The National Academies Press: Bridges For Service Life Beyond 100 Years: Service Limit State Design (2014)

THE NATIONAL ACADEMIES PRESS
This PDF is available at http://nap.edu/22441 SHARE

   
Bridges for Service Life Beyond 100 Years: Service Limit

State Design (2014)
DETAILS
0 pages | 8.5 x 11 | PAPERBACK

ISBN 978-0-309-43375-4 | DOI 10.17226/22441
CONTRIBUTORS
GET THIS BOOK Modjeski and Masters, Inc.
FIND RELATED TITLES
SUGGESTED CITATION
National Academies of Sciences, Engineering, and Medicine 2014. Bridges for

Service Life Beyond 100 Years: Service Limit State Design. Washington, DC: The
National Academies Press. https://doi.org/10.17226/22441.

Visit the National Academies Press at NAP.edu and login or register to get:
– Access to free PDF downloads of thousands of scientific reports

– 10% off the price of print titles
– Email or social media notifications of new titles related to your interests
– Special offers and discounts

Distribution, posting, or copying of this PDF is strictly prohibited without written permission of the National Academies Press.
(Request Permission) Unless otherwise indicated, all materials in this PDF are copyrighted by the National Academy of Sciences.
Copyright © National Academy of Sciences. All rights reserved.

Bridges for Service Life Beyond 100 Years: Service Limit State Design
The Second
S T R A T E G I C H I G H W A Y R E S E A R C H P R O G R A M
REPORT S2-R19B-RW-1
Bridges for Service Life Beyond 100 Years:

Service Limit State Design
Modjeski and Masters, Inc.
with
University of Nebraska, Lincoln
University of Delaware
NCS Consultants, LLC
T R A N S P O R TAT I O N R E S E A R C H B O A R D
WASHINGTON, D.C.
2015
www.TRB.org
Copyright National Academy of Sciences. All rights reserved.

Subject Areas
Bridges and Other Structures
Highways
Maintenance and Preservation Materials
Figure 2.1 Credit

With the permission of the Canadian Standards Association (operating as CSA Group), material is repro-
duced from CSA Group standard CAN/CSA S6-06, “Canadian Highway Bridge Design Code,” which is
copyrighted by CSA Group, 5060 Spectrum Way, Suite 100, Mississauga ON, L4W 5N6. This material is
not the complete and official position of CSA Group on the referenced subject, which is represented solely
by the standard in its entirety. While use of the material has been authorized, CSA is not responsible for
the manner in which the data is presented, nor for any interpretations thereof. For more information or to
purchase standards from CSA Group, please visit http://shop.csa.ca/ or call 1-800-463-6727.

The Second Strategic Highway SHRP 2 Report S2-R19B-RW-1

Research Program
ISBN: 978-0-309-27375-6
America’s highway system is critical to meeting the mobility and
economic needs of local communities, regions, and the nation. © 2015 National Academy of Sciences. All rights reserved.
Developments in research and technology—such as advanced
materials, communications technology, new data collection tech-
nologies, and human factors science—offer a new opportunity Copyright Information
to improve the safety and reliability of this important national
resource. Breakthrough resolution of significant transportation Authors herein are responsible for the authenticity of their materials and for
obtaining written permissions from publishers or persons who own the copy-
problems, however, requires concentrated resources over a short right to any previously published or copyrighted material used herein.
time frame. Reflecting this need, the second Strategic Highway The second Strategic Highway Research Program grants permission to repro-
Research Program (SHRP 2) has an intense, large-scale focus, duce material in this publication for classroom and not-for-profit purposes.
Permission is given with the understanding that none of the material will be
integrates multiple fields of research and technology, and is used to imply TRB, AASHTO, or FHWA endorsement of a particular prod-
fundamentally different from the broad, mission-oriented, uct, method, or practice. It is expected that those reproducing material in this
document for educational and not-for-profit purposes will give appropriate
discipline-based research programs that have been the mainstay acknowledgment of the source of any reprinted or reproduced material. For
of the highway research industry for half a century. other uses of the material, request permission from SHRP 2.
Note: SHRP 2 report numbers convey the program, focus area, project number,
The need for SHRP 2 was identified in TRB Special Report 260: and publication format. Report numbers ending in “w” are published as web
documents only.
Strategic Highway Research: Saving Lives, Reducing Congestion,
Improving Quality of Life, published in 2001 and based on a
study sponsored by Congress through the Transportation Equity Notice
Act for the 21st Century (TEA-21). SHRP 2, modeled after the
The project that is the subject of this report was a part of the second Strategic
first Strategic Highway Research Program, is a focused, time- Highway Research Program, conducted by the Transportation Research Board
constrained, management-driven program designed to com- with the approval of the Governing Board of the National Research Council.
plement existing highway research programs. SHRP 2 focuses The members of the technical committee selected to monitor this project and
to review this report were chosen for their special competencies and with regard
on applied research in four areas: Safety, to prevent or reduce the for appropriate balance. The report was reviewed by the technical committee
severity of highway crashes by understanding driver behavior; and accepted for publication according to procedures established and overseen
Renewal, to address the aging infrastructure through rapid design by the Transportation Research Board and approved by the Governing Board of
the National Research Council.
and construction methods that cause minimal disruptions and
The opinions and conclusions expressed or implied in this report are those of
produce lasting facilities; Reliability, to reduce congestion through the researchers who performed the research and are not necessarily those of the
incident reduction, management, response, and mitigation; and Transportation Research Board, the National Research Council, or the program
sponsors.
Capacity, to integrate mobility, economic, environmental, and
The Transportation Research Board of the National Academies, the National
community needs in the planning and designing of new trans- Research Council, and the sponsors of the second Strategic Highway Research
portation capacity. Program do not endorse products or manufacturers. Trade or manufacturers’
names appear herein solely because they are considered essential to the object
of the report.
SHRP 2 was authorized in August 2005 as part of the Safe,
Accountable, Flexible, Efficient Transportation Equity Act: A
Legacy for Users (SAFETEA-LU). The program is managed by the
Transportation Research Board (TRB) on behalf of the National
Research Council (NRC). SHRP 2 is conducted under a memo-
randum of understanding among the American Association of
State Highway and Transportation Officials (AASHTO), the
Federal Highway Administration (FHWA), and the National
Academy of Sciences, parent organization of TRB and NRC.
The program provides for competitive, merit-based selection SHRP 2 Reports
of research contractors; independent research project oversight; Available by subscription and through the TRB online bookstore:
and dissemination of research results. www.mytrb.org/store
Contact the TRB Business Office:
202-334-3213
More information about SHRP 2:

www.TRB.org/SHRP2

The National Academy of Sciences is a private, nonprofit, self-perpetuating society of distinguished scholars
engaged in scientific and engineering research, dedicated to the furtherance of science and technology and
to their use for the general welfare. On the authority of the charter granted to it by Congress in 1863, the
Academy has a mandate that requires it to advise the federal government on scientific and technical matters.
Dr. Ralph J. Cicerone is president of the National Academy of Sciences.
The National Academy of Engineering was established in 1964, under the charter of the National Academy
of Sciences, as a parallel organization of outstanding engineers. It is autonomous in its administration and
in the selection of its members, sharing with the National Academy of Sciences the responsibility for
advising the federal government. The National Academy of Engineering also sponsors engineering programs
aimed at meeting national needs, encourages education and research, and recognizes the superior achieve-
ments of engineers. Dr. C. D. (Dan) Mote, Jr., is president of the National Academy of Engineering.
The Institute of Medicine was established in 1970 by the National Academy of Sciences to secure the
services of eminent members of appropriate professions in the examination of policy matters pertaining
to the health of the public. The Institute acts under the responsibility given to the National Academy of
Sciences by its congressional charter to be an adviser to the federal government and, on its own initiative,
to identify issues of medical care, research, and education. Dr. Victor J. Dzau is president of the Institute
of Medicine.
The National Research Council was organized by the National Academy of Sciences in 1916 to associate
the broad community of science and technology with the Academy’s purposes of furthering knowledge and
advising the federal government. Functioning in accordance with general policies determined by the
Academy, the Council has become the principal operating agency of both the National Academy of
Sciences and the National Academy of Engineering in providing services to the government, the public, and
the scientific and engineering communities. The Council is administered jointly by both Academies and
the Institute of Medicine. Dr. Ralph J. Cicerone and Dr. C. D. (Dan) Mote, Jr., are chair and vice chair,
respectively, of the National Research Council.
The Transportation Research Board is one of six major divisions of the National Research Council. The
mission of the Transportation Research Board is to provide leadership in transportation innovation and
progress through research and information exchange, conducted within a setting that is objective, interdisci-
plinary, and multimodal. The Board’s varied activities annually engage about 7,000 engineers, scientists, and
other transportation researchers and practitioners from the public and private sectors and academia, all of
whom contribute their expertise in the public interest. The program is supported by state transportation
departments, federal agencies including the component administrations of the U.S. Department of Transporta-
tion, and other organizations and individuals interested in the development of transportation. www.TRB.org
www.national-academies.org

SHRP 2 STAFF
Ann M. Brach, Director

Stephen J. Andrle, Deputy Director
Cynthia Allen, Editor
Kenneth Campbell, Chief Program Officer, Safety
Jared Cazel, Editorial Assistant
JoAnn Coleman, Senior Program Assistant, Capacity and Reliability
Eduardo Cusicanqui, Financial Officer
Richard Deering, Special Consultant, Safety Data Phase 1 Planning
Shantia Douglas, Senior Financial Assistant
Charles Fay, Senior Program Officer, Safety
Carol Ford, Senior Program Assistant, Renewal and Safety
James Hedlund, Special Consultant, Safety Coordination
Alyssa Hernandez, Reports Coordinator
Ralph Hessian, Special Consultant, Capacity and Reliability
Andy Horosko, Special Consultant, Safety Field Data Collection
William Hyman, Senior Program Officer, Reliability
Linda Mason, Communications Officer
David Plazak, Senior Program Officer, Capacity and Reliability
Rachel Taylor, Senior Editorial Assistant
Dean Trackman, Managing Editor
Connie Woldu, Administrative Coordinator
ACKNOWLEDGMENTS
This work was sponsored by the Federal Highway Administration in cooperation with the American Asso-
ciation of State Highway and Transportation Officials. It was conducted in the second Strategic Highway
Research Program (SHRP 2), which is administered by the Transportation Research Board of the National
Academies. At various times the project was managed by Monica Starnes, Mark Bush, and Jerry DiMaggio.
Because of the close relationship between this project and NCHRP Project 12-83, information and report
sections were freely exchanged between the two projects. NCHRP 12-83 is being managed by Waseem
Dekelbab, and the principal investigator is Wagdy G. Wassef of Modjeski and Masters, Inc.
The research described in this report was performed by Modjeski and Masters, Inc., supported by
the University of Nebraska, Lincoln; the University of Delaware; and NCS Consultants, LLC. John M.
Kulicki of Modjeski and Masters, Inc., was the principal investigator. Other authors are Wagdy G. Wassef of
Modjeski and Masters, Inc.; Andrzej S. Nowak of the University of Nebraska, Lincoln; Dennis R. Mertz
of the University of Delaware; Naresh C. Samtani of NCS Consultants; and through his participation in
NCHRP Project 12-83, Hani Nassif of Rutgers University. The following graduate students contributed
to this project at different times: Marek Kozikowski, Przemyslaw Rakoczy, Krzysztof Waszczuk, and
Anna Maria Rakoczy of the University of Nebraska; Dustin M. Schopen and Benjamin Berwick of the
University of Delaware; and Dan Su of Rutgers University.

FOREWORD
Jerry A. DiMaggio, D.GE, PE, SHRP 2 Senior Program Officer, Renewal
This report, Bridges for Service Life Beyond 100 Years: Service Limit State Design, describes
research, outcomes, and products on the basis of the R19B project objectives. The objectives
were to develop design and detailing guidance and calibrated service limit states (SLSs) to
provide 100-year bridge life and to develop a framework for further development of cali-
brated SLSs. The products of this study are expected to be directly usable by the American
Association of State Highway and Transportation Officials (AASHTO) and departments of
transportation (DOTs) and include
• Provisions needed to implement SLSs and the associated load and resistance factors
necessary to produce calibrated bridge components and systems expected to have a
predictable service life.
• Detailed design and detailing provisions required to design and construct the calibrated
component or system.
• Databases used in the calibration, as well as instructions for a calibration spreadsheet, for
use by DOTs to track and adjust service-based reliability with time.
Consideration of SLSs requires different input data from the previously calibrated Strength
Limit State I (also known as “ultimate or strength limit states,” or ULSs). In ULSs, the limit
state function is defined by resistance, which is considered constant in time, and loads. For
SLSs, a different approach is needed because
• Exceeding a service limit state does not lead to a clear, immediate loss of functionality.
• Acceptable performance can be subjective (full life-cycle analysis is required).
• Resistance and load effects can be and often are correlated.
• Load must be considered to be a function of time, described by magnitude and frequency
of occurrence.
• Resistance may be strongly affected by quality of workmanship, operation procedures,
and maintenance.
• Resistance is subject to changes in time, mostly but not only by deterioration.
• Resistance can depend on geographical location (e.g., climate, exposure to industrial pollu-
tion, or deicing agents).
The topic of limit state design, also known as load resistance factor design (LRFD), within the
United States has been under development and implementation for more than 25 years. The
benefits of this design platform are now well understood by the bridge and structures com-
munity as well as by transportation decision makers. Generally, it has been assumed that main-
tenance activities will be sufficient to prevent significant loss of the strength and stiffness that
would result in unsatisfactory service level performance. It has been recognized that advance-
ments and further maturity of the LRFD platform need to focus on quantification and calibra-
tions of the SLSs. Although previous work has been published in this area, the R19B study serves
as a foundational reference to partially fill knowledge gaps and, perhaps more importantly, for
direct application and reference for future study in this emerging technical area of design.

CONTENTS
1 Executive Summary
4 CHAP T ER 1 Purpose of Report and Relation to Scope
4 1.1 Objectives of Project R19B
5 1.2 Scope
6 1.3 Research Team
7 1.4 Relationship of Project R19B to Project R19A
7 1.5 Relationship of Project R19B to NCHRP Project 12-83
7 1.6 Special Challenges Related to SLSs
9 1.7 Serviceability Versus Deterioration
9 1.8 Durability
12 1.9 Initial Coordination with FHWA Long-Term Bridge Performance Program
13 1.10 Dialogue with AASHTO HSCOBS and Others
14 CHAP T ER 2 Current State of the Art
14 2.1 Approach
14 2.2 Summary of Literature Survey
18 2.3 Serviceability Requirements in Several Modern Bridge Design Specifications
44 2.4 Surveys of Current Practice
47 2.5 SLSs to Be Considered in This Report
48 CHAP T ER 3 Overview of Calibration Process
48 3.1 Introduction
49 3.2 Calibration by Determination of Reliability Indices
54 3.3 “Deemed to Satisfy”
55 3.4 Customizing the Process
56 CHAP T ER 4 Deterioration
56 4.1 Introduction
56 4.2 Bolukbasi et al. (2004)
60 4.3 Jiang and Sinha (1989)
62 4.4 Hatami and Morcous (2011)
62 4.5 Comparison of Equations from Bolukbasi et al. (2004), Jiang and Sinha
(1989), and Hatami and Morcous (2011)
69 4.6 Agrawal and Kawaguchi (2009)
78 4.7 Stukhart et al. (1991)
88 4.8 Massachusetts DOT
94 CHAP T ER 5 Live Load for Calibration
94 5.1 Development of Live Load Models for Service Limit States
99 5.2 Initial Data Analysis
111 5.3 Statistical Parameters for Service Limit States Other than Fatigue

121 5.4 Development of Statistical Parameters of Fatigue Load

143 5.5 Development of Overload (Service II) Parameters
145 C H AP TER 6 Calibration Results
145 6.1 Foundation Deformations, Service I: Lifetime
171 6.2 Cracking of Reinforced Concrete Components, Service I Limit State:
Annual Probability
176 6.3 Live Load Deflections, Service I: Annual Probability
180 6.4 Overload, Service II: Annual Probability
188 6.5 Tension in Prestressed Concrete Beams, Service III Limit State: Annual
Probability
208 6.6 Fatigue Limit States: Lifetime
217 C H AP TER 7 Proposed Changes to AASHTO LRFD
218 7.1 Foundation Deformations—Service I
233 7.2 Live Load Response
238 7.3 Premature Yielding and Slip of Bolts—Service II
240 7.4 Cracking of Prestressed Concrete—Currently Service III
243 7.5 Fatigue
260 C H AP TER 8 Purpose and Contents of Appendix F
261 C H AP TER 9 Summary and Recommendations
261 9.1 Summary
262 9.2 Recommendations
262 9.3 Implementation
263 References
269 Appendix A. SLS Requirements in the Eurocode
292 Appendix B. SHRP 2 R19B Survey of Bridge Owners
304 Appendix C. Comparison of Crack Width Prediction Equations for Prestressed
Concrete Members
310 Appendix D. Derivation of the Resistance Prediction Equation of Prestressed
Concrete Bridge Girders
313 Appendix E. Normal Probability Plots of Fatigue Data for the Various Detail
Categories
322 Appendix F. Data Used for Calibration

Executive Summary
The objectives of SHRP 2 Project R19B were to develop design and detailing guidance and cali-
brated service limit states (SLSs) to provide 100-year bridge life and to develop a framework for
further development of calibrated SLSs. Generally, it has been assumed that maintenance activi-
ties will be sufficient to prevent significant loss of the strength and stiffness that would result in
unsatisfactory service-level performance.
Consideration of SLSs requires different input data from the previously calibrated Strength
Limit State I (or “ultimate or strength limit states,” ULSs). In ULSs, the limit state function is
defined with two variables: resistance, which was considered constant in time, and loads. For SLSs,
a different approach is needed because of the following factors:
• Exceeding SLSs does not lead to a clear, immediate loss of functionality, so defining the resis-
tance is very subjective.
• Acceptable performance can be subjective (full life-cycle analysis is required).
• Resistance and load effects can be and often are correlated.
• Load must be considered to be a function of time, described by magnitude and frequency of
occurrence.
• Resistance may be strongly affected by quality of workmanship, operation procedures, and
maintenance.
• Resistance is subject to changes in time, mostly but not only deterioration, with difficulty
predicting initiation time and time-varying rates of deterioration (e.g., corrosion, accumula-
tion of debris, cracking).
• Resistance can depend on geographical location (e.g., climate, exposure to industrial pollu-
tion, exposure to deicing agents, proximity to the ocean).
On the basis of a survey of owners and a literature review that included other national and
international bridge design specifications, a set of possible SLSs was developed. Those SLSs were
reviewed to determine which could be calibrated using reliability theory. Calibrated, reliability-
based load factors or resistance factors, or both, were developed for
• Foundation deformations;
• Cracking of reinforced concrete components;
• Live load deflections;
• Permanent deformations;
• Cracking of prestressed concrete components; and
• Fatigue of steel and reinforced concrete components.
The calibration process produced target reliability levels much different from those used for
the strength calibration. This outcome was expected because, in general, the consequences of

exceeding SLSs are an order, or even several orders, of magnitude smaller than those associated with
ULSs. Thus, an acceptable probability of exceeding an SLS is much higher than for a ULS. Although
the strength calibration was based on a target reliability index of about 3.5 for a 75-year life, with
some exceptions, most of the SLS calibration was generally done with a target reliability index
around 1.0 to 1.5 on the basis of an annual probability. Once the target reliabilities were deter-
mined, changes to load factors, resistance factors, or other design parameters were developed.
Extensive use was made of WIM (weigh in motion) collected at 32 sites around the country.
The raw data consisted of 65 million vehicle records consisting of axle weights, axle spacing,
speed, and vehicle classification. After filtering, about 35 million records formed the database
from which live load biases and coefficients of variation (CVs) were computed for a variety of
span lengths, average daily truck traffic, and time periods. To acknowledge the assumption that
the 32 sites were representative of the whole country, project live load biases were set at the mean
plus 1.5 standard deviations.
WIM data processing for fatigue consisted of passing the complete string of vehicles over influ-
ence lines for simple spans and two-span continuous units of various spans. The resulting histo-
ries of live load moments were processed using rainflow counting and cumulative damage
methodologies to provide damage-equivalent moments and axle loads and cycles per design truck
passage. Fatigue test results for the American Association of State Highway and Transportation
Officials (AASHTO) fatigue categories were reassessed using a new damage accumulation model
combined with fitted distributions to calculate biases and CVs for each category. Reliability indi-
ces for current designs were also evaluated. All this information was used in calibrating new load
factors and new constants for some of the categories.
A general procedure has been developed for calibration of SLSs based on an evaluation of the
effect of vertical or horizontal foundation deformations on bridge structures. The procedure is dem-
onstrated by using measured field data for immediate settlements of spread footings on soils. The
prediction accuracy for several methods of calculating immediate vertical movement of spread foot-
ings on soils was developed from the measured field data. This prediction accuracy is expressed in
terms of the probability of exceeding a deformation criterion (performance criterion) chosen by the
bridge designer. Using the correlation between probability of exceedance and the reliability index,
the prediction accuracy of calculation methods can be expressed in terms of the reliability index. The
target reliability index for the current calculation method by Hough (1959) recommended in the
AASHTO LRFD Bridge Design Specifications was determined for comparison with several other cal-
culation methods. The conservatism of Hough’s method is demonstrated, and use of the method by
Schmertmann et al. (1978) is recommended.
Given the regional nature of geotechnical engineering, a step-by-step process is included to
enable the process to be repeated with local data. This step-by-step process can be applied to both
vertical and horizontal deformations of all structural foundation types such as footings, drilled
shafts, and driven piles. Also recommended for consideration by the AASHTO Highway Sub-
committee on Bridges and Structures (HSCOBS) is adoption of the construction point and d-0
concepts for calculating vertical deformations, which take into account the angular distortions
within bridge spans in the context of construction stages.
The criteria used by the Canadian Highway Bridge Design Code, which is based on deflection,
frequency, and perception, were evaluated and calibrated as a possible alternative to the current
deflection criteria. In this case, the reason for change was not found to be quite as compelling as
some of the other SLSs, but the rationale has been presented. Several other SLSs were either
found unsuitable for calibration, or the difference in application did not justify a change in the
AASHTO LRFD. In these cases, the rationale is also presented.
On the basis of the reported research and calibrations, draft agenda items required to imple-
ment the findings of this project through changes to the AASHTO LRFD Bridge Design Specifica-
tions are presented for consideration by the AASHTO HSCOBS. Most of these items are
evolutionary, but several would require changes to the typical American bridge design process if
they are adopted. These required changes are detailed in the report.

Owners who make exceptions to some of the AASHTO LRFD requirements will have to evaluate
the findings of this research and decide their jurisdiction-specific requirements. The effect of the
proposed specifications revisions on specific types of components will be debated by AASHTO’s
technical committees and HSCOBS when the revisions are considered.
The products of this research, which are expected to be directly usable by AASHTO and
departments of transportation, include the following:
• Provisions needed to implement SLSs and the associated load and resistance factors necessary
to produce calibrated bridge components and systems expected to have a predictable service
life. When practical, the provisions are based on a 100-year life; if a component or system can-
not reasonably be expected to last 100 years, the expected life is given.
• Some detailed design and detailing provisions required to design and build the calibrated
component or system.
• Appendix F, containing the databases used in the calibration, as well as instructions for a cali-
bration spreadsheet for use by departments of transportation to track and adjust service-
based reliability with time. It is expected that implementing owners will track deterioration
and changes to load regimes with time and adjust built-in models and assumptions over time.
The lack of quality information regarding the change in serviceability over time for bridges in
different environments and traffic conditions was a continual challenge during this project.
There is a national database summarizing the results of the National Bridge Inspection system
that could provide a wealth of information. However, no well-accepted direct link between the
National Bridge Inventory (NBI) condition data and the types of unsatisfactory performance
related to the SLSs calibrated in this study has been found. Several locally developed predictors
of changes in the NBI condition number over time have been presented to provide guidance to
owners on possible changes to the resistance side of the limit states used in this report within the
context of the caveat above. This lack of correspondence between the NBI condition numbers
and quantifiable changes in behavior or resistance as they relate to service response and, ulti-
mately, to the strength limit states, limits the use of this information to both owners and research-
ers. One of the recommendations coming out of this project is to initiate work to close this gap.
Finally, there is much interest nationally and internationally on the improved implementation
of SLSs that should be considered in any continued development of the AASHTO LRFD.

Chapter 1
Purpose of Report and Relation to Scope
1.1 Objectives of Project R19B was defined as the process of determining values of load and
resistance factors so that the designed components will satisfy
The request for proposal for SHRP 2 Project R19B stated the the selected reliability-based criterion (i.e., the reliability of the
following objectives: structure is close to the target value). Calibration involved the
development of statistical models for load and resistance, selec-
• Develop new design codes that incorporate a rational tion of the target reliability index, and reliability analysis.
approach based on service limit states (SLSs) for durability The products of this research, expected to be directly usable
and performance of bridge systems, subsystems, compo- by the American Association of State Highway and Transporta-
nents, and details that are critical to reaching the expected tion Officials (AASHTO) and departments of transportation
service life and assuring an actual life beyond 100 years. (DOTs), include the following:
Special focus should be given to problematic systems, sub-
systems, components, and details. The proposed SLSs will • Provisions needed to implement SLSs and the associated
include data sets related to durability, fatigue, fracture, and load and resistance factors necessary to produce calibrated
redundancy as integral issues of service life as reported in bridge components and systems expected to have a predict-
SHRP 2 Project R19A. able service life. When practical, the provisions are based on
• Develop performance measures incorporating predefined a 100-year life; if a component or system cannot reasonably
component classifications that will use full probability- be expected to last 100 years, the expected life is given.
based service life design criteria to maximize the actual life • Some detailed design and detailing provisions required to
of the system. Consider material performance (including design and build the calibrated component or system.
durability); structural performance of systems, subsystems, • Appendix F, which contains the databases used in the cali-
and components (optimum joints and bearings); and design bration, as well as instructions for a calibration spread-
practices leading to longer and more predictable service life. sheet for use by DOTs to track and adjust service-based
• Develop comprehensive design procedures, proposed speci- reliability with time. It is expected that implementing own-
fication changes, and implementation tools that include ers will track deterioration and changes to load regimes
durability design in addition to structural design. The with time and adjust built-in models and assumptions
development should also consider structural and material over time.
redundancy, and system, subsystem, and component per-
formance measures that will use service life design criteria It is assumed that the AASHTO LRFD Bridge Design Speci-
to maximize the actual life of the system. The adjustments fications (AASHTO LRFD) requirements are a package that
to SLSs should not adversely affect ultimate or strength has to be considered together. Owners who make exceptions
limit states (ULSs) and extreme event limit states. to some AASHTO LRFD requirements will have to evaluate
the findings of this research and decide their jurisdiction-
To best accomplish the project objectives, the project team specific requirements. The effect of the proposed specifica-
first developed a list of the applicable SLSs for various com- tions revisions on specific types of components will be debated
ponents. A framework for calibration that accommodates aging by AASHTO’s technical committees and Highway Subcom-
and deterioration models, applicable loads, and other design mittee of Bridges and Structures (HSCOBS) when the revisions
parameters for the components was developed. Calibration are considered.

