Evolution of the AASHTO

Bridge Design
Specifications
A Story in Two Parts
John M. Kulicki, Ph.D.,
Chairman/CEO
Modjeski and Masters, Inc.

OHBDC Especially 1983

Edition

NCHRP 20-7/31 Development of

Comprehensive Bridge Specs and
Comm.
Task 1 - Review of other specifications for coverage
and philosophy of safety.
Task 2 - Review AASHTO documents for possible
inclusion into specification.
Task 3 - Assess the feasibility of a probability-based
specification.
Task 4 - Prepare an outline for a revised AASHTO
specification.

Part 1
The AASHTO LRFD Bridge
Design Specifications:
Yesterday, Today and
Tomorrow

Spring 1986 Gang of Four

Name
James E. Roberts
H. Henrie Henson
Paul F. Csagoly
Charles S. Gloyd

1986 Affiliation
Caltrans
CODOH
FLDOT
WashDOT

May 1987 HSCOBS - A Turning

Point
Findings of NCHRP Project 20-7/31 presented.
Seven options for consideration.
Funding requested to initiate an NCHRP project
to develop a new, modern bridge design
specification.
NCHP Project 12-33 - Development of
Comprehensive Specification and Commentary.
Modjeski and Masters, Inc. began work in July,
1988.

Getting Organized
Editorial Team
Frank Sears
Paul Csagoly
Dennis Mertz
John Kulicki
Code Coordinating Committee
Task Forces Essentially by Section and
Calibration
56 Members Only 1 defector in 5 years

Development Objectives

Technically state-of-the-art specification.

Comprehensive as possible.
Readable and easy to use.
Keep specification-type wording do not
develop a textbook.
Encourage a multi-disciplinary approach
to bridge design.

Not always peace in the valley!

Major Changes

Constraints
Do not allow for further deterioration.
Do not explicitly allow future increase in
truck weights.
No requirement to make bridges uniformly
heavier or lighter.

A new philosophy of safety - LRFD

The identification of four limit states
The relationship of the chosen
reliability level, the load and resistance
factors, and load models through the
process of calibration
new load factors
new resistance factors

Allowable Stress Design

Load Factor Design

Q i R E / FS

i Qi R

where:
Qi =
RE =
FS =

a load
elastic resistance
factor of safety

where:
i =
Qi =
R =
=

a load factor
a load
resistance
a strength reduction factor

Load and Resistance Factor

Design

i i Qi Rn = Rr

in which:
i = D R I 0.95 for loads for max

= 1/(I D R) 1.0 for loads for min

where:
i = load factor: a statistically based multiplier
on force effects
= resistance factor: a statistically based
multiplier applied to nominal resistance

i
D
R
I
Qi

LRFD (Continued)
=
=
=
=
=

Rn =
Rr =

load modifier
a factor relating to ductility
a factor relating to redundancy
a factor relating to importance
nominal force effect: a
deformation stress, or stress
resultant
nominal resistance
factored resistance: Rn

Some Algebra

LRFD - Basic Design Concept

(R -Q) = 2R + Q2
=
R= Q+

Reliability Calcs Done for M and V

Simulated Bridges Based on Real Ones

2R + Q2 = R =

i xi
Q + 2R + Q2

R - Q

2R + Q2
i xi

Reliability of Std Spec vs. LRFD

175 Data Points

25 non-composite steel girder bridge

simulations with spans of 30,60,90,120,and 200
ft, and spacings of 4,6,8,10,and 12 ft.
Composite steel girder bridges having the same
parameters identified above.
P/C I-beam bridges with the same parameters
identified above.
R/C T-beam bridges with spans of 30,60,90,and
120 ft, with spacing as above.

