels was not suitable to
express overall environmental performance. They show low environmental impact coverage and few life cycle
stages of the product. These results are consistent with the literature on ecolabels that confirmed that a relevant
assessment of environmental performance of a product should not be based on only few impact indicators and/or life
cycle stages. Strong convergence was observed along the idea that the holistic approach of the LC ecolabels was
better than the fragmented approach of MA ecolabels to support the development of environmentally friendly
product. It has been supported by both the life cycle stages and environmental indicators analysis and the literature
review (Cobut et al. 2012). The literature review on ecolabels again supported these findings as the LCA
methodology is considered to be the most complete approach to environmental appreciation of product. Even
though, life cycle thinking has shown that LC-ecolabelled appearance wood products are more likely to demonstrate
a more accurate environmental performance than the three concurrent categories of ecolabels (i.e. FO, IA and MA
ecolabels), LC ecolabels yet appear to be more difficult to implement due to the necessities and complexities of the
LCA methodology.
2.1 THOUGHTS ON ECOLABELS DRIVEN ECODESIGN FOR INTERIOR WOODEN DOORS
Appearance wood products, such as interior wooden doors, are mainly assemblies of other wood products that go
from fiber boards to veneer sheets. Such components may already possess environmental certifications like FO
ecolabels or compliance certification for indoor air standards. However, it is difficult to integrate and value those
ecolabels at a product level, since it becomes an assembly of ecolabelled and non-ecolabelled products. Single
attribute ecolabels cannot represent real environmental performances of a door, while MA and LC categories are
little by little disgarded for EPDs.
If the practice of ecodesign in this industry sector is addressed relatively to the presence of previously mentioned
ecolabels, it seems essential to consider the previous findings. While, it has been seen that single attribute ecolabels
(FO and IA ecolabels) are not sufficient to propose a holistic approach of ecodesign, MA and LC ecolabels should
bring better grounds for thoughts for ecodesign implementation in a company. Despite their relevance, MA and LC
ecolabels seem not to be as well suited for a proper and transparent method for ecodesign as EPDs. Certification
criteria in MA and LC ecolabels are still too subjective for a rigorous application of LCA. Whereas results presented
in an EPD are stemmed necessarily from LCA results and undergo third-party evaluations. Hence, the significance
of performing a LCA for interior wood doors for addressing ecodesign.
3. LIFE CYCLE ASSESSMENT AS GROUNDED ENVIRONMENTAL ANALYSIS FOR WOODEN DOORS ECODESIGN
3.1 METHODOLOGY
The functional unit is a standard interior door with a surface of 2.1 by 0.9 meter and a thickness of 4.5 cm. The
door is assumed to stay in the building as long as its function is needed. According to the manufacturer, the life span
of the door is life-guaranteed but the faces have a life span of 40 years (Cobut et al. 2013a).
The entire life cycle of the product is included in the system. Meaning that, all the steps are considered from the
acquisition of the raw materials to the end-of-life. This is a cradle-to-grave LCA. Since the manufacturer includes
317
Human health
Ecosystem quality
Climate change
Resources
(DALY)
(PDF.m!.yr)
(MJ primary)
Raw Materials
69.2 %
82.4 %
70.7 %
78.6 %
Manufacturing
0.449 %
- 0.379 %
0.358 %
0.464 %
Packaging
1.63%
5.58%
1.76%
2.60%
Transport to site
13.1%
7.74%
10.7%
8.13%
Usage
0.623%
0.184%
0.657%
0.408%
End-of-life
15.0%
4.44%
15.8%
9.80%
Total
100 %
100 %
100 %
100 %
318
319
320
321
ABSTRACT
Wood is natural material with unique features. Such unique features have led to high demand for wood and wood
products. Production of wood based material (wbm) including wood composites having unique features is in the
rise. Most often the production of wbm involves using materials other than wood which affect the recyclability and
eco-compatibility of end-life products. European Union has regulated values for elements and substances allowed
in recycled wood fibers for the production of wbm. Monitoring heavy metals in wbm is usually done with Atomic
Absorption Spectroscopy or Inductively Coupled Plasma Spectrometry, accurate but laborious and expensive
methods. ED-XRF Energy Dispersive X-ray Fluorescence technique is fast, portable, relatively inexpensive, pretty
accurate in analysing presence and concentration of heavy metals. This technology was used in analysis of wood
and wood based products either virgin material or material from recycling. Specimens were classified based on
type of material and evidence of pollution. Their elemental concentration was compared with natural and
normative values. The wbm more compliant to heavy metals normative were not evidently polluted; among polluted
materials more compliant were plywood and fibreboard, while less compliant were particleboard and solid wood.
Average concentrations of Cd, Cl, Cr, Pb and partially Cu were over the acceptable limit. Very high values of Cl,
Pb and Cr were found. Cd, Ba, Ti and Zr showed as well high concentration, far more the natural. Technique has
proved its efficiency in monitoring heavy metals and other elements, although with certain limitations.
1. INTRODUCTION
Wood as a raw material has many advantages. Sustainability, no carbon fingerprint and recyclability are just some
to mention. Common manufacturing wood processes involve applications of some materials like adhesives,
preservatives, coatings, or application of overlays or veneers which may not be eco-friendly. The present European
standard (EU Commission Decision 2009/894/EC limit values of elements and substances allowed in recycled wood
fibers for the production of wood-based materials for the award of the Community Ecolabel for wooden furniture)
limits the concentration of most hazardous molecules/elements, namely As, Cr, Hg (25 mg/kg), Cu (40 mg/kg), Cd (50
mg/kg), Pb (90 mg/kg), F (100 mg/kg), Cl (1000 mg/kg), Pentachlorophenol (5 mg/kg), Tar oils-benzo(a)pyrene (0,5
mg/kg). Wood waste has three main destinations: combustion, landfill and re-cycling in other wbm such as
particleboards. Uncontrolled pollutants present in wood waste may be therefore transferred from recycling facilities to
atmosphere, soil or other wbm, according to the respective final destination. Monitoring the contaminants of wooden
materials/waste allows managing and controlling these transfers. In this research the efficiency of a handheld ED-XRF
device in monitoring elemental concentration of wood residues was assessed. The correct measurements by ED-XRF,
in principal designed for metals, involve several preliminary corrections based on type of matrix (metal alloys, wood...)
and background (Sorensen, 1981). These parameters, calibration and procedures were previously investigated (Fellin
et al. 20131 and 2).
* Corresponding author: Tel.: (+39) 0461-660215; Fax: (+39) 0461-650045; E-mail: fellin@ivalsa.cnr.it
322
323
concentration
range [mg/kg] As Ba Br Ca Cd Cl Cr Cu
0-10
100 nd 67 nd nd nd nd 49
11-25
0 nd 20 nd 0 nd nd 38
26-50
0 nd 7 nd 9 nd 0 4
51-100
0 nd 0 nd 45 nd 0 3
101-250
0 10 7 nd 27 nd 54 1
251-1000
0 31 0 18 18 nd 15 5
1001-3k
0 38 0 46 0 nd 31 0
3k-10k
0 18 0 24 0 nd 0 0
10k-50k
0
3 0 12 0 14 0 0
up to 500k
0
0 0 1 0 86 0 0
chemical element
Fe Hg Mn Ni Pb
nd 0 nd 9 7
0 50 5 73 18
0 50 29 18 9
0 0 38 0 28
15 0 26 0 16
43 0 2 0 16
26 0 0 0 4
12 0 0 0 2
4 0 0 0 2
0 0 0 0 0
Sb Se Sn Sr Ta Ti
nd 0 nd 37 0 nd
nd 0 0 38 0 nd
100 100 59 12 67 nd
0
0 17 8 33 1
0
0 22 4 0 6
0
0 2 1 0 10
0
0 0 0 0 8
0
0 0 0 0 25
0
0 0 0 0 43
0
0 0 0 0 7
Zn
17
40
12
12
9
6
2
1
1
0
Zr
0
21
21
21
36
0
0
0
0
0
Comparing the results with maximum natural values (Table 2) was useful to find the unusual concentrations of the
elements. The highest concentrations were seen for Cd, Cl, Sb, Ti and Zr. These elements were detected in
concentration from 10 to 1000 times higher than the natural in about 80% of positive specimens (or rather with
concentrations higher than limit of detection for that element). Sb and Zr reached extremely high values (up to 150k
times the natural) due to extreme low natural values. Ba, Br, Cr, Hg and Pb ranged up to 100 times more the natural
levels; this fact also indicates how far wbm may be from natural wood. Fe, Sn and Zn are ranging up to 10X natural
values, and this indicates potential source of pollution. For the remaining elements As, Ca, Cu, Mn, Ni, Se and Sr the
vast majority of specimens range up to 2X the natural values, and consequently are not considered a source of concern.