5
1.2 Scope The INC included appropriate experts from AASHTO, the

Federal Highway Administration (FHWA), state DOTs,
1.2.1 Original Scope industry, and academia (a minimum of seven volunteers is
As originally scoped, the project was broken into two phases expected for this committee).
containing the tasks described below.
Phase 2
Phase 1 Task 6
Task 1 Conduct analytical trial runs as appropriate for evaluating
Conduct a review of the literature to review and identify cur- the performance of systems, subsystems, components, and
rent practices of using SLS principles for determining struc- details developed under Project R19A, as well as existing sys-
tural service life approaches. tems, subsystems, components, and details that are critical to
reaching the expected service life beyond 100 years. Submit
Task 2 an interim report to SHRP 2 and to the approved INC.
Supplement the interim report provided by Project R19A at
the beginning of this contract with a follow-up survey to Task 7
identify successful systems, subsystems, components, and Plan a working session with the INC to review and gather
details that have lasted 100-plus years. Systems, subsystems, feedback on the interim report. Submit to SHRP 2 an updated
components, and details that have proven to or have the interim report based on the input from the INC.
potential to solve common bridge durability and structural
performance problems were of special interest for this task. In Task 8
addition, identify problematic components and the nature Incorporate the framework for the long-term bridge perfor-
and cause of failures that resulted in reduced service life, and mance program (LTBPP), and use the LTBPP framework to
document any available maintenance and rehabilitative costs. validate performance expectations. Ensure that the ULSs and
Compile any existing data regarding loadings and accelerated extreme event limit states are not compromised. The approach
environmental testing results as documented by others for from the research team was to include recommendations for
evaluating the performance of developed systems, sub- future bridge practitioners on how to adjust the SLSs to include
systems, components, and details. The research team was not semiprobabilistic assumptions developed under this project.
aware of significant relevant information, so this type of data
was not incorporated into the SLS calibration approach. Task 9
Owner-supplied information can be used to supplement the Develop a data set format that will be adaptable for future
calibration. maintenance by the AASHTO Technical Committee on Loads
and Loads Distribution (T-5).
Task 3
Develop an SLS approach that can be used to calibrate Task 10
100-plus-year service life. A benchmark in the calibration will Work with the Project R19A team and industry to develop
evaluate the suitability of the existing 75-year load and resis- recommendations for AASHTO-formatted LRFD design and
tance factor design (LRFD) approach presented in the cur- load rating specifications and analysis methods, including
rent specifications. detailed examples for bridge systems, subsystems, compo-
nents, and details that incorporate results from Project R19A.
Task 4
Prepare an interim report documenting Tasks 1 through 3 Task 11
and a detailed work plan for executing the SLS calibration. Develop an implementation plan suitable for adoption and
The research team proposed and the sponsor approved maintenance by AASHTO that is based on the findings of this
moving Task 5 into Phase 1. The work of an independent work.
committee takes place in later tasks still left in Phase 2. Task 5
deals with identifying the members of the committee. Task 12
Prepare a final report, including recommendations for future
Task 5 research.
The research team proposed, for SHRP 2 approval, an inde- The first five tasks were developed in the Phase 1 report
pendent national committee (INC) of experts to review and previously submitted and reviewed by the project expert
critique the suitability of the data set and the SLS approach. technical group.

1.2.2 Revised Scope for Tasks 5 and 6 parameters. At the completion of a proof-of-concept partial
calibration, the research team developed an interim report on
Task 5 was completed. Members of the INC were selected
findings. Once the calibration procedures were coded into
and approved. However, as the project unfolded, the expert spreadsheets, the value of trial runs or partial calibrations
technical group was augmented with additional experts in became insignificant. That report has been folded into the
the areas of calibration and deterioration, and the value of present report.
the INC decreased accordingly. Copies of the Phase 1 report A database of bridges was useful during the calibration
were provided to members of the INC in August 2011, but process to assess current reliability versus the reliability
little response was received. Written comments received from resulting from proposed changes in design equations and
one reviewer were very supportive of the approach outlined methodologies, as well as selecting load and resistance factors.
in the Phase 1 report, and those comments were submitted For this project, the database compiled under National Coop-
to the SHRP 2 staff. erative Highway Research Program (NCHRP) Project 12-78
Task 6 of the research plan for Project R19B, as modified (Mlynarski et al. 2011) was selected as the source of sample
and submitted in November 2011, initially required the bridges. The database contains information on over 18,000
submission of an interim report documenting a proof-of- bridges suitable for analysis using AASHTOWare’s Bridge Rat-
concept demonstration of the proposed calibration of SLSs. ing analytical software.
After consulting with SHRP 2 staff, the research team decided The NCHRP 12-78 database was sorted to select relatively
that the interim report served little purpose and that, due to modern bridges, potential candidate bridges were identified,
the unforeseen difficulty in finding suitable data for calibration and a partial list of candidate bridges was submitted to SHRP 2.
and other analytical work, the resources originally pro- As the calibration procedures were more fully developed
grammed for the trial calibration runs, the interim report, and and it was determined that insufficient data were available to
the working sessions with the INC (i.e., Tasks 6 and 7) could fully calibrate some of the SLSs, the sample bridge population
be better used to advance the other tasks. The specific require- was used with three SLSs: cracking of prestressed concrete
ments for a revised Task 6 are described below. beams, settlement, and deflections. The sample bridge popula-
tion was also used to investigate the ramifications of potential
Revised Task 6 changes to two SLSs: cracking of prestressed concrete beams
and overload of steel bridges. Subsets of the candidate bridge
Conduct analytical trial runs as appropriate for evaluating database used for these purposes are included in Appendix F.
the performance of systems, subsystems, components, and
details developed under Project R19A, as well as existing sys-
tems, subsystems, components, and details that are critical to
1.3 Research Team
reaching the expected service life beyond 100 years. Submit The organization and relationship of the primary team mem-
an interim report to SHRP 2 and to the approved INC. bers are shown in the organizational chart in Figure 1.1.
The research team developed serviceability provisions based The functional lead responsibilities for leading individual
on the findings and calibration approach outlined in Chapters tasks were as follows:
3 and 6. These provisions include improvements to the exist-
ing service and fatigue limit states as shown below: • Task 1. The two universities.
• Task 2. University of Nebraska–Lincoln (UNL) (Atorod
• Load-induced fatigue of steel and concrete details and Azizinamini took the lead).
components; • Task 3. All components of the team participated in the
• Live load deflection; work, but it was Modjeski and Masters’ role to see that the
• Permanent deformation of compact steel components; group reached a conclusion and a product.
• Cracking of reinforced-concrete components; • Task 4. Modjeski and Masters, with help from the other
• Tension in prestressed concrete components; team members.
• Settlement of foundations; • Task 5. As with Task 3, this task was done by the entire
• Horizontal movements of abutments; and group with Modjeski and Masters seeing that a successful
• Slip of slip-critical bolted connections. outcome occurred.
• Task 6. UNL took the lead.
Initially, the calibration was to proceed in two stages: a • Task 7. Modjeski and Masters took the lead.
proof-of-concept stage involving a subset of the SLSs and a • Task 8. University of Delaware took the lead.
subset of parameters (random variables), followed by a pro- • Task 9. University of Delaware took the lead, with assis-
duction calibration involving all SLSs and a wider range of tance from UNL.

7
SHRP 2
R19B TCC
J. M. Kulicki, PhD, PE
Principal Investigator
W. G. Wassef, PhD, PE D. R. Mertz, PhD, PE A. S. Nowak, PhD

CO-P.I. CO-P.I. CO-P.I.
Management Limit States Limit States

Limit States Calibration Calibration
QA/QC Review LTBPP Coordination A. Azizinamini, PhD, PE
Recommendations Recommendations R19A Coordination
N. Samtani, PhD, PE
Geotech
Figure 1.1. Project R19B organizational chart.
• Task 10. Modjeski and Masters took the lead, working pri- 1.5 Relationship of Project R19B
marily with the University of Delaware. to NCHRP Project 12-83
• Task 11. Modjeski and Masters took the lead.
• Task 12. Modjeski and Masters took the lead. Several members of the research team were also involved
with NCHRP Project 12-83, Calibration of LRFD Concrete
Bridge Design Specifications for Serviceability. The goal of
1.4 Relationship of Project R19B the NCHRP 12-83 project was to calibrate the concrete-related
to Project R19A SLSs currently in the AASHTO LRFD (2012) and, as needed, to
Projects R19A and R19B combined should have resulted in develop new calibrated concrete-related limit states for incor-
the development of AASHTO-formatted provisions for design poration into the AASHTO LRFD. Significant overlap exists
of bridges capable of providing more than 100 years of ser- between the SHRP 2 R19B and NCHRP 12-83 projects in the
vice life. The provisions should address both existing and new area of concrete structures. All aspects of the work under
bridges. The procedures have to be quantifiable for both NCHRP 12-83 are fully applicable to SHRP 2 R19B. Most of the
existing and new bridges. concrete-related aspects of this report were originally developed
One of the major tasks within R19A was identifying prom- in NCHRP 12-83 and are incorporated here.
ising systems, subsystems, components, details, and retrofit
concepts capable of prolonging the service life of bridges at
1.6 Special Challenges
optimal total costs. R19A was to have developed details and
Related to SLSs
subsystems requiring calibration or development (or both)
of new limit state design provisions. The R19B work depended The ULSs of the AASHTO LRFD are calibrated through
on R19A’s developing these details or subsystems. For selected structural-reliability theory to achieve a certain level of
ideas, R19A was also to have developed deterioration models. safety. They are intended to achieve similar component
Incorporating these deterioration models into a general SLS proportions to those of the Standard Specifications for High-
design provision framework was to be a major undertaking way Bridges. These ULSs do not consider the integration
within R19B. The choice of a general SLS design framework of the daily, seasonal, and long-term service stresses that
was an important issue that affected the research directions directly affect long-term bridge performance and subsequent
of both the R19B and R19A projects. service life.
As of this writing, no new details and subsystems have been The current SLSs of the AASHTO LRFD are intended to
recommended to R19B. One existing system, integral and ensure a serviceable bridge for the specified 75-year design life.
semi-integral abutments, has been identified by R19A for These limit states are based on the traditional serviceability
calibration, but no limit states have been suggested or data- provisions of the Standard Specifications for Highway Bridges.
bases identified. The SLSs are not calibrated using reliability theory to truly

achieve a determined life with a specific level of certainty the service life of steel structures. Should this limit state
because the tools and data to accomplish this calibration were now be transferred to the SLSs? In many ways, fatigue is one
not available to the AASHTO LRFD code writers. of the more quantifiable and calibratable of the SLSs com-
The current AASHTO LRFD SLSs include limits on the pared with those that may be developed dealing with dete-
following: rioration of joints, bearings, coatings, and similar structural
features.
• Live load deflection of structures; • Does it provide a method to evaluate the significance of
• Cracking of reinforced-concrete components; interventions in extending the service life of the struc-
• Tensile stresses in prestressed concrete components; ture component? Can the proposed limit states distinguish
• Compressive stresses in prestressed concrete components; between interventions that slow deterioration and those
• Permanent deformations of compact steel components; that effectively halt deterioration for some period of time
• Slip of slip-critical friction bolted connections; and before it starts again? Can they respond to repairs that rein-
• Settlement of shallow and deep foundations, among others. state or increase load-carrying capacity?
The background of the current AASHTO LRFD (2012) Consideration of SLSs requires different input data than
SLSs is presented in Chapter 2. Some of these SLSs may relate ULSs require. In ULSs, the limit state function is defined with
to a specified design life; others do not. Many are presently two variables: resistance (which is considered constant in
deterministic, such as limiting the tensile stresses in pre- time) and loads. For SLSs, a different approach is needed for
stressed concrete components to a level thought to result in the following reasons:
a crack-free component. This SLS could be calibrated to
achieve a certain probability of a crack-free component, but • The definition of resistance is very difficult.
the calibration would include a service life only in determin- • Acceptable performance can be subjective (full life-cycle
ing the live load the component must resist (e.g., a 75-year analysis is required).
live load). • Resistance and load effects can be and often are correlated.
To achieve the objective of developing the appropriate • Load is considered as a function of time, described by mag-
tools, candidate SLSs were evaluated against a set of criteria. nitude and frequency of occurrence.
This evaluation applied both to the retention of some of the • Resistance is strongly affected by quality of workmanship,
existing SLSs in the AASHTO LRFD and any new limit states operation procedures, and maintenance.
developed as part of this project and Project R19A. The crite- • Resistance is subject to changes (mostly but not only dete-
ria include the following: rioration) in time, with difficulty predicting initiation time
and time-varying rates of deterioration (e.g., corrosion,
• Is the limit state quantitatively and qualitatively meaning- accumulation of debris, cracking).
ful? Does it tell us something that we can use to maintain a • Resistance can depend on geographical location (climate,
structure in service and continue or extend its service life? exposure to industrial pollution, exposure to salt as a deicing
• Can the limit state be calibrated? Can we develop limit agent, or proximity to the ocean).
state functions, such as indicated in Task 3, and develop a
means either through the resources of Project R19B or by In general, the consequences of exceeding SLSs are an
leveraging the results of Project R19A or the LTBPP to order, or even several orders, of magnitude smaller than those
determine the data necessary to do a calibration? (When associated with ULSs. Thus, an acceptable probability of
no such data existed, expert elicitation [Delphi process] exceeding an SLS is much higher than for a ULS. If the target
was used to determine the range of data and the relative reliability index (bT) for ULS is bT = 3.5 to 4.0, then for SLS,
importance of certain characteristics in the data, including bT = 0.0 to 1.0 might be quite acceptable.
uncertainty, so that some calibration could proceed.) The current AASHTO LRFD (2012) considers foundation
• Does a limit state really relate to the service life rather settlement as an SLS. Foundation SLSs were probably the
than to some other characteristic? For example, the Model most difficult issue dealt with in R19B because of the wide
Code for Service Life Design specifically states that it excludes range of physical parameters, numerous analytical solutions,
fatigue as part of the SLSs (Fédération Internationale du and the regional nature of the practice of geotechnical engi-
Béton 2006). This exclusion may be in part because this neering. Bridge foundations and other appurtenant struc-
document was developed primarily for concrete structures. tures such as approach embankments should be designed so
The current AASHTO LRFD contains fatigue requirements that their deformations will not damage the bridge super-
under a separate limit state, the fatigue-and-fracture limit structure or other structural elements or ancillary elements
state. The assessment of fatigue life is very much related to such as utilities, which are often attached to bridge structures.

9
Various aspects of deformations that should be considered in a reduced design requirement rather than a new or more strin-
the design of bridges include gent requirement, there is an enhanced need to thoroughly and
prudently evaluate the design implications.
• The effect of uneven settlement between various support
elements;
1.7 Serviceability Versus
• The rotation and horizontal movements of the foundation
Deterioration
system affecting movements at the bridge-seat level; and
• Serviceability problems near a bridge abutment, in par- Various researchers have considered deterioration of highway
ticular the ubiquitous “bump at the end of the bridge” that bridges and tried to track change over time for various types
affects joint serviceability and abutment performance. of bridge and service conditions (i.e., type of roadway) by
using National Bridge Inventory (NBI) condition numbers or
The cumulative effect of these deformations may generate a similar state-specific index. Others have tried to relate dete-
uneven deformations and stresses across a bridge system and rioration to bridge type as the primary variable. Although the
its subsystems. In the case of an irregular pattern of settle- general deterioration of the bridge inventory is important
ment, a reversal of stresses may occur in a bridge deck, result- from an administrative point of view, the specific impact on
ing in the deck cracking at various locations. Cracking allows load-carrying capacity that might reduce the service life is a
moisture ingress, initiation of corrosion, and degradation of microlevel consideration. The various deterioration models
various bridge elements, resulting in reduced structural are of limited value in that context. Nevertheless, they are part
integrity. Thus, foundation deformations affect not only the of the current state of the art and can inform an owner’s
quality of ride and the safety of the traveling public, but also effort to account for the effects on resistance over time. With
the structural integrity of the bridge and its various compo- that in mind, several deterioration models found in the litera-
nents. In addition, such deformations often lead to costly ture that use information currently available to owners are
maintenance and repair measures. The service life of a bridge reviewed in Chapter 4 of this report.
structure, its components, or ancillary elements such as utili- As discussed in Chapter 2, a survey of bridge owners was
ties attached to the bridge can be significantly affected by the conducted to identify which bridge components required
deformation characteristics of the foundation system. sufficient periodic maintenance to be a significant factor in
In addition, it may be found that changes to material or their maintenance budgets. The number of times 23 compo-
construction specifications are a more effective way to deal
nents were cited is shown in Chapter 2, Figure 2.6.
with apparent serviceability issues than codified SLSs. This
could be the case, for example, with deck cracking, for which
changes to mix proportions or the use of curing practices 1.8 Durability
designed to reduce shrinkage may be as effective as limit
states based on strain calculations. Design Guide for Bridges for Service Life (Azizinamini et al.
The conservative nature of bridge engineering practice leads 2013), a product of SHRP 2 Project R19A, contains guid-
to one final special challenge for the development and calibra- ance for selecting system, subsystems, and components of
tion of SLSs. This challenge is that the concern for public safety bridges believed to promote long life. That information is not
and the stewardship of public funds often results in a long repeated here.
institutional memory of past unsatisfactory experience. It is Producing more durable bridges is best achieved through
often a slow process to recognize when advances in technology a holistic approach starting with type and location decisions
or codification have addressed a past problem. In the case of through the entire bridge life to decommissioning. A study
SLSs in particular, which are often subjective, it is difficult to for the Alabama DOT that addressed virtually every aspect of
ascertain whether changes to design provisions have resulted in the bridge delivery and maintenance system identified 57 fac-
the desired improvement. It is analogous to the axiom that one tors needed to provide more durable bridges (Ramey and
cannot prove a negative. For example, several years ago major Wright 1994). The DuraCrete report (2000) describes how
changes were made to the provisions for the design of modular the ability to quantify the durability and service life of a bridge
expansion joints, particularly in regard to fatigue. Have these changes during the design phase, construction phase ending
changes solved the problem so that the service life of these with transfer (handing over), subsequent inspection and assess-
joints has been increased to the point that they need not be ment phase, and possible repurposing phase. It is pointed out
considered in this project? Has enough experience been gained that the designer has the least accurate information about
to know? How much good experience is needed to alter any environmental loads, material properties, and quality of the
lingering perceptions based on earlier designs? To the research constructed facility than at any other time in the service life
team, these issues imply that when the results of R19B lead to of the facility.

10
Rostam (2005) describes two overarching strategies for 1. Acceptable level of reliability with respect to the natural
addressing durability: environment and human-made loads. The particular mod-
ification to the AASHTO LRFD proposed in Chapter 7 of
• Strategy A, avoidance (such as use of corrosion-resistant this report relates to cracking of reinforced concrete, con-
rebar); and trol of stresses in prestressed concrete, control of fatigue
• Strategy B, selection of materials and details to resist dete- cracking in steel and concrete construction, and settlement.
rioration for a given time. This recommended modification, as well as the endorse-
ment of current practice for limiting stresses due to over-
Reliability modeling of deterioration is relevant only for loads, will contribute to reduced damage and hence extended
Strategy B, and European researchers have developed some service life.
deterioration models and reliability applications for concrete 2. Properly constructed with suitable materials. The benefits of
components (DuraCrete 2000). In particular, models have quality construction are self-evident. Every DOT has con-
been proposed for corrosion of rebar from salt intrusion and struction and material specifications, as well as field and
carbonation. Some of the necessary data have been accumu- plant inspections, intended to ascertain that those require-
lated in Europe. ments are achieved in the completed project. It is outside
Freyermuth (2009) lists the following options for achiev- the scope of this project to critique those processes. Gener-
ing extended service life of concrete bridges, although the ally, concrete structures are adversely affected by ingress of
extension is not quantified: salt, which leads to corrosion of embedded steel; chemical
attack, such as alkali–silica reactivity and sulfate attack; and
• Use of high-performance concrete to decrease permeability; scaling, such as that associated with freeze–thaw cycles.
• Use of prestressing to reduce or control cracking; Langley (1999) details some of the steps taken to address
• Use of jointless bridges, or bridge segments, and integral these issues on the Confederation Bridge. Mirza (2007)
bridges; summarizes the concrete durability provisions of various
• Use of integral deck overlays on precast concrete segmental Canadian Standards Association specifications.
bridges in aggressive environments; and Some well-accepted durability-enhancing materials and
• Selective use of stainless steel reinforcing. processes are described below. The cost of these enhance-
ments and the benefit achieved vary from state to state and
These important strategies may be regarded as high-level even within a state. Environmental regulations and mainte-
decisions that should be made before the detailed numerical nance and protection of traffic can add tremendously to the
design proceeds. Use of noncorrosive deicing and fixed anti- total cost of maintenance operations, and these associated
icing spray technology was also noted as an in-service strat- costs also vary widely. Therefore, no attempt has been made
egy for enhancing concrete deck life, in particular. to quantify cost–benefit characteristics.
It is of interest to consider the number of railroad bridges • Salt intrusion is slowed by drainage control; providing
that have served for over 100 years with minimal main- suitable cover; use of dense, low-permeability concrete
tenance. Although corrosion is often evident in railroad (such as high-performance concrete (HPC) and ultra-
bridges, the severe attack of structural steel and reinforcing high-performance concrete); and control of cracking.
steel from deicing salt, in particular, is a major distinction in • The effects of salt intrusion and depassivation due to
the deterioration of highway bridges, as is the pounding from carbonation can be mitigated by using corrosion inhib-
truck traffic. itors, coated reinforcing, bimetallic reinforcement,
Structures intended to provide at least a 100-year service stainless steel reinforcing, or nonmetallic reinforce-
life must have the following four attributes, which are dis- ment such as fiber-reinforced plastic composites. With-
cussed in detail below: out citing cost–benefit specifics, it will generally be
found that cost increases with each step in the reinforc-
• Be conceived, sited, and designed to provide an acceptable ing path above.
level of reliability with respect to the natural environment • Aggregate reactivity issues such as alkali–silica reactivity
and human-made loads. are usually handled by prescreening possible sources by
• Be properly constructed with suitable materials and details. using laboratory tests to identify susceptibility. Most states
• Be provided with adequate control of deck drainage, espe- have approved sources that largely eliminate aggregate
cially in areas where deicing or environmental salt is reactivity. Use of low-alkali cement can also reduce sus-
applied. ceptibility of a concrete mix.
• Be treated with timely preventative maintenance of protec- • Sulfate attack is a result of the growth of minerals
tive coatings, drainage systems, joints, and bearings. caused by reaction of chemicals in the cement with

11
sulfates in the mix; usually these sulfates are in the Fatigue is an in-service design issue that is virtually
water, but they may be in the aggregate. Sulfate attack unaffected by material choice and seems to be well
debonds the aggregate and creates expansive pressure addressed from the resistance side by current design
leading to crack or delimitation. The causes and criteria.
effects are similar to alkali–silica reactivity. Use of Properly constructed, the key descriptor in the second
Type II, Type V, or blended cement is often indicated attribute contributing to 100-year bridge service life, car-
as well as use of approved material sources. Factors ries with it the requirement to provide sufficient field mon-
that reduce permeability are also usually helpful itoring of construction to ensure that the work is executed
Detwiler (2008) states that “Maximum limits on the within tolerances that are consistent with those assumed in
water-cementitious materials ratio, combined with the design. For example, the concrete cover is one of the
good concreting practices—especially good curing— major factors related to the rate of chloride intrusion and
are even more important to sulfate resistance than the carbonation, yet it is difficult to control in the field unless
right cement.” ACI 201.2R-08 (ACI Committee 201 suitable spacers are provided and the rebar cage, tendon
2008) provides recommended mix practices for vari- ducts, and so forth are sufficiently tied to maintain their
ous sulfate concentrations. position during concrete placement. Proper consolidation,
• Freeze–thaw cycles can lead to scaling of the concrete curing, and water control are critical, especially in regard to
surface due to pressure caused by the expansion of cracking and permeability. This requires vigilance by
water in the concrete. Use of air-entraining, high- laborers, supervisors, and inspectors.
strength mixes and low permeability are effective 3. Providing adequate control of deck drainage. Damage caused
countermeasures, although air entrainment can result by deck drainage, particularly salt-laden drainage, has been
in reduced strength and may not be compatible with a major cause of deterioration in both steel and concrete
HPC or high-strength concrete. Use of fly ash can be bridges. The reduction in the number of deck joints through
the use of continuous construction, combined with the
counterproductive if delayed strength gain exposes the
widespread use of coated reinforcing, has reduced the
concrete to freezing before sufficient strength has been
impact of this problem. Although fully integral bridges have
developed.
eliminated all deck joints, many bridges are still designed
• Prestressing contributes to control of salt intrusion by
with some joints. In addition, most existing bridges contain
reducing in-service cracking. Although some cracking
joints, and they will be in use for a long time.
may result from overloads, thermal gradients, and shrink-
4. Timely preventative maintenance of protective coatings,
age, the cracks will generally close when the causative
drainage systems, joints, and bearings. Preservation of coat-
effect is reduced or eliminated. Beam ends exposed to
ing systems is probably the most important step in the
salt-laden deck drainage have been found to be suscep-
preservation of painted steel bridges and contributes to
tible to corrosion damage resulting from water entering reduced permeability of concrete surfaces. Weathering
the beam via the strand ends. Tabatabai et al. (2004) steel bridges often have the area under and adjacent to
documented the benefit of coating the end 2 ft with deck joints coated, in which case preservation of that coat-
sealing materials and concluded that of four tested ing is as important as maintaining the coating system on
materials, a polymer resin coating was most effective painted steel bridges.
and easiest to apply. Maintenance of joints, troughs, and drainage hardware
Modern bridge steels are produced to tight tolerances helps to control the flow of deck drainage to reduce dete-
of strength and element sizes. Toughness varies some- rioration of bearings, girders pier caps, and abutments.
what more than other properties, but it usually exceeds The use of continuity and integral and semi-integral abut-
the minimum specified values, sometimes substantially. ments has been found to be effective in drainage control.
Corrosion can have a significant effect on service life if Washing of bridges is usually thought to be a cost-
not addressed. From a material point of view there are five effective means of bridge maintenance. However, Klaiber
general ranges of corrosion susceptibility provided by et al. (2004) found that for bridges on secondary roads,
conventional steel, weathering steel, high-performance after 10 years deck washing did not produce significant
steel, 1035 steel (sometimes referred to as semistainless improvement in deck durability.
steel), and stainless steel. Improved corrosion resistance The effects of degradation were not included in the
and cost increase with each step along the product line. reported calibrations. It is assumed that maintenance will
Stainless steel has received little use in bridge construc- take place before deterioration significantly affects service
tion primarily due to cost. However, it has seen more use load response. Further, the deterioration that might affect
in recent years. service response (other than appearance issues, which