2006 Monte Carlo Reanalysis of

1993 Beta

2006 Monte Carlo Reanalysis of 1993 Beta

4 0r 5 data points each

2006 Monte Carlo Analysis of Beta

for New Bridge Data Base

2006 Monte Carlo Analysis

Monte Carlo Analysis: Beta vs Span Length

Computed Beta Factor

Bridge Database: Beta Factors Using Monte Carlo Analysis

4.5
4.0

3.0
2.5
2.0

124 Bridges

1.5
1.0

CIP Boxes

0.5

50

100

150

0.0
1

11

16

21 26

31 36

41 46

51

56 61

66 71

76

81 86

200

250

300

350

Span Length (FT)

91 96 101 106 111 116 121

Full Set of 124 Database Bridges

Major Changes

2006 Monte Carlo Analysis

Revised calculation of load distribution

Monte Carlo Analysis: D/L Ratio vs. Beta

Computed Beta Factor

Beta Factor

3.5

4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0

0.6

S
g = 0.075 +

2900

4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0.0

1.0

2.0

3.0

4.0

5.0

0.2

S

L

0.1

Kg
3
Lt s

Circa
1990

Dead to Live Load Ratio (Unfactored)

Distribution Factors Revisited (2005)

On-Going Work NCHRP 12-62
y = 1.8817x + 0.0388
2
R = 0.4005

Major Changes (Continued)

Moment in the Interior Girder, 1 Lane Loaded, Location 104.00

AASHTOStandard Distribution Factor

1.4

1.2

0.8

0.6

0.4

0.2

0
0

0.2

0.4

0.6

0.8

Rigorous Distribution Factor

1.2

1.4

y = 0.9729x + 0.1378
2
R = 0.3521

Moment in the Interior Girder, 1 Lane Loaded, Location 104.00

1.4

LRFD Distribution Factor

1.2

Courtesy of
Prof. Jay Puckett

0.8

0.6

0.4

0.2

0
0

0.2

0.4

0.6

0.8

1.2

Rigorous Distribution Factor

Major Changes (Continued)

Changes to the earthquake provisions to
eliminate the seismic performance
category concept by making the method of
analysis a function of the importance of the
structure.
Guidance on the design of segmental
concrete bridges from Guide Spec.
The development of a parallel
commentary.
New Live Load Model HL93
Continuation of a long story

1.4

Combine plain, reinforced and prestressed

concrete.
Modified compression field/strut and tie.
Limit state-based provisions for foundation
design.
Expanded coverage on hydraulics and scour.
The introduction of the isotropic deck design.
Expanded coverage on bridge rails.
Inclusion of large portions of the
AASHTO/FHWA Specification for ship collision.

1912 Article Published in

Transactions of ASCE,
Henry B. Seaman
It would thus seem that 80 lb/sf would
be a maximum load, if indeed it should
not be much less, for long spans.

Bridge Engineering, Published in

1916, J.A. Waddell

L.R. Manville and R.W. Gastmeyer,

Engineering News, September 1914

Waddell discusses the source of distributed

load used in the design of bridges:

The customary loading assumed for the design of

highway bridges in the past has been a certain
uniform live load alone, possibly a typical heavy
wagon or road-roller, or a uniform live load with a
concentration.
But these older types of loading are inadequate
for purposes of design to take care of modern
conditions; they should be replaced by some
types of typical motor trucks.

Some people have the idea that a herd of

cattle will weigh more per square foot than a
crowd of people, but such is not the case, as
the actual limit for the former is about 60 lb/ft2.

L.R. Manville and R.W. Gastmeyer,

Engineering News, September 1914
4-Ton
10-Ton
12-Ton
14-Ton
17-Ton
20-Ton

3k
4k
6k
4k
14k
12k

5k
16k
18k
24k
20k
28k

8'
12'
12'
12'
12'
12'

5'
6'
6'
6'
6'
6'

1923 AREA Specification

10-Ton
15-Ton
20-Ton

4k
6k
8k

16k
24k
32k
5.5

14'

'

VERY CLOSE!!

1923 AREA Specification

8k

32k
14'

19'

8k

32k
14'

19'

8k

1924 AREA Specification

32k
14' 15'

8k
15'

20-Ton

20-Ton

20-Ton

800 lb per ft

Shoemakers Truck Train and

Equivalent Load (Bold Added)-1923
The 7.5-ton [capacity] truck, which weighs
about 15 tons when loaded to capacity and
which can be overloaded to weigh about
20 tons, is the heaviest commonly used.
A large part of traffic, however, is carried in
trucks of 5 tons capacity or less. It would
not appear to be necessary, therefore, to
provide for a succession of 20-ton loads.