Table 2: Multielement analysis of wooden material, % of specimens classified in relation with maximum natural elemental
concentration, in mg/kg.
frequency on
positives [%]
Chemical element
max natural conc. [mg/kg]
positive specimens
[n]
below natural conc.
up to 2X natural
2-5X natural
5-10X natural
10-100X natural
100-1000X natural
up to 150000X natural
As
4
1
100
0
0
0
0
0
0
Ba
254
39
10
10
28
18
33
0
0
Br
Ca Cd Cl Cr Cu
Fe Hg
50 20000 5 500 40 200 2000 2
15 284 11 14 13 130 114 2
93
93
nd nd 0 95
77 nd
0
6
nd nd 0
2
9
nd
7
1
0 nd 54 4
11
0
0
0
9 nd 0
0
2
50
0
0
82 14 46 0
2
50
0
0
9 86 0
0
0
0
0
0
0
0
0
0
0
0
Mn Ni Pb
Sb
840 80 30 0,013
126 11 57
4
100 100 25
nd
0
0 14
nd
0
0 28
nd
0
0 14
nd
0
0 16
nd
0
0
4
100
0
0
0
0
Se
253
3
100
0
0
0
0
0
0
Sn Sr Ta Ti
30 101 6 55,3
41 235 3 114
0 95 nd nd
66 4
0
1
22 1 33
6
12 0 67
6
0
0
0
21
0
0
0
61
0
0
0
5
Zn
Zr
200 0,002
306
14
89
nd
4
nd
4
nd
1
nd
2
nd
1
nd
0
100
If the Ecolabel limits (Table 3) were considered as the criteria for comparing, the most problematic elements were
Cl and Cr, with 100% of measurement over the limits. It is important to notice that for both elements the LOD does not
allow measurements below the Ecolabel limit. Cd and Pb have 81% and 41% respectively of positive specimens over
the limits, with an uneven distribution up to 100X the limits. 10% of specimens positive to Cu resulted to be out of
Ecolabel limit, 1 case out of the limit for Hg, while the only specimen positive to As shows a concentration within the
limit.
324
Chemical element As
Ecolabel limit [mg/kg] 25
positive specimens
[n]
1
up to 0,1X limit
nd
0,1-0,25X
100
0,25-0,5X limit
0
0,5-1X limit
0
1-2X limit
0
2-5X limit
0
5-10X limit
0
10-100X limit
0
100-1000X limit
0
1000-10000X limit
0
Cd
Cl
50 1000
11
14
nd
nd
nd
nd
0
nd
9
nd
45
nd
27
nd
9
nd
9
14
0
86
0
0
Cr
25
13
nd
nd
nd
nd
0
31
23
46
0
0
Cu
40
130
0
49
33
8
4
0
2
4
0
0
Hg
25
2
nd
0
0
50
50
0
0
0
0
0
Pb
90
57
5
12
16
26
9
18
7
5
2
0
325
ACKNOWLEDGEMENTS
Rilegno (Consorzio Nazionale per la raccolta, il recupero e il riciclaggio degli imballaggi di legno), ASIA (Azienda
Speciale per l'Igiene Ambientale) and CRB-Ulaval (Centre de recherche sur le bois-Universit Laval) for allowing
sampling, Anna and Jakub Sandak for help and suggestions, Alireza Kaboorani for reviewing the article.
REFERENCES
Ellis E. L., 1965. Cellular ultrastructure of woody plants. Wilfred A. Cot, Syracuse University press.
European Union Commission Decision of 30 November 2009 2009/894/EC On establishing the ecological criteria for the award
of the Community eco-label for wooden furniture, C(2009) 9522 OJ L 320, 5.12.2009, p. 2332.
European Standard EN 14961-1, 2010. Solid biofuels - Fuel specifications and classes - Part 1: General requirements, European
Committee For Standardization, Avenue Marnix 17, B-1000 Brussels.
European Standard EN 15297, 2011. Solid biofuels, Determination of minor elements As, Cd, Co, Cr, Cu, Hg, Mn, Mo, Ni, Pb, Sb,
V and Zn, European Committee For Standardization, Avenue Marnix 17, B-1000 Brussels.
Fellin M., Negri M., Zanuttini R., 2011. Monitoring pollutants on wood residues using FT-IR-ATR technology, ISCHP 11
Virginia Tech, Blacksburg, VA, USA 16-18 October 2011, pp. 187-195, ISBN 978-0-9837700-0-8.
Fellin M., Negri M., Zanuttini R., 20131, unpublished, Multi-elemental analysis of wood waste using Energy Dispersive X-Ray
Fluorescence (ED-XRF) analyzer, submitted to Eur. J. of Wood and Wood Prod. - Manuscript IDHRW-13-0094.
Fellin M., Negri M., Zanuttini R., 20132, unpublished, Characterization of ED-XRF technology applied to wood matrix, submitted
to Wood Research.
Hasan R., Schindler J., Solo-Gabriele H. M., Townsend T. G., 2010. Online sorting of recovered wood waste by automated
XRF-technology. Part I: detection of preservative-treated wood waste, Waste manag., 31, 688-694, doi:
10.1016/j.wasman.2010.11.010.
Pendias A. K., 2011. Trace elements in soil and plants, IV edition, CRC Press, Taylor & Francis Group, ISBN 978-1-4200-9368-1.
Sorensen I., 1981. A calculated background correction method for XRF. Spectrochim. Acta, 36B (3), 243-249, Pergamon Press
Ltd.
Valerio V., Villone A., Nanna F., Barisano D., 2009. Chemical characterisation of biomass feedstocks for the gasifier test,
deliverable D5.3 of UNIQUE project, Seventh Framework Programme Theme Energy.
Williams C., 1976. The Rapid Determination of Trace Elements in Soils and Plants by X-ray Fluorescence Analysis. J. Sci. Fd
Agric. 27, 561-570.
326
Biomaterials
Virginia Tech
ABSTRACT
Life Cycle Assessment (LCA) is becoming an internationally accepted tool for product engineers, manufacturing
engineers, and consumers to assess environmental impacts associated with products life cycles: from material
selection, through product engineering, production, product use, and product end of life options. There is a critical
need to develop further LCA models for the forest products field to compare the environmental credentials of
similar products and services as the forest products industry is embracing sustainability concepts. The overall goal
is to provide better quantitative measures for responsible utilization of renewable natural resources and to develop
the optimal forest products (in this case wooden furniture) while balancing economic, environmental and social
aspects. Presented research outcomes are: assessing the environmental footprint of available natural materials;
presenting LCA models of the production process of a wooden chair with variable parameters; benchmarking the
environmental impacts associated with a wooden chair produced from various resources (plantation vs. natural
forest); and comparing different scenarios based on the effects of location, transportation, manufacturing,
finishing, use, and disposal. The research goal is to develop simplified guidelines for engineering, manufacturing,
use and after use life of sustainable forest products intended to be used by product engineers, manufacturing
engineers, and consumers. The LCA tool is also part of teaching concepts. A methodology to teach furniture
design based on aesthetic, strength, manufacturing and environmental concepts was developed.
1. INTRODUCTION
Life Cycle Assessed (LCA) is an analytical tool that is used to systematically characterize environmental impacts
associated with the life cycle of a product or process. Life cycle stages that may be accounted for in a LCA include raw
material extraction, transportation, manufacturing or processing, transportation, and disposal. After defining the scope
of life cycle assessment, the first step in performing a LCA is to account material and energy inputs and outputs in each
life cycle phase of the product and process by completing a life cycle inventory (LCI). The LCI data is then assessed to
better characterize the extent of ecological and human health impacts associated with the product or process. This
assessment is often done through the use of a weighting method wherein certain impact categories are assessed as
having a greater consequence on ecological and human health than other impact categories. One distinct advantage of
life cycle assessment is that it systematically and objectively quantifies environmental impacts of a product or process
and allows multiple products or processes to be compared.
* Corresponding author: Eva Haviarova, Tel.: (765) 494 3619; Fax (765) 496 1344; E-mail:ehaviar@purdue.edu
327
328
Process Name
Hardwood, standing, under bark, in forest/RER
S
Eucalyptus ssp. standing, under bark, u=50%,
in plantation
Electricity, low voltage, Production BR,
at grid/BR S
Heat from LPG FAL
Heat from LPG FAL
Quantity
168.0 in3
(0.002753 m3)
342.4 in3)
(0.005611 m3)
1.4 kWh
2058.8 Btu
440.6 Btu
329
330
331
332
4 DISCUSSIONS
4.1 OBSERVATIONS AND FOLLOW UP STUDY
There are several weaknesses of the Sima Pro model that were identified. One is that the model is not universally
applicable as parameters such as electricity grid mix, forest management techniques, and local natural resources will
vary and potentially impact the universal applications of this analysis. A second weakness is that end of life processes,
especially in terms of carbon sequestration, still remain ambiguous. To address these concerns, a unit process life cycle
assessment of the carbon dioxide equivalent units emitted or sequestered from cradle to grave is currently being
performed.
An emission rate, also known as emission intensity is defined as the amount of a constituent emitted for a given unit
output (Us Energy Info). Throughout the study emission rate was quantified as kilograms CO2 per board-foot,
kilograms CO2 per mile, kilograms CO2 per kilowatt hour. Our final functional unit was converted to kilograms CO2
per the life of one chair. It is important to note that this is different from a greenhouse gas footprint. To examine the
effects on climate change, a greenhouse gas footprint requires the quantification of all greenhouse gasses including
methane, carbon dioxide, and nitrous oxide. After the mass of each greenhouse gas emission is quantified, this value is
then converted using a greenhouse gasses global warming potential (GWP) weighting factor into equivalent units of
Carbon Dioxide (CO2eq) to get what is known as the absolute emission value (US Energy Info). For example, methane,
a greenhouse gas less common than CO2, is generally accepted to have 21-23 times the global warming effect per gram.
There was also an attempt to evaluate changes in emissions that vary with the geographical location of production.
Discrepancies in an LCA performed using aggregate data caused by manufacturing a wood product at a different
geographical location include different energy grid mixes, different wood resources available, and different wood
growth strategies used in different geographical locations (i.e., monoculture plantation, multi-species plantation,
unmanaged forest stand, and managed forest stand.)
333
334
Poster
335
Forest products
Swedish University of Agricultural Sciences
Uppsala, 750 07, Sweden
1. INTRODUCTION
The market for forest products is an international business. This in turn sets high requirements on the possibilities of
transferring information along the wood value chain (WVC). Several authors have pointed out that a key component in
handling the complexity of the WVC is the efficient management of information (Haartveit et al. 2004, Lancioni and
Chandran 2009, Nordstrm and Mller 2009). To reach optimal solutions for the whole chain, new systems are needed
that can simultaneously include, for example, industrial orders, stand, tree and log characteristics and availability, as
well as transport and logging costs in the analysis (Eriksson et al. 2003).