12
could be considered an SLS for some bridges) could be project coordination meeting was held in the autumn of
quite different from that which affects deflection, vibra- 2009. Both teams recognized the benefit that could result
tions, concrete cracking, and so forth. The condition of from an open sharing of information and data needs. An
the bridge can be included as a change to resistance at a initial list of worthwhile types of data that the LTBPP team
given point in time, and reliability indices can be recalcu- might consider measuring in the bridges they will be
lated on that basis. instrumenting was presented to the LTBPP team for their
consideration:
Although details, materials, and techniques that are antici-
pated to increase service life can be identified, the quantifica- • Put survey targets on substructures (piers and abutments),
tion of that increased life, or the change in reliability, is not preferably starting with a bridge under construction, using
generally possible at this time. An exception may be the rates some sort of laser monitoring to determine displacements
of chloride ingress and carbonization for uncracked concrete and rotations with time. Possible foundation monitoring
under conditions similar to those in laboratory testing. For points were discussed with the LTBPP team.
example, Fick’s Second Law of Diffusion has been used to • Try to measure the relative and absolute movement between
estimate the time until chlorides reach a threshold value at substructure and superstructure.
reinforcement in the Confederation Bridge given cover, a dif- • Collect data on the rate of aging of joints and bearings,
fusion constant, and a chloride content. including debris collection and initiation of leaking.
Rostam (2005) lists the following parameters required to • Collect data on vehicular damage to joints.
determine mix design qualities to provide a target service life: • Try to collect data on traffic patterns, including convoying
and lane usage.
• The design surface chloride concentration; • Measure relative movements at and across joints, similar to
• The background chloride concentration foreseen in the what bridge inspection teams sometimes measure (some-
concrete mix; times hard to relate to temperature).
• The chloride diffusivity; • Try to monitor longitudinal forces in structures. This could
• The critical chloride concentration triggering corrosion apply to the design of joints and bearings, as well as col-
of the reinforcement (the threshold value); and umns and foundations.
• The aging factor, represented by a decreasing diffusion • Try to determine if there is any in-service way to monitor
coefficient with increasing age. change in friction with age and wear of expansion joints
and bearings.
Procedures are available (DuraCrete 2000) to develop a • Try to monitor pavement growth and effectiveness of cycle
distribution of time to reach the threshold for a given cover control joints.
from which the criteria exceedances, and hence a reliability • Assess corrosion loss or other elements of resistance change.
index, can be found. • Monitor coating deterioration.
To date this approach has not been widely used. Develop- • Try to measure something on jointless bridges involving
ment of various parameters for regional or local material the potential for pier damage or movements of integral
sources and concrete mixes would probably be needed for and semi-integral abutments, pressure behind abutments,
wide application. and movements and stresses in the piles of integral and
semi-integral abutments.
• Start to assemble data on the variability of prestress
1.9 Initial Coordination with
camber.
FHWA Long-Term Bridge
• Monitor possible development of cracks in prestressed
Performance Program
concrete beams and relate to overloads and environmental
As indicated above, very little usable data have been found for factors.
use in developing and calibrating SLSs. FHWA recently initi- • Monitor regional thermal gradients in superstructures, as
ated the long-term bridge performance program (LTBPP), well as large exposed box members such as tie girders and
which is intended to measure response factors for in-service ribs. Verify Imbsen’s NCHRP study and extend to steel.
bridges for as long as 20 years. This project could collect data • Monitor salt ingress regionally and relate to application
needed for future development and improved calibration of rate and structural parameters.
SLSs, possibly even a full probabilistic approach. Task 8 in
Phase 2 of this project requires that the R19B research team In order to remain in active contact with the LTBPP team,
establish a dialogue with the LTBPP research team. SHRP 2 the R19B principal investigator accepted a position on
staff asked that this dialogue be started earlier, and a joint the Transportation Research Board’s Long-Term Bridge

13
Performance Committee. In addition, one of the coprincipal 2012 and 2013; T-14 (Steel Design) in 2010, 2011, 2012, and
investigators has been associated with LTBPP virtually 2013; and T-15 (Foundations) in 2012, as well as the full
throughout the project and is now technical director of the HSCOBS meeting in 2009, 2010, 2011, and 2012
project. • The 2012 Annual Meeting of the National Transportation
Board in Washington, D.C.
• The 2010 Annual Meeting of the Prestressed Concrete
1.10 Dialogue with AASHTO
Institute in Philadelphia, Pennsylvania
HSCOBS and Others
• The International Association for Bridge Maintenance and
Numerous presentations were made to keep the bridge com- Safety meeting, 2012
munity apprised of issues related to Project R19B, to seek • The U.S.–China Seminar on Highway Technology in
information for the project, and to gauge reactions of owners Beijing, 2012
to potential new design requirements. • The SHRP 2–Forum of European National Highway
Venues included the following: Research Laboratories Joint Symposium in Brussels, Bel-
gium, 2010
• The AASHTO Technical Committees: T-5 (Loads and Load • The poster session at the 2009 Annual Meeting of the
Distribution) in 2011 and 2012; T-10 (Concrete Design) in National Transportation Board in Washington, D.C.

Chapter 2
Current State of the Art
2.1 Approach 2.2.1 Serviceability, SLS, Deterioration,

and Maintenance in the
As part of Phase 1, an assessment of the current state of the Technical Literature
art related to service limit states (SLSs) was conducted as
follows: A limited survey was made of sources readily available at
Modjeski and Masters and on the Internet to investigate
• A review of technical literature was conducted and is sum- the range of issues and phenomena various organizations
marized in Section 2.2. associate with the terms serviceability, SLS, deterioration,
• A survey was made of the requirements for SLSs in several and maintenance. The results are listed in the following
modern bridge design specifications, including the Amer- subsections.
ican Association of State Highway and Transportation
Officials’ (AASHTO) AASHTO LRFD Bridge Design Speci- Serviceability
fications. This survey included reconstructing the back-
ground of the existing SLS requirements. Much of the • Merriam-Webster (2010)—Fit for use, of adequate quality
detail on concrete requirements was developed under (comes from definition for serviceable).
NCHRP Project 12-83 after initial identification was made • Wikipedia (2010)—Conditions under which a structure is
in SHRP Project R19B. This approach is consistent with still considered useful.
the relationship between these projects as introduced in • Manual for Bridge Evaluation (2008)—A term that denotes
Chapter 1. The requirements of the Eurocode and the restrictions on stress, deformation, and crack opening under
Canadian Highway Bridge Design Code (CHBDC) (2006) regular service conditions.
were also reviewed, and significant clauses are summa- • Steel Construction Manual (2011)—A state in which the
rized here. function of a building, its appearance, maintainability, dura-
• A survey of owners and some industry groups was con- bility, and comfort of its occupants are preserved under
ducted by the R19B research team. Surveys were also normal usage.
conducted by the R19A and NCHRP 12-83 teams. The • ASCE/SEI 7-10: Minimum Design Loads for Buildings and
survey results obtained by the R19B team are discussed Other Structures (2010)—Structural systems, and members
in detail in Section 2.4.1, and the results obtained during thereof, shall be designed to have adequate stiffness to limit
NCHRP 12-83 and those obtained during R19A that deflections, lateral drift, vibration, or any other deforma-
relate to R19B are summarized in Sections 2.4.2 and tions that adversely affect the intended use and performance
2.4.3, respectively. of buildings and other structures.
• 2006 Seattle Building Code (International Code Council
2007)—Structural systems and members thereof shall be
2.2 Summary of Literature
designed to have adequate stiffness to limit deflections and
Survey
lateral drift as set by code writing bodies such as the Ameri-
Results of the literature survey as they relate to the current can Concrete Institute (ACI), the American Institute of
requirements in various design specifications are summarized Steel Construction (AISC), and the International Building
in Section 2.3. Code (IBC).
14

15
• Eurocode (EN 1992-2 2005)—Perform adequately under • ISO 2394 (1998)—A state that corresponds to conditions
all expected actions. beyond which specified service requirements for a structure
• ISO 2394 (1998)—Ability of a structure or structural ele- or structural element are no longer met.
ment to perform adequately for normal use under all 44 Local damage (includes cracking) that may reduce the
expected actions. working life of the structure or affect the efficiency or
appearance of structural or nonstructural elements;
44 Unacceptable deformations that affect the efficient use
Service Limit State
or appearance of structural or nonstructural elements
• Wikipedia (2010)—Fails to meet technical requirements or functioning of equipment; and
for use while remaining strong enough to stand (service- 44 Excessive vibrations that cause discomfort to people
ability limit). or affect nonstructural elements or functioning of
• Manual for Bridge Evaluation (2008)—Limit state relating equipment.
to stress, deformation, and cracking. • International Federation for Structural Concrete (fib) Bulle-
• AASHTO LRFD (2012) tin 34 (Fédération Internationale du Béton 2006)—States
44 Service I—Deflection control, crack-width control in that correspond to conditions beyond which specified ser-
R/C members, slope stability; vice requirements for a structure or structural member are
44 Service II—Control yielding of steel structures, slip of no longer met.
slip-critical connections; • Louisiana Department of Transportation and Development
44 Service III—Crack control in prestressed concrete LRFD Bridge Design Manual (2006)—Stress, deformation,
members; and crack width are limited under service conditions.
44 Service IV—Relating to tension in prestressed concrete 44 Pile foundations—settlement and horizontal movement.
• West Virginia Department of Transportation (DOT), Divi-
columns with the objective of crack control;
44 Deformations—Article 2.5.2.6; sion of Highways, Bridge Design Manual (2006)—Covers
44 Concrete—Cracking, deformation, and concrete stresses cracking, deformations, deflections and concrete stresses.
specified by Articles 5.7.3.4, 5.7.3.6, and 5.9.4;
44 Steel—Permanent deformations due to localized yield- Deterioration
ing that would impair rideability under severe traffic
• Merriam-Webster (2010)—Action or process of deterio-
loadings as specified by Articles 6.10.4.2 and 6.11.4; and
rating (deteriorating defined as “to make inferior in quality
44 Decks—Deck deformation (9.5.2).
or value”).
• AISC Steel Design Guide 3 (2003)—Define the functional • Wikipedia (2010)—To make worse.
performance of the structure (should be met), involve • CHBDC (2006)—Includes corrosion.
response of people and objects to the behavior of the struc- • U.S. Army Corps of Engineers Coastal Engineering Manual
ture under load. (2002)—Gradual aging of the structure and or its compo-
• AISC Steel Construction Manual (2011)—Limiting con- nents over time.
dition affecting the ability of a structure to preserve its • Bridge Inspector’s Reference Manual (2002)—Definition:
appearance, maintainability, durability or the comfort of decline in quality over a period of time due to chemical or
its occupants or function of machinery, under normal physical degradation.
usage. • Bridge Inspector’s Reference Manual (2002)—Types of dete-
• ASCE/SEI 7-10: Minimum Design Loads for Buildings and rioration for different materials.
Other Structures (2010)—Conditions in which the functions 44 Timber.
of a building or other structure are impaired because of local ▪▪ Natural defects—checks, splits, shakes, fungi, and
minor damage or deterioration of building components or insects;
because of occupant discomfort or annoyance. ▪▪ Chemical—acids, bases or alkalis; and
• 2006 Seattle Building Code (International Code Council ▪▪ Other types—delamination, loose connections, sur-
2007)—A condition beyond which a structure or member face depressions, fire, impact or collisions, abrasion or
becomes unfit for service and is judged to be no longer mechanical wear, overstress, weathering or warping,
useful for its intended function. protective coating failure.
• Eurocode (EN 1992-2 2005)—Associated with conditions of 44 Concrete.
normal use, concerned with the performance of structure ▪▪ Reinforced concrete—cracking, scaling, delamina-
or part of structure, comfort of people, and appearance of tion, spalling, chloride contamination, efflorescence,
structure. ettringite formation, honeycombs, pop-outs, wear,
• CHBDC (2006)—See Section 2.3.3. collision damage, abrasion, overload damage;

16
▪▪ Prestressed concrete—structural cracks, exposed pre- There is broad similarity in the use of the terms investi-
stressing tendons, corrosion of tendons in bond zone, gated, especially for maintenance and deterioration, with the
loss of camber due to creep or prestress losses; Bridge Inspector’s Reference Manual providing much more
▪▪ Reinforcement—corrosion; and detail, as would be expected. The term serviceability gener-
▪▪ Causes—temperature fluctuation, chemical attack, ally relates to high-level statements on structural behavior.
moisture absorption, differential foundation move- SLS ranges from generalities to very specific quantitative
ment, design and construction deficiencies, fire. requirements, although most of the surveyed sources deal
44 Steel—corrosion, fatigue cracking, overloads, collision, with vibrations, deflections including foundation settlement,
heat, paint failures; user comfort, and cracking.
44 Concrete decks—cracking, scaling, delamination, spall- There is little mention of appearance-related issues such as
ing, efflorescence, honeycombs, pop-outs, wear, collision rusting of steel or cracking or discoloration of concrete in
damage, abrasion, overload damage, reinforcement corro- relation to serviceability.
sion, prestressed concrete deterioration; and Generally, the SLSs currently specified in AASHTO LRFD
44 Steel decks—bent, damaged, or missing members; corro- consider most of the behaviors found in this part of the litera-
sion, fatigue cracks, other stress-related cracks. ture survey. This does not preclude the improvement of func-
tionality through calibration, nor the possibility that new limit
Maintenance states might be identified through other aspects of the litera-
ture search reported in this section; more extensive evaluation
• Merriam-Webster (2010)—The upkeep of property or
of the state of the art summarized in Sections 2.2 and 2.3; the
equipment.
results of surveys reported in Section 2.4; or the experience of
• Wiktionary (2010)—Actions performed to keep a system
the research team.
or machine functioning or in service.
• Eurocode (EN 1992-2 2005)—Under “Use and Maintenance”:
Monitoring performance, inspection for deterioration or 2.2.2 Search for SLSs not yet Implemented
distress, investigation of problems, and certification of work.
• Ontario Traffic Manual: Book 5: Regulatory Signs (2000)— Several reports were reviewed to determine whether any
The upkeep of highways, traffic control devices, other trans- additional SLSs should be considered when designing bridges.
portation facilities, property, and/or equipment. The additional information was meant to supplement the
• CHBDC (2006)—Under “Inspection and Maintenance” of literature review performed as part of SHRP 2 Project R19A.
commentary: Without routine inspection, maintenance, Reports were gathered from sources such as the National
repair or rehabilitation it is unlikely that any structure will Cooperative Highway Research Program, the Federal Highway
achieve its design life. Administration, the ACI Structural Journal, American Con-
• U.S. Army Corps of Engineers’ Coastal Engineering Manual crete Institute (ACI) committee documents, and conference
(2002)—Recognize potential problems and take appropri- proceedings of the Structures Congress and the American
ate action to assure project continues to function at accept- Society of Civil Engineers (ASCE).
able level. The investigated reports pertained to establishing foun-
• Bridge Inspector’s Reference Manual (2002)—Basic repairs dation limit states, concrete cracking of beams and bridge
performed on a facility to keep it at an adequate level of decks, concrete shrinkage, fatigue of prestressed concrete
service. members, and methods of controlling vibration. Each report
• ISO 2394 (1998)—Total set of activities performed during was reviewed to determine the usefulness of the information.
the design working life of a structure to enable it to fulfill Any methods that could potentially be used in creating new
the requirements for reliability. SLSs were noted and investigated further.
• fib Bulletin 34 (Fédération Internationale du Béton 2006)— Many sources provided information that was too general
Set of activities that [is] planned to take place during the to be useful, with many of the discussed methods for reduc-
service life of the structure in order to fulfill the require- ing serviceability issues relating to nonstructural aspects of
ments of reliability. the design process, which would not be useful in calibrating
• Maryland Manual on Uniform Traffic Control Devices for limit states. Some of the sources, however, provided useful
Streets and Highways (2006)—Activities performed to methods of anticipating and determining the effects of ser-
retain the legibility and visibility of the device, and to retain viceability issues such as crack width, crack spacing, and pre-
proper functioning of the device. stressed concrete fatigue.
• Ohio DOT Bridge Design Manual (2007)—Keeping all por- Bridge-related research problem statements are reviewed
tions in good condition with regard to strength, safety, and annually by Technical Committee 11 of the Highway Subcom-
rideability. mittee on Bridges and Structures. It was thought that a review

17
of these documents could show a need for additional SLSs that • High-load multirotational—For pots, 10+ years experi-
were not approved for funding but may still be worthwhile in enced; for other high-load multirotational, 15 to 40 years
the context of this project. However, there is apparently no experienced, 30 to 75 expected.
archive of old research problem statements. • Fabricated steel—15 to 100 years experienced, 50 to 75
expected.
2.2.3 Joints and Bearings In addition to the suggested or expected design service life
The design lives of bearings and expansion joints are impor- of bearings proposed by various design manuals or industry
tant with regard to the serviceability of bridges. With the publications, bearing manufacturers also provided expected
exception of deck deterioration, poor performance of these life for their products. The expected life of the bearing depends
components probably results in most of the deterioration on the manufacturer and the quality of installation, but is
and maintenance activities on typical bridges. Even a cursory typically within the range of 20 to 80 years. Maurer Söhne
investigation into the design life of these components showed (2011) suggests that their MSM sliding bearings provide a
widely varying results. service life of up to 80 years. Agom International, srl. (2013)
Several codes and design guides included expected lives of and D. S. Brown (Kaczinski 2008) suggest a service life of more
bearings: than 50 years for their pot bearings and steel-reinforced elas-
tomeric bearings, respectively. Technoslide (2013) provides
• fib Bulletin 34—10 to 25 years (Fédération Internationale
documentation on plain bearings manufactured by Bearing
Technologies that suggests a service life of 20 to 40 years for
du Béton 2006).
• Queensland Government Main Roads Specification MRS81:
elastomeric bearings; stainless steel, polytetrafluoroethylene,
and CSB-10 bearings have a life that is assumed to match the
Bridge Bearings (2012), Section 6.3—100 years for Expo-
life of the bridge. CSB-10 is a proprietary material manufac-
sure Classification B1, design life for bearings in Second
tured by CSB Bearings Co.
Gateway Bridge (Outokumpu 2013). The steps involved to
The service life for expansion joints has been examined by
achieve a specified service life include
at least two agencies within the United States. Reports sum-
44 Definition of the characteristics of the environment;
marizing estimated service life along with a minimum and
44 Identification of the potential deterioration mechanisms
maximum estimate were developed. The results are shown in
in that environment;
Table 2.1 (Indiana sent out two surveys; the results of both
44 Determination of the likely rate of deterioration; surveys are included in the table).
44 Assessment of the material life; Several other organizations and projects have also looked
44 Definition of the required material performance; into the service life of expansion joints. The Bridge Joint
44 Consideration of a probabilistic approach to the vari- Association (2010) suggests that the service life of the expan-
ability of the relevant parameters; and sion joint should equal the service life of adjacent surfaces.
44 Assessment and definition of the need for further The same life-cycle cost analysis completed in Japan (Itoh
protection. and Kitagawa 2001) for bearings also suggested the service
• Steel Bridge Bearing Selection and Design Guide (1996)— life of expansion joints as 15 to 25 years, with the average
Shorter than that of other bridge elements. being 20 years. NCHRP Synthesis 319 (Purvis 2003) noted
• Japan (Itoh and Kitagawa 2001)—25 to 35 years with an that in Florida elastomers used in joint seals must provide a
average replacement at 30 years (these values are estimated service life warranty for a minimum of 5 years. Research
to determine life-cycle cost). completed as part of Project R19A (AASHTO 2013) shows
• Indian Railways Institute of Civil Engineering (Bridge Bear- the following for estimates of service lives:
ings 2006)—Attempt to specify a bearing with an expected
life similar to that of the bridge. • Field molded joints—1 to 3 years.
• Strip seal joints—3 to 30 years.
Project R19A completed a survey of state DOTs with regard • Compression seal joints—3 to 30 years (also listed as 2 to
to their experiences with bearings. The results from their 20 years).
interim report are summarized below: • Finger plate joints—10 to 50 years.
• Modular expansion joints—10 to 50 years.
• Elastomeric—15 to 50 years experienced, 50 to 75 expected.
• Polytetrafluoroethylene—20 to 50 years experienced, 20 to As joints and bearings typically have service lives much less
75 expected. than the 100-year criterion for this project, these elements
• Cotton duck—35 to 50 years experienced, 75 expected. were not calibrated, but instead should be designed to be

18
Table 2.1. Service Life of Expansion Joints
Arizona Indiana
Joint Type Mean Min Max Mean Min Max
Pourable seals 11.5 4 30 5.2, 5.6 1, 0 15, 20

Compression seals 12.7 5 25 11.7, 10.3 0, 2 20, 20
Strip seals 18.0 8 30 11.9, 10.9 0, 1.5 20, 25
Finger or slide plate joints 28.1 10 75 — — —
Modular joints 19.2 10 25 — — —
Integral abutments 50.9 15 100 8.7, 7.3–9.8 0, 1.5 20, 15–20
Polymer-modified asphalt — — — 3.5, 5.7–5.8 1, 0–1.5 10, 10–20
Note: — = not available.
Sources: For Arizona, Evaluation of Various Types of Bridge Deck Joints 2006; for Indiana, Chang and
Lee 2001.
replaceable. Expansion joint manufacturers also provided esti- As decisions were made as to the retention or modification
mates of service life for their products. One modular expan- of the current AASHTO LRFD provisions, background infor-
sion joint manufactured by Maurer provides an estimated mation for the current SLSs is provided below.
service life of 40 years and 20 years for replaceable components.
D. S. Brown (Kaczinski 2008) notes that soft joints (silicone–
Settlement of Shallow and Deep Foundations
urethane and asphaltic plug joints) have a life expectancy of
less than 5 years. Miska (2013) noted that their neoprene com- Serviceability aspects of foundations and walls are related to
pression seal has over 30 years of proven durability. the deformation characteristics of geomaterials and struc-
tural elements. In the current AASHTO LRFD (2012), Sec-
2.3 Serviceability tion 10 (Foundations) and Section 11 (Abutments, Piers, and
Requirements in Several Walls) present a variety of formulations for estimating defor-
Modern Bridge Design mation of foundations and walls. These formulations are not
Specifications consistent in the sense that they range from theoretical,
semiempirical formulations to charts based on measured
2.3.1 AASHTO LRFD deformations. For example, the vertical settlement of spread
The current AASHTO LRFD (2012) SLSs include limits on footings is based on Hough’s (1959) method, which is largely
a theoretical method, but the settlement of pile groups is based
• Live load deflection of structures. on a choice of one of four idealized cases that use Hough’s
• Fatigue of steel and concrete details. method. In contrast, the lateral deformations of retaining
• Cracking of reinforced-concrete components. walls are based on semiempirical methods and charts. Such
• Tensile stresses in prestressed concrete components. approaches are adopted for all types of foundations and walls
• Compressive stresses in prestressed concrete components. in Sections 10 and 11 of AASHTO LRFD. Although this wide
• Settlement of shallow and deep foundations. range of approaches is understandable given that foundation
• Permanent deformations of compact steel components. design is more an art based on observations than a science, it
• Slip of slip-critical friction bolted connections. created a challenge in the context of SLS calibration using a
consistent basis. The calibration processes proposed for geo-
Design provisions are specified either in the resistance sec- technical features in Chapter 6 could be considered to estab-
tions or Section 3 of AASHTO LRFD. The design load combi- lish a consistent framework for foundations and walls. An
nations in AASHTO LRFD are presented in Table 3.4.1-1. As example of this approach is demonstrated for vertical settle-
stated in Chapter 1, these SLSs and the associated load and ment of spread footings.
resistance factors are based on apparent successful past practice
and have not been subject to a reliability-based calibration. Tolerable Vertical Deformation Criteria
There are no consistent performance levels associated with From the viewpoint of serviceability of a bridge structure,
these limit states, although some are associated with differences the geotechnical limit states relate to foundation deforma-
in environmental or traffic exposure. tions. Uneven displacements of bridge abutments and pier

19
foundations can affect the ride quality, functioning of deck of 2 to 4 in. (50 to 100 mm), depending on span length, appear
drainage, and the safety of the traveling public, as well as the to be acceptable, assuming that approach slabs or other provi-
structural integrity and aesthetics of the bridge. Such move- sions are made to minimize the effects of any differential move-
ments often lead to costly maintenance and repair measures. ments between abutments and approach embankments. Finally,
horizontal movements in excess of 2 in. (50 mm) appear likely
However, overly conservative criteria can be wasteful. Deter-
to cause structural distress. The potential for horizontal move-
mination of deformation criteria should be a collaboration
ments of abutments and piers should be considered more care-
between the geotechnical engineer and the structural engi- fully than is done in current practice. (Wahls 1983)
neer to find the optimum solution. Within the context of
foundation deformation, the geotechnical limit states can be AASHTO LRFD used data from Moulton et al. (1985) and
broadly categorized into vertical and horizontal deforma- Wahls (1983) to produce the guidance summarized in Table 2.2
tions for any foundation type (e.g., spread footings, driven for the evaluation of tolerable vertical movements in terms of
piles, drilled shafts, or micropiles). angular distortions.
Agencies often limit the deformation to values of 1 in. or The criteria in Table 2.2 suggest that for a 100-ft span, a
less without any rational basis. The literature survey revealed differential settlement of 4.8 in. is acceptable for a continu-
that the only definitive rational guidance related to the effect ous span, and 6 in. is acceptable for a simple span. These
of foundation deformations on bridge structures is based relatively large values of differential settlement create con-
on a report by Moulton et al. (1985). From an evaluation of cern for structural designers, who often arbitrarily limit tol-
314 bridges nationwide, the report offered the following erable movements to one-half to one-quarter of the values
conclusions: listed in Table 2.2 or develop guidance such as that shown in
Table 2.3.
The results of this study have shown that, depending on type of
Another example of the use of more stringent criteria is
spans, length and stiffness of spans, and the type of construc-
tion material, many highway bridges can tolerate significant from Chapter 10 of the Arizona Department of Transportation
magnitudes of total and differential vertical settlement without (ADOT) Bridge Design Guidelines (2009), which states the
becoming seriously overstressed, sustaining serious structural following:
damage, or suffering impaired riding quality. In particular, it
was found that a longitudinal angular distortion (differential The bridge designer should limit the total settlement of a
settlement/span length) of 0.004 would most likely be tolerable foundation per 100 ft span to 0.5 in. Linear interpolation
for continuous bridges of both steel and concrete, while a value should be used for other span lengths. Higher total settlement
of angular distortion of 0.005 would be a more suitable limit for
simply supported bridges (Moulton et al. 1985). Table 2.2. Tolerable Movement Criteria for
Highway Bridges (AASHTO LRFD 2012)
Another study states the following:
Limiting Angular
In summary, it is very clear that the tolerable settlement criteria Distortion, d/L (radians) Type of Bridge
currently used by most transportation agencies are extremely
0.004 Multiple-span (continuous-span)
conservative and are needlessly restricting the use of spread footbridges
ings for bridge foundations on many soils. Angular distortions
0.008 Simple-span bridges
of 1/250 of the span length and differential vertical movements
Table 2.3. Tolerable Movement Criteria for Highway Bridges

(Geotechnical Design Manual 2012)
Differential Settlement over 100 ft

Within Pier or Abutments and
Differential Settlement Between
Total Settlement at Piers (Implied Limiting Angular
Pier or Abutment Distortion, radians) Action
d ≤ 1 in. d100 ft ≤ 0.75 in. Design and construct