800 lb per ft

32k
14'

8k
19'

20-Ton

32k
14'

20-Ton

8k
19'

32k
14' 15'

20-Ton

800 lb per ft

Shoemakers Truck Train and

Equivalent Load - 1923
6k 24k
14'
30'

6k 24k
14'
30'

8k 32k
14'
30'

6k 24k
14'
30'

6k 24k
14'

15-Ton

15-Ton

20-Ton

15-Ton

15-Ton

28,000 lb
600 lb/ft

By 1929 lane load becomes what we have today

1928-1929 Conference
Specification
6k 24k
14'
15-Ton

30'

6k 24k
14'
30'
15-Ton

8k 32k
14'
30'
20-Ton

6k 24k
14'
30'
15-Ton

6k 24k
14'

1941 AASHTO----HS20
(Almost)
8k

32k

32k
14 '

14 '

15-Ton

H20 - S16
18,000 lb for Moment
26,000 lb for Shear

640 lb/ft

1941-1944 Rebellion & Chaos!!

Much disagreement over HS Loading
An Analysis of Highway Loads Based on
the Special Loadometer Study of 1942 by
Dr. A.A. Jakkulka
Recommended HS20 Truck Because it
was the more common stress producing
truck on the road

1944 Agreement
No HS Lane Load---use H20 Lane Load
Variable axle spacing adopted more
closely approximates the tractor tailors
now in use
HS20-S16-44..44 added to reduce
confusion from so many changes

640 lb/ft

32,000 lb for Moment

40,000 lb for Shear

1941-1944 Rebellion & Chaos!!

At the December 1944 Bridge Committee
Group Meeting, a progress report on a truck
loading study conducted at Texas A and M
College was presented. The minutes of the
meeting state that:
the discussion that followed .soon
developed into a free for all over them
good old fighting words what design loading
should be used. After the meeting got down to
normalcy again, Mr. Paxon presented

Live Load Continued to be

Debated
Early 1950s Discussion to remove the
lane load as too heavy and wasteful for
continuous spans
Throughout 50s there are discussions
about increasing the design truck
1958 Decision to do nothing until after
AASHO Road Tests are completed.

Live Load Continued to be

Debated

Exclusion Loads Based on TRB

Special Report 225, 1990

Late 60s H40, HS25 and HS30 discussed

1969 SCOBS states unanimous opposition to
increasing weight of design truck wasteful
obsolescence of existing bridges
1978 HS25 proposed again
1979 HS25 again commentary
need for heavier design load seems unavoidable
HS25 best present solution
5% cost penalty
Motion soundly defeated

Selected Notional Design Load

EXCL/HS20 Truck or Lane or 2

110 kN Axles @ 1.2 m

HL-93

EXCL/HL 93 Circa 1992

BIN

!!
O
G

NCHRP 12-33 Project Schedule

First Draft - 1990 general coverage

Second Draft - 1991 workable
Third Draft - 1992 pretty close
Two sets of trial designs - 1991 and 1992
Fourth Draft - 1993 ADOPTED!!
12,000 comments
Reviewed by hundreds
Printed and available - 1994

Implementation (Continued)

Implementation Starts Slowly

Lack of software.
Early lack of training but several thousand have
taken NHI courses with more to come.
Perceived difficulties
Load distribution
Shear in concrete
Foundations
Load cases seemed numerous but that may
be because they are all stated
Continual changes more later
Similar story with EUROCode plus national
issues.

Down size, right size, capsize.

To SI or not to SI? Thats the question.
But things are moving, especially
compared to other major changes.
Federal deadline: 2007.
By 2007:
5,000 LRFD bridges
More than half of states doing part or all
LRFD

Upgrades and Changes to 1990

Technology

First LRFD Major Bridge Opened

1997

1996 foundation data reinserted.

New wall provisions ongoing upgrade.
2002 upgraded to ASBI LFRD Segmental Guide
Specs.
MCF shear in concrete simplified and clarified
several times major update in 2002.
Load distribution application limits expanded
several time in 1990s due to requests to liberalize.
More commentary added.