The overall objective of the WoodApps-project is to develop a pilot "Information and Communication Service
Platform" (ICSP) to link process partners within the international hardwood value chain, from forest to customer. To
optimize the supply of relevant data gained outside the system, cutting-edge services realized as mobile phone
applications will be connected to the platform (Anon. 2011). Furthermore, different tools, e.g. a visual assessment tool
for hardwood quality (logs and sawn timber) that has been developed in Sweden, could make communication between
actors easier (Karlsson et al 2011). Focus in the first step of the project has been to collect information about the
processes and participants in the WVC in order to develop an idea concept for the WoodApps ICSP.
To give an idea of what type of modules are needed in the ICSP, and to identify the terminology to be used in the
platform service descriptions, this study presents a description of how the companies in the WVC interact. The
description concerns kind of relationships, kinds of information shared and how information is exchanged.
2. MATERIALS AND METHODS
A qualitative approach with open questions was chosen to collect information about the targeted issues. A total of
18 representatives from 11 companies covering different activities in the WVC were interviewed. Main business areas
of the responding companies were (figures in brackets indicate number of respondents with a specific business area)
forest management (5), forest owners associations (4), forest service providers (6), transport providers (2) and wood
processing industries (4). Most companies deal with several business areas. The survey was carried out in Germany,
Ireland, Slovenia and Sweden during 2012; the results have been documented in two final reports (Donovan 2013,
Woxblom 2013).
3. RESULTS AND DISCUSSION
Today business partners communicate in a number of ways. Although digital ways of communication are steadily
increasing, it is important to remember that personal meetings are still needed. It is probably not possible, or even
desirable, to digitalize all kinds of communication. An important issue that should be highlighted is that there is a large
spread of users knowledge; not all actors are used to digital information and written information is still needed.
336
The respondents experienced an information gap through the WVC. A problem that was highlighted during the
interviews was that a lot of information was gathered but only a minor part of it was forwarded to the relevant actors and
shared between business actors. Already existing data is therefore recorded again, or reconstructed. The main reasons
identified are that there are many actors along the entire WVC, and that information could be modified or lost on its way
from sender to final receiver. Information and data is very often collected manually or by different kinds of mobile
devices. Often data is transferred into electronic systems and exchanged within a company, but there is a lack of
possibility for exchanging information between different systems and between actors in the WVC. Different actors
need different kinds of information to be able to perform their work, but several examples of information was often
missing in the interviews. For example, haulage contractors can get information about where to both find and deliver a
load of logs, but the data needed to choose an adequate vehicle for the transport (e.g. the type of logs to be carried, the
volume with or without bark, species) is often missing. Other examples of missing information are GPS-data,
cross-cutting instructions, just in time information on daily stocks, required assortments, impact of growth conditions
of logs on final wood quality, and origin of logs.
The respondents think that there is a huge potential for improvement in the area of communication and exchange of
information both within companies and between different actors along the WVC. The respondents requested solutions
for registering data, tools for operational planning, standardized and integrated systems and various forestry related
solutions. It is important that new solutions are easy to use and cost effective. A system of communication in the WVC
should make it possible to digitize data in a structured and secure way with a communication language adjusted to the
different partners and related to existing systems in the WVC. Among the suggestions for new solutions, we can
identify some that already exist and are in use by some of the responding companies. There seems to be a difference
mainly depending on company size. Large companies have financed and developed their own systems, smaller
companies do not have those kinds of resources (knowledge and money) needed. Until now the forest sector as a whole
has not been particularly good at co-operating in the development of standardized solutions.
4. CONCLUSIONS
In order to make systems work along the WVC, co-operation to develop standard solutions is important to secure
consistency of information. Experience and knowledge about IT-solutions differ between the respondents, and
therefore implementation of novelties should be done carefully. The respondents on the one hand describe available
solutions for the handling of information that facilitates their daily work today. On the other hand they mean that there
is a huge potential for improvement in the area of communication tools in the WVC.
ACKNOWLEDGEMENT
The authors acknowledge the financial support from the WoodWisdom-Net Research Programme which is a
trans-national Research and development programme jointly funded by national funding organizations within the
framework of the ERA-NET WoodWisdom-Net 2.
REFERENCES
Anon. 2011. WoodApps Improvement in collaboratioin along the wood value chain through knowledge-based methods and
mobile applications. Application WoodWisdom-Net.
Donovan, P. 2013. State of the Art in Mobile Technologies. Work Package 2:Mobile Applications. Milestone 1b of the WoodApps
project (Confidential report).
Eriksson, I, Arlinger, J. Forsberg, M. Granlund, P. Lfrot, C. Norin, K. and Forsberg, D. 2003. InForChain Information tools
for a more efficient forestry-wood chain. Work package 2 Requirements 2005 and beyond. Report QLRT-1999-01477.
SkogForsk - The Forestry Research Institute of Sweden, Uppsala, Sweden.
Haartveit, E.Y. Kozak, R.A. and Maness, T.C. 2004. Supply Chain Management Mapping for the Forest Products Industry: Three
Cases from Western Canada. Journal of Forest Products Business Research 1(5).
Karlsson, R. Palm, J. Woxblom, L. and Johansson, J. 2011. Konkurrenskraftig kundanpassad affrsutveckling fr lvtr - metodik
fr samordnad affrs- och teknikutveckling inom leverantrskedjan fr bjrkmnen. Report no. 19, Department of Forest
products, Swedish University of Agricultural Sciences (SLU), Uppsala, Sweden.
Lacioni, R.A. and Chandran, R. 2009. Managing knowledge in industrial markets: new dimensions and challenges. Industrial
337
Prerequisites
for the Conference
Realization of
Transnational
Communication
Platform WoodApps
International
Scientific
onthe
Hardwood
Processing
(ISCHP 2013)
Marketing Management 38:148-151.
Nordstrm, M. and Mller, J.J. 2009. Den skogliga digitala kedjan Fas 1. Working paper no. 676, Skogforsk, Uppsala, Sweden.
Woxblom, L. 2013. Survey and user requirements. Work Package 1: Business Optimization. Milestone 1a of the WoodApps project.
(Confidential report).
338
339
decreased with increasing seed origin latitude and sapling height. According to the analysis of covariance model, the
proportion of browsed trees decreased by 7.3 percentage units per 1 increase (110 km) in seed origin latitude at
the mean latitude (61.2N) and mean sapling height (1.4 m). The proportion of trees with stem breakage (TSB) averaged between 2 and 64 % and the proportion of repeatedly browsed trees (RBT) 0 and 48 % among origins. A
similar pattern of variation among the origins was detected as shown for average BT values, i.e. TSB and RBT also
decreased with increasing latitude and sapling height. TSB decreased by 4.9 percentage units per 1 increase (110
km) in seed origin latitude at the mean latitude (61.2N) and mean sapling height (1.4 m). The number of browsed
branches in the browsed trees also decreased with increasing seed origin latitude. In origins with a high proportion
of browsed trees, individual trees were also more heavily browsed. At the age of 5 years, when no browsing had
occurred in the field trial, the average height of the birch saplings was 1.4 m. The average height of the origins varied between 1.0 and 1.7 m, and the height of individual saplings between 0.3 and 3.5 m. At the age of 10 years, the
saplings with no stem breakage had reached an average height of 3.8 m.
4. DISCUSSION
Differences in the degree of moose browsing among the origins may be related to differences in their autumn
phenology. Birches from southern latitudes cease their height growth and remain green later in the autumn than the
northern origins (Velling 1979, Viher-Aarnio et al. 2005). Thus, the moose probably preferred the southern origin
trees, because they were still green and leafy compared to the northern ones. In addition, seasonal variation has been
shown in the digestibility of birch twigs (Palo et al. 1985). Organic matter digestibility and crude protein contents
are lowest in winter twigs, whereas the proportion of hydrophilic phenols and cell walls (cellulose, hemisellulose
and lignin) is highest. Both phenolic substances and high fibre content decrease the digestibility of birch twigs (Palo
1985, Palo et al. 1985). Thus, it is suggested that the differences in palatability and browsing between seed origins
observed in this study may be explained by the differences in timing of growth cessation and switch from the active
growth to leafless state, and subsequent seasonal changes in digestibility, including fibre content and secondary
metabolites.
In conclusion, the use of birch regeneration material from more southern latitudes, e.g. from the Baltic countries,
and northward seed transfers within the country in Finland cannot be recommended due to the higher risk of browsing, especially due to the higher risk of stem breakage, a severe type of moose damage. The possible benefit of increased growth of southern origins would be easily lost as a result of moose browsing.
REFERENCES
Aarne, M. and E. Ylitalo (eds.). Finnish Statistical Yearbook of Forestry 2012. Finnish Forest Research Institute. SVT. Agriculture, forestry and fishery 2012. 454 p.
Cederlund, G., H. Ljungqvist, G. Markgren, and F. Sthlfelt. 1980. Foods of moose and roe-deer at Grims in Central Sweden.
Results of rumen content analyses. Swedish Wildlife Research 11(4):169-247.
Heikkil, R. and J. Raulo. 1987. Moose damage in plantations of Betula pendula established in 1976-77.
Metsntutkimuslaitoksen tiedonantoja 261. 16 p. In Finnish with English Abstract.
Heikkil, R., A. Lilja and S. Hrknen. 1993. Recovery of young Betula pendula trees after stem breakage. Folia Forestalia 809.
10 p. In Finnish with English summary.
Hjeljord, O., F. Sundstl and H. Haagenrud. 1982. The nutritional value of browse to moose. Journal of Wildlife Management
46(2):333-343.