(0.000625)
1 in. < d ≤ 4 in. 0.75 in. < d100 ft ≤ 3 in. Ensure structure can tolerate settlement
(0.000625–0.0025)
d > 4 in. d100 ft > 3 in. Need departmental approval
(>0.0025)

20
limits may be used when the superstructure is adequately flexible joints along the utility such that it does not control
designed for such settlements. The designer shall also check the bridge design.
other factors such as rideability and aesthetics. Any total set-
tlement that is higher than 2.5 in, per 100 ft span, must be Tolerable Horizontal Deformation Criteria
approved by the ADOT Bridge Group.
Horizontal deformations cause more severe and widespread
problems for highway bridge structures than do equal magni-
Although from the viewpoint of structural integrity there
tudes of vertical movement. Tolerance of the superstructure to
are no technical reasons for structural designers to set arbi-
horizontal (lateral) movement will depend on bridge seat or
trary additional limits to the criteria listed in Table 2.2, there
joint widths, bearing type(s), structure type, and load distri-
are often practical reasons based on the tolerable limits of
bution effects. Moulton et al. (1985) found that horizontal
deformation of other structures associated with a bridge, such movements less than 1 in. were almost always reported as
as approach slabs, wingwalls, pavement structures, drainage being tolerable, while horizontal movements greater than 2 in.
grades, utilities on the bridge, and deformations that adversely were typically considered to be intolerable. On the basis of this
affect ride quality. Thus, the relatively large differential settle- observation, Moulton et al. (1985) recommended that hor-
ments based on Table 2.2 should be considered in conjunc- izontal movements be limited to 1.5 in. The data presented
tion with functional or performance criteria not only for the by Moulton et al. (1985) show that horizontal movements
bridge structure but also for all associated facilities. Samtani resulted in more damage when accompanied by settlement
and Nowatzki (2006) suggest the following steps in this regard: than when occurring alone.
1. Identify all possible facilities associated with the bridge

structure and the tolerance of those facilities to movement. Limitations on the Live Load
An example of a facility on a bridge is a utility (e.g., gas, Deflection of Structures
power, or water). The owners of the facility can identify The current requirements for live load deflection limits in the
the tolerance of their facility to movements. Alternatively, AASHTO LRFD have their roots in the corresponding provi-
the facility owners should design their facilities for the sions of the Standard Specifications for Highway Bridges, 17th ed.
movements anticipated for the bridge structure. (2002). These provisions have been reviewed repeatedly. Sum-
2. Due to the inherent uncertainty associated with estimated maries by Wright and Walker (1972), Roeder et al. (2002), and
values of settlement, determine the differential settlement by Barker and Barth (2007) are often referenced.
using conservative assumptions for geomaterial properties The ASCE Committee on Deflection Limitations of Bridges
and prediction methods. It is important that the estimation of the Structural Division (1958) reported on their examina-
of angular distortion be based on a realistic evaluation of tion of the live load deflection limits and depth-to-span ratios
the construction sequence and the magnitude of loads at in the 1953 American Association of State Highway Officials
each stage of the construction sequence. (AASHO) Standard Specifications for Highway Bridges. The
3. Compare the angular distortion from Step 2 with the various earliest deflection limits were adopted in 1871 by the Phoenix
tolerances identified in Step 1 and in Table 2.2. Using this Bridge Company, which limited deflection to 1/1,200 of the
comparison, identify the critical component of the facility. span length for a train moving 30 mph. The American Rail-
Review this critical component to check if it can be relo- way Engineering Association (AREA) adopted depth-to-span
cated or if it can be redesigned to more relaxed tolerances. ratios in the early 1900s, although the limits were without
Repeat this process as necessary for other facilities. In basis. Depth-to-span ratios for highway bridges were initially
some cases, a simple resequencing of the construction of set forth in 1913 and adopted by AASHO in 1924. Vibrations
the facility based on the construction sequence of the became an issue in the 1930s, and the Bureau of Public Roads
bridge structure may help mitigate the issues associated attempted to provide a correlation between the bridges with
with intolerable movements. vibration problems and bridge properties. The result was limit-
ing deflections to L/800 for simple and continuous spans with-
This three-step approach can be used to develop project- out pedestrians, L/1,000 for simple and continuous spans with
specific limiting angular distortion criteria that may differ pedestrians, and L/300 for cantilevered spans. The ASCE
from the general guidelines listed in Table 2.2. For example, if Committee surveyed state highway departments to obtain data
a compressed-gas line is fixed to a simple-span bridge deck on the behavior of bridges and the views of experienced bridge
and the gas line can tolerate an angular distortion of only designers. The conclusions of the survey included the follow-
0.002, then the utility will limit the angular distortion value ing: maximum oscillations occur with passage of medium-
for the bridge structure, not the criterion listed in Table 2.2. weight vehicles, not heavy vehicles; reports of objectionable
However, this problem is typically avoided by providing vibrations came from continuous-span bridges more often

21
than simple-span bridges; and there is no defined level of any. Some of us in designing a girder that is very shallow in
vibration that constitutes being undesirable. The vibration of proportion to its length decrease the unit stress or increase
the bridge is affected by the following quantities: the section according to some rule which we guess at. We put
that in there so that a man would have a warrant for using
• Bridge flexibility and associated natural frequency. whatever he pleased.”
• Flexibility of vehicle suspension and associated natural The report concluded that the reasons for the two criteria,
frequency. deflection limit and depth-to-span ratio, are of different ori-
• Relative weight of vehicles and bridge. gin. The deflection limit is to limit undesired vibration, but
• Vehicle speed. the depth-to-span ratio is a result of economics. In addition,
• Profile of approach roadway and bridge deck. the report writers could not provide recommendations as to
• Frequency of load application. what constitutes undesirable deflection or vibration or how
• Motion caused by loads in adjacent spans of continuous- best to limit deflections or vibrations. The ASCE Committee
span structures. had minor modifications, but due to the empirical nature of
• Damping characteristics of bridge and vehicle. the current limits, they believed that they could not suggest
the revisions. They also believed that the then-current limits
The use of depth-to-span ratios began in the early 1900s were sufficient until further test data became available, but
with the American Railway Engineering and Maintenance of that girders with composite action should be limited to
Way Association (AREMA) (at that time AREA) specification smaller deflections.
that pony trusses and plate girders should have a depth not In U.S. practice, the deflection of bridges supporting vehic-
less than 1/10 of the span length. These ratios have changed ular traffic is generally limited to the span length divided by
little over the years. The current depth-to-span limits are 1/10 800 for simple spans and continuous spans and divided by
for trusses and 1/12 for rolled shapes and plate girders. 300 for cantilever arms. The specifications have placed fur-
The early specifications for highway bridges adopted with ther limits on bridges also intended to carry pedestrian and
some modification the depth-to-span ratios from AREMA. bicycle traffic. There is little technical support for the efficacy
The changes in depth-to-span ratios for highway bridges are of the current deflection provisions. They are simple to use,
shown in Table 2.4 for selected time periods. but they do not directly relate to the actual issue of concern,
Both AREMA and AASHTO specifications included state- namely, the vibration response under live load. Although
ments that required flanges to be strengthened if section depths the quasistatic deflection and dynamic response both involve the
smaller than those required by the limiting depth-to-span ratio stiffness of the bridge, the dynamic response also involves the
were used. mass, damping, and the characteristics of the forcing function,
The use of depth-to-span ratios was primarily to limit which is in turn related to the surface roughness, suspension
deflections, but it was also driven by economics. The limiting characteristics of the vehicle, and other parameters.
values of depth-to-span ratios have decreased with time, Wright and Walker (1972) developed a summary of the
while allowable stresses have increased. This would result in experience with the deflection limitation provisions in the era
shallower sections being used, which would result in larger during which the bulk of the steel structures were of non-
deflections. This result confused the ASCE Committee on
composite construction. Roeder et al. (2002) revisited the
Deflection Limitations of Bridges of the Structural Division,
subject decades later and suggested that
which was tasked with investigating the origins of the deflec-
tion and depth-to-span limits. The committee quoted the
• the current AASHTO limits are insufficient for control of
1905 AREA Committee’s explanation of their depth-to-span
vibrations and should ultimately be removed;
ratios: “We established the rule because we could not agree on
• the current limit of L/800 for bridges without pedestrians
is not always sufficient to control vibrations, but should
Table 2.4. Historic Depth-to-Span Ratios not be removed as there is insufficient documentation to
for Highway Bridges warrant removing it from the design specifications; and
• the applied loading and use of load factors and distribu-
Year Trusses Plate Girders Rolled Shapes tion factors should be clarified.
1913, 1924 1/10 1/12 1/20
Roeder et al. (2002) also suggested immediately removing the
1931 1/10 1/15 1/20
L/1,000 deflection limit for bridges with pedestrian access.
1935, 1941, 1/10 1/25 1/25
As alternatives to the deflection limits (L/800 and L/1,000)
1949, 1953
and until a method for controlling vibration frequency and
2012 1/10 1/25 1/25
amplitude is approved by AASHTO, they suggest using the

22
equations developed by Wright and Walker (1972) or the cri- the stresses in the deck may also increase due to interac-
teria provided in the CHBDC (2006) for simple-span bridges. tion between the deck and girder. The increased stresses
Barker and Barth (2007) have compared the procedure in may lead to additional cracking or deterioration of the
AASHTO LRFD, which was intended to provide some unifor- bridge deck. Dynamic response of the bridge is affected
mity in application, to the specific procedures used in several minimally by increases in flexibility; the increased flexibil-
states. They found wide variations in load, load distribution, ity leads to more lateral distribution of the load to adja-
and deflection limits. In some states, the individual interpreta- cent girders.
tion is severe enough to frequently control the design, particu- • Krauss and Rogalla (1996) examined available literature;
larly of steel bridges, by a significant margin. A sample of the surveyed 52 transportation agencies in the United States
reported variation follows: and Canada; and performed research using analytical
methods, as well as field and laboratory measurements. The
• Bridges without pedestrian access survey was used to develop an understanding of how often
44 L/1,600 (one state); transverse cracks are noted in new bridge decks, as well as
44 L/1,100 (one state); how they are believed to form. More than 18,000 bridges
44 L/1,000 (five states); and were analyzed to examine the stresses in the concrete deck.
44 L/800 (40 states). Laboratory testing indicated that concrete mix, environ-
• Bridges with pedestrian access mental conditions during concrete placement, and con-
44 L/1,600 (one state); struction practices significantly affected the formation of
44 L/1,200 (two states); transverse cracks. It was also determined that bridge charac-
44 L/1,100 (one state); teristics such as deck geometry and girder type, spacing, and
44 L/1,000 (39 states); and size significantly affect the formation of transverse cracks. It
44 L/800 (three states). was determined that continuous multigirder steel spans are
• Loads used based on AASHTO load factor design (LFD) more susceptible to transverse cracks due to restraint of the
requirements deck. Krauss and Rogalla also noted that longer spans are
44 HS20 truck only (one state); more susceptible to cracking than shorter spans.
44 HS20 truck plus impact (16 states); • Goodpasture and Goodwin (1971) evaluated whether any
44 HS20 lane load plus impact (one state); relationship existed between deck deterioration and live
44 HS20 truck plus lane load without impact (one state); load deflection. They examined 27 bridges to determine
44 Larger of HS20 truck plus impact or HS20 lane load plus which bridge type had the most cracking. Bridge types
impact (seven states); included plate girders, rolled shapes, concrete girders, pre-
44 HS20 truck plus lane plus impact (17 states); stressed girders, and trusses. Ten continuous steel girder
44 Military or permit vehicles (four states); and bridges were evaluated to determine if the stiffness of the
44 HS25 truck (eight states). bridge influenced transverse cracking. The results indi-
cated no correlation between girder flexibility and amount
Live load deflection is sometimes postulated to be a con- of transverse cracking.
tributor to the cracking of concrete decks. A sample of the • Walker and Wright’s (1971) analysis indicated that spalling,
conflicting literature on this issue follows: scaling, and longitudinal cracking are not associated with
girder flexibility. Transverse deck moments result in tension
• Fountain and Thunman (1987) conducted a study in which along the top surface of the deck, possibly resulting in deck
they examined the live load deflection criteria for steel cracking. Increased girder flexibility results in larger posi-
girder bridges with concrete decks and how the deflection tive transverse moments and smaller negative moments
criteria are associated with cracking of the concrete deck. resulting in reduced likelihood of deck cracking.
Cracking can be caused by numerous factors, including • Nevels and Hixon (1973) examined 195 girder bridges
plastic shrinkage, deck restraint, drying shrinkage, long- consisting of simple- and continuous-span steel plate and
term flexure due to service loads, and repetitive vibrations. rolled girders and prestressed concrete girders. Span lengths
The results indicated that the live load deflection criteria ranged from 40 to 115 ft. They concluded that there was
did not meet the desired goals, which were strength, dura- no relationship between flexibility and deck deterioration.
bility, and safety of steel bridges. Fountain and Thunman Similarly, the Portland Cement Association (1970) pre-
questioned the applicability of the live load deflection cri- sented results of a study in which substantial evidence was
teria as a majority of steel girder bridges are built with collected that indicated flexible bridges, typically steel
composite decks, and composite decks lead to small tensile girder bridges, do not have a greater tendency to exhibit
stresses in the deck. In addition, as bridge stiffness increases, deck cracking damage than other bridge types.

23
• Barker et al. (2008) examined deflection limits and deflec- Some modern specifications, such as the Ontario High-
tion loadings from various states for a suite of 10 bridges for way Bridge Design Code (1979) and its successor the CHBDC,
both LFD and load and resistance factor design methods. use a combination of frequency, perception levels, and deflec-
The results indicate that states using larger loads and more tion limits to distinguish between acceptable and unacceptable
restrictive deflection limits end up with designs controlled response. Figure 2.1, taken from CHBDC (2006), illustrates
by deflection. To meet the more restrictive deflection limits, this approach, which has the benefit of directly addressing
a significantly stiffer bridge would be needed. Furthermore, the design issue of vibration control. This is similar to the
it was noted that the 10 bridges were performing well and procedure for building design developed by Murray et al.
had not demonstrated any detrimental effects, either user (2003).
comfort or structural damage, due to excessive deflections. In the Eurocode, live loads include a vibration factor to
The suite of 10 bridges would not satisfy the deflection cri- account for stresses caused by vibration; no checks for fre-
teria in several states and would require additional steel be quency or displacement are required (EN 1990 2002). In New
added; the additional steel would not be required for Zealand, vertical velocity is limited to 0.055 m/s (2.2 in./s)
strength but rather to meet the deflection criteria. under two 120 kN (27 kip axles) of one HN unit if a bridge
carries significant pedestrian traffic or where cars are likely
The literature reviewed above indicates that transverse to be stationary. Previous versions included span-to-depth
deck cracking can be affected by many factors. In addition, ratios and deflection limits, but these have been removed.
there is disagreement on whether limiting static live load Several proposed dynamics-based approaches in the litera-
deflections (girder flexibility) is a satisfactory method to pre- ture are summarized below:
vent deck cracking. Of the articles reviewed, the conclusions
are equally divided between those that concluded that girder • Wright and Walker (1972) recommended limits based on
flexibility affects deck cracking and those that concluded that vertical acceleration to control vibration; this includes
girder flexibility does not affect deck cracking. As indicated composite action.
by some of the studies presented above, concrete material fac- 44 ds = static deflection caused by live load with a wheel line
tors may be more important to reduce the formation of early distribution factor of 0.7 on one stringer acting with its
age deck cracks. share of deck
1000
500
UNACCEPTABLE
200
without sidewalks
100
with sidewalks,
static deflection, mm
occasional pedestrian use

50
with sidewalks,
frequent pedestrian use
20.0
10.0
5.0
ACCEPTABLE
2.0
1.0
0 1 2 3 4 5 6 7 8 9 10
first flexural frequency, Hz
Source: Canadian Standards Association.
Figure 2.1. Deflection provisions in CHBDC (2006).

24
44 Natural frequency for simple or equal spans: Fatigue-and-Fracture Limit States

π Eb I b g General
fb =
2 L2 w The fatigue-and-fracture limit state is divided into two load
v
44 Speed parameter: α = combinations: Fatigue I for infinite-life fatigue resistance and
2 fb L
Fatigue II for finite-life fatigue resistance. These relatively
44 Impact factor: DI = a + 0.15
new provisions appeared in the 2009 interim changes to load
44 Dynamic component of acceleration = a = DI × δ s (2 πfb )
2
provisions in Section 3 of the AASHTO LRFD published in
must be less than 100 in./s2
early 2009. The fatigue resistance provisions for concrete and
• Barth and Wu (2007) provide equations to estimate the
steel bridges in Sections 5 and 6 of the AASHTO LRFD,
natural frequency of continuous-span steel I-girder bridges
respectively, were modified accordingly.
44 f = λ 2 f sb for continuous spans
Loads
π Eb I b g
44 f sb = for simple spans The fatigue load of AASHTO LRFD Article 3.6.1.4 and the
2 L2 w fatigue live load load factors of AASHTO LRFD Table 3.4.1-1
Ic are based on extensive research of structural steel highway
44 λ 2 = a b
Lmax bridges. The fatigue load is the AASHTO LRFD design truck
(HS20-44 truck of the Standard Specifications for Highway
where Bridges) but with a fixed rear-axle spacing of 30 ft. The live
L = span length; load load factors for the fatigue limit state load combinations
EbIb = flexural rigidity of composite steel girder; are summarized in Table 2.5.
g = acceleration due to gravity;
w = weight per unit length of composite steel girder; Infinite-Life Fatigue
Lbmax = maximum span length; and The Fatigue I load factor of 1.50, used to design highway
Ic = average moment of inertia of composite girder bridges with higher traffic volumes for infinite fatigue life, is
section. based on a 1-in-10,000 rate of exceedance (Dexter and Fisher
44 For two-span bridges 2000). The infinite-life fatigue or constant amplitude fatigue
▪▪ a = 0.95 (1.44 for metric units) threshold stress range is the stress range below which the
▪▪ b = 0.046 inherent flaws in steel do not propagate significantly during
▪▪ c = 0.032 the design life of the bridge. If all the stress ranges experienced
44 For three- or more span bridges by a detail are below this value, the detail is assumed to have
▪▪ a = 0.88 (1.49 for metric units) infinite life. Thus, this stress range represents a maximum
▪▪ b = -0.033 limit to achieve infinite life. This stress range is revisited in
▪▪ c = 0.033 Section 6.6 through simulation using weigh-in-motion data.
Finite-Life Fatigue
Presently, specifications based on determining the fre-
NCHRP Report 267 (Fisher et al. 1983) established that the
quency have not received wide acceptance in U.S. practice.
root mean cube of the stress ranges experienced by a steel-
There has been a perceived difficulty in determining the first
bridge detail characterizes accumulated fatigue damage well
fundamental frequency of the bridge. Equations for simple-
when portions of the stress range distribution exceed the
span structures have been available for decades [e.g., Biggs
constant amplitude fatigue threshold more often than the
(1964)]. Similarly, formulas for frequency have been devel-
1-in-10,000 rate cited above, no matter how small these por-
oped for continuous structures of regular geometry. His-
tions exceeding the threshold are. Thus, the effective stress
torically, frequencies could be calculated using the Rayleigh
method typically implemented through Newmark’s numer-
ical integration. Roeder et al. (2002) summarized empirical
equations that are based not only on theoretical structural Table 2.5. Fatigue Live Load
Load Factors
dynamics but also have adjustments for apparent behavior
in the field. Modern refined computational methods make Fatigue Limit State Live Load
the determination of frequencies and mode shapes relatively Load Combination Load Factor
straightforward. Thus, there does not seem to be any imped-
Fatigue I 1.50
iment to adopting an approach similar to that specified in
Fatigue II 0.75
the CHBDC.

25
range for estimating accumulated fatigue damage may be fatigue threshold (indicated by a horizontal S-N curve) up to
taken as shown by Equation 2.1: that point. Further, it suggests that the compression stress
limit of 0.40fc′ is based on a target fatigue life of 10 million
(∆σ)effective = 3 ∑ (∆σ )
i
3
(2.1) cycles. For highway bridges, a target fatigue life of 10 million
cycles is significantly less than the design life. A highway bridge
i
with average daily truck traffic of 2,000 trucks per day would
The Fatigue II load factor produces a force effect that rep- experience over 50 million cycles during its 75-year design life.
licates the fatigue damage due to the entire spectrum of stress For this study, the research by Ople and Hulsbos (1966)
ranges experienced by the bridge detail. In other words, the used to define these S-N curves was reevaluated to estimate
fatigue damage due to passage of the effective truck over the the fatigue resistance to about 108 (100 million) cycles, a prac-
bridge for a total number of cycles, equal to the average daily tical upper bound for highway bridges. The uncertainty of
truck traffic averaged over the 75-year life span, is assumed the fatigue resistance is quantified in terms of bias, mean, and
equal to the fatigue damage due to the actual truck traffic coefficient of variation.
crossing the bridge in 75 years.
Nonprestressed Reinforcement
Recommendations As used here, nonprestressed reinforcement includes straight
The stress ranges represented by both load factors (the root reinforcing bars and welded-wire reinforcement. AASHTO
mean cube and the exceedance of 1 in 10,000) are based on LRFD (Article 5.5.3.2) specifies the fatigue resistance of these
observations of steel highway bridges and structural steel types of reinforcement.
laboratory specimens. Extending these stress ranges to steel The fatigue resistance of straight reinforcing bars and
reinforcement, both nonprestressed and prestressed, is quite welded-wire reinforcement without a cross weld in the high-
appropriate as the stress ranges represent fatigue damage accu- stress region (defined as one-third of the span on each side of
mulation in steel. It is assumed that these fatigue damage the section of maximum moment) is specified by Equation 2.2:
accumulation models apply to concrete in compression, as
well as steel reinforcement. This approach is proposed for this (∆F )TH = 24 − 0.33 f min (2.2)
study. A validation of these principles for concrete highway
bridges is far beyond the scope and funding of this study. where fmin is the minimum stress; TH is threshold.
For welded-wire reinforcement with a cross weld in the
Fatigue Resistance of Concrete Structures high-stress region, the fatigue resistance is specified by
The fatigue resistance values of concrete, nonprestressed Equation 2.3:
reinforcement and prestressing tendons in the AASHTO
LRFD are based on ACI 215R-74(92), Considerations for (∆F)TH = 16 − 0.33 f min (2.3)
Design of Concrete Structures Subjected to Fatigue Loading
(ACI Committee 215 1974). This reference includes an exten- Equations 2.2 and 2.3 implicitly assume a ratio of radius to
sive bibliography on fatigue resistance of concrete and its height (i.e., r/h) of the rolled-in transverse bar deformations
reinforcement. of 0.3.
These fatigue resistances are defined as constant amplitude
Concrete fatigue thresholds in AASHTO LRFD. ACI Committee report
The compressive stress limit of 0.40fc′ for fully prestressed ACI 215R-74(92) and the supporting literature indicate that
components in other than segmentally constructed bridges in nonprestressed reinforcement exhibits a constant amplitude
Article 5.5.3.1 of AASHTO LRFD applies to a combination of fatigue threshold, yet it is unclear that these equations are in
the Fatigue I limit state load combination (which includes fact the threshold values. ACI 215R-74(92) suggests that the
only live load) plus one-half the sum of the effective prestress resistances are “a conservative lower bound of all available test
and permanent loads after losses (a load combination derived results.” In other words, a horizontal constant amplitude
from a modified Goodman diagram). This suggests that com- fatigue threshold has been drawn beneath all the curves.
pressive stress limit represents an infinite-life check, as the The studies used to define the fatigue resistance of non
Fatigue I limit state load combination corresponds with infi- prestressed reinforcement (Fisher and Viest 1961; Pfister and
nite fatigue life. Hognestad 1964; Burton and Hognestad 1967; Hanson et al.
ACI 215R-74(92) indicates that the fatigue resistance of 1968; Helgason et al. 1976; Lash 1969; MacGregor et al. 1971;
concrete in the form of an S-N curve (stress range versus Amorn et al. 2007) were reanalyzed to estimate constant ampli-
number of cycles) is approximately linear between 100 and tude fatigue thresholds for each case (analogous to the vari-
10 million cycles. It does not exhibit a constant amplitude ous detail categories defined for steel details) that could be

26
identified in the research and to determine their uncertainty in response to these various uncertainties, it is proposed that this
terms of bias, mean, and coefficient of variation. The various fatigue limit state not be calibrated.
thresholds were grouped together to make design practical and
more rational than the single threshold currently defined. Welded and Mechanical Splices of Reinforcement
The AASHO Road Test (1962) demonstrated that a bridge In AASHTO LRFD Article 5.5.3.4, constant amplitude fatigue
does not necessarily collapse due to fracture following fatigue thresholds are given in Table 5.5.3.4-1. These values are used
of nonprestressed reinforcement. Such nonprestressed rein- in the general fatigue limit state equation (AASHTO LRFD
forcement fracture results in distress such as excessive deflec- Equation 5.5.3.1-1) for the design of welded or mechanical
tion and wide cracks, which facilitate detection and subsequent splices of reinforcement for infinite fatigue life.
repair. This consequence suggests that a target reliability index Review of the available test data in NCHRP Research
(bT) less than that for ultimate limit states (ULSs) is acceptable Results Digest 197 (1994) suggests that any splice capable of
(in other words, bT < 3.5). developing 125% of the yield strength of the bar will sustain
1 million cycles of a 4-ksi constant amplitude stress range.
Prestressing Tendons This fatigue limit is a close lower bound for the splice fatigue
Fully prestressed components satisfying the tensile stress lim- data contained in NCHRP Research Results Digest 197 (1994).
its specified in AASHTO LRFD Table 5.9.4.2.2-1 at the Ser- NCHRP Research Results Digest 197 (1994) found that
vice III limit state load combination are exempt from fatigue there is substantial uncertainty in the fatigue performance
considerations. (The Service III limit state load combination of different types of welds and connectors, much as in struc-
and its calibration are discussed in Chapter 6.) This exemption tural steel details. However, all types of splices appeared
acknowledges that tendons in uncracked prestressed beams to exhibit a constant amplitude fatigue limit for repetitive
do not experience stress ranges resulting in fatigue cracking. loading exceeding about 1 million cycles. The stress ranges
Most prestressed concrete bridge members are covered by this for over 1 million cycles of loading given in AASHTO LRFD
exemption. Table 5.5.3.4-1 are based on statistical tolerance limits to
For segmentally constructed bridges, AASHTO LRFD Arti- constant amplitude staircase test data, such that there is a
cle 5.5.3.3 specifies the fatigue resistance of prestressing ten- 95% level of confidence that 95% of the data would exceed
dons as given in Table 2.6. Reductions in constant amplitude the given values for 5 million cycles of loading. These values
fatigue threshold limits for fretting fatigue are not included in may, therefore, be regarded as a fatigue limit below which
the tabulated values. fatigue damage is unlikely to occur during the design life-
In-service fatigue cracking of prestressing tendons has not time of the structure. This is the same basis used to establish
been observed, thus justifying the exemption. The majority of the fatigue design provisions for unspliced reinforcing bars
research on fatigue cracking of prestressing strands is based in AASHTO LRFD Article 5.5.3.2, which is based on fatigue
on testing of tendons in air. Application of the resultant fatigue tests reported in NCHRP Report 164 (Helgason et al. 1976).
resistance to concrete members with prestressing tendons is
questionable (Hanson et al. 1970; Tachau 1971; Warner and Steel Structures
Hulsbos 1966). Thus, the uncertainty of the fatigue resistance Finite-Life Fatigue
of prestressing tendons in concrete members is not well doc- The statistical bias and coefficient of variation of finite-life
umented. In addition, the determination of stress ranges in steel fatigue resistances are relatively well defined. NCHRP
cracked prestressed concrete members is complicated and Report 286 (Keating and Fisher 1986) summarizes the mean
beyond the normal prestressed concrete member design pro- finite-life fatigue resistance curves for the AASHTO detail
cedure (Abeles et al. 1969, 1974; Abeles and Brown 1971). The categories A through E′ and their standard deviations. The
uncertainty of this determination is also not well defined. In AASHTO nominal finite-life fatigue resistance curves,
defined in log-log space, are illustrated in Figure 2.2 (Fig-
ure C6.6.1.2.5-1 of AASHTO LRFD). The finite-life fatigue
Table 2.6. Prestressing Tendon resistances are represented by the sloping portions of the
Fatigue Resistance curves. The nominal fatigue resistance curves are determined
by subtracting two standard deviations from the mean curves.
Radius of Constant Amplitude
Curvature (ft) Fatigue Threshold (ksi)
The finite-life fatigue resistance (i.e., the allowable stress
range to reach a certain number of cycles) is defined by Equa-
>30 18 tion 2.4:
≤30 and >12 Linear interpolation
()
1
between 18 and 10
A 3
≤12 10 ∆σ = (2.4)
N