Upgrades and Changes

2004 major change in steel girder design
in anticipation of
2005 seamless integration of curved
steel bridges ending three decade quest

Curved Girder Leaders

Dr. Bill Wright

Dr. Don White
Mr. Mike Grubb
Dr. Dennis Mertz
Mr. Ed Wasserman

HSCOBS Asserts Ownership

Upgrades and Changes (Continued)

2005 P/C loses updated
2006 complete replacement of Section 10
Foundation Design
2006 more concrete shear options
2007 - big year
Streamline MCF for concrete shear design
1,000 year EQ maps and collateral changes
Seismic Guide Spec - displacement based
Pile construction update
2008 - Coastal bridge Guide Spec

LRFD Oversight Committee Circa 2002

The mission is to promote LRFD as the

national standard for bridge design and develop a
strategic plan to successfully implement LRFD by
2007 for all new bridge designs.
to develop a strategic plan to identify and
prioritize educational and training needs.

Future as Seen in 1993 Continued

Development

Where Do We Go From Here?

Calibration of Service Limit State

Deformation, cracking, service stress limits.
What quantitative criteria can be established?
What is the structural penalty for violating a non-strength
limit state?
How often can the limit state be exceeded in the design
life?
What is an appropriate reliability index?
What is the appropriate loading in terms of magnitude,
configuration, and placement? How does this relate to
multiple presence factors?
Should permit loads and illegal loads be considered?
Will SHRP 2 and NCHRP 12-83 do it??

Quantifying Redundancy.
Expanded database of loads, etc.
Refinement of foundation provisions.
Simplification of load distribution.
Improvements in reliability procedures.
Joint probability procedures
LL with EQ?
Ship and scour?
EQ and scour?
Ice and wind?
Etc.

Other Limit States??

Does current design address the real culprits?

Where are owners spending maintenance $$?
Do we know the impact of changes?

Will FHWA LTBP Tell Us??

10

Rehabilitation
Applying new standards to existing bridges
has always been a challenge.
Are other limit states or load combinations
or reliability targets appropriate for rehab?
Do we need and Application Manual for
rehab?

Quantification of Redundancy
2005 T-5 commits to work with
results of:
NCHRP 406 redundancy of super
NCHRP 458 redundancy of sub
Goals:
Multiplier table for routine girder bridges.
Process for evaluating more complex
bridges for a reliability index in damaged
state.

Fatigue and Fracture

Bridge Security
Per 2003 BRC recommendations, T1
formed several years ago
Much research ongoing
ASCE Committee on Bridge Security
formed James Ray, Chair
First fledgling steps towards specifications
on ballot NCHRP 12-72

Joint Probability of Occurrence

2005 T-5 also commits to continued review

of:
FHWA Synthesis Report on Extreme
Loading Combinations by Nowak, Knott and
Dumas, August, 1996
NCHRP 489 extreme events, 1999
2005 T-5 presentation by Sue Hida on
CALTrans in-house study of joint probability of
scour and EQ-----non-issue.
Focus shifting to all hazard approach.

Perfection is Still an Illusive Goal

Should new load histograms be obtained?

Traffic changes after 1970s oil embargo
Increases in legal loads
CBs, etc.
Load bandwidth increase

Having said that still seeing little load

induced damage
Have we given up on F and F Spec
changes for HPS?

But Improvement is Possible and

Demanded by Society

11

Summary
The object was to switch to a more robust, more
expandable, more adaptable platform---------like
Windows vs. DOS.
As with the switch to Windows, there were some
transitional learning curves and headaches----but
many developers can see benefits, users can see
the logic.
It is unrealistic to expect the LRFD Specs to become
static-----researches will always have new ideas,
nature will continue to teach us lessons.
But LRFD was intended to adapt and grow!

Net Effect So Far

04
06

09
94

Accessory For 5th Edition???

Thank You
And A Special Thank You
To All Who Helped Over The
Last Two Decades!!

But Some Must Be Mentioned

NCHRP Ian Friedland, Scott Sabol,
Dave Beal
SCOBS Bob Cassano, Clellon Loveall,
Jim Siebels, Dave Pope, Mal Kerley
Panel Jim Roberts, Chairman
AASHTO Kelley Rehm, Ken Kobetsky
Modjeski and Masters Dennis Mertz,
Wagdy Wassef, Diane Long

12

First ProblemDefining Failure

Part 2
Bridge Failures and
Design Specifications

First ProblemDefining Failure

Collapse event easy to identify and agree on
cause is something else
Inability to serve intended function
Lack of sufficient service strength
Misalignment

Unsightly defect
Cracks
Misalignment
Discoloration

Disproportionate future maintenance

Shortened service life

Sunshine Skyway 1980

The easy definition:

Any unplanned occurrence in construction
or service
Now lets get realistic

How Does Profession React?