Hjeljord, O., N. Hvik and H.B. Pedersen. 1990. Choice of feeding sites by moose during summer, the influence of forest structure and plant phenology. Holarctic Ecology 13:281-292.
Lavsund, S. 1987. Moose relationships to forestry in Finland, Norway and Sweden. Swedish Wildlife Research. Supplement
1:229-244.
Lilja, A. and R. Heikkil. 1995. Discoloration of birch trees after wounding or breakage. In: Aamlid, D. (ed.). Forest Pathology
Research in the Nordic Countries 1994. Aktuelt fra Skogforsk 4(95):30-32.
Marttila, V., H. Granholm, J. Laanikari, T. Yrjl, A. Aalto, P. Heikinheimo, J. Honkatukia, H. Jrvinen, J. Liski, R. Merivirta
and M. Paunio. 2005. Finlands National Strategy for Adaptation to Climate Change. Publications of the Ministry of Agriculture and Forestry 1/2005. 276 p. In Finnish with English Abstract.
Palo, R.T. 1985. Chemical defense in birch: Inhibition of digestibility in ruminants by phenolic extracts. Oecologia 68:10-14.
Palo, R.T., K. Sunnerheim and O. Theander. 1985. Seasonal variation in phenols, crude protein and cell wall content of birch
(Betula pendula Roth) in relation to ruminant in vitro digestibility. Oecologia 65:314-318.
Velling, P. 1979. Initial development of different Betula pendula Roth provenances in the seedling nursery and in field trials.
Folia Forestalia 379. 14 p. In Finnish with English Summary.
340
341
Saeid Eshaghi*, Hossein Hosseinkhani2, Mohammad Mahdi Barazandeh2 and Masoud Reza
Habibi2
1- Department of Wood Science and Technology, Faculty of Natural Resources, University of Tehran, Karaj, Iran
2- Member of Scientific board. Wood and Paper Science Research Division, Research Institute of Forests and
Rangelands, P.O. Box: 13185-116, Tehran-Iran, Fax: 0098(21)44196575
*Corresponding author: hhosseinkhani@gmail.com
Abstract
The formaldehyde emission from particleboard made of saxaul wood (Haloxylon persicum) by using urea
formaldehyde (UF) was evaluated. For this purpose, desiccators' method is used according to ASTM D 5582-00.
Results indicated that formaldehyde emission increased with the increase in the board density (%51). Increase in the
moisture gradient also resulted in a decrease in the formaldehyde emission.
Keywords: Formaldehyde emission, Particleboard, saxaul wood, desiccators' method, Density, Moisture gradient
1. Introduction
However Formaldehyde-based resins are inexpensive, it has the advantages of superb bonding properties. Thus, they
are used extensively as the adhesives in manufacturing various household products. One prominent application of
UF resin is in the manufacture of particleboard (PB), plywood, and Fiber board (FB). Several thin sheets of wood
are glued together by UF resin to produce plywood, whereas PB and FB are manufactured by mixing wood chips
and fibers with the resin and then pressing the mixture into its final form at a high temperature (Kim and Kim,
2005b, 2006a).
Many wooden flooring products containing formaldehyde-based resins release formaldehyde vapors, thus causing
consumers dissatisfaction and health-related complaints. Various symptoms such as the most common of which is
irritation of the eyes and the upper respiratory tract are attributed to the emission of formaldehyde gases. Also,
formaldehyde has been found to produce nasal carcinomas in mice and rats after a prolonged exposure at 14.1 and
5.6 mg/L level, respectively. These findings have led to an intensified interest in the indoor environment. Consumer
products, especially construction materials, are a major source of formaldehyde emissions in the indoor environment
(Kim et al., 2007a; Brown, 1999). Martinez and Belanche (2000) carried out a research on the effect of wood species
on formaldehyde emission in plywood. They found that amount of formaldehyde emission has more related to
342
343
3.
The results derived from variance analyses showed that the independent effect of board density on the formaldehyde
emission is statistically significant at the level of 5%. According to figure 2a and 2b, with increasing of board
density, formaldehyde emission increased by 100%. Formaldehyde emissions increase at higher density boards can
be attributed to the chemical structure of the wood. Because of its chemical structure and composition of the wood
connections crystalline urea formaldehyde is destroyed. So much more of this type of wood is used in making the
board more of the crystalline compound used in making particleboard.
Figure 2b illustrates the independent effect of moisture gradient on the formaldehyde emission. The results showed
that with increasing of moisture gradient, formaldehyde emission decreased. Moisture gradient of 2 and 4 % showed
37 and 69% of increase, respectively, in comparison to Moisture gradient of 0%. In fact, the increase in the moisture
gradient caused Condensation wood particle in the surface of board and formaldehyde emission decreased.
Figure 2c illustrates the corresponding effect of board density and moisture gradient on the formaldehyde emission.
The interaction of density and moisture gradient on formaldehyde emissions, figure 2c shows that the formaldehyde
emission rate increases with increasing density gradient with increasing moisture gradient and reducing
formaldehyde emission displays. The release rate of 4% moisture gradient and density of 0.75 g/cm3, contrary to
normal procedure, increasing the formaldehyde emission displays, which can be attributed to the distribution of poor
glue boards have these conditions.
344
4.
Conclusions
Formaldehyde emissions related to factors such as board density, moisture gradient, and the type of adhesive. In this
study, the effects of moisture gradient and board density were analyzed on formaldehyde emissions. The results
showed that with increasing board density, which increases formaldehyde emissions because it can be attributed to
the chemical structure of the wood. The gradient of humidity on formaldehyde emissions showed a gradient of
increasing moisture gradient, the amount of formaldehyde emissions significantly (69%) decreased. Therefore, to
reduce formaldehyde emissions, moisture gradients can be a cost-effective proposal.
5.
Reference
ASTM D 5582-00, 2000. Standard Test Method for Determining Formaldehyde Levels from Wood Products Using a Desiccator.
American Society of Testing and Materials
345
Aydin, I. Colakoglu, G. Colak, S. and Demirkir, C. 2006, Effects of moisture content on formaldehyde emission and mechanical
properties of plywood, Building and Environment 41: 13111316
Brown, S.K. 1999. Occurring of volatile organic compounds in indoor air. In: Salthammer, T. (Ed.). Organic Indoor Air
Pollutants, Occurrence Measurement. Wiley-VCH. Weinheim. pp. 71184.
Formaldehyde Institute Technical Committee. 1980. Tentative test method for emission of formaldehyde from wood products,
Desiccators' method: June 24.
Kim, S. and H. J. Kim. 2005b. Effect of addition of polyvinyl acetate to melamine formaldehyde resin on the adhesion and
formaldehyde emission in engineered flooring. International Journal of Adhesion and Adhesives 25 (5): 456461.
Kim, S. and H. J. Kim. 2006a. Initial tack and instantaneous adhesion of MF/PVAc hybrid resins used as adhesives for composite
flooring materials. Journal of Adhesion Science and Technology 20 (7), 705722.
Kim, J. S., Y.G. Eom, S. Kim, and H.-J. Kim. 2007a. Effects of natural-resource-based scavengers on the adhesion properties and
formaldehyde emission of engineered flooring. Journal of Adhesion Science and Technology 21 (34): 211225
Martinez, E. and Belanche, M. I. 2000, Influence of veneer wood species on plywood formaldehyde emission and content, Holz
als Roh- und Werkstoff 58: 31-34
National particleboard association. 1981. Tentative test method for emission of formaldehyde from wood products: desiccator
method. National particleboard association. Silver spring. Md.
URL-1. EPA, Engineered wood products manufacturing. http://www.epa.gov/ttn/chief/ap42/ch10/final/c10s09.pdf (Last visited:
26.11.2004).
Wiglusz, R. Sitko, E. Nikel, G. Jarnuszkiewicz, I. and Igielska, B. 2002. The effect of temperature on the emission of formaldehyde
and volatile organic compounds (VOCs) from laminate flooring. Building and Environment 37 (1), 4144.
346
A. Jahan-Latibari
INTRODUCTION
INTRODUCTION
Medium density fiberboard (MDF), production has been expanding at very fast rate and new plants have been
INTRODUCTION
Medium
densityinfiberboard
(MDF),regions.
production
been expanding
at very
fast rate
and new
plants
have
been
erected mainly
the developing
Thehas
unique
specifications
of MDF
opened
its way
into
various
erected
mainly
ingenerated
the developing
regions.
The unique
specifications
of MDF opened
its
way
intofurniture.
various
applications
and
interest
among
different
consumer
sectors
especially
cabinet
and
home
Medium density fiberboard (MDF), production has been expanding at very fast rate and new plants have been
applications
andexcellent
generated
interest
among different
consumer
sectors to
especially
cabinet
and home
furniture.
High price
and
return
on investment
enabled
producers
good
quality
rawinto
material
to
erected
mainly
in the developing
regions. The
uniqueMDF
specifications
of utilized
MDF opened
its way
various
High price
and excellent
return
on investment
enabled
MDF producers
to utilized
goodsatisfaction.
quality raw As
material
to
fulfill
the
requirements
of
the
product
and
respond
to
consumer
needs
very
well
to
their
a
result,
applications and generated interest among different consumer
sectors especially cabinet and home furniture.
3
fulfill the
requirements
of
theraised
product
and respond
to consumer
needs(FAO,
very well
to their satisfaction. As a result,
world
MDF
production
was
to
almost
58.9
million
m
in
2009
2010).
High price and excellent return on investment enabled MDF
producers to utilized good quality raw material to
world MDF production was raised to almost 58.9 million m3 in 2009 (FAO, 2010).
fulfill the requirements of the product and respond to consumer needs very well to their satisfaction. As a result,
Interest in production and consumption of MDF is not limited
to industrial countries, and other regions are not
world MDF
productionand
wasconsumption
raised to almost
58.9 is
millionlimited
m3 in 2009
(FAO, 2010).