27
Source: American Association of State Highway and Transportation Officials.
Figure 2.2. Nominal fatigue resistances in AASHTO LRFD.
where A is a constant defined for each detail category, and

Table 2.8. Statistical Parameters
N is the number of cycles to failure.
for Finite-Life Fatigue Resistance
The current constant, A, is tabulated for each detail category
in Table 2.7 for the mean finite-life fatigue resistance. Coefficient of
The current estimates of uncertainty for finite-life fatigue Detail Category Bias Variation
resistances are tabulated in Table 2.8. A 2.8 0.59
B 2.0 0.71
Infinite-Life Fatigue
The uncertainty of statistical parameters for infinite-life B′ 2.4 0.67
fatigue resistances is not well defined. The infinite-life fatigue C 1.3 0.83
resistance is defined by a constant amplitude fatigue thresh- C′ 1.3 0.83
old for each detail category. These thresholds, used for design, D 1.6 0.77
are tabulated in Table 2.9 (Table 6.6.1.2.5-3 of the AASHTO
E 1.6 0.77
LRFD).
These threshold values were not determined as rigorously E′ 2.5 0.63
as the finite-life curves discussed above because experimental
Table 2.9. Nominal Constant

Table 2.7. Constant A Amplitude Fatigue Thresholds
for Mean Fatigue
Resistance Nominal Constant
Amplitude Fatigue
Detail Category A (108) Detail Category Threshold (ksi)
A 700 A 24
B 240 B 16
B′ 146 B′ 12
C 57 C 10
C′ 57 C′ 12
D 35 D 7
E 18 E 4.5
E′ 10 E′ 2.6

28
testing near the threshold is time consuming and costly. Con- been agreed that the bar diameter is not a major variable. For
servative thresholds were estimated graphically by using limited engineering practice, equations in the ACI 318-08 Code (ACI
experimental test observations. Thus, the uncertainty of these Committee 318 2008) and AASHTO LRFD (2012) are used
threshold values is not defined. to control cracking. The corresponding provisions are dis-
A Delphi process was employed to investigate the un- cussed below.
certainty of the infinite-life fatigue resistances represented by
the constant fatigue thresholds. At the winter 2010 meeting of Crack Control Reinforcement
the Bridge Task Force, in conjunction with the winter 2010 This section reviews previous research studies on control of
meeting of AASHTO Technical Committee T-14, the topic of cracking and predicting crack width in concrete members. A
the uncertainty of the thresholds was discussed. As the same significant amount of research has been conducted to investi-
characteristics that influence the uncertainty of the finite-life gate crack control in concrete members. The research resulted
fatigue resistance of welded details influence the uncertainty in the development of numerous equations to predict the
of the infinite-life fatigue resistance, the Bridge Task Force crack width on the tension surface and the side faces at the
concluded that the statistical parameters associated with the level of reinforcement. Equations available to predict crack
well-defined finite-life fatigue resistance (i.e., the bias and width were developed for concrete members with cover less
coefficient of variation) would be assumed appropriate for than 2.5 in. and are not applicable for beams with larger con-
the infinite-life fatigue resistance, as well. crete cover. Different equations have been adopted by differ-
With this assumption, the mean values of infinite-life ent codes. However, for calibration purposes, these equations
fatigue resistance are tabulated in Table 2.10 below. were evaluated with regard to accuracy and applicability. The
The statistical parameters for infinite-life fatigue resis- results from various equations were compared and validated
tance are those tabulated in Table 2.8 for finite-life fatigue using data collected from available literature.
resistance. One of the early studies by Clark (1956) included testing
58 specimens and collecting over 105 crack-width readings.
Clark concluded that the average crack width is closely related
Cracking in Concrete Structures to the following parameters: (1) the diameter of the reinforcing
Cracking in concrete structures is controversial but must bar, (2) the total reinforcement ratio, (3) the area of the beam
be controlled for aesthetic purposes, durability, and corro- section, and (4) the distance from the bottom reinforcement
sion resistance. Cracking is primarily caused by flexural and to the beam bottom surface. Clark stated that the average
tensile stresses, but also from temperature, shrinkage, shear, width was also proportional to the stresses in the reinforcing
and torsion. Although researchers do not agree on any single bars beyond the cracking stress. He suggested that the width
crack-width spacing, the most significant parameters to con- of the cracks can be reduced by using a large number of
small-diameter bars and by increasing the amount of the steel
trol cracking are widely agreed on. The most sensitive factor
reinforcement. On the basis of these results, Equation 2.5 was
is the reinforcing steel stress, followed by concrete cover, bar
developed to predict the average crack width of the concrete
spacing, and the area of concrete surrounding each bar. It has
beams. The maximum crack width was estimated by multi-
plying the average crack width by 1.64 (Clark 1956).
Table 2.10. Mean Infinite-Life D 1 
Fatigue Resistance w ave = C1  f s − C2  + n  (2.5)
p  p 
Constant
Amplitude Fatigue where
Detail Category Threshold (ksi) wave = average width of cracks (in.);
A 67 C1, C2 = coefficients that depend on distribution of bond
B 32
stress, bond strength, and tensile strength of con-
crete; for Clark’s study, C1 = 2.27 × 10-8 (h - d)/d,
B′ 29
C2 = 56.6;
C 13 D = diameter of reinforcing bar (in.);
C′ 16 p = As/Ae = cross-sectional area of reinforcement/
D 11 cross-sectional area of concrete;
Ae = bd (in.2);
E 7
b = width of component (in.);
E′ 6
fs = computed stress in reinforcement (psi);

29
n = ratio of modulus of elasticity of steel to concrete The maximum concrete cover tested in this study was
(assumed to be 8 in Clark’s study); 3.31 in.; however, only three test specimens over 2.5-in. cover
h = overall depth of beam/slab (in.); and were tested.
d = distance from compressive face of beam/slab to In the study by Frosch (1999), crack widths were determined
centroid of longitudinal tensile reinforcement. from an equation developed from a physical model. Results
were compared with the test data used in Kaar and Mattock
Kaar and Mattock (1963) also developed a well-known (1963) and Gergely and Lutz (1968). The crack-width model
crack-width equation for bottom face cracking, as given by developed in this study showed that the crack spacing and
Equation 2.6: width are functions of the distance between the reinforcing
steel bars. Crack control can be achieved by limiting the spac-
wb = 0.115βf s 4 A (2.6) ing of these reinforcing bars. On the basis of these research
findings, Frosch (1999) suggested that limiting the maximum
where bar spacing would prevent large cracks in concrete beams.
wb = maximum crack width (taken as 0.001 in.); The equation to calculate the maximum crack width for
b = ratio of distances to neutral axis from extreme tension uncoated reinforcement was developed on the basis of the
fiber and from centroid of reinforcement; physical model as shown by Equation 2.8 (Frosch 1999):
fs = steel stress calculated by elastic cracked-section the-
()
ory (ksi); and 2 f s  2 s 2
wc = β  dc + (2.8)
2 
A = average effective concrete area around reinforcing bar, 
Es
having same centroid as reinforcement (in.2).
Broms (1965) conducted tests on 37 tension and 10 flexural where
members to analyze crack width and crack spacing. Broms s = maximum permissible bar spacing (in.);
observed that crack spacing decreased rapidly with increasing wc = limiting crack width (in.) [0.016 in., based on ACI
load, and a number of primary tensile cracks formed on the 318-95 (ACI Committee 318 1995)];
surface of flexural and tension members. Secondary tensile Es = elastic modulus of steel reinforcement (can be taken
cracks were confined to the surrounding area of reinforce- as 29,000 ksi);
ment. The study concluded that the absolute minimum visible b = 1.0 + 0.08dc;
crack spacing is the same as the distance from the surface to dc = bottom cover measured from center of lowest bar
(in.); and
the center of the reinforcing bar located nearest to the surface
fs = stress in steel reinforcement.
of the member. Thus, the theoretical minimum crack spacing
is equal to the thickness of the concrete cover (Broms 1965). Frosch (1999) suggested that for epoxy-coated reinforce-
Gergely and Lutz (1968) developed an equation to predict ment, Equation 2.8 (for uncoated reinforcement) should be
the crack width based on a detailed statistical assessment of multiplied by a factor of 2. Equation 2.8 has been rearranged
experimental data available in the literature at the time. Gergely to solve for the allowable uncoated bar spacing, as shown in
and Lutz identified various parameters, such as reinforcing bar Equation 2.9:
locations, stresses in the reinforcement, concrete cover depth,
 w E  
2
and spacing of the reinforcement, as the controlling factors
affecting the crack width. The Gergely and Lutz equation is s = 2   c s  − dc2 (2.9)
  2 f sβ  
presented as shown in Equation 2.7:
The following design recommendation, which was based
wb = 0.076βf s 3 Adc (2.7) on the physical model and addresses the use of both uncoated
and coated reinforcement, was presented. The equation to
where calculate the maximum spacing of reinforcement was given
wb = maximum crack width (taken as 0.001 in.); as shown by Equation 2.10 (Frosch 1999):
b = ratio of distances to neutral axis from extreme tension
fiber and from centroid of reinforcement;  d 
fs = steel stress calculated by elastic cracked-section the- s = 12α s 2 − c  ≤ 12α s (2.10)
 3α s 
ory (ksi);
dc = bottom cover measured from center of lowest bar where
(in.); and
A = average effective concrete area around reinforcing bar, 36
αs = γc
having same centroid as reinforcement (in.2). fs

30
dc = thickness of concrete cover measured from extreme The study of the physical model showed that sections with
tension fiber to center of bar or wire located closest an effective depth of 36 in. and covers up to 3 in. can be
thereto, in.; designed without skin reinforcement. For thicker covers, the
s = maximum spacing of reinforcement (in.); maximum effective depth not requiring skin reinforcement
as = reinforcement factor; should be decreased. Maximum effective depth decreases for
gc = reinforcement coating factor: 1.0 for uncoated reinforce- covers thicker than 3 in. for Grade 60 reinforcement, resulting
ment, 0.5 for epoxy-coated reinforcement, unless test in the maximum depth (d = 36 in.).
data can justify a higher value; and To prevent excessive cracks throughout the depth of the
fs = calculated stress in reinforcement at service load (ksi). section, maximum spacing of the reinforcement should be
determined. According to Frosch (2002), the placement of
The calculated stress in reinforcement at service load ( fs)
the first bar is the most critical for the spacing of the skin
should be computed as the moment divided by the product
of steel area and internal moment arm; fs should not exceed reinforcement. The maximum crack width (ws) was calcu-
60% of the specified yield strength fy. lated halfway between the primary reinforcement and the
Frosch (2001) summarized the physical model for cracking first skin reinforcement bar at a distance x = s/2, yielding
and illustrated the development and limitations of the pro- Equation 2.12:
posed design method. He recommended formulas for calculat-
()
2
ing the maximum crack width for uncoated and epoxy-coated fs s
ws = 2 ds2 + (2.12)
reinforcement, as well as the design recommendation for their Es 2
use, similar to those in Frosch (1999).
In general, the largest crack widths are expected at the For sections with skin reinforcement, it is necessary to
extreme tensile face of the beam. However, Beeby (1979) con- determine the location in the section at which the reinforce-
ducted studies that showed the largest crack widths in the ment can be discontinued. As crack widths are controlled by
web along the beam side face occurred at about midheight. skin reinforcement below its end point, it is necessary to calcu-
Frosch (2002) conducted research on the modeling and con- late the maximum distance where the skin reinforcement can
trol of cracking on the side face of concrete beams. The study be eliminated. The maximum crack width will occur approxi-
showed that to provide adequate crack control, the maximum mately halfway between the neutral axis and the location of
skin reinforcement spacing is a function of the side cover. It the first layer of skin reinforcement at a distance x = sna/2 from
was also shown that a maximum bar spacing of 12 in. pro- the neutral axis (Frosch 2002). The maximum crack width can
vides reasonable crack control for up to 3 in. of concrete be calculated with Equation 2.13 based on the physical model
cover. The crack model developed by Frosch (2002) allows for developed by Frosch (2002):
the calculation of the crack width at any location along the
(dε− c) d + (s2 )
2
cross section. A profile of the crack width through the depth
of the section is more easily created and allows for informa- w s = sna s 2
s
na
(2.13)
tion regarding optimum locations for placing skin reinforce-
ment for the purpose of controlling side face cracks. where sna is the maximum distance where the skin reinforce-
Frosch (2002) showed that the crack spacing and crack ment can be eliminated.
width along the side face are functions of the distance from Frosch (2002) recommended that the design formula
the reinforcement, so the crack can be controlled by adding should be based on a physical model to address the control
skin reinforcement and limiting the reinforcement spacing. of cracking in reinforced-concrete structures and to unify
As the maximum crack width was observed halfway between the design criteria for controlling cracking in side and ten-
the reinforcement and neutral axis, Equation 2.11 can be used sion faces. Frosch (2002) recommended the maximum
to solve for crack width wc at x = (d - c)/2: spacing of flexural tension reinforcement as given by Equa-
tion 2.14:
( )
2
1
w c = ε s ds2 + (d − c) (2.11)  d 
2 s = 12α s 2 − c  ≤ 12α s (2.14)
 3αs 
where
es = strain in steel reinforcement = fs/Es; where
ds = concrete cover for skin reinforcement (in.);
d = effective depth (in.); and 36
αs = ;
c = depth of neutral axis from compression face (in.). fs

31
dc = thickness of concrete cover (in.) (for bottom face average and maximum crack widths of reinforced-concrete
reinforcement, measured from extreme tension fiber members with thicker concrete covers at various locations.
to center of bar, and for skin reinforcement, mea- Choi and Oh (2009) studied crack widths in transversely
sured from side face to center of bar); posttensioned concrete deck slabs in box girder bridges. They
s = maximum spacing of reinforcement (in.); tested four full-scale concrete box girder segments and derived
as = reinforcement factor; and the maximum-crack-width equation from the testing data, as
fs = calculated stress in reinforcement at service load (ksi). given by Equations 2.17 and 2.18:
The fs value should be computed as the moment divided by 0.75
 At,eff  h−x
the product of steel area and internal moment arm; fs should w max = 3 × 10−6 ( f s − f0 ) φs  (2.17)
not be more than 60% of the specified yield strength fy.  Ast + ξApt  d−x
Skin reinforcement is required along both side faces of a
member for a distance d/2 from the nearest flexural tension τ ap π + (n − 1) φs
ξ= (2.18)
reinforcement if the effective depth exceeds the depth calcu- τ as nπ φp
lated by Equation 2.15:
where
d = 42α s − 2dc ≤ 36α s (2.15)
Ast = total area of reinforcing bars (mm2);
Apt = total area of prestressing tendons (mm2);
Epoxy-coated reinforcement is widely used to increase the
At,eff = effective tensile concrete area (mm2);
durability of structures. The epoxy coating has been shown to
d = effective depth (mm);
decrease bond strength, which can decrease crack spacing
fs = increment of reinforcing bar stress after decompres-
and increase crack widths when compared with uncoated
sion (MPa);
reinforcement (Blackman and Frosch 2005). Blackman and
f0 = steel stress at initial occurrence of crack (MPa);
Frosch investigated crack widths in concrete beams by using
h = height of cross section (mm);
epoxy-coated reinforcement. The primary variables used in
n = number of strands in a flat duct;
the study included epoxy coating thickness and reinforcing
x = depth of neutral axis (mm);
bar spacing. Blackman and Frosch designed 10 slab specimens
wmax = predicted maximum crack width (mm);
to examine the effect of epoxy coating on cracks and con-
fs = diameter of reinforcing bar (mm);
cluded that the epoxy coating thickness did not significantly
fp = diameter of prestressing tendons (mm); and
affect the concrete cracking behavior. Frosch (1999, 2001, τ
ap = 0.465 for grouted posttensioned tendons.
2002) and Blackman and Frosch (2005) presented an equa- τ as
tion, given here as Equation 2.16, to compare the average
measured crack spacing for the uncoated and epoxy-coated Control of Cracks in Current Code Provisions
bars with the calculated values: The current code provisions specifying the distribution of
reinforcement are reviewed in this section.
Sc = ψs d p (2.16) ACI requirements for flexural crack control in beams and
thick one-way slabs are based on the statistical analysis of
where maximum-crack-width data from several sources (Gergely
Sc = crack spacing (in.); and Lutz 1968). ACI maintains that crack control is particu-
d* = controlling cover distance (in.); and larly important when reinforcement with yield strength over
ψs = crack spacing factor (1.0 for minimum crack spacing, 40,000 psi is used. Good detailing practices such as concrete
1.5 for average crack spacing, and 2.0 for maximum cover and spacing of reinforcement should lead to adequate
crack spacing). crack control even when reinforcement with a yield strength
Cracking of structures is rather common and is not always of 60,000 psi is used. ACI 318-08 Article 10.6 (ACI Commit-
damaging to the structure. However, when considering a tee 318 2008) does not distinguish between interior and
bridge deck, moderately sized cracks can be detrimental to exterior exposure because corrosion is not clearly correlated
the longevity of the structure due to the harsh environmental with surface crack widths in the range normally found at
exposure. Recently, increased concrete cover coupled with service-load levels. ACI 318-08 only requires that the spacing
high-performance concrete has become increasingly popular of reinforcement closest to the tension face (s) does not
because of its durability. However, this practice results in exceed that given by Equation 2.19
unrealistically small bar spacing and prevents the use of con-
temporary crack control practices that are based on statistical  40,000
s = 15  − 2.5cc (2.19)
studies. Thus, it is desirable to develop methods to predict  f s 

32
 40,000 Unlike ACI, AASHTO specifies exposure conditions to meet

but not greater than 12  , where cc is the least dis-
 f s  the needs of the authority having jurisdiction. The Class 1
tance from the surface of reinforcement or prestressing steel to exposure condition is based on a maximum crack width of
the tension face. If there is only one bar or wire nearest to the 0.017 in. and applies when cracks can be tolerated due to
extreme tension face, s in Equation 2.19 is the width of the reduced concerns of appearance or corrosion. This exposure
extreme tension face. These provisions are not sufficient for class can be thought of as an upper bound in regard to crack
structures subject to very aggressive exposure or designed to width for appearance and corrosion. The Class 2 exposure con-
be watertight. dition generally applies to decks and substructures exposed to
Special investigation is required for structures subject to water and any other components exposed to corrosive envi-
very aggressive exposure or designed to be watertight. ACI ronments. AASHTO LRFD (2008) also specifies requirements
318-99/318R-99 (ACI Committee 318 1999) limited the for skin reinforcement based on ACI 318-11 (ACI Committee
maximum spacing to 12 in., but this limitation was removed 318 2011). AASHTO Equation 5.7.3.4-1 (given here as Equa-
in ACI 318-08 (ACI Committee 318 2008). ACI 318-08 also tion 2.20) applies to both reinforced and prestressed concrete,
recommends the use of several bars at moderate spacing with specifications on the steel stresses used. In general, if the
rather than fewer bars at larger spacing to control cracking. AASHTO Class 2 exposure condition is used, AASHTO spac-
These provisions were updated recently to reflect the higher ings are less than those derived by the ACI equation. How-
service stresses that occur in flexural reinforcement with the ever, if the Class 1 exposure condition is used, ACI spacing
use of the load combinations introduced in ACI 318-02/ becomes more conservative.
318R-02 (ACI Committee 318 2002). The maximum bar
spacing to directly control cracking is specified. Similar rec- Principal Stresses in Webs of Segmental Concrete Bridges
ommendations have been stated for deep beams with the Okeil (2006) studied the allowable tensile stress for webs of
requirement of skin reinforcement. prestressed segmental concrete bridges by using a reliability-
AASHTO LRFD (2012) also provides provisions of rein- based approach. In this study, six prestressed segmental con-
forcement spacing to control flexural cracking. Like ACI, crete bridge designs were analyzed. Okeil stated that by
AASHTO emphasizes the importance of reinforcement detail- complying with the allowable tensile stresses, flexural crack-
ing and that smaller bars at moderate spacing tend to be more ing at the top and bottom fibers is controlled. However, for
effective than an equivalent area of larger bars. AASHTO the webs, cracks might develop due to a biaxial stress state
LRFD also agrees with ACI 318-08 on the most important resulting from a combination of shear and normal stresses.
parameters affecting crack width and specifies a formula To control shear cracking, the principal stress must be limited
for the distribution of reinforcement to control cracking. to an allowable tensile stress ( ft,all). This issue was addressed
The equation in AASHTO LRFD (2008) is based on the phy by the Florida DOT (Structures Manual 2013) and resulted in
sical crack model of Frosch (2001) rather than the statisti- a recommendation for the allowable tensile stresses to be
cally based model used in previous editions. The equation used in checking web tensile principal stress (s1). However,
(given here as Equation 2.20) limits bar spacing rather than the recommendation ignored the accompanying compressive
crack width: principal stress (s2), which has a significant effect on the ten-
700 γe sile strength of concrete. The objective of Okeil’s study was
s≤ − 2dc (2.20) to develop an allowable stress limit under which cracking in
βs f ss
webs of prestressed segmental bridges under service-load
conditions can be controlled.
where
Three equations were considered: ACI (ACI Committee
dc
bs = 1 + (the geometric relationship between 318 2005), Kupfer and Gerstle (1973), and Oluokun (1991),
0.7 (h − dc ) as shown in Equations 2.21 to 2.23, respectively:
crack width at tension face versus crack width at
reinforcement level);
f tu = 6.7 ( f c′)
0.5
(2.21)
ge = exposure factor (1.00 for Class 1 exposure, 0.75 for
Class 2 exposure);
f tu = 1.59 ( fc′)
0.67
dc = thickness of concrete cover measured from extreme (2.22)
tension fiber to center of the flexural reinforcement
f tu = 1.38 ( fc′)
0.69
located closest thereto, in.; (2.23)
fss = tensile stress in steel reinforcement at the SLS (ksi);
and where ftu is uniaxial tensile strength of concrete (psi), and f c′ is
h = overall thickness of depth of the component (in.). concrete compressive strength (psi).

33
Okeil (2006) concluded that Equation 2.23 provides a better Stress Limitations for Prestressing Tendons
estimate of the tensile strength over a wider range of concrete AASHTO LRFD (2012) provides stress limits for prestressing
compressive strengths. Using a biaxial state of stress and regres- tendons at various service conditions. These stress limits are
sion analysis, Okeil developed a relationship between the listed in Table 2.11.
tensile strength and the corresponding compressive strength, ACI 318-08 provides similar limits on the tensile stress in
as shown in Equation 2.24: prestressing tendons and rebars (ACI Committee 318 2008).
Major revisions to the limits were made in the 1983 version
σ tu σ of ACI 318 to incorporate the higher yield strength of low-
= 1 + 0.85 cu (2.24)
f tu fc′ relaxation wire and strand (ACI Committee 318 1983). The
ACI 318-08 stress limits for prestressing steel are listed as fol-
where scu and stu are the ultimate strengths of concrete under lows (ACI Committee 318 2008):
a compression–tension biaxial state of stress (psi).
Due to prestressing steel jacking force: 0.94fpy but not greater
By combining Equations 2.23 and 2.24, Equation 2.25 is than the lesser of 0.80fpu and the maximum value recom-
obtained: mended by the manufacturer of prestressing steel or anchor-
age devices.
 σ cu  Immediately after prestress transfer: 0.82fpy but not greater
σ tu = 1.38 ( fc′)
0.69
1 + 0.85 f ′  (2.25)
than 0.74fpu. Post-tensioning tendons, at anchorage devices
c
and couplers, immediately after force transfer: 0.70fpu.
After a detailed parametric study and reliability analysis, EN 1992-2 (Eurocode 2): Design of Concrete Structures
Okeil (2006) recommended an expression, given in Equa- (EN 1992-2 2005) restricts inelastic deformation of the steel
tion 2.26, for estimating the allowable tensile stress in the in concrete structures at the SLS to prevent large, perma-
webs of posttensioned segmental bridges under biaxial nently open cracks. In EN1992-2, at the SLSs, the stress limit
stresses: for prestressing steel is 0.75fpk after allowance for losses, where
fpk is the characteristic tensile strength of prestressing steel.
0.7  σ 
fct = 0.60 ( fc′) 1 + 0.85 2  (2.26) The exact meaning of characteristic tensile strength is not
 fc′ defined in EN1992-2 and is interpreted here as the specified
strength. This limit of 0.75fpk is listed in EN1992-2 Section 7.
where s2 is the principal stress in the centroidal stress block
in the web of a posttensioned segmental bridge. Concrete Tension Stresses
The findings of this study are limited to the range of con- The early discussion of cracking control is diverse. At the First
crete compressive strengths between 5 and 8 ksi. United States Conference on Prestressed Concrete in 1951,
Table 2.11. Stress Limits for Prestressing Tendons (AASHTO LRFD 2012)
Tendon Type
Stress-Relieved Deformed
Strand and Plain Low-Relaxation High-
Condition High-Strength Bars Strand Strength Bars
Pretensioning
Immediately before transfer (fpbt) 0.70fpu 0.75fpu —
At SLS after all losses (fpe) 0.80fpy 0.80fpy 0.80fpy
Posttensioning
Before seating, short-term fpbt may be allowed 0.90fpy 0.90fpy 0.90fpy
At anchorages and couplers immediately after 0.70fpu 0.70fpu 0.70fpu
anchor set
Elsewhere along length of member away from 0.70fpu 0.74fpu 0.70fpu
anchorages and couplers immediately after
anchor set
At SLS after losses (fpe) 0.80fpy 0.80fpy 0.80fpy
Note: — = not applicable.