Research
Additions to design spec to design avoidance
Changes to material specs
Changes to fabrication or construction specs
Add to non-spec body of knowledge
Operational changes
Policy changes
Retrofits
Mixtures

Lets look at some examples!!

Sunshine Skyway
Vessel hit side span pier which was
not protected
Response design
1994 Guide Specification
Written and adopted
Partly incorporated into
AASHTO LRFD 1st Edition

Webber Falls

Sunshine Skyway
Basic parameters
AF = (N ) ( P A) ( P G ) ( PC )
P A = ( B R ) ( R B ) ( R C ) ( R XC ) ( R D )

Human factors and serendipity

hard to eliminate

Webber Falls

Webber Falls

Pier 1
Pier 4

Pier 3

Pier 2

Pier 1

MM-60

MM-62
West Abutment

Pier 4

Pier 3

Pier 2

Pier 1

Webber Falls

Webber Falls - Retrofit

Special structural features

Barge reversed-back exerts more load
Corner hit-less energy absorption
Hit weakest pier

Reactions
Ops and Policy-Recommendation for alarms if
active controls not sense in some period of time
Retrofit-similar bridges got pier protection added

2009 - Now There Is A Choice

Silver Bridge

Silver Bridge

Silver Bridge

Reaction

Reaction (cont.)

Fracture Control Plan

Materials-fracture toughness
Fabrication-welder quals and testing
Documentation
Weld repairs
Design Specifications
Identification of FCM and tension components
Toughness requirements identified but not used
in design calcs

I-794 Hoan Bridge

Policies
NBIS with special requirements for FCMs
Redundancy stressed
Permit numerical demonstration of redundancy
Retrofits
Sister bridge demolished and replaced
Some redundancy enhancement

I-794 Hoan Bridge

(December 2000)

I-794 Hoan Bridge

Welds Touching or Overlapping
in Web Gap
Web Plate

Gusset Plate

I-794 Hoan Bridge

All fractures were brittle (cleavage) initiation
at the intersecting welds
No fatigue crack growth was detected in the
web at any shelf plate detail
The shelf plate lack of direct connection to the
transverse connection plate created a large
geometric crack like condition
Failure analysis showed that the crack like
geometry and the high triaxiality resulted in
fracture on the upper shelf

Stress Concentration + Triaxial Constraint = HIGH RISK

I-794 Hoan Bridge

Static Wind - Tay Bridge

Response
FHWA memorandum cites two criteria
that can indicate fracture vulnerability
1) Intersecting / overlapping welds
2) Evidence of rapid crack growth

Body of knowledge-detailing guidance

RULE: Intersecting weld toes must
have at least of clear separation toe-totoe to allow relief of constraint

Dynamic Wind - First Tacoma

Narrows

Wind - Response
Research - Identify quasi-static pressure
and understand dynamic phenomena
Design Spec static wind pressure and
overturning line load
Body of knowledge on dynamic actions
Section models
Aeroelastic models
Terrain models
Computational methods
A variety of potential fixes

Section Model

Japanese Aeroelastic Tests

Wind/Rain - Cables

What Causes Cable Vibration?

Classic analysis says that a steady wind
will impose a steady force/constant
deflection of the cable. However,
General experience - cable stay vibrations build up
to violent movements under the combined effect of
rain and wind, both being light to moderate in their
intensities.
Some bridges show excessive vibration with
no rain, possibly due to vortex shedding.
There are other various phenomena
which may also be responsible.

Reducing Cable Movement

Various methods used to
control cable movement
Use of energy absorbing pads
between cables and end of
anchor pipes.
Cross-tying the cable together.
Modifying the cable shape to
change the aerodynamic
conditions.
Installation of energy absorbers
(or dampers) on the cables to
reduce the cable vibration.

Cable Damper Retrofit

Cable dampers are double-acting,
linear, fluid viscous mechanical devices:
Dampers mounted to the cables in the tower head.
Steel collar at cable attachment point.
Steel bracket assembly attaches other end of
damper to stationary steel anchor pipe.