Interest
in such
production
MDF
industrial
countries,
andtheir
otherMDF
regions
are not
exempt to
developments.
Countriesof
like
China not
and Turkeyto
have
been able
to expand
production
exempt
to
such
developments.
Countries
like
China
and
Turkey
have
been
able
to
expand
their
MDF
production
at very fast
rates than industrial
countries.
these
regions
are faced with
the severe
shortage
of wood
as
Interest
in production
and consumption
of However,
MDF is not
limited
to industrial
countries,
and other
regions
are not
at very
fast
rates
than industrial
countries.
However,
theseasregions
are faced
with
the severe
shortage
of wood as
the
main
raw
material
and
therefore
these
regions
such
Iran
have
been
searching
alternative
procedures
to
exempt to such developments. Countries like China and Turkey have been able to expand their MDF production
the
main
raw
material
and
therefore
these regions
such as
Iran
havespecies
been searching
alternative
procedures
to
fulfill
the
needed
wood,
among
them
plantation
of
exotic
fast
growth
such
as
poplar
and
eucalyptus
trees
at very fast rates than industrial countries. However, these regions are faced with the severe shortage of wood as
fulfill
the
needed
wood,
among
them
plantation
of
exotic
fast
growth
species
such
as
poplar
and
eucalyptus
trees
have
beenraw
in the
centraland
pointtherefore
of research
andregions
developments.
the main
material
these
such as Iran have been searching alternative procedures to
have been in the central point of research and developments.
fulfill the needed wood, among them plantation of exotic fast growth species such as poplar and eucalyptus trees
Therefore MDF production must depend on lower quality raw material. In the era that countries around the world
have been MDF
in the production
central point
of research
andlower
developments.
Therefore
must
depend
on
quality raw
material.for
In the
eraraw
thatmaterial
countries
around
the world
desire
to implement
local capacities
for MDF
production,
the situation
MDF
supply
is becoming
desire due
to implement
local capacities
for MDFFast
production,small
the situation
MDFand
raweucalyptus
material supplywere
is becoming
tense,
to unavailability
rawdepend
material.
diameterfor
poplar
first
Therefore MDF
productionofmust
on lowergrowth,
quality raw material.
In
the era
that countrieswood
around thetheworld
tense, due to(Schneider,
unavailability
of 2000).
raw material.
Fast
growth,
small
diameterthepoplar
and eucalyptus
woodP.were
the first
candidates
et
al.
Roffael
et
al.
(1992)
investigated
potential
of
fast
growth
trichocarpa
desire to implement local capacities for MDF production, the situation for MDF raw material supply is becoming
candidates
(Schneider,
et of
al. MDF
2000). Roffael
et al.that
(1992)
investigated
the potential
of old
fasttrees
growth
P. trichocarpa
wood
for the
production
expressed
fibers
produced
16 and
years
produced
tense, due
to unavailability
of raw and
material.
Fast growth,
small
diameterusing
poplar
eucalyptus
wood
were stronger
the first
wood boards
for the than
production
of MDF
and5 years
expressed
that fibers
produced
using
16used
years
old
trees
produced
stronger
MDF
similar
fibers
from
old
trees.
Wu-Zhang
et
al.
(2000)
the
wood
fibers
combined
with
candidates (Schneider, et al. 2000). Roffael et al. (1992) investigated the potential of fast growth P. trichocarpa
MDF boards
than
similar
fibers
from
5 as
years
old
trees.
Wu-Zhang
et al.
(2000) in
used
the
wood
fibers both
combined
with
bamboo
fibers
and
wood
reported
that
the
ratio
of
bamboo
fibers
increases
the
MDF
furnish,
MOE
and
wood for the production of MDF and expressed that fibers produced using 16 years old trees produced stronger
bamboo
fibers and
wood
reported
that asPine
the ratio
ofwere
bamboo
fibers
increases
in the MDF
furnish,
both
MOE and
MOR
improves
and
the
IB
deteriorates.
fibers
studied
for
MDF
production
to
find
a
way
to
reduce
the
MDF boards than similar fibers from 5 years old trees. Wu-Zhang et al. (2000) used the wood fibers combined with
MOR improves and the IB deteriorates. Pine fibers were studied for MDF production to find a way to reduce the
bamboo fibers and wood reported that as the ratio of bamboo fibers increases in the MDF furnish, both MOE and347
MOR improves and the IB deteriorates. Pine fibers were1studied for MDF production to find a way to reduce the
1
1
348
20
18.77
17.08
14.73
MOR (MPa)
15
10
5
0
170
180
190
Press Temperature ( C)
Figure 1: Illustration of the influence of the press temperature on MOR of the MDF boards
The combined effect of increasing press temperature and steaming time is illustrated in Figure 2.Longer
steaming time produced higher MOR but lower MOE except press temperature of 190 C. It can be anticipated
that longer steaming time intiated the thermo-hydrolysis of the wood substances especially hemicelluloses and
weakens the structure of the wood itself. The longer press timedeteriorated the MOE of the boards and the
boards pressed at 3 minutes showed higher MOE. The results indicates that either shorter press time or lower
press temperature ought to be used for pressing urea formaldehyde bonded MDF borads. The combined effect of
steaming time and press temperature on the MOE of the MDF boards was statistically significant at 95%
confidence level.
20
13.77
10
7.5 Min
5
170
15 Min
180
1802
1800
MOE (MPa)
MOR (MPa)
1950
16.11
15.7
15
2100
18.8 18.7
18.04
1601
Press Temperature ( C)
1643
1500
1200
900
190
1809
1662
7.5 Min
170
15 Min
180
190
Press Temperature ( C)
Figure 2: Illustration of the influence of press temperature and time on MOR and MOE of MDF boards
349
0.8
0.71
IB (MPa)
0.66
3 Min
0.74
4 Min
0.62 0.63
0.6
0.61
0.4
0.2
170
180
190
Press Temperatur ( C)
Figure 3: Illustration of the influence of the press temperature and steaming time on the internal bond (IB) of the MDF boards
We have investigated different press temperatures and times to find the best combination of variables. Optimum
combination was found as 4 minutes pressing at 190 C. Of course these values are valid for 10 mm thick board. It
can be expressed that at lower press time and temperature, sufficient heat is not transferred to the board core and
resin curing is not complete. Apparently, press temperature is more influential on all measured properties than press
time. Even though, the strength of MDF from uncommon raw material such as eucalyptus wood do not compare
with softwoods, but it can be utilized as raw material in fiber deficient countries.
5.93
5.64
5.69
5.5
5.07
4.53
7.5 Min
170
15 Min
180
190
Press Temperature ( C)
Figure 4: Illustration of the influence of the presstemperature and steaming time on the thi ckness swelling after 2 hours
immersion in water of the MDF boards
350
REFERENCES
1234-
Cristiane Inacio de Campos; Francisco Antonio Rocco Lahr. (2004). Production and characterization of MDF using
Eucalyptus fibers and castor oil-based polyurethane resin. Material Reaserch J. Vol.7, no.3, 421-425.
EN 310, 1996. Wood based panels, determination of modulus of elasticity in bending and bending strength. European
Standardization Committee, Brussell.
EN 317, 1996. Particleboards and fiberboards, determination of swelling in thickness after immersion. European
Standardization Committee, Brussell.
EN 319, 1996. Wood based panels, determination of tensile strength perpendicular to plane of the board. European
Standardization Committee, Brussell.
5-
6-
Eleoterio , JR. ; Tomazello-Filho, M. ; Bortoletto-Junior , G .. (2000). Mechanical and physical properties of MDF
panels of different densities and resin content.Departamento de Engenharia , Fundacao Universidade de Blumenau ,
7-
89-
58:1-2,123-124.
10- Wu-ZhangKang; Zhang-Hong Jian; Huang-Su Tong; Yuan-YongSheng; Wu-ZK; Zhang-HJ; Huang-SY; Yuan-YS.
(2000). Effect of manufacturing technology on properties of MDF from bamboo and wood.China - Wood Industry.
2000, 14:3, 7-10; 4 ref.
351
Secondary Manufacturing
FPInnovations
Qubec, Qubec, G1P 4R4, Canada
ABSTRACT
The objectives of the project was to define the cupping trend line of the EWF constituted with a BBP substrate and
hard maple top layer according to the strips width. EWF constructions were realized using three different top layer
thicknesses (2, 3 and 4 mm), three numbers of plies of Baltic birch plywood as substrate (5, 7 and 9 plies) and three
strip widths (82, 133 and 184 mm). The 27 EWF constructions were submitted to 3 weeks at 20C and 20% RH
and 3 weeks at 20C and 80% RH and cupping deformation was measured on each product. The ANOVA test was
conducted to compare the cupping performance of the various EWF constructions. All three main effects, 2-way
and 3-way interaction were highly significant. The combination of a 2 mm thick top layer, 9-ply BBP substrates
and 82 mm wide strips represent t he best combination with cupping deformation of 0.107 mm. On the other
way, the combination of 4 mm thick top layer, 5-ply BBP substrate and 184 mm wide strips represents t he less
efficient construction with a cupping deformation of 1.166 mm. The average cupping performance construction is
made with a 3 mm thick top layer and a 7-ply BBP substrate for each of the three widths.
1. INTRODUCTION
Baltic birch plywood (BBP) (Betula pendula) from 6 to 11 mm thick (5 to 9 plies) is the most popular substrate used
by the Canadian wood flooring manufacturers for engineered wood flooring (EWF) products owing to its high mechanical
properties and low cost. Bouffard et al. (2007) tested EWF product from many manufacturers to define the average
cupping deformation of the flooring strips. The wide variability between EWF construction in terms of top layer
thickness and type and thickness of substrate made it difficult to determine the cupping limit. However, Bouffard et al.