34
some experts opined that a completely crackless concrete In the current ACI 318-11, Section 18.4.1 specifies the
member is only better for the specific purpose, but others allowable tensile stress in concrete immediately after prestress
thought that cracking of prestressed concrete beams is as transfer (before time-dependent prestress losses) as follows
important as yielding. In 1958, the Tentative Recommenda- (ACI Committee 318 2011):
tions for Prestressed Concrete proposed by ACI-ASCE Joint
Committee 323 suggested that prestressed concrete, before Where computed concrete tensile stresses, ft, exceeds 6 fci′
losses due to creep and shrinkage, should meet the following at ends of simply supported members, or 3 fci′ at other
limits (note units in the following provisions are in pounds locations, additional bonded reinforcement shall be pro-
per square inch for the allowable tensile stress): vided in the tensile zone to resist the total tensile force in
concrete computed with the assumption of an uncracked
3 fci′ for members without nonprestressed reinforcement; section.
6 fci′ for members with nonprestressed reinforcement pro-
vided to resist the tensile force in concrete; computed on the The AASHTO Standard Specifications for Highway Bridges
basis of an uncracked section. (1992) specified the allowable tensile stresses, before losses
due to creep and shrinkage, as follows:
The 1963 Building Code Requirements for Reinforced Con-
crete (ACI Committee 318 1963) included the recommenda- 200 psi or 3 fci′ for members in tension areas with no bonded
tion for the tensile stress limits proposed by ACI-ASCE Joint reinforcement;
Committee 323 (1958), with some modifications: [w]here the calculated tensile stress exceeds this value,
reinforcement shall be provided to resist the total tension
3 fci′ for members without auxiliary reinforcement in the force in the concrete computed on the assumption of
tension zone; uncracked section. The maximum tensile stress shall not
[w]hen the calculated tension stress exceeds 3 fci′ , rein- exceed 7.5 fci′ .
forcement shall be provided to resist the total tension force
in the concrete computed on the assumption of uncracked Table 2.12 shows the tensile stress limits and provisions of
section. AASHTO LRFD (2008).
The 1977 Building Code Requirements for Reinforced Con- Existing Limit States That Are Deterministic
crete modified the allowable tensile stress limit as follows or Represent Detailing Requirements
(ACI Committee 318 1977): The following limit states exist in AASHTO LRFD. Reviewing
the background of these limit states revealed that they are
6 fci′ for the extreme fiber stress in tension at ends of simply either deterministic or represent detailing requirements that
supported members; cannot be calibrated. No calibration is anticipated for these
3 fci′ for the extreme fiber stress in tension at other locations. limit states.
Table 2.12. Tensile Stress Limits in Prestressed Concrete at SLS After Losses,
Fully Prestressed Components (AASHTO LRFD Table 5.9.4.2.2-1 [2008])
Bridge Type Location Stress Limit
Other Than Segmentally Tension in the precompressed Tensile Zone Bridges, Assuming Uncracked Sections
Constructed Bridges For components with bonded prestressing tendons or reinforcement that are subjected to not 0.19 fc′ (ksi)
worse than moderate corrosion condition.
For components with bonded prestressing tendons or reinforcement that are subjected to severe 0.0948 fc′ (ksi)
corrosive conditions
For components with unbonded prestressing tendons No tension
Segmentally Con- Longitudinal Stresses Through Joints in the Precompressed Tensile Zone
structed Bridges Joints with minimum bonded auxiliary reinforcement through the joints sufficient to carry the calcu- 0.0948 fc′ (ksi)
lated longitudinal tensile force at a stress of 0.5 fy; internal tendons or external tendons
Joints without the minimum bonded auxiliary reinforcement through joints No tension
Transverse Stress Through Joints
Tension in the transverse direction in precompressed tensile zone 0.0948 fc′ (ksi)
Principal Tensile Stress at Neutral Axis in Web

All types of segmental concrete bridges with internal and/or external tendons, unless the Owner 0.110 fc′ (ksi)
imposes other criteria for critical structures.

35
Fatigue in Concrete Deck Slabs and Culvert Top Slabs Birrcher et al. (2009) concluded that “[c]rack control rein-
(AASHTO LRFD Article 5.5.3.1) forcement shall be distributed evenly near the side faces of
Stresses measured in concrete deck slabs of bridges and top the strut. Where necessary, interior layers of crack control
slabs of box culverts in service are far below infinite fatigue reinforcement may be used.”
life, most probably due to internal arching action.
AASHO Standard Specifications for Highway Bridges (1975) Control of Permanent Deformation
includes the background that led to waiving fatigue require-
ments for these components. Steel structures are subject to requirements intended to pre-
vent changes in riding quality and appearance resulting from
Fatigue of Reinforcement of Fully Prestressed Components permanent deflections in service. Starting with specifications
(AASHTO LRFD Article 5.5.3.1) for LFD in the early 1970s, steel structures have been subject
For fully prestressed components designed to have extreme to two limitations to guard against these undesirable behav-
fiber tensile stress due to a Service III limit state within the ten- iors. There is a requirement that the service-load stress under
sile stress limit specified in AASHTO LRFD Table 5.9.4.2.2-1, an overload be less than 95% of yield in a composite girder or
the fatigue limit state load factors, the girder distribution fac- 80% of yield in a noncomposite girder and that slip-critical
tors, and dynamic load allowance cause fatigue limit state stress connections be designed for the same overload requirement.
to be considerably less than the corresponding value deter- In LFD, the overload requirement was dead load plus 5/3 of
mined from Service III. For fully prestressed components, the the HS20 loading. Due to the increased demand of the HL-93
net concrete stress is usually significantly less than the concrete live load, the corresponding provisions in the AASHTO LRFD
tensile stress limit specified in AASHTO LRFD Table 5.9.4.2.2-1. are investigated at the Service II limit state, which involves a
As a result, the calculated flexural stresses are significantly load factor on live load of 1.30.
reduced. For this situation, the calculated steel stress range, The response of girder structures to excessive overloads
which is equal to the modular ratio times the concrete stress was one of several issues explored during the AASHO Road
range, is almost always less than the steel fatigue stress range Test of the late 1950s and early 1960s and documented in a
limit specified in AASHTO LRFD Article 5.5.3.3. series of reports issued by the Highway Research Board
(AASHO Road Test 1962), the predecessor of the Transpor-
Fatigue of Prestressing Tendons (AASHTO LRFD Article 5.5.3.3) tation Research Board. The structures of the AASHO Road
With fatigue in fully prestressed components waived (see Test were designed to undergo many repetitions to relatively
above), these provisions are only applicable to segmental high stresses.
bridges. Little data are available on the randomness of load Table 2.13 shows a summary of the initial stresses in steel
and resistance of segmental bridges. There is no evidence of bridges of composite and noncomposite construction. The
fatigue damage on these structures, so no changes are recom- nominal yield stress for the material in these bridges was
mended, and calibration will not be made. 33 ksi, so it can be seen that in many cases these bridges were
subjected to loads beyond the yield stress.
Crack Control Reinforcement for Components Designed Using
Table 2.13 also indicates the number of live load passages
Strut and Tie Model (AASHTO LRFD Article 5.6.3.6)
to which these structures were subjected. After the repetitions
Birrcher et al. (2009) proposed new provisions regarding
of actual truck loading, some of the structures were subjected
crack control reinforcement as follows: “The spacing of the
to further cycles of load to investigate fatigue through the use
bars in these grids shall not exceed the smaller of d/4 and 12.0
of eccentric mass dampers.
in.” Moreover, they continued, “The reinforcement in the ver-
In American Iron and Steel Institute Bulletin 15, Vincent
tical and horizontal direction shall satisfy the following
(1969) summarizes the basis for LFD of steel structures.
[shown here as Equation 2.27]”:
Bulletin 15 contains the following statement: “There is, how-
ever, a definite need for a control on the possibility of perma-
Av A nent deformations under infrequent overloads which may
≥ 0.003, h ≥ 0.003 (2.27) impair the riding quality of the bridges.” The establishment
bw sv bw sh
of the 80% and 95% criteria is demonstrated in Figure 2.3,
where taken from Bulletin 15, which shows the permanent set at
Av, Ah = total area of vertical and horizontal crack control midspan of several of the bridges from the AASHO Road
reinforcement within spacing sv and sh, respectively; Test and the corresponding ratio between test stress and
bw = width of member web (in.); and the actual measured yield point of the steel in the bridges.
sv, sh = spacing of vertical and horizontal crack control The two criteria for composite and noncomposite struc-
reinforcement, respectively. tures are seen to produce an accumulated displacement of

36
Table 2.13. Data from AASHO Road Test (1962)
Summary of Initial Stresses in Steel Bridges

Design Stress (ksi) Actual Stress (ksi) No. of Vehicle Passages
Center Exterior Interior Center Exterior To First
Bridge Beam Beam Beam Beam Beam Cracking Total
Noncomposite Bridges
1A 27.0 — 25.3 27.7 30.1 536,000 557,400
1B 34.8 — 32.5 35.4 40.5 — 235
2A 35.0 — 35.0 39.4 41.1 — 26
3A 27.3 — 28.6 30.9 35.4 — 392,400
4A 34.7 — 35.9 38.9 41.1 — 106
4B 34.7 — 39.1 42.1 42.3 — 106
9A — 27.0 22.9 24.7 25.5 477,900 477,900
9B — 27.0 24.0 24.6 26.0 477,900 477,900
Composite Bridges
2B 35.0 — 30.2 33.8 35.8 531,500 558,400
3B 26.9 — 26.0 28.8 31.0 535,500 557,800
Note: — = not available.
approximately 1 in. at the midspan of bridges of an approxi- providing generally the same, or even higher, level of over-
mately 50-ft span. Deflection measurements at various times load performance as was provided by LFD in most cases.
during the road test indicate that most structures accumu- Consider Figure 2.4, which shows the ratio of the HL-93
lated most of the eventual permanent set in the very early loading to the HS20 loading in the Standard Specifications.
repetitions of loading. The load factor in the Standard Specifications for this case
The provisions for control of permanent deformations in was 1.67; the current load factor for the AASHTO LRFD Ser-
steel structures were incorporated into AASHTO LRFD with vice II load combination is 1.3. That means that whenever
an adjustment for the increased live load with the intent of the moment ratio in Figure 2.4 is greater than 1.28, then the
Source: Vincent (1969). Reproduced with permission from the American Iron & Steel Institute.
Figure 2.3. Development of service stress limits.

37
Definition of SLS
The Eurocode (EN 1990 2002) defines SLSs as those concerning
• The functioning of the structure or structural members

under normal use;
• The comfort of users; and
• The appearance of the construction works.
The Eurocode (EN 1990 2002) includes requirements call-

ing for
• The serviceability requirements to be agreed on for each

individual project;
Source: American Association of State Highway and
Transportation Officials. • A distinction to be made between reversible and irrevers-
ible serviceability limit states; and
Figure 2.4. Ratio of HL-93 moment to • The verification of SLS based on criteria concerning the
HS20 moment.
following aspects:
a. Deformations that affect
current demand is higher than that required by the Standard –– The appearance,
Specifications. –– The comfort of users,
Several issues arose regarding retention of, or revisions to, –– The functioning of the structure (including the
the provisions related to control of permanent deformations; functioning of machines or services), or
these are discussed in Section 6.4. –– That cause damage to finishes of nonstructural
members.
b. Vibrations
2.3.2 Eurocode –– That cause discomfort to people, or
–– That limit the functional effectiveness of the
The Eurocode contains the following sections, to which refer-
ence is made in subsequent sections of this report: structure.
c. Damage that is likely to adversely affect
–– The appearance,
• EN 1990 (Eurocode 0): Basis of Structural Design
–– The durability, or
• EN 1991 (Eurocode 1): Actions on Structures
–– The functioning of the structure.
• EN 1992 (Eurocode 2): Design of Concrete Structures
• EN 1993 (Eurocode 3): Design of Steel Structure
In the context of serviceability, the Eurocode considers the
• EN 1994 (Eurocode 4): Design of Composite Steel and Concrete
term appearance to be concerned with such criteria as high
Structures
deflection and extensive cracking, rather than aesthetics
• EN 1995 (Eurocode 5): Design of Timber Structures
(EN 1990 2002).
• EN 1996 (Eurocode 6): Design of Masonry Structure
• EN 1997 (Eurocode 7): Geotechnical Design
• EN 1998 (Eurocode 8): Design of Structures for Earthquake
Background on the Eurocode’s Reliability Basis
Resistance
• EN 1999 (Eurocode 9): Design of Aluminum Structures The Eurocode specifies that structures be designed for a par-
ticular design working life (EN 1990 2002). The design work-
These Eurocode sections allow the user countries to incorpo- ing life is defined as the period for which a structure is assumed
rate country-specific requirements through the incorporation to be usable for its intended purpose with anticipated mainte-
of a national annex. nance but without major repair being necessary. Examples of
The Eurocode replaced most previous country specifica- design working life are given in Table 2.14.
tions, such as the German Institute for Standardization and The levels of reliability relating to ULS and SLS can be
the British BS5400, and it is expected to eventually replace all achieved by suitable combinations of protective measures
other European Union member country specifications. It is (e.g., protection against fire or corrosion), measures relating
assumed that the requirements of the Eurocode encompass to design calculations (e.g., choice of partial factors), mea-
those of the previous specifications and, thus, no other Euro- sures relating to quality management, measures aimed to
pean specifications were reviewed. reduce errors in design (e.g., project supervision), and

38
Table 2.14. Design Working Lives Table 2.16.

Multiplication
Design Factor (KF1)
Working Life Design Working for Reliability
Category Life (years) Examples
Differentiation
1 10 Temporary structures
Reliability Class KF1
2 10–25 Replaceable structural parts
(e.g., gantry girders, RC1 0.9
bearings)
RC2 1.0
3 15–30 Agricultural and similar
structures RC3 1.1
4 50 Building structures and other

common structures
5 100 Monumental building struc- Table 2.17 summarizes the probabilities of failure (Pf) inher-
tures, bridges, and other ent to the Eurocode and the AASHTO LRFD for ULSs, along
civil engineering structures with the corresponding reliability indices (b) below them in
Source: Adapted from Table 2.1 of EN 1990 (EN 1990 2002). italics. The defining probabilities of failure in the case of the
Eurocode and the defining reliability indices for the AASHTO
LRFD are shown in boldface.
execution (construction) of the structure (e.g., inspection
during execution) and other kinds of measures. SLS Reliability
The Eurocode defines three levels of consequences classes The SLSs of the Eurocode are categorized as reversible and irre-
(CC1, CC2, and CC3), as defined in Table 2.15. Three reli- versible. Reversible SLSs are those for which no consequences
ability classes (RC1, RC2, and RC3) may be associated with remain once a load is removed from a structure. For example,
the three consequence classes. a crack-width limit state with a sufficiently small size is a revers-
The vast majority of bridges are designed to CC2, with ible limit state, but one defined by a high width (e.g., 0.5 mm)
CC3 a possibility only for those bridges with very high con- is irreversible because, if the crack width is high enough, once
sequences of failure, such as a signature bridge. the live load is removed the crack does not close completely.
The provisions of the Eurocode, specifically EN 1990 The irreversible SLSs, which do not concern the safety of
(EN 1990 2002) with the partial factors given in Annex A1 and the traveling public, are calibrated to a higher probability of
EN 1991 to EN 1999, yield designs consistent with reliability failure and corresponding reliability index than the strength
class RC2. The Eurocode uses the multiplication factors (KF1) limit states, as shown in Table 2.18.
given in Table 2.16 applied to load factors to differentiate the
three reliability classes. Other measures (e.g., differing levels of SLS Load Combinations
quality control) in lieu of modifying the load factors are some- EN 1990 (2002) includes three types of load combinations for
times preferred. the SLSs: characteristic combination, frequent combination,
and quasipermanent combination. Table 2.19 summarizes
the Eurocode SLS load combinations.
Table 2.15. Eurocode Consequence Classes
Consequence Description Related to Reliability

Serviceability Design Basic Approach
Class Consequences Class
Basic Equation
CC1 Low consequence for loss of RC1 The basic equation in the Eurocode (EN 1990 2002) for verify-
human life; economic, social, or ing that an SLS is satisfied is
environmental consequences
small or negligible
Ed ≤ Cd
CC2 Moderate consequence for loss of RC2
human life; economic, social, or
environmental consequences where
considerable Cd = is the limiting design value of the relevant service-
CC3 Serious consequences for loss of RC3
ability criterion and
human life or for economic, Ed = is the design value of the effects of actions specified in
social, or environmental concerns the serviceability criterion, determined on the basis
Source: Adapted from Table B1 of EN 1990 (EN 1990 2002). of the relevant combination.

39
Table 2.17. Target Probabilities of Failure (Pf ) and Target Reliability Indices (bT)
Reference Period (years)
Code 1 50 75 100 120
Eurocode CC2 1.00E-06 5.00E-05 7.50E-05 1.00E-04 1.20E-04

(KF1 = 1.0)
4.75 3.89 3.79 3.72 3.67
CC3 1.00E-07 5.00E-06 7.50E-06 1.00E-05 1.20E-05
(KF1 = 1.1)
5.20 4.42 4.33 4.26 4.22
AASHTO LRFD Typical bridges 2.67E-06 1.33E-04 2.00E-04 2.67E-04 3.20E-04
(hI = 1.0)
4.55 3.65 3.50 3.46 3.41
Important bridges 9.60E-07 4.80E-05 7.20E-05 9.60E-05 1.15E-04
(hI = 1.05)
4.76 3.90 3.80 3.73 3.68
Table 2.18. Irreversible SLS of construction works or agreed with the client or the national
Target Probabilities of Failure and authority.
Corresponding Reliability Indices
Combination of Actions (Load Combinations)
Reference Period The combinations of actions (load combinations) for service-
(years)
ability limit states in the Eurocode are defined symbolically by
Reliability Class 1 50 Equation 2.28, which is the characteristic (rare) combination;
RC2 1.00E-03 1.00E-01 Equation 2.29, which is the infrequent combination; Equa-
2.9 1.5
tion 2.30, which is the frequent combination; and Equation 2.31,
which is the quasipermanent combination. The characteristic
Source: Adapted from Table C2 of EN 1990
(Eurocode 0) (EN 1990 2002). combination (Equation 2.28) is normally used for irreversible
limit states; the frequent combination (Equation 2.30) is nor-
mally used for reversible limit states.
Serviceability Criteria
Specific serviceability criteria such as crack width, stress or  
strain limitation, and slip resistance exist in separate sections Ed = E ∑ Gk , j + Pk + Qk ,1 + ∑ c0,i i Qk ,i  (2.28)
of the Eurocode (EN 1991 to EN 1999). In addition to these  j ≥1 i >1 
requirements, project-specific deformations to be considered
 
in relation to serviceability requirements are required to be as Ed = E ∑ Gk , j + Pk + c1,1
′ i Qk ,1 + ∑ c1,i i Qk ,i  (2.29)
detailed in relevant code annexes in accordance with the type  j ≥1 i >1 
Table 2.19. SLS Combinations
SLS Load Combination
Acceptance of
Type Description Type Infringement Example
Reversible Limit states that will not be Frequent Specified duration and Crack-width limit state of a prestressed con-
exceeded when the actions frequency of infringe- crete beam with bonded tendons charac-
that caused the infringement ments are accepted terized by a 0.2-mm crack width
are removed
Quasipermanent Specified long-term Crack-width limit state for a reinforced-
infringement is concrete or prestressed-concrete beam
accepted with unbonded tendons characterized by
a 0.3-mm crack width
Irreversible Limit states that remain per- Characteristic No infringement accepted Crack-width limit state characterized by a
manently exceeded after the (5% probability 0.5-mm crack width, because such a wide
actions that caused the of exceedance) crack cannot completely close once the
infringement are removed loads that caused it are removed

40
  be exceeded is a large fraction of the reference period. It

Ed = E ∑ Gk , j + Pk + c1,1 i Qk ,1 + ∑ c1,i i Qk ,i  (2.30) may be expressed as a determined part of the characteristic
 j ≥1 i >1  value by using a factor (2 ≤ 1.0).
 
Ed = E ∑ Gk , j + Pk + ∑ c 2,i i Qk ,i  (2.31) The Eurocode allows some of the above expressions to be
 j ≥1 i >1  modified and gives detailed rules in relevant sections of the
code (parts of EN 1991 to EN 1999). Each Eurocode country
where has its own national annex in which country-specific require-
Gk,j = characteristic (extreme) value of permanent ments are placed; thus, the Eurocode allows each country to
action j; specify its own serviceability criteria in its national annex.
Gkj,sup/Gkj,inf = upper/lower value of permanent action j; Recommended values of the  factors for different types of
P = relevant prestressing value of prestressing structures (e.g., buildings, highway bridges, or railway bridges)
action; are tabulated in the Eurocode. Table 2.20 shows the recom-
Qk,1 = characteristic value of the leading (domi- mended values for highway bridges.
nant) Variable Action 1;
Qk,i = characteristic value of the accompanying Note 1: The  values may be set by the National Annex. Rec-
Variable Action 1; ommended values of  factors for the groups of traffic loads
0 = factor for characteristic value of a variable and other more common actions are given in
action;
1 = factor for frequent value of a variable action; • Table A2.1 for road bridges;
and • Table A2.2 for foot bridges; and
2 = factor for quasipermanent value of a variable • Table A2.3 for railway bridges.
action.
The terms in Equations 2.28 through 2.31 are further defined Note 2: When the National Annex refers to the infrequent
as follows: combination of actions for some serviceability limit states of
concrete bridges, the National Annex may define the values
• effect of action (E): Effect of actions (or action effect) on of 1infq. The recommended values of 1infq are
structural members (e.g., internal force, moment, stress,
strain) or on the whole structure (e.g., deflection, rotation). • 0,80 for gr1a (LM1), gr1b (LM2), gr3 (pedestrian loads),
• permanent action (G): Action that is likely to act through- gr4 (LM4, crowd loading), and T (thermal actions);
out a given reference period and for which the variation in • 0,60 for FW in persistent design situations; and
magnitude with time is negligible, or for which the varia- • 1,00 in other cases (i.e., the characteristic value is substi-
tion is always in the same direction (monotonic) until the tuted for the infrequent value).
action attains a certain limiting value.
• variable action (Q): Action for which the variation in Note 3: The characteristic values of wind actions and snow
magnitude with time is neither negligible nor monotonic. loads during execution are defined in EN 1991-1-6 (2005).
• characteristic value of a variable action (0 Qk): Value When relevant, representative values of water forces (Fwa)
chosen (insofar as it can be fixed on statistical bases) so may be defined for the individual project.
that the probability that the effects caused by the combi-
nation will be exceeded is approximately the same as by
the characteristic value of an individual action. It may be Existing Limit State
expressed as a determined part of the characteristic value by A summary of the SLS requirements in the Eurocode is in
using a factor (0 ≤ 1.0). Appendix A.
• frequent value of a variable action (1 Qk): Value deter-
mined (insofar as it can be fixed on statistical bases) so that
either the total time within the reference period during 2.3.3 Canadian Highway Bridge Design Code
which it is exceeded is only a small given part of the refer-
Background
ence period, or the frequency of its being exceeded is lim-
ited to a given value. It may be expressed as a determined The CHBDC (2006) and earlier Ontario Highway Bridge Design
part of the characteristic value by using a factor (1 ≤ 1.0). Code (1991) cover ULS and SLS. The serviceability limit states
• quasipermanent value of a variable action (2 Qk): Value in the CHBDC include fatigue, deflection, cracking, and com-
determined so that the total period of time for which it will pressive stress in concrete. The SLS acceptability criterion was

41
Table 2.20. Eurocode Recommended Values of  Factors for

Highway Bridges
Action Symbol 0 1 2
Traffic loads (EN gr1a (LM1 + pedes- TS 0,75 0,75 0

1991-2 Table 4.4 trian or cycle-track
UDL 0,40 0,40 0
[EN 1991-2 2003]) loads)a
Pedestrian + cycle-track 0,40 0,40 0
loadsb
gr1b (single axle) 0 0,75 0
gr2 (horizontal forces) 0 0 0
gr3 (pedestrian loads) 0 0 0
gr4 (LM4—crowd loading) 0 0,75 0
gr5 (LM3—special vehicles) 0 0 0
Wind forces Fwk
Persistent design situations 0,6 0,2 0
Execution 0,8 — 0
F *W 1,0 — —
Thermal actions Tk 0,6 c 0,6 0,5
Snow loads Qsnk (during execution) 0,8 — —
Construction loads Qc 1,0 — 1,0
a
The recommended values of 0, 1, and 2 for gr1a and gr1b are given for roads with traffic corresponding to
adjusting aQi, aqi, aqr, and bQ equal to 1. Those relating to unified distribution load (UDL) correspond to the most
common traffic scenarios, in which an accumulation of lorries can occur, but not frequently. Other values may be
envisaged for other classes of routes, or of expected traffic, related to the choice of the corresponding a factors.
For example, a value of 2 other than zero may be envisaged for the UDL system of LM1 only, for bridges
supporting a severe continuous traffic. See also EN 1998-2 (2005).
b The combination value of the pedestrian and cycle-track load, which is mentioned in Table 4.4a of EN 1991-2
(2003), is a “reduced” value. 0 and 1 factors are applicable to this value.

c The recommended  value for thermal actions may in most cases be reduced to zero for ULSs EQU, STR, and
0
GEO. See also the design Eurocodes.