NEW U-BOLT

CL NEW
CABLE COLLAR
ANCHOR PIPE
NEW DAMPER
BRACKET ASSEMBLY

NEW CABLE
DAMPER
TOWER CORE
CL STAY CABLE

Fatigue

Fatigue
Yellow Millpond Bridge

Fatigue

Response research on resistance and

loads
Much added to body of knowledge

Distortion Fatigue

Response Design Spec

CAFT (Typ.)

( f ) ( F )n

( F )n =

N = ( 365)( 75) n ( ADTT ) SL

(F )n = (F )TH

A 3

Distortion Fatigue

Earthquakes - San Fernando 1971

Response
Research
Body of knowledge from case studies
Specification verboten details
No quantification in spec so far

San Fernando 1971

San Fernando 1971

Response
Research
Lab tests, particularly on column cyclic behavior

San Fernando 1971

San Fernando 1971

Response (cont.)

Response (cont.)

Capitalize on building research ATC 2, NEHRP

Shear failures
Bond and development

Design Specifications
ATC 6 Document lead to
Div. I-A of Std Spec
SSPC
Design Spectrum
Site factors
R factors

Csm =

1.2 AS
2.5 A
Tm 2 / 3

Soil Profile Type

Site
Coefficient

II

III

IV

1.0

1.2

1.5

2.0

Northridge 1994

San Fernando 1971

Response (cont.)

Design Specifications
Methods of analysis
Seat widths
Confinement
Plastic hinging
Bond and development
Multispan Bridges
Other Bridges
Seismic
Zone

Single-Span
Bridges

1
2
3
4

No seismic
analysis
required

Essential Bridges

Critical Bridges

regular

irregular

regular

irregular

regular

irregular
*

SM/UL

SM

SM/UL

MM

MM

MM

SM/UL

MM

MM

MM

MM

TH

SM/UL

MM

MM

MM

TH

TH

Northridge 1994

Recent Observations
Apparent good behavior
of retrofits
Column Wrapping
Longitudinal Restrainers
Isolation

Improvements to Seismic Spec

2007 AASHTO changes to LRFD Spec
About 1000 yr event
New Maps-PGS and
Peak horiz spectral
response acceleration
coefficient

2009 - Now There Is A Choice

Improvements to Seismic Spec

2007 AASHTO Changes to LRFD Spec
New response spectrum
Revised site factors
Uniform load method
More on P
New provisions for
columns, foundations
and much more

Coastal Storms

New Guide Spec-Displacement based

TASK ORDER DTFH61-06-T-70006 (2006)

DEVELOPMENT OF GUIDE SPECIFICATIONS
AND HANDBOOK OF RETROFIT OPTIONS
FOR BRIDGES VULNERABLE TO COASTAL
STORMS
Bay Saint Louis, MS

Pensacola, FL

Biloxi, MS

Biloxi, MS

Wave Tank Studies

Past Reactionary Cycle Has Often

Been Iterative
Event

Application

Research

Research

Event

Application

Modjeski and Masters, Inc.

Moffatt & Nichol, Inc.
Ocean Engineering Associates, Inc.
Dennis R. Mertz
DAppolonia

Status Report in 2007 General Session

2009 Coastal Guide Specification

Is There A Better Way?

Can we make a summary of past events
more accessible?
HDR FHWA report
On line summaries like crash tested railings?

Can we be proactive?
What iffing by think tanks?
Would this uncover yet unseen phenomena?
Can we deduce behavior w/o experiments?
Can we calibrate design or analysis model w/o
experiential data?

Is There A Better Way?

The problem:
So far nature has been asking the
questions and we have been trying to find
answers.
For proactive response we would have to
know the fundamental questions!

Difficult ------ but maybe we should at

least try to get ahead of the need some
how.

A Modest Proposal
Maybe it would be worth a trial
workshop to see what might evolve!
High level, innovative thinkers
Not just bridge people
Meet several times give time to think

Objective: What big thing are we

missing?
We have spent money on more
speculative ideas!!

Lets at least try to get ahead

of events!!
Thank you
Some graphics provided by:
Dr. J. Fisher
Mr. D. Barrett
Dr. D. Mertz
Mr. J. OConner
Dr. R. Connor
Dr. Z. Savor
Dr. Z Prucz
Dr. D. Sheppard

10