(2007 and 2009) found a significant impact of the strip width and the top layer thickness on the cupping performance. The
objective of this project was to define the cupping trend line of the EWF strips made up of BBP substrate and hard
maple top layer according to the strips width, top layer thickness, and the number of plies in the BBP construction.
2. M ATERIAL AND METHOD
The engineered wood flooring (EWF) constructions used in this study consisted of flooring strips made up of two
components: Baltic birch plywood (BBP) as substrate and a sawn top layer of sugar maple (Acer saccharum). The
number of plies in the construction (5, 7, or 9) was a facto r in the stud y as was the thickness of the hard maple
top layer (2, 3 and 4 mm) and the width of the EWF strips : 82 mm (3 in.), 133 mm (5 in.) and 184 mm (7 in.).
The adhesive selected to bond the layers was a polyvinyl acetate (PVA) Type I adhesive. Press time was 30 minutes
at 200 psi and at room temperature (21 C). After the pressing operation, EWF strips were reconditioned at 20C and
50% RH for two weeks.
To evaluate the performance of EWF constructions, strips were glued onto 610 by 1220 mm cement boards
(Permabase) with a one-part urethane adhesive (Bostik Best). For each type of EWF construction, 12 strips were
installed. Installations were placed in a conditioning room kept at 20C and 20% RH for 3 weeks followed by 3 weeks
* Corresponding author: Tel.: (418) 781-6723; Fax: (418) 659-2922; E-mail: jean-francois.bouffard@fpinnovations.ca
352
at 20C and 80% RH. Cupping deformation was tracked throughout each climatic condition cycle according to the
following schedule: days 0, 1, 2, 3, 7, 14 and 21.
For each EWF construction, cupping deformation values were plotted as a function of time, and deformation
amplitude was determined. The cupping measurement method is described in Blanchet et al. (2003).
An analysis of variance (ANOVA) was performed using the Statistical Analysis System (SAS) software. The
ANOVA test was conducted to compare the cupping of the various EWF constructions. The experimental design
consisted of three factors: 3 top layer thicknesses, 3 BBP substrate thicknesses, and 3 strip widths. It resulted in
27 combinations, each replicated 12 times.
3. RESULTS AND DISCUSSION
The analysis of variance results for the experimental design shows that all main effects were found to be highly
significant with a significance level of 0.01. The effect of the width is the most important with F-value of 641.32.
The 2-way and the 3-way interactions were also highly significant. The interaction between the substrate number of
ply, top layer thickness and strips width on the amplitude of cupping deformation is presented at figure 1.
Figure 1: Effect of the interaction of substrate number of plies, top layer thickness and strips width on the amplitude
of cupping deformation
The analysis indicates that the three main effects generate a significant impact on the cupping deformation of EWF
strips. Increasing the strip widths creates a higher cupping deformation followed by a reduction of the number of BBP
substrates plies and then by an increase of the top layer thickness. The combination of a 2 mm-thick top layer, 9 ply
BBP substrate, and 3 inch-wide strips represents the best combination, with cupping deformation of 0.107 mm. The
combination of 4 mm thick top layer, 5-ply BBP substrate, and 7 inch wide strips represents the least efficient
construction with a cupping deformation of 1.166 mm. For each of the three strip widths, intermediate cupping
performance is achieved with a 3 mm thick top layer and a 7-ply BBP substrate.
4. CONCLUSION
The statistical analysis of the cupping performance of all 27 EWF constructions shows a significant impact of the
three main effects (width, substrate thickness and top layer thickness), 2-way and 3-way interactions. The average
cupping measurement varied from 0.107 mm to 1.166 mm. Strip width has the most important impact on the cupping,
followed by substrate thickness and then top layer thickness. The intermediate cupping performance trend line is
represented by a 3 mm thick top layer with a 7-ply BBP substrate.
REFERENCES
Blanchet, P., Beauregard,R., Cloutier, A., Gendron, G., Lefebvre, M. 2003. Evaluation of various engineered wood flooring
constructions. Forest Prod. J. 53(5):3037.
Bouffard, JF.,Blanchet, P., 2009. Use of fiberboard as substate in floating engineered wood flooring. Forest prod. J.
59(11/12):6-10.
Bouffard, JF., 2007. Performance of wide engineered wood flooring strips. Report number: 5342.FPInnovations, xx pp..
353
1. INTRODUCTION
The wastes, which remain after hardwood processing, are commonly used as a fuel. However, it wood be more
economically attractive to process them so that they could be then used in composites with other materials to improve
their properties.
Recently, lignocellulosic materials have aroused considerable interest for their use as fillers with a thermoplastic
matrix (see, for example, Karmarkar et al. 2007). We have also established that wood processing wastes can be used for
reinforcing of paper (Treimanis et al. 2011). In the present work, hardwood processing wastes were used for obtaining
nanoparticle fillers for their use in papersheets.
2. EXPERIMENTAL
Handsheets were produced from bleached kraft pulp as well as office and newsprint waste paper. Nanoparticles
were obtained from grey and black alder bark, and birch sawdust. Non-extracted bark and that extracted in biorefinery
were used. These materials were destructed by a thermocatalytic method developed by us (Laka and Chernyavskaya
1996). The destructed wastes were dispersed in water medium in a ball mill, and at sufficient concentrations, gel-like
dispersions were obtained, which contained nanoparticles. The dispersions were dissolved and introduced in the paper
furnish in different amounts. Nanoparticle content was 5-20% (from dry matter). The mechanical properties, water
vapour sorption and swelling in water were investigated.
3. RESULTS AND DISCUSSION
Using Zetasizer Nano ZS 90 it was established that most part of nanoparticles have size ~300 nm.
Figure 1 shows the tensile index in a dry state and the burst index of papersheets from wood kraft pulp versus the
content of the nanoparticle filler from non-extracted and extracted grey alder bark. It can be seen that, with increasing
filler content, tensile index and burst index increase. The nanoparticles fill the submicropores and enhance formation of
hydrogen bonds in the paper. The nanoparticles from extracted grey alder bark increase the indices to a greater extent.
The smaller effect of the nanoparticle filler obtained from non-extracted bark can be explained by the fact that the
*
Corresponding author: Tel.: +371 67545122; Fax: +371 67550635; E-mail: lamar@edi.lv
354
2a
2
1a
2a
2
Figure 1: Tensile index in the dry state (1) and burst index (2)
versus the content of the nanoparticle filler from non-extracted
and extracted (1a and 2a) grey alder bark.
1a
Figure 2 shows the tensile index in the wet state of the papersheets produced from wood kraft pulp versus the
content of the nanoparticle filler from non-extracted and extracted grey and black alder bark. Just as in a dry state,
nanoparticle fillers increase also the tensile index in a wet state.
Investigating the effect of the nanoparticle fillers on handsheets, produced from office and newsprint waste paper,
it has been established that, in this case, the influence is smaller. This can be connected with the fact that the strength of
the papersheets from office and newsprint waste paper is greater than that of the papersheets from unbeaten wood pulp.
The swelling and sorption slightly decrease with increasing the filler content. At a 20% filler content from birch
sawdust, the sorption and swelling of papersheets from wood pulp decrease by 16 and 14%, respectively. The effect of
other investigated fillers is smaller.
4. SUMMARY
Nanoparticle fillers from grey and black alder bark, and birch sawdust, which improve the mechanical properties of
papersheets and can be used as reinforcing fillers, were obtained. Nanoparticle fillers, obtained from wood bark after
extracting in biorefinery cycle, improve the mechanical properties to a higher extend. Nanoparticle fillers increase the
water resistance of papersheets.
ACKNOWLEDGEMENT
The study was financed by the State Programme Nat Res and the Ministry of Agriculture, Republic of Latvia, in
the framework of the WoodWisdom-Net Project Lilo.
REFERENCES
Karmarkar, A., Chauhan, S.S., Modac, J.M., and M. Chanda. 2007. Properties of wood-fiber reinforced polypropylene
composites. Composites, Part A, 38: 227-233.
Treimanis, A.., Laka, M., Vikele, L., Chernyavskaya, S., and M. Skute. 2011. Nano-micro particles from wood and cellulose used as
paper additives. In: Abstract Book of the 241st American Chemical Society National Meeting. March 27-31, 2011, Anaheim,
CA, USA, p. 48.
Laka, M., and S. Chernyavskaya. 1996. Patent LV 11184 (In Latvian).
355
International
Scientific
Conferenceon
onHardwood
Hardwood Processing
2013)
International
Scientific
Conference
Processing(ISCHP
(ISCHP
2013)
Department of Forest and Ecosystem Science, The University of Melbourne, Burnley Campus, 500
Yarra Boulevard
c
356
International
Scientific
ConferenceononHardwood
Hardwood Processing
2013)
International
Scientific
Conference
Processing(ISCHP
(ISCHP
2013)
3. Test and develop new processing methods and products from veneer: new techniques will
be tested in order to increase the recovery and the quality as well as to improve product
added value produced from Acacias and Eucalyptus.
4. Monitor and analyse economic impact to smallholders and SMEs during the project, and to
investigate potential new value creation.
5. Implement education and training related to veneer production and use in manufacturing.
In the early stages of this project, an analysis of the existing planation resource has been
completed. Also, the current plantation veneer processing practices and efficiencies in Vietnamese
companies have been identified and characterised through a series of onsite surveys. The outcome of
these studies provide a broad assessment of opportunities to improve current primary wood processing
practices and efficiencies in order to optimise the research direction required for subsequent activities.