Source: Adapted from Table A2.1 of EN 1990 (EN 1990 2002).
determined by reference to past practice. As an example of period of 3 weeks is acceptable. However, the group did not
this process, special consideration was given to the tensile feel strongly about it, so they agreed that the target probabil-
stress limit state in prestressed concrete girders. The accept- ity of exceeding this limit state is 50%, which corresponds to
ability criterion was formulated in terms of the minimum the target reliability index (bT = 0).
return period for exceeding the decompression moment. It
was assumed that the girders will crack due to shrinkage
Existing Limit States
before installation or under exceptionally heavy trucks and
that the crack will reopen each time the decompression In general, the SLSs in the CHBDC are very similar to the
moment is exceeded. An open crack, even for a fraction of a SLSs currently specified in AASHTO LRFD. There are some
second, is assumed to allow water with salt or other pollutants differences in application, but the general phenomena being
to penetrate and eventually reach the rebar and prestressing treated are basically the same. No new limit states that do not
steel, resulting in corrosion, delamination, spalling of con- exist in AASHTO LRFD were found in the 2006 CHBDC.
crete, and girder failure. The minimum acceptable return CHBDC Clause 3.5.1 and Table 3.1, in particular, contain
period for exceeding the decompression moment was then the requirements for load factors and load combinations.
determined by a group of experts invited by the Code Con- Table 3.6.1(a) lists only two load combinations for service-
trol Committee using a process of expert elicitation (Delphi ability limit states. Service-load combinations use a load fac-
process). The group was asked to provide their expert opin- tor of 0.9 for the live load that is based on the CL-W-625 truck
ion. They deliberated and came to a conclusion that a return (140.5 kips, 59 ft long) or lane loading. The CL-W-625 truck

42
is considerably larger than the HL-93 truck alone (i.e., without Clause 8.12.3.2 provides guidance on calculating the crack
the uniform distributed load). Load Combination 2 applies to width and spacing based on parameters that include the aver-
superstructure vibration only. The CHBDC also specifies a age strain in the reinforcing. A distinction is made for epoxy-
lane load that consists of 80% of the axles of the CL-W truck coated reinforcement, for which the calculated crack width is
superimposed on a UDL of 9 kN/m, which is similar to the increased 20%.
UDL used with the HL-93 loading. Clause 10.5.3.1 specifies serviceability limit states for steel
CHBDC Clause 6.4.1.3 deals with serviceability limit states structures; these include deflection, yielding, slipping of bolted
and foundations. Three criteria are noted: joints, and vibrations.
Clause 10.5.3.2 for deflections is a cross reference for
• Foundation deformations that cause SLS limits to be Clause 10.16.4, which applies to orthotropic decks only.
exceeded; Clause 10.5.3.3 deals with the prevention of general yield-
• Deformations that cause the riding surface or transitions ing at the SLSs, which appears to pertain to Clause 10.11.4
between the approaches and the bridge to become un- (permanent deflections for composite sections). The latter is
acceptable; and similar to the AASHTO overload requirements, except that
• Deformations that cause unacceptable structural mis the CHBDC load factor for live load is 0.9 as opposed to the
alignment, distortion, or tilting. AASHTO load factor of 1.3. As discussed in Chapter 6, the net
result is probably similar because of the heavier CHBDC live
Clause 7.6.5.2 deals with construction requirements for load. This is not a new limit state, although the numerical
pipe arches and limits to downward deflection. The commen- values might differ somewhat from AASHTO.
tary reinforces that this is a construction requirement rather Clause 10.11.3 is an SLS for differential shrinkage between
than a design control. restrained and free shrinkage of concrete and steel composite
Clause 7.7.5.2 speaks to upward or downward crown deflec- members.
tion during construction of metal box structures and pro-
vides a 1% requirement. Little additional information is
provided in the commentary, which notes that AASHTO 2.3.4 Japanese Geotechnical Society
Article 12.8.5.3 has limits for live load deflection.
Foundation Design Guideline
Clause 8.5.1 states that cracking, deformation, stress, and The Japanese Geotechnical Society (JGS) prepared a draft foun-
vibration SLSs should be considered. dation design guideline in 2002. This document attempts to
Clause 8.5.2 specifies serviceability limit states for concrete phrase the structural and geotechnical design principles follow-
structures and indicates that these are cracking, deforma- ing the general requirements of ISO 2394. Three limit states are
tions, stress, and vibration. defined on the basis of the following functional statements:
Clause 8.5.2.2 deals with a cross reference to Clause 8.12
with some limits on earth cover. • With respect to the various magnitudes and frequencies of
Clause 8.5.2.3 deals with deformation provisions and indi- loading during the expected service life, the structures shall
cates that short-term and long-term deformations may affect satisfy structural performance as characterized by structural
the function of the structure. strength, stability, deformability, and durability, including
Clause 8.5.2.4 deals with stresses in the component not serviceability, repairability, and safety with appropriate levels
exceeding certain values of Clauses 8.7.1, 8.8.4.6, and 8.23.7. of reliability.
Clause 8.5.2.5 deals with vibrations and refers back to • The structures shall be designed to be sufficiently safe so as
clauses in Section 3 on loads. to prevent serious injury to occupants and surrounding
The commentary for Clause 8.5.2.1 speaks to the fact that, in personnel during all possible design situations through the
general, nonprestressed and partially prestressed components design working life. This functional statement is related to
are expected to crack under the service loads. The commentary the topic of safety.
indicates that it is generally a good practice to provide sufficient • The structures may be designed, by judgment of the owner
prestress so that under permanent loads, any cracks previously based on the importance of the structure, such that normal
caused due to the application of live load are closed under per- functions are preserved (serviceability) or damage is lim-
manent loads to enhance durability. ited within a certain tolerable level (repairability) against
Clause 8.12 deals with control of cracking by specifying specified loading conditions during the design working life
distribution requirements and a tensile strain limit. with appropriate reliability.
Clause 8.12.3.1 specifies limits on crack width for non- • It is not prohibited for owners of the structures to specify
prestressed and prestressed components for several types of additional functional statements other than those stated
exposure. above based on their own judgment.

43
The above functional statements are in the context of design repaired at a reasonable cost and in a relatively short period
working life. The JGS document indicates that the design work- of time. This limit state, therefore, can be interpreted as a
ing life may be determined by considering various factors state in which the majority of the value of the structure is
including life-cycle cost, durability, deterioration, and the func- preserved. Furthermore, this limit state sometimes implies
tional life of the structure. The document notes that care a state in which marginal use of the structure is possible for
should be taken to ensure that the safety margin (i.e., reli- rescue operations right after an extraordinary event such
ability) introduced to each limit state is strongly related to the as a large earthquake. This limit state corresponds to the
design working life of the structure. The structural perfor- repairability defined in the functional statement above.
mance requirements of the structure are specified by several • SLS—Damage to the structure is limited to a level at which
limit states according to the load levels classified according to all common functions of the structure are preserved and do
their frequencies, as follows: not influence structural durability. This limit state corre-
sponds to the serviceability defined in the functional state-
• High-frequency variable actions are those expected to ment above.
occur once or a few times with significantly high probabil-
ity during the design working life of the structure. The JGS document indicates that additional performance
• Low-frequency variable actions are those that may or may requirements and limit states other than those defined above
not occur during the design working life of the structure may be defined as deemed necessary. With respect to the three
(i.e., a low-frequency variable action is an event with a very limit states identified above, the JGS document provides a
low occurrence probability). conceptual view of a performance matrix for describing the
performance requirements of a structure. In the performance
Using the preceding concepts, the JGS presents three major matrix, design situations and limit states are taken as the axes
limit states in the following qualitative manner: of the coordinate system, and performance requirements are
coordinated according to the importance of the structure. The
• ULS—The structures may sustain considerable damages example performance matrix presented in Figure 2.5 consists
but not to the extent of collapse that would result in serious of three levels of design situations and limit states. It reflects a
injury or loss of life. This limit state corresponds to the seismic design situation, which for most structures in Japan is
functional statement of safety as noted above. the critical design situation.
• Repairable limit state—Damage to the structure, although The performance requirements defined above are required to
it may influence durability, is limited to a level that can be be verified by two approaches: Approach A and Approach B.
Damage to a Structure
Repairable Limit
Magnitude of Actions
SLS ULS
State
Important, ordinary
High frequency, and easily

— —
low impact repairable
structures
Medium
Important and Easily repairable
frequency, —
ordinary structures structures
medium impact
Ordinary and
Low frequency,
— Important structures easily repairable
high impact
structures
Figure 2.5. Conceptual view of a performance matrix. Note: — 

performance requirement not specified by Japanese Geotechnical
Society.

44
Approach A does not require any specified method for perfor- was concluded that other specifications do not include “new”
mance verification of the structure. It requires, however, that the SLSs that need to be added to AASHTO LRFD. However, the
designer prove the structure satisfies the specified performance review resulted in some concepts that were of interest. These
requirements with an appropriate level of reliability. A designer concepts include
who uses Approach A is required to submit the necessary design
report and documentation for examination to the administra- • The target reliability index for SLSs may have different val-
tive organization or local government responsible for control- ues for different limit states. Furthermore, the target reli-
ling the safety of the structure. In contrast, in Approach B the ability for a certain limit state may vary depending on the
verification of performance requirements is based on specific consequences of exceeding that limit state.
design codes specified by the owner. The JGS document recom- • To differentiate between different limit states according to
mends use of the partial-factors format for design. the consequences of exceeding the limit state, the following
factors were considered:
44 Whether the limit state is reversible or irreversible: Irre-
2.3.5 Overarching Characteristics of Other versible limit states may have higher target reliability
Specifications to Be Considered than reversible limit states.
Reversible Versus Irreversible Limit States 44 Relative cost of repairs: Limit states that have the poten-
tial to cause damage that is costly to repair may have
SLSs may be categorized as reversible and irreversible. Revers- higher target reliability than limit states that have the
ible SLSs are those for which no consequences remain once a potential of causing only minor damage.
load is removed from a structure. Irreversible SLSs are those
for which consequences remain.
Due to their reduced safety implications, irreversible SLSs, 2.4 Surveys of Current Practice
which do not concern the safety of the traveling public, are cali-
2.4.1 Summary of R19B Survey
brated to a higher probability of failure and a corresponding
lower reliability index than the strength limit states. Reversible A focused survey was sent to 31 bridge owners, four industry
SLSs are calibrated to an even lower reliability index. representatives, and one university. A copy of the survey
questionnaire and a summary of responses are included in
Appendix B. The state bridge engineers who received the sur-
Load-Driven Versus Non-Load-Driven Limit States vey specifically included the chairs of the AASHTO Technical
The difference between load-driven and non-load-driven limit Committees for joints and bearings, culverts, steel design,
states is basically in the degree of involvement of externally concrete design, loads, and foundations. Sixteen responses
applied load components in the formulation of the limit state were received. The survey consisted of two parts: one was
function. In the load-driven limit states, the damage occurs due superstructure oriented, and the other was substructure and
to accumulated applications of external loads, usually live load foundation movement oriented.
(trucks). Examples of load-driven limit states include decom- Although there were only 16 responses, some consistency
pression and cracking of prestressed concrete and vibrations or was apparent in the most significant items in structural main-
deflection. The damage caused by exceeding SLSs may be tenance budgets, as seen in Figure 2.6.
reversible or irreversible and, therefore, the cost of repair may The most-cited responses confirmed that serviceability
vary significantly. However, in non-load-driven SLSs, the dam- issues relating to expansion joints and deck cracking are
age occurs due to deterioration or degradation as a function of widespread. Responses highlighted the following: deteriora-
time and aggressive environment or as inherent behavior due to tion and section loss of beam ends, painting of steel mem-
certain material properties. Examples of non-load-driven SLSs bers, problems with bearings, corrosion of reinforcement,
include penetration of chlorides leading to corrosion of rein- and deck overlays. Although there were many serviceability
forcement, leaking joints leading to corrosion under the joints, issues, many responses indicated that the SLSs are adequate
and shrinkage cracking of concrete components. In these exam- in their current form or would be adequate with some addi-
ples, the external load occurrence plays a secondary role. tional limit states, such as
• Foundation settlement;
2.3.6 Lessons Learned from Review of
• Guidance on stress checks based on corrosion-reduced
Existing Design Specifications
section properties;
Review of existing design specifications revealed that the SLSs • Better crack control reinforcement provisions and stress
covered by different specifications are somewhat similar. It limits for concrete flexural members; and

45
Responses to SHRP 2 R19B Survey Question 1

14
12
10
# of Responses
Fatigue
Built-up Steel Corrosion
Scour
Slope Maintenance
Expansion Joints
Abutments
Movable Bridges
Concrete Coating
Steel Paints
Timber Components
Bearings
Deck Drains
Concrete Decks
Deck Overlays
Anchor Cables
Railings and Curbs
Reinforcement
Header Joint
Weld Cracking
Repair/Replacement
Repair/Replacement
Approach Slabs
Durability
Concrete
Steel
Figure 2.6. Survey responses indicating the most significant structural maintenance budget items.
• Additional limit states for connections, expansion joints, that were received focused on structures typically built within
and bearings. the last two decades. All the structures mentioned in the
responses were continuous spans with integral or stub abut-
Despite suggestions for additional limit states, there was a ments. In addition, the approach fill was either a mechani-
common theme that additional limit states would not have cally stabilized earth wall or fill with side slope. The responses
affected or prevented the observed reduced serviceability. regarding tolerable movements were split almost evenly
Approximately half the responses indicated that the respon- between acceptable and not acceptable. The responses appeared
dents did not use deterioration models other than Pontis, while to be specific to the structures described in the section about
the other half used engineering judgment, had developed their bridge movement and observed distress rather than a general-
own models, or were collecting data to develop their own ized response indicative of a population of bridges. The final
model. In addition, approximately one-half of the responses questions dealt with the allowable movement of new struc-
indicated that no additional assessments were completed tures, with a majority of agencies noting that they are not
beyond those that are a part of Pontis, and one-quarter indi- following the guidance on tolerable movements found in
cated that they complete additional qualitative assessments but AASHTO LRFD Article C10.5.5.2. Agencies differed in what
no additional quantitative assessments. The other one-quarter their criteria for allowable movements were, with some deter-
of the responses indicated that they complete additional quali- mining criteria on a case-by-case basis, and others using
tative and quantitative assessments, including condition sur- general-purpose quantitative requirements.
veys, chloride penetration depth measurements, ultrasonic
testing, and condition scales for each component combined
2.4.2 Summary of NCHRP Project 12-83
with figures and notes that show the overall condition and defi-
Survey Related to Concrete Design
ciencies. The qualitative assessments have indicated a correla-
tion between deterioration and reduced serviceability, but the A survey of current practices related to the SLSs of concrete
reduction in serviceability was not quantified. structures was developed in NCHRP Project 12-83. The sur-
There were few responses to the second questionnaire vey was sent to major bridge owners across North America,
about bridge movement and observed distress. The responses including all 50 state DOTs, the Ministry of Transport in all

46
Canadian provinces, the District of Columbia, and many service conditions suggest a need exists for a service-load
turnpike authorities, bridge authorities, and commissions. combination akin to the Strength II limit state that applies to
The survey included 20 questions covering the following permit (overload) vehicles. The load factors for live load for
topics: such a load combination can be determined using the same
principles used for calibrating other SLSs. However, the sta-
• Modifications to the specification loading (HL-93 loading) tistical parameters to be used for permit vehicles differ from
for SLSs; those for random traffic.
• Checking SLSs under the effects of legal loads as part of the One important modification to the existing limit state is
normal design procedure; the load factor for live load in the Service III limit state in
• Revisions to the SLS stress limits for prestressed concrete AASHTO LRFD. One state, Louisiana, uses a load factor of 1.0
components; for live load to check tension in prestressed concrete under
• Revisions to existing SLSs for concrete structures; the Service III limit state instead of the 0.8 specified in
• Method used for designing for control of cracking by dis- AASHTO LRFD. The higher load factor addresses an issue
tribution of reinforcement; that has gained importance with AASHTO’s adoption of
• Checking concrete superstructure and substructures for newer prestress loss equations in 2005. Some engineers are of
any additional service-load combinations beyond those in the opinion that the lower load factor compensated for the
AASHTO LRFD; conservatism in the older prestressing loss equations and,
• Checking concrete structures for SLSs under overloads; thus, guarded against excessive conservatism in the design.
• Cracking of pretensioned concrete beams immediately The use of the new equations, which are believed to provide
after prestressing force release; a more accurate estimate of the prestressing losses, may have
• Observations of cracking of prestressed concrete beams in eliminated the need for the 0.8 load factor.
service;
• Damage to ends of prestressed beams under expansion
2.4.3 Summary of the R19As Survey
joints;
as it Relates to R19B
• Use of the deck empirical design method and the perfor-
mance of these decks in service; One of the main objectives in Phase 1 of Project R19A was the
• Observations of deck cracking; identification and ranking of the problematic areas preventing
• Type of reinforcement bars used in newer decks (e.g., black bridges from providing long service life. The research team
bars, epoxy-coated, galvanized, stainless steel); considered two alternatives for ranking of the performance:
• Average life span of concrete decks and the main reasons
decks are replaced; • Ranking based on quantitative performance data that are
• Types of concrete superstructures in use; obtained from experimental investigations or field obser-
• Problems with bearings in concrete structures; vations of bridges that are currently in service; and
• Cracking of abutments and piers; • Ranking based on qualitative opinion data (an expert elici-
• Average service life span of the concrete substructures; tation or Delphi process).
• Fatigue problems in concrete superstructures; and
• Use of coatings in concrete substructures. The R19A research team concluded that despite the avail-
ability of some experimental data, it is very difficult to quan-
Responses from 27 state DOTs and the Ontario Ministry of tify the performance of actual in-service components. The
Transportation were received. The responses to the ques- majority of the reported tests were performed using acceler-
tionnaire indicated that most bridge owners apply the SLSs ated testing methods, which are not easily correlated with
included in AASHTO LRFD with few or no revisions. The field conditions. They often focused only on the effect of one
additional limit states used by bridge owners appear to be degradation process, while experience shows that reality is
related either to owner-specified vehicles or to address a spe- more complex, and often several degradation processes inter-
cific issue that does not seem to be shared by other bridge act with the environmental loads. The combined effects and
owners, as evident by the lack of use of these additional limit complexity of deterioration processes and the uncertain nature
states by other owners. It is expected that some of the other of environmental loads complicate the prediction of the ser-
agencies that did not respond to the questionnaire also use vice life for both new and existing structures.
permit vehicles in checking some aspects of the design under The other source of quantitative data is from long-term
service loads. The use of permit vehicles to check some ser- monitoring of bridges in service, but such data are not avail-
vice conditions and the desire expressed by some bridge able at this time. In summary, the research team concluded
designers to have guidance on applying permit vehicles to that there are no available data for quantitative evaluation

47
Table 2.21. SLSs Identified for Development
LRFD Article Reversible No. of Lanes MPF
2.5.2.6.2 Criteria for Deflection Yes Single —

3.4.1 Load Factors and Load Combinations for Fatigue No Single —
5.5.3.1 General—Compressive Stress Limit for Concrete— No Single No
A Fatigue Criterion
5.5.3.2 Fatigue of Reinforcing Bars No Single —
5.5.3.4 Fatigue of Welded or Mechanical Splices of No Single —
Reinforcement
5.6.3.6 Crack Control Reinforcement—To be revised but not No — —
calibrated—Deemed to satisfy
5.7.3.4 Control of Cracking by Distribution of Reinforcement— No na —
Not calibratable—Deemed to satisfy
5.9.3 Stress Limitations for Prestressing Tendons No Multiple Yes
5.9.4.2.2 Tension Stresses Yes Single No
6.10.4.2 Permanent Deformations of Steel Structures No Single No
6.13.2.8 Slip Resistance of Bolts No Single No
10.6.2.4 Settlement Analysis of Shallow Foundations No for footing, possible Multiple for sands, none —
for superstructure for clays
10.8.2.2 Settlement (related to drilled shaft groups) No Multiple Yes
10.8.2.4 Horizontal Movement of Shaft and Shaft Groups No — —
Note: MPF = multiple presence factor; — = current criteria do not specify whether or not the MPF is applicable; na = not applicable.
and ranking of existing or promising strategies to quantify In summary, the R19A research team stated that due to the
the reduction in service life due to deterioration. lack of quantitative data with respect to almost all bridge ele-
As there are no quantitative data for ranking and selection ments, it was difficult to propose or develop new design
of the problematic areas, the R19A research team prioritized methodologies that are based on deterioration models.
the research topics based on the qualitative opinion of experts.
Obtained information was organized and presented in tech-
2.5 SLSs to Be Considered
nology, strategy, and research tables that provide the informa-
in This Report
tion on the potential service life issues and available solutions.
The tables also provide information on the advantages and Potential limit states and possible calibration approaches for
disadvantages of different design concepts, along with other general requirements, concrete structures, steel structures, geo-
relevant data necessary to evaluate different strategies address- technical issues, joints, and bearings have been reviewed. Some
ing durability. Four major problem research areas were identi- of the potential limit states have since been determined to be
fied: decks, joints, bearings, and durability. uncalibratable. For example, some are deterministic or are based
The major product from the R19A research effort is Design on judgment and experience. The SLSs believed to be calibrat-
Guide for Bridges for Service Life (Azizinamini et al. 2013), able are listed in Table 2.21 along with whether the phenomena
which is intended to complement AASHTO LRFD specifica- being addressed are reversible or irreversible and whether the
tions and incorporate the design for durability and enhanced live load involves single-lane or multiple-lane loading.
service life. The document provides a basis for the selection, Note that SLS references to partial prestressing have been
design, fabrication, construction, inspection, management, removed. AASHTO no longer accepts partial prestressing as a
and maintenance of bridge systems. design strategy.

Chapter 3
Overview of Calibration Process
3.1 Introduction methods are more viable approaches for structural design. In

Level I design methods, safety is measured in terms of a safety
The new generation of bridge design codes is based on proba- factor, or the ratio of nominal (design) resistance to nominal
bilistic methods. Load and resistance (load-carrying capac- (design) load. In Level II, safety is expressed in terms of a reli-
ity) parameters are treated as random variables, and structural ability index (b). The Level II approach generally requires
performance is quantified in terms of the reliability index iterative techniques best performed using computer algo-
(Nowak and Collins 2013). This approach allows for a ratio- rithms. For simpler cases, closed-form solutions to estimate
nal comparison of different materials and load combinations. b are available. Closed-form analytical procedures to estimate
An increased degree of uncertainty causes a reduction in reli- load and resistance factors should be considered approxi-
ability, and strict control of structural parameters results in a mate, with the exception of very simple cases for which an
safer structure. The probabilistic analysis requires statistical exact closed-form solution exists. Alternatively, spreadsheet
models of load and resistance parameters. The load models programs can be used to estimate load and resistance factors
for bridges can be based on truck surveys and other field tests. using the more rigorous and adaptable Monte Carlo simula-
Resistance models for structural components (e.g., bridge tion technique, which in turn can be used to accomplish a
girders) can be derived from material tests, lab tests, and ana- Level II probabilistic analysis.
lytical simulations. The goal of Level I or II analyses is to develop factors that
With the advent of limit states design methodology in North increase the nominal load or decrease the nominal resistance
American design specifications, there has been an increasing to give a design with an acceptable and consistent reliability.
demand to obtain statistical data to assess the reliability of To accomplish this, an equation that incorporates and relates
designs. Reliability depends on load and resistance factors that all the variables that affect the potential for failure of the
are determined through calibration procedures using available structure or structural component must be developed for each
statistical data. Methodologies that can be used to determine limit state.
load and resistance factors, including the basic reliability con- For load and resistance factor design (LRFD) calibration
cepts and detailed procedures that can be used to characterize purposes, statistical characterization should focus on the pre-
data to develop the statistics and functions needed for reliabil- diction of load or resistance relative to what is actually mea-
ity analysis, are described in NCHRP Report 368 (Nowak sured in a structure. Thus, this statistical characterization is
1999) and TRB Circular E-C079 (Allen et al. 2005). typically applied to the bias, the ratio of the measured to pre-
The code calibration procedure can include closed-form dicted value. The predicted (nominal) value is calculated using
solutions for estimating load and resistance factors that can the design model being investigated. The degree of variation is
be used for simple cases, as well as more rigorous probabilis- measured in terms of the coefficient of variation (CV), which
tic analysis methods such as the Monte Carlo method, which is the ratio of the standard deviation to the mean value.
is described in Section 3.2.3. There are three levels of proba- Regardless of the level of probabilistic design used to per-
bilistic design: Levels I, II, and III (Nowak and Collins 2013). form LRFD calibration, the steps needed to conduct a cali-
The Level I method is the least accurate, and Level III is the bration are as follows:
only fully probabilistic method. However, Level III requires
complex statistical data beyond what are generally available • Develop the limit state equation to be evaluated so that
in engineering practice. Level I and Level II probabilistic the correct random variables are considered. Each limit
48

49
state equation must be developed on the basis of a pre- 3.2 Calibration by

scribed failure mechanism. The limit state equation Determination of
should include all the parameters that describe the failure Reliability Indices
mechanism and that would normally be used to carry
out a deterministic design of the structure or structural 3.2.1 Basic Framework
component. Expanding on the four basic steps outlined above, the frame-
• Statistically characterize the data on which the calibration work for calibration of service limit state (SLS) using reliabil-
is based (i.e., the data that statistically represent each ran- ity indices is summarized as follows:
dom variable in the limit state equation being calibrated).
Key parameters include the mean, standard deviation, and 1. Formulate the limit state function and identify basic vari-
CV, as well as the type of distribution that best fits the data ables. Identify the load and resistance parameters and for-
(often normal or lognormal). mulate the limit state function. For each considered limit
• Select a target reliability value based on the margin of state, the acceptability criteria were established. In most
safety implied in current designs, considering the need for cases, it was not possible to select a deterministic bound-
consistency with reliability values used in the development ary between what is acceptable and unacceptable. Some of
of other AASHTO LRFD specifications, the consequence of the code-specified limit state functions do not have a
exceeding the limit state, cost, and the levels of reliability physical meaning (e.g., allowable compression stress in
for design as reported in the literature for similar struc- concrete).
tures. If the performance of existing structures that were 2. Identify and select representative structural types and design
designed using the current code provisions is acceptable, cases. Select the representative components and structures
then there is no need to increase the safety margin in the to be considered in the development of code provisions
newly developed code. Furthermore, the acceptable safety for the SLSs.
level can be taken as corresponding to the lower tail of dis- 3. Determine load and resistance parameters for the selected
tribution of betas. design cases. Identify the design parameters on the basis
• Determine load and resistance factors by using reliability of typical structural types, loads, and locations (climate,
theory consistent with the selected target reliability. exposure to harsh environment). For each considered ele-
ment and structure, the values of typical load components
The accuracy of the results of a reliability theory analysis is must be determined.
directly dependent on the adequacy, in terms of quantity and 4. Develop statistical models for load and resistance. Gather sta-
quality, of the input data used. The final decision made tistical information about the performance of the consid-
regarding the magnitude of the load and resistance factors ered types and models in selected representative locations
selected for a given limit state must consider the adequacy of and traffic. Gather statistical information about quality of
the data. If the adequacy of the input data is questionable, the workmanship. Ideally, for a given location and traffic, the
final load and resistance factor combination selected should required data include general assessment of performance,
be weighted toward a level of safety that is consistent with assumed time to initiation of deterioration, assumed dete-
past successful design practice, using the reliability theory rioration rate as a function of time, maintenance, and repair
results to gain insight as to whether past practice is conserva- (frequency and extent). Develop statistical load and resis-
tive or unconservative. tance models (as a minimum, determine the bias factors
The calibration procedure can be different depending on and CVs). The parameters of load and resistance are deter-
the type of limit state. In the case of serviceability limit mined not only by magnitude, as is the case with strength
states, it is much more complex, mostly due to difficulties in limit states, but also by frequency of occurrence (e.g., crack
formulation of the limit state equation. The parameters of opening) and as a function of time (e.g., corrosion rate,
load and resistance are determined not only by magnitude, chloride penetration rate). The available statistical param-
as is the case with strength limit states, but also by frequency eters were used, but the database is limited, and for some
of occurrence (e.g., crack opening) and as a function of serviceability limit states there is a need to assess, develop,
time (e.g., corrosion rate, chloride penetration rate). Accept- or derive the statistical parameters. The parameters of time-
ability criteria are not well defined as they are subjective varying loads were determined for various time periods.
(e.g., deflection limit, allowable tensile stress), and the The analyses were performed for various traffic parameters
code-specified limit state function does not necessarily have (average daily truck traffic, legal loads, multiple presence,
a physical meaning (e.g., allowable compression stress in traffic patterns). The load frequencies served as a basis for
concrete). determination of acceptability criteria.