Plantation resource analysis
The exisiting plantation resource was analysed via 2011 data from the Vietnam Ministry of
Agriculture and Rural Development (MARD) and a series of standing tree and tree core
measurements representing the different agricultural zones, provinces, and plantation ages for the
highest volume species throughout Vietnam. The species investigated were Acacia mangioum, Acacia
auriculiformis x mangium and Eucalyptus urophylla. Measurements included breast height diameter,
tree height, ovality, heartwood proportion and basic density. The main results from this analysis are:
The majority of Vietnams forests are concentrated in the Northeast, South Central Coast
and North Central Coast regions.
The largest proportion of natural (52%) and plantation (75%) forests is used for producing
forest products. 47% of forest plantations are owned by smallholder farmers indicating
that improvements in the use, productivity and value of this resource should have
significant benefit to farmers incomes.
72% of the plantation forests are purely acacia species plantations. 20% are purely
eucalyptus plantations. In the past two years 77% of forest plantations in Vietnam have
been solely acacia species.
Generally, the average breast height diameter, tree height, heartwood proportion and basic
density of trees increase with age from 2-11 years. Anomalies in this trend tend to occur in
higher stocked plantations where growth is stunted.
At 7-years of age, larger diameter and taller trees can be expected to grow in areas with
alluvial soils followed by basaltic and ferralitic soils.
Faster grown trees tend to have lower basic densities at 7-years of age.
Average tree diameter, heartwood proportion and basic density of A. mangium were
highest at 7 and 10/11 years of age compared with the other species grown in the same
location.
Eucalyptus urophylla trees have significantly lower average basic density than the acacia
species at the same age when grown in the same location on the sites investigated.
In conclusion, the wood properties at age 7-years have attained a reasonable level to be suitable for
veneer production. Initial stocking rates may impact wood growth and properties development and
higher stocking might not equate to higher returns to growers at the end of the rotation. The data
generated will allow further investigation into the economic impact of harvesting age on smallholder
profitability; an objective of this project.
Plantation veneer company surveys
Eight companies were selected across the northern, central and southern regions of Vietnam to
evaluate current capacity via a benchmarking survey of representative plantation veneer processing
facilities and strength, weaknesses, oportunities and threats (SWOT) analysis to identify key areas of
improvement and optimisation. Selection criteria for companies were developed and selected so a
good overview of the veneer processing and product manufacturing industry in the country could be
established.
The assessment was based on survey questionaires which identified and listed important data and
information to be collected. The Data was collated and recommendations of possible improvements,
357
International
Scientific
ConferenceononHardwood
Hardwood Processing
Processing (ISCHP
International
Scientific
Conference
(ISCHP2013)
2013)
modifications and changes were provided to each company. The main points generated from this
survey are:
Peeled veneer uses plantation timber but logs for sliced veneer are either imported or
natural forest timber.
There are no grading rules for plantation logs used for the production of peeled and sliced
veneer available in Vietnam. There are grading rules for logs from old growth native forest
available but these are not applicable for small diameter logs from young plantation trees.
Generally, logs are ordered or purchased according to each individual companys
specification which is mainly based on customer requirements.
The thickness variation of peeled and sliced veneers is generally too high according to
international standards.
Information regarding optimal peeling and slicing of plantation acacias and eucalypts is
not available.
Drying and storage of veneers needs to be improved to reduce moisture content variation
in veneers and veneer-based products.
National standards to grade plantation sliced and peeled veneer for different product
applications needs to be developed.
Production layout improvements are required in some companies to increase processing
and production efficiencies.
Training should be provided for the assessed companies and the industry as a whole. The
training should address all aspects of veneer processing and manufacturing, such as the
production of peeled and sliced veneers, production of veneered panels, veneered products
(plywood, veneered furniture, engineered flooring, moulded veneered components, etc)
and related topics, such as production management practices, including production flow,
quality control, machinery maintenance and safe working conditions.
Occupational, health and safety conditions in the factories should be improved.
In summary, the assessed companies have good potential to improve the quality of products and
become more efficient by making modifications and changes to the veneer processing and
manufacturing procedures and by applying rigorous quality control systems in their production
processes.
Although improvements and developments are necessary to make the companies more efficient
and competitive in national and international markets, numerous opportunities can be identified which
indicate that the industry has a good chance of succeeding. One of the indicators is a strong
commitment from the companies which are open to suggestions and willing to implement
recommended changes and improvements. This commitment, combined with the enthusiasm of the
companies to change and modernise their production processes promises positive outcomes.
Improving the competitiveness of hardwood plantation processors and manufacturers in Vietnam
through improved efficiency and profitability should improve the livelihhoods of farmers and
processing workers.
Acknowledgements: The authors acknowledge the financial and in-kind support of all
collaborators involved in the project and funding provided by the Australian Centre for International
Agricultural Research (ACIAR), Canberra, Australia.
358
International
Scientific
Conference
Hardwood Processing
Processing (ISCHP
International
Scientific
Conference
ononHardwood
(ISCHP2013)
2013)
Faculty of Forestry, National University of Laos, Dong Dok Street, Xaythany District, Vientiane
Capital, Lao PDR
Department of Forest and Ecosystem Science, The University of Melbourne, Burnley Campus, 500
Yarra Boulevard
c
359
International
Scientific
ConferenceononHardwood
Hardwood Processing
Processing (ISCHP
International
Scientific
Conference
(ISCHP2013)
2013)
processing practices and efficiencies in order to optimise the research direction required for
subsequent activities.
A number of companies were selected from the Vientiane and LuangPrabang areas of Lao to
evaluate current capacity via a benchmarking survey of representative primary processing facilities
and strength, weaknesses, oportunities and threats (SWOT) analysis to identify key areas of
improvement and optimisation. Selection criteria for companies were developed and selected so a
good overview of the sawmilling and furniture manufacturing industry in the country could be
established. The selected companies specialise in green sawmilling hardwood plantation material,
solid wood processing, the production of furniture or a combination of these. A total of 13 companies
were selected.
The assessment was based on survey questionaires which identified and listed important data and
information to be collected. The questionaire covers the following topics: company profile,
infrastructure, log and species data, processing and manufacturing systems, workplace health and
safety, expertise and training, grading, storage and SWOT analysis. Data was collated and
recommendations of possible improvements, modifications and changes were provided to each
company.
Companies vary in size from small to large depending on the size of the workforce. Most
companies operate for 8 hours per day and 28 days per month. Some companies are in partnership
with overseas companies who also provide export markets.
All of the assessed companies are using plantation species of various ages, different dimensions
and wood quality. The predominent species is teak (Tectona grandis) followed by a number of other
exotic and native plantation species.
The majority of companies purchase timber as squared flitches from various locations. Some
sawmills grade logs/flitches for price based on size but not quality.
Generally, most sawmills have some production inefficiencies which can be easily rectified by
training, machine maintenance, optimised processing and production layout. Most sawmills are using
equipment that is not new but usable.
The majority of companies have drying facilites but most are old designs with little control over
drying conditions. Better drying regimes are required to optimise dried quality and to meet
expectations of export markets.
Most manufacturing sites require better knowledge and training in the best aplication and types of
glues and finishes to meet the desired needs of current and future markets.
Ocupational health and safety in most factories could be improved easily using signage, better
waste management, improved use of personal protective equipment and safety systems.
The assessed companies provided valuable comments on thier current problems and most urgent
issues. The majority of concerns involve: more financial support from Governement to improve
infrastructure, in-house training in the areas of production mangement, machinery utilisation, product
quality and marketing, and support for export market development and product design.
The companies stated that technical advice would be beneficial in the areas of kiln drying,
machinery use and maintenance, production flow, product quality, marketing and design.
The outcomes of this survey have allowed the project team to target priority areas to improve the
process and manufacturing effeciencies in the companies surveyed and similar companies throughout
Laos. The next phase of this project objective is to assist companies to make the recommended
changes to improve their productivity and quality to add value to their products. Improving the
competitiveness of hardwood plantation processors and manufacturers in Laos through improved
efficiency and profitability should improve the livelihhoods of farmers and processing worker.
Acknowledgement: The authors acknowledge the financial and in-kind support of all
collaborators involved in the project and funding provided by the Australian Centre for International
Agricultural Research (ACIAR), Canberra, Australia.
360
References
1. Midgley, S. et al. Scoping Study: Payments for Environmental Services and Planted Log
Value Chains in Lao PDR. Part 3: Opportunities for R&D across the value chain for
plantation-grown wood in Lao PDR. Report for ACIAR. Pp 53-90, 2011
361
Communication
Assessment,
management
and conservation
of forest
resources:
Improving
Hardwood
Production Yields
and
Conserving Resources
with Proper End Sealing
the FAO perspective
Author: Norman E. Murray
Chief Executive Officer
UC Coatings Corporation
1
2
Eduardo Mansur
, Alberto
Del 14215
LungoUSA
, Cesar Sabogal2
Buffalo,
New York
2
2
Director,
Forestry Officer,ABSTRACT
Forest Assessment,
Forestry Officer, Forest Assessment,
Forest Assessment, Management and
Management
and
Conservation
Division,
Management and Conservation
Losses
during hardwood
Conservation
Division, production due to end checking (end splits caused by too rapid drying) can be from 5% Forestry
Department,
Division,
Forestry Department,
20%, depending
on
the
refractory
nature
of
the
individual
species.
This
loss
of
yield is economically
important to the
Forestry Department,
Food and Agriculture
Organization
of theof forest
Foodresources.
and Agriculture
Organization
hardwood
producers,
and is ofofenvironmental
concern due
to the waste
Properly
sealingofthe ends
Food
and Agriculture
Organization
United
Nations,
Rome,
Italy
the
United
Nations,
Rome,
Italy yields and
ofthe
hardwood
logs
and
lumber
can
reduce
or
eliminate
end
checking
losses,
thereby
improving
production
United Nations, Rome, Italy
1
ABSTRACT
definition: End checks, end checking the splits and cracks that are caused by uneven, too rapid
FAO is drying
a specialized
agency
of the
United Nations
leading
efforts to eradicate
on theinternational
ends of pieces
of wood,
especially
regarding
logsglobal
and lumber
hunger, eliminate rural poverty and promote the sustainable management of natural resources. In a
worldthe
withquality
an expected
population
of 9 billion
people
by 2050, demand
and otherAllowing
products and
Maintaining
of valuable
hardwoods
gives
significant
benefitsfor
tofood
everyone.
unnecessary
services
from
nature
will
be
greater
than
ever.
This
pressure
will
change
both
the
extent
and
quality defects
losses during processing is expensive and wasteful of natural resources so it is important tothe
eliminate
of a substantial
portion of
Understanding
changes
is essential
if investments
whenever possible.
End checking
is the
oneworlds
of the forests.
most common
drying these
defects
and can
cause losses
of 5% - 20% of
are
to
be
made
to
improve
forest
management
to
both
provide
for
a
growing
human
population
and toof freshly
the wood being processed. End checking is also one of the easiest defects to prevent. Coating the ends
maintain
crucial
forest with
biodiversity.
FAOsealer
is committed
to generating
information
on forests
and
cut hardwood
logs and
lumber
a proper
is an effective
and reliable
economical
method
of preventing
end
checking. forestry practices to promote the conservation and sustainable use of forest resources worldwide.
Clearly, additional investment is needed in research, whilst technical support, and field activities are still
Logs
needed to slow the loss and degradation of forest area, particularly in the tropics.
All hardwood production starts with the logs. Proper harvesting techniques are important to minimize damage
to the logs when they are felled. Immediately after cutting the logs to their required length, the ends of the logs are
quickly susceptible to damage and degrade from end checking, which can be greatly reduced or prevented by
1. INTRODUCTION
applying
an effective end sealer. The longer the logs are stored before sawing or slicing, the greater the risk of
significant The
damage
if the
are left unprotected.
Food
andends
Agriculture
Organization of the United Nations (FAO) is a specialized international
Beforeleading
discussing
the causes
end checking
in logs,
is an additional
problem
logs that
agency
global
efforts and
to solution
eradicateforhunger,
eliminate
ruralthere
poverty
and promote
the with
sustainable
3
can
be addressed
applying
an effective
sealer.
species of
hardwoods
many forum,
softwoods
get infected
management
ofby
natural
resources.
FAOend
serves
its Some
192 member
countries
as and
a neutral
retaining
the
with
fungus
spores
that
cause
blue
stain.
following core functions: 1) facilitate and support countries in the development and implementation of
normative
and standard-setting
instruments
suchor as
international
agreements,
of conduct,
Blue Stain:
Blue stain is difficult
to reduce
retard
without using
fungicidecodes
chemicals.
Whentechnical
the local
temperatures
are between
20C
and 50C,
the fungus
thatand
causes
blue stain
grows
in theand
endsinformation,
of different in
species
standards and
others; 2)
assemble,
analyse,
monitor
improve
access
to data
areasof
logs.
These
warm conditions
+ water
+ food
sourceand
+ airsupport
are necessary
for the fungus
to grow
rapidly.
water
related
to FAOs
mandate; 3)
facilitate,
promote
policy dialogue
at global,
regional
and The
country
and
the
food
source
(sugars
in
the
sapwood)
are
already
present
in
the
wood.
Air
is
the
one
ingredient
that
can
levels; 4) advise and support capacity development at country and regional level to prepare, implement, monitorbe
limited with an effective end coating that seals the ends completely.
and evaluate evidence-based policies, investments and programmes; 5) advise and support activities that
Therefore,
to reduceand
blueimprove
stain damage
in the
of logs,technologies
it is important
seal practices
the ends immediately
assemble,
disseminate
the uptake
of ends
knowledge,
andtogood
in the areas after
of
cutting.
The
longer
the
ends
are
not
sealed,
the
more
air
enters
the
log
and
the
faster
the
blue
stain
grows.
Sealing
FAOs mandate; 6) facilitate partnerships for food and nutrition security, agriculture and rural development
the ends quickly after cutting can reduce the speed of growth. In separate experiments done at the University of
between governments, development partners, civil society and the private sector; 7) advocate and communicate
Wisconsin-Madison, Department of Forestry and by the Chinese Academy of Forestry, it was shown that properly
at national,
regional
and global
levels in twice
areas as
of long
FAOs
sealed
logs could
be stored
approximately
as mandate.
uncoated logs, and the coated logs had 50% less damage
from blue stain.
End Checking: As soon as the logs are cut, the ends begin drying. The fastest drying occurs at warmer
temperatures and lower humidity. The end of a log will not dry uniformly: one area will dry quickly, another area
will dry more slowly; areas with crushed fibers or having dirt or mud in the end may not dry at all. The uneven
drying causes one area to shrink faster than another area and uneven shrinkage causes splits and cracks in the end of
3 log end checking.
the
FAO has 192 member states along with the European Union, Faroe Islands and Tokelau which are associate members
364
365
Lumber lost
to end checking
3 cm
2.0%
5 cm
3.3%
10 cm
6.7%
15 cm
10.0%
End checking losses in logs and lumber are expensive and an unnecessary waste of natural resources. Someone
eventually pays this cost: the logger who receives a lower price for his logs; the sawmill that receives less money for
lower quality lumber; the furniture manufacturer who loses yield during production.
Since end coating is recognized to be an effective method of reducing end checking, the answers to six basic
questions about end coating are important to consider.
1. What is end coating? End coating is the application of any coating material to seal the end grain of
freshly cut logs or lumber. The some common end coatings are hot paraffin wax, solvent-based resin sealers,
bituminous coatings and water-based wax emulsion sealers.
The most widely used end coatings are the water-based wax emulsions. They are safe and easy to apply with
brushes, rollers or sprayers. They do not have any of the problems connected to other products: hot wax is more
difficult to apply and can burn the user or catch fire; solvent-based materials create hazardous fumes; bituminous
coatings are messy and abrasive, causing saws and veneer knives to dull a little bit each time they saw through the
coating.
2. How does it work? End coating puts a moisture barrier on the ends of the wood so it may dry more
evenly from end to end. According to the USDA Dry Kiln Operator's Manual , "end checks occur because moisture
moves much faster in the longitudinal direction than in either transverse direction. Therefore, the ends of boards dry
faster than the middle and stresses develop at the ends. These stresses are eventually relieved by the ends of the
board splitting apart.
3. What should be end coated? Any log or lumber specie that is prone to end checking should be coated,
particularly hardwoods. Veneer logs, thick lumber and wide lumber are especially prone to severe checking.
4. When should end coating be applied? End coating should be applied to the unchecked ends freshly cut
logs and lumber. The sooner end coating is applied after cutting, the more effective it will be in preventing losses
caused by end checking.
5. How is it applied? End coatings are commonly applied by brush, roller or sprayer. Spraying equipment
that is specifically designed to apply end coating is available from end coating manufacturers. Typically, a package
of lumber can be coated easily in 30 - 60 seconds with the proper equipment.
6. How much does it cost to end seal lumber? The most commonly used end coating, the wax emulsions,
will cover approximately 2.5 m2 per liter. Depending on the length of the lumber, the cost will be approximately
USD $0.75 - $1.00 per m3. (You can calculate your own cost and saving using the End Sealer Usage and Savings
Calculator available at www.uccoatings.com.)
Quality end coatings are available with color or without color (clear). Many lumber producers end coat their
lumber with a certain color to improve the appearance of the lumber, creating a good first impression. Other
producers use a colored end coating to help them distinguish logs or lumber in their own inventory by age or by
species.
The value created by end coating freshly cut logs and lumber or the value received by having suppliers end
seal the logs and lumber which you purchase is significant in terms of cost savings and for the conservation of raw
material and natural resources. Lumber that is end coated at the sawmill enhances the quality of the lumber as well
as the image and reputation of the lumber producer. For the lumber manager and kiln operator, end coating is a tool
366
Norman Murray is CEO of UC Coatings Corporation, USA, manufacturer of wood protection products for the
hardwood industry, world-wide. He has over 30 years experience in the wood protection field, lecturing for
numerous universities and trade associations. In many of the research studies cited, the product ANCHORSEAL
was used in the testing. ANCHORSEAL is the most widely used end sealer in the world.
References:
Denig, J., E. Wengert and E. Simpson 2000. Drying Hardwood Lumber, USDA FPL-GTR-118
Linares-Hernandez, A. and E. Wengert 1997. End Coating Logs to Prevent Stain and Checking. Forest Products Journal, 47(4):
65-70.
QIN Li, ZHAO You-ke, HUANG Rong-feng, LU Jiang-xiong, 2005. End Coating Wood Logs to Prevent Stain and Checking.
Chinese Academy of Forestry.
Reitz, R. and R. Page, 1971, rev. 1999. Air Drying of Lumber, USDA FPL-GTR-117
Simpson, W. ed., 1991. Dry Kiln Operators Manual, USDA Handbook #188
Wengert, E. and F. Lamb, 1990. Is End Coating Effective? Virginia Tech
Wengert, E. 1990. Drying Oak Lumber, Madison, WI: University of Wisconsin, Department of Forestry.
(USDA = United States Department of Agriculture)
367
ISCHP 2015
The next edition of the International Scientific
Conference on Hardwood Processing will be held
in Quebec City, Canada, in 2015.
Quebec City area will show its most beautiful
colors as the conference will take place in fall. The
format will remain two days of conference followed
by a technical tour. The conference will be jointly
organized by Universit Laval, FPInnovations and
Quebec Wood Export Bureau (QWEB). You are all
welcome to join us in Quebec City in two years to
celebrate the fifth edition of the conference!
Conference Chair: Pierre Blanchet
ISBN 9788890927300