50
5. Develop the reliability analysis procedure. Reliability can be accurately it can be represented as a normal distribution. In
calculated using either a closed-form formula or the addition, the curve representing the CDF of any other type of
Monte Carlo method. The reliability index for each case random variable can be evaluated, and its shape can provide
can be calculated using closed formulas available for par- an indication about the statistical parameters, such as the
ticular types of probability distribution functions in the maximum value and type of distribution for the whole CDF
literature or the Monte Carlo method. In this study, all the or, if needed, only for the upper or lower tail of the CDF. Fur-
reliability calculations were based on Monte Carlo analy- thermore, the intersection of the CDF with the horizontal
sis. The Monte Carlo method is a stochastic technique that axis (zero on the vertical scale) corresponds to the mean. The
is based on the use of random numbers and probability slope of the CDF determines the standard deviation, or sx as
statistics to simulate a large number of computer-based shown in Figure 3.2. A steeper CDF on probability paper
experiments. The outcome of the simulation is a large indicates a smaller standard deviation. Further information
number of solutions that takes into account all the ran- about the construction and use of probability paper can be
dom variables in the resistance equation. found in textbooks (e.g., Nowak and Collins 2013).
6. Calculate the reliability indices for current design code and
current practice. Calculate the reliability indices for selected
3.2.2 Closed-Form Solutions
representative bridge components corresponding to cur-
rent design and practice. The reliability index (b) is defined as shown by Equation 3.1:
7. Review the results and select the target reliability index. Use
the calculated reliability indices to select the target reliabil- β = Φ −1 (Pf ) (3.1)
ity index (bT). Select the acceptability criteria (i.e., perfor-
mance parameters) that are acceptable and the performance where F-1 is the inverse of the standard normal distribution,
parameters that are not acceptable. and Pf is the probability of failure.
8. Select potential load and resistance factors. Prepare a rec- If the limit state function (g) can be expressed in terms of
ommended set of load and resistance factors. The objec- two random variables, R representing resistance and Q repre-
tive is that the design parameters (load and resistance senting the load effect, then g is given by Equation 3.2:
factors) have to meet the acceptability criteria for the con-
sidered design situations (location and traffic). The design g = R−Q (3.2)
parameters should provide reliability that is consistent,
uniform, and conceivably close to the target level. and the probability of failure is expressed by Equation 3.3:
9. Calculate reliability indices. Calculate the reliability indices
corresponding to the recommended set of load and resis-
Pf = Prob (g < 0) (3.3)
tance factors for verification. If the design parameters do
not provide consistent safety levels, modify the parameters
and repeat Step 8. b can then be calculated using a closed-form formula in
two cases: when both R and Q are normal random variables
Figure 3.1 presents the flowchart for the basic calibration or when both R and Q are lognormal random variables. In all
framework described in the nine steps above. other cases, the available procedures produce approximate
Step 4 requires the analysis of data describing load and results.
resistance. Normal probability paper is a special scale that If both R and Q are normal random variables, b can be
facilitates the statistical interpretation of data. The horizontal calculated using Equation 3.4:
axis represents the variable (e.g., gross vehicle weight, mid-
span moment, or shear) for which the cumulative distribu- R−Q
β= (3.4)
tion function (CDF) is plotted. The vertical axis represents σ 2R + σ Q2
the number of standard deviations from the mean value,
which is often referred to as the standard normal variable, or where_
the Z-score. The vertical axis can also be interpreted as the R =
mean or expected value of the distribution of
probability of being exceeded; for example, one standard _ resistance;
deviation corresponds to 0.159 probability of being exceeded. Q = mean or expected value of the distribution of load;
The most important property of normal probability paper sR = standard deviation of the distribution of resistance;
is that the CDF of a normal random variable is represented and
by a straight line. The straighter the plot of data, the more sQ = standard deviation of the distribution of load.

51
Calibration
framework
Formulate the limit state

function
Select structural types

and design cases
Determine load Determine resistance

Load survey parameters for selected parameters for selected
design cases design cases
Experimental
data
Develop statistical Develop statistical
Experimental
models for loads and models for resistance
data
load combinations variables
Engineering
judgment
Statistical Statistical
Observations parameters of parameters of
loads resistance
Develop the reliability

analysis procedure
Calculate the reliability

indices for current design
code or current practice
Review the results and

select the target reliability
index, β T
Select potential load and

resistance factors
Calculate reliability indices
Modify potential load and No Yes

β ≈ βT
resistance factors
List obtained load
and resistance
factors
End of the
calibration
procedure
Figure 3.1. Flowchart of basic calibration framework.

52
0.995 Normal
probability
0.990 scale
0.980
2
0.950
0.900
0.841
Standard normal variable
1
0.800
0.700
σx
0.600
0 0.500
0 1 2 3 4 5 6 7
0.400
σx
0.300
0.200 0.159 µx
-1
0.100
0.050
-2 0.020
0.010
0.005
-3
Figure 3.2. Use of normal probability paper.
Sometimes, R - Q is termed M, the margin of safety. Using is well known. Application of Monte Carlo simulation follows
this terminology, b is given by Equation 3.5: these steps:
M • It is assumed that dead load is normally distributed and

β= (3.5)
σM live load CDF is as shown on the probability paper [directly
from WIM (weigh in motion) data]. The statistical param-
For the case in which both distributions are lognormally eters of live load depend on the time period. For longer
distributed, a more complete derivation of the closed-form time periods, the statistical parameters are obtained by
solutions and how they can be applied to LRFD calibration extrapolation of the available WIM data. The total load is
is shown by Allen et al. (2005). Although closed-form solu- a sum of dead load and live load and, therefore, in practice
tions are useful for illustrative purposes, in practice either it can be treated as a normal variable. This assumption is
load or resistance or both are not normally distributed, partly justified by the central limit theorem, and it is
which limits the use of closed-form solutions in code acceptable if the load components are of similar magni-
calibration. tude (Nowak and Collins 2013).
• Resistance is assumed to be lognormally distributed. The
resistance side of the LRFD equation is a product of terms.
3.2.3 Using Monte Carlo Simulation
• The minimum statistical parameters needed for each
in the Calibration Process
random variable are the CV (V) and the bias (l). Using
The typical application of Monte Carlo simulation, refer- the reported statistics of load and resistance along with
enced in Step 5 for bridge structural reliability and as reported computer-generated random numbers, the distributions
in the literature (Allen et al. 2005; Nowak and Collins 2013), of load and resistance are developed, and values are chosen

53
randomly from these distributions. For example, for the respectively, which are accepted equations for the range of
simple load combination of dead load plus live load, ran- values that were considered:
dom values of dead load and live load are chosen from the
normal distributions fitted in the region of interest. A ran- µ R = Rn i µ M i µ F i µ P (3.7)
dom value of resistance is chosen from the lognormal dis-
tribution of resistance. VR = VM2 + VF2 + VP2 (3.8)
• The simulation is run by selecting random values from
both the load and resistance distributions. The limit state The statistical parameters of resistance were determined
function [Ri - (Di + Li)] is calculated for each set of random using the test results available before 1990, special simula-
variables. If the value is equal to or greater than zero, the tions, and engineering judgment. They were developed for
function is satisfied, and the individual case is safe. If the noncomposite and composite steel girders, reinforced con-
value is negative, the criterion is not satisfied, and the case crete T-beams, and prestressed concrete AASHTO-type g irders.
represents a failure. Bias factors and CVs were determined for material factor M,
• After a large number of iterations, the failures are counted, fabrication factor F, and analysis factor P. Factors M and F
and the failure rate is determined. For the sampling to be were combined.
significant at least 10 failures should be observed; other- For structural steel, the statistical parameters are found in
wise, more iteration is necessary. If the expected probabil- papers by Ravindra and Galambos (1978), Yura et al. (1978),
ity of failure is very low, then the number of iterations can Cooper et al. (1978), and Hansell et al. (1978), which are sum-
be prohibitively large. An alternative way to determine the marized in Ellingwood et al. (1980). The information included
reliability index is to generate a smaller number of limit the mean values and CV for the yield strength of steel, tensile
state function values, plot the results on normal probabil- strength of steel, and modulus of elasticity for hot-rolled
ity paper, and extrapolate the obtained lower tail of the beams and plates. In addition, they provided the statistical
distribution function. The extrapolated lower tail will parameters (mean value and CV) for the fabrication factor
allow for assessment of the reliability index and probability and the professional factor. In the very last phase of calibration
of failure (or failure rate). for AASHTO LRFD, the American Iron and Steel Institute
• By using the failure rate, the reliability index is determined as provided the upgraded bias factors and CVs for yield strength
the inverse of the standard normal cumulative distribution. of structural steel. These values were then used in Monte
Carlo simulations to determine the parameters of resistance
A step-by-step procedure for implementing the Monte for noncomposite and composite girders for the moment-
Carlo method by using statistical functions commonly avail- carrying capacity and shear. [More recent data gathered after
able in spreadsheet programs is presented in Appendix F. the Northridge earthquake by Dexter et al. (2000) and Dexter
and Melendrez (2000), and data reported by Bartlett et al.
(2003), show improved statistics, although Bartlett et al. rec-
3.2.4 Statistical Parameters for Resistance ommend no resistance factor changes until more is known. In
and Other Loads the case of the steel SLSs calibrated in this study, the newer
The discussion in this section is excerpted from Kulicki et al. data could affect only the overload limit state, making the reli-
(2007). ability analysis somewhat conservative. Given the paucity of
resistance data on which this limit state is based, the analysis
was not updated for the more recent data.]
Resistance Models For concrete components, the material parameters were
Resistance was considered as a product of a nominal resis- taken from Ellingwood et al. (1980). As in the case of struc-
tance (Rn) and three factors: M, or material factor (strength tural steel, the statistical parameters were obtained, but no
of material, modulus of elasticity); F, or fabrication factor raw test data. The basis for these parameters was research by
(geometry, dimensions); and P, or professional factor (use of Mirza and MacGregor (1979a, 1979b). The data included
approximate resistance models; e.g., the Whitney stress block, mean value and CV for the compressive strength of concrete,
idealized stress and strain distribution model). Resistance (R) yield strength of reinforcing bars, and prestressing strands. In
is given by Equation 3.6: addition, the data included the statistical parameters of fab-
rication factor and professional factor.
R = Rn i M i F i P (3.6) The material data, combined with the statistical parame-
ters of the fabrication factor and professional factor, were
The mean value of resistance (µR) and the CV of resistance used in Monte Carlo simulations that resulted in the statisti-
(VR) may be approximated by Equations 3.7 and 3.8, cal parameters of resistance for steel girders (noncomposite

54
Table 3.1. Statistical Parameters of Table 3.2. Statistical Parameters of Dead Load
Component Resistance
Dead Load Component Bias Factor CV
Material and
Factory-made members, 1.03 0.08
Fabrication
DL1
Factors Professional Resistance
(M and F) Factor (P) (R) Cast-in-place, DL2 1.05 0.10
Type of
Structure l V l V l V Wearing surface, DL3 3 in. (mean thickness) 0.25
Noncomposite steel girders Miscellaneous, DL4 1.03 ~ 1.05 0.08 ~ 0.10
Moment 1.095 0.075 1.02 0.06 1.12 0.10

(compact)
Moment 1.085 0.075 1.03 0.06 1.12 0.10
(noncompact) DL2, respectively (Lind and Nowak 1978). However, there is
no report available to support these data. The CVs used in
Shear 1.12 0.08 1.02 0.07 1.14 0.105
calibration were taken from the National Bureau of Stan-
Composite steel girders
dards Special Publication 577 (Ellingwood et al. 1980) and
Moment 1.07 0.08 1.05 0.06 1.12 0.10 include other uncertainties (also human error).
Shear 1.12 0.08 1.02 0.07 1.14 0.105 The parameters of DL3 were calculated using the survey
Reinforced concrete data provided by the Ontario Ministry of Transportation in
conjunction with calibration of the OHBDC (1979).
Moment 1.12 0.12 1.02 0.06 1.14 0.13
Shear with 1.13 0.12 1.075 0.10 1.20 0.155
steel 3.3 “Deemed to Satisfy”
Shear without 1.165 0.135 1.20 0.10 1.40 0.17 When you can measure what you are speaking about, and
steel
express it in numbers, you know something about it, when
Prestressed concrete you cannot express it in numbers, your knowledge is of a mea-
Moment 1.04 0.045 1.01 0.06 1.05 0.075 ger and unsatisfactory kind; it may be the beginning of knowl-
edge, but you have scarcely, in your thoughts advanced to the
Shear with 1.07 0.10 1.075 0.10 1.15 0.14
steel
stage of science.—William Thomson, Lord Kelvin
Source: Nowak (1999).
The least rigorous process for establishing design require-
ments, and load and resistance factors in particular, is referred
to as “deemed to satisfy.” In this process, experience and
empirical observation are used to define the boundary
and composite), reinforced concrete T-beams, and pre-
between satisfactory performance and unsatisfactory perfor-
stressed concrete girders, for moment and shear, as shown in
mance. It provides no quantifiable way of assessing the pro-
Table 3.1 (Nowak 1999). The statistical parameters include
vided margin of adequacy, such as safety or reliability. As
three factors representing uncertainty in materials, dimen-
there is no way to quantify the performance margin, there is
sions and geometry, and analytical model.
no way to assess the benefit of a change in requirement other
It was assumed that resistance is a lognormal random
than a general knowledge that changing this or that should
variable.
move in the direction of higher performance. The obvious
corollary is that cost–benefit cannot be quantified. An exam-
Statistics of Loads Other Than Live Load ple of “deemed to satisfy” is the specification of concrete
cover requirements in U.S. practice, which is based only on
The data presented below were developed in support of experience and has no consistent mathematical basis.
strength calibrations, but they are equally applicable to load Nonetheless, “deemed to satisfy” has a place in the pan-
calculations related to SLS calibration (see Table 3.2). theon of engineering tools. It is often the basis of detailing
The bias factors for DL1 and DL2 were provided by the requirements and may serve as the beginning of design speci-
Ontario Ministry of Transportation and were based on sur- fication development, as in “experience shows that if we do
veys of actual bridges in conjunction with calibration of the (or do not do) this or that the results are generally accept-
Ontario Highway Bridge Design Code (OHBDC 1979; Lind able.” Expert elicitation (Delphi process) or an experimental
and Nowak 1978). The CVs provided by the Ministry of program may provide insight into the adequacy of “deemed
Transportation for dead load were 0.04 and 0.08 for DL1 and to satisfy.”

55
3.4 Customizing the Process it is expected that at some point in time corrosion will have
resulted in a 10% reduction in resistance of a class of bridges.
The data used in the calibration described in this report are Referring to the box marked Statistical Parameters of Resis-
provided in Appendix F. The key to providing the ability for tance in Figure 3.1, the resistance would be adjusted (low-
owners to adjust the calibration of the SLSs for their own ered) by 10%. If it is determined that not only is it expected
experience is to either adjust the data in Appendix F or supply that the average resistance will be lowered, but that the values
state-specific information of the same type. of resistance are becoming more diverse (random), then the
The following attributes were identified as necessary to bias and CV of resistance can also be adjusted based on that
allow bridge owners to customize the calibration process and experience, as they are simply input variables. Rerunning the
to develop spreadsheets for their particular needs. calculations for the affected population of the originally pro-
The ability of the process to address these issues is provided bridges, or an owner-supplied set, will allow the owner
vided as follows: to track the change in reliability indices. If one wanted to esti-
mate the effect on reliability-based ratings or postings, one
• Ability of the Monte Carlo procedure to produce a probability could keep or modify the target reliability index and repeat
of criteria exceedance and the associated reliability index. This the lower iteration loop by using revised trial load factors
ability is at the core of applying the Monte Carlo procedure. until sufficient convergence of the reliability indices was
If 100,000 trial calculations of a given limit state function are found. This would essentially be a recalibration.
produced using randomly generated loads and resistances • Ability to react to user intervention as reflected in an improved
that are consistent with the mean values and CVs for resistance (also user supplied). This is basically the opposite
that limit state function, and the function is not satisfied of the process of downgrading resistance discussed above.
100 times, then the failure rate is 0.001 and the success rate • Ability to accept either a user-supplied database from which the
is 0.999. The corresponding reliability index from a hand- product will determine a new bias and CV or a user-supplied
book of probability functions or inverse standard normal bias and CV from an external calculation. As discussed above,
CDF available in many computer applications is 3.09. the bias and CV are input variables that a bridge owner is able
• Ability to accept a user-supplied deterioration of load-carrying to adjust.
capacity. A possible approach to downgrading the resistance • Ability to accommodate a user-supplied resistance model. This
with time is discussed in Chapter 4 by using condition num- is especially important for the geotechnical community due
ber as a surrogate for deterioration. For example, assume that to the regional nature of practice in that discipline.

Chapter 4
Deterioration
4.1 Introduction including an estimate of deterioration in recalibrating the

service limit states by using the framework described in
Deterioration and a degradation of serviceability indicated by Chapter 3 of this report. One simple way to do this would be
a reduction in usable capacity (either by a change in rating or to accept the premise that until more usable data on the
a change in reliability index) are not interchangeable terms, change of resistance with time are available, it is reasonable to
although they may be related for some elements or systems. treat the percentage change in condition number as a surro-
For example, a steel girder, especially a rolled beam, may have gate for change in resistance and adjust the resistance in the
visible corrosion suggesting that the element is deteriorating calibration spreadsheets accordingly. In some cases, the use of
and in need of maintenance, but it may have little or no per- the equations included in this chapter may be beyond the
ceptible change in deformations, stresses, or rider comfort. range for which they were developed.
Sometimes a corrosion hole may exist in a place that does not
control an evaluation. Similar observations can be made about
the early stages of damage to prestressed beams resulting from
4.2 Bolukbasi et al. (2004)
poor drainage control at expansion joints. The resulting spall- Bolukbasi et al. (2004) used historic NBI rating data for 2,601
ing and possible rusting of rebar and strands may be unsightly bridges from Illinois to determine regression equations relating
but have relatively little structural effect until the damage is the bridge age to the condition rating of the deck, superstruc-
well advanced. For these types of elements or systems, deterio- ture, and substructure. No distinction between cast-in-place
ration models or databases geared to predicting maintenance (CIP) decks and precast panels was indicated in the reference.
budget needs are not especially useful to the calibration pro- The data were adjusted such that bridges with a sudden rating
cess needed for Project R19B. Degradation first leads to loss of increase are excluded from the study (a sudden increase in
service, and if left untreated, can lead to loss of load-carrying rating indicates performance of maintenance). The resulting
capacity resulting in failure. equations suggest the rating and a corresponding service life if
However, for some elements and subsystems, a high cor- no maintenance occurs. Equations are provided for the follow-
relation may exist between loss of serviceability and deterio- ing categories: all bridges; steel bridges; reinforced-concrete
ration as indicated by a change in National Bridge Inspection (RC) bridges; prestressed concrete bridges; Interstate bridges;
Standards (NBIS) sufficiency ratings or the results of a dete- non-Interstate bridges; bridges with annual average daily traffic
rioration model. Such might be the case for decks and bear- (AADT) <5,000; bridges with 5,000 < AADT < 10,000; and
ings, for which a deteriorated state could be considered in the bridges with AADT >10,000. Within each bridge category,
calibration by owner adjustment of nominal resistance based equations are provided to estimate the rating for the deck,
on bridge-specific knowledge or deterioration modeling. superstructure, and substructure.
Several researchers have proposed algorithms predicting Table 4.1 shows the rating prediction equations for the nine
the change in condition number, either the National Bridge categories, as well as a graph showing the condition rating
Inventory (NBI) condition number or a variation, over time versus time. The end of service life is typically defined as when
for bridge details or complete structures. Five of these pro- a rating of 3 is achieved, and maintenance would be required
posals are reviewed in this chapter. Bridge owners could use to continue using the structure.
one or more of these proposals, or others that may be found The equations were also plotted by component (deck,
to more accurately reflect local conditions, as a basis for superstructure, substructure), allowing an investigation into
56

57
Table 4.1. Rating Prediction Equations and Graphs for Nine Categories of Bridges
All Bridges Deck: R = 8.960814 - 0.20144T + 0.006719T 2

- 9.67 × 10-5T 3
10.00
Deck
9.00 Superstructure: R = 8.854089461 - 0.144890772T
8.00 Superstructure + 0.003122716T 2 - 2.91 × 10-5T 3
7.00 Substructure
Substructure: R = 8.767383274 - 0.127816817T
Rating
6.00 + 0.002736488T 2 - 2.57 × 10-5T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Steel Bridges Deck: R = 8.922947 - 0.19861T + 0.006735T 2

- 9.77 × 10-5T 3
10.00
Deck
8.00 Superstructure + 0.004406448T 2 - 5.36 × 10 -5T 3
7.00 Substructure
Substructure: R = 8.822326892 - 0.148338077T
Rating
6.00 + 0.004166181T 2 - 4.83 × 10 -5T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Reinforced Concrete Bridges Deck: R = 8.605268604 - 0.1277358696T +

0.0023501188T 2 - 3.643 × 10 -5T 3
10.00
Deck
7.00 Substructure
Substructure: R = 8.624414481 - 0.123890228T
Rating
6.00 + 0.002486843T 2 - 2.21 × 10 -5T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Prestressed Concrete Bridges Deck: R = 9.243165 - 0.25857T + 0.01004T 2

- 1.5 × 10 -4T 3
10.00
Deck
7.00 Substructure
Substructure: R = 9.075226897 - 0.19604399T
Rating
6.00 + 0.006203563T 2 - 7.49 × 10 -5T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
(continued on next page)

58
Table 4.1. Rating Prediction Equations and Graphs for Nine Categories of Bridges (continued)
Interstate Bridges Deck: R = 8.920346 - 0.21323T + 0.007687T 2

- 1.5 × 10 -4T 3
10.00
Deck
7.00 Substructure
Substructure: R = 8.956002854 - 0.205796117T
Rating
6.00 + 0.008041095T 2 - 1.31981 × 10 -4T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Non-Interstate Bridges Deck: R = 8.981204 - 0.20173T + 0.007319T 2

- 1.1 × 10 -4T 3
10.00
Deck
7.00 Substructure
Substructure: R = 8.745705917 - 0.113435369T
Rating
6.00 + 0.002153535T 2 - 2.01 × 10 -5T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Bridges with AADT < 5,000 Deck: R = 8.974903 - 0.20009T + 0.007589T 2

- 1.1 × 10-4T 3
10.00
Deck
7.00 Substructure
Substructure: R = 8.688714213 - 0.10308987T +
Rating
6.00 0.001890448T 2 - 1.87 × 10 -5T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Bridges with 5,000 < AADT < 10,000 Deck: R = 8.887719688 - 0.1873850501T
+ 0.0047333447T 2 - 7.2279 × 10 -5T 3
10.00
Deck
7.00 Substructure
Substructure: R = 8.791762968 - 0.139058986T
Rating
6.00 + 0.0026443T 2 - 2.02 × 10 -5T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)

59
Table 4.1. Rating Prediction Equations and Graphs for Nine Categories of Bridges (continued)
Bridges with AADT > 10,000 Deck: R = 9.047654 - 0.24391T + 0.009357T 2

- 1.7 × 10 -4T 3
10.00
Deck
7.00 Substructure
Substructure: R = 9.045892953 - 0.218679597T
Rating
6.00 + 0.008359897T 2 - 1.31336 × 10 -4T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Source: Bolukbasi et al. (2004).
how other criteria affect the service life of the bridge. Figure 4.1, shortest service life (45 to 50 years). Steel and prestressed con-
Figure 4.2, and Figure 4.3 show the predicted deck, super- crete bridges also have shorter service lives (55 to 65 years), but
structure, and substructure condition rating, respectively, this is likely due to the fact that many of these are located on
versus time. The graphs in Table 4.1 show that the service life the Interstate and are subject to high traffic counts. The basis
of the deck is typically shorter than the service life of either for this difference, 55 years and 65 years, could not be found
the superstructure or the substructure. All categories have a in the reference. Non-Interstate, RC, and low-traffic bridges
deck service life less than 55 years. For Interstate bridges and have an estimated service life of approximately 75 years.
high-traffic-volume bridges, Figure 4.1 shows the deck ser- Bridges with AADT between 5,000 and 10,000 are shown to
vice life is much closer to 40 years. High-traffic-volume have the longest service life, greater than 80 years. The AADT
bridges have an estimated service life of 41 years, medium- between 5,000 and 10,000 category may have a longer esti-
traffic-volume bridges have an estimated service life of mated life due to less traffic than the Interstate and AADT
47 years, and low-traffic-volume bridges have an estimated ser- >10,000 categories, combined with routine maintenance
vice life of 54 years. The medium- and high-traffic bridge deck and repair, resulting in a longer service life.
condition ratings decrease at a faster rate than the low-volume Figure 4.3 shows the substructure condition rating versus
bridges. For the first 30 years, the medium and high traffic have time for the nine bridge categories. As with the other condi-
nearly identical deck condition ratings. After the first 30 years tion ratings, bridges falling into the high-traffic and Interstate
they split, with the high-traffic bridges decreasing faster. categories have the shortest service life (an estimated 50 years).
Figure 4.2 shows the superstructure rating versus time for The substructures of prestressed concrete bridges and steel
the nine bridge categories. Similar to the deck condition rating, girder bridges have estimated service lives of approximately
Interstate bridges and bridges with AADT >10,000 have the 62 and 67 years, respectively. The basis of this difference could
Deck Ratings Superstructure Ratings

10.00 10.00
All Bridges All Bridges
9.00 Steel Bridges 9.00 Steel Bridges
8.00 RC Bridges 8.00 RC Bridges
Prestressed Concrete Bridges PS Concrete Bridges
7.00 Interstate Bridges 7.00 Interstate Bridges
Rating
Rating
6.00 Non-Interstate Bridges 6.00 Non-Interstate Bridges

AADT < 5,000 AADT < 5,000
5.00 5,000 < AADT < 10,000
5.00 5,000 < AADT < 10,000
4.00 AADT > 10,000 4.00 AADT > 10,000
3.00 3.00
2.00 2.00
0 20 40 60 80 100 120 0 20 40 60 80 100 120 140
Time (years) Time (years)
Figure 4.1. Deck condition rating versus time. Figure 4.2. Superstructure condition rating versus time.

60
Substructure Ratings 4.3 Jiang and Sinha (1989)

10.00
All Bridges
9.00 Steel Bridges
In their 1989 report on bridge performance and optimization,
RC Bridges Jiang and Sinha discussed the results of regression analysis and
8.00 PS Concrete Bridges
7.00 Interstate Bridges Markov chain analysis to estimate the average rating of a group
Non-Interstate Bridges
Rating
6.00
of bridges. They considered Interstate and non-Interstate
AADT < 5,000
5.00 5,000 < AADT < 10,000 bridges, as well as steel and concrete bridges; no distinction was
AADT > 10,000 made between reinforced or prestressed concrete construction.
4.00
3.00
Geographic location and traffic volume were initially consid-
2.00
ered, but because they did not appear to influence the regres-
0 20 40 60 80 100 120 140 sion analysis, they were not considered as separate categories.
Time (years) A relatively small sample (several hundred bridges) was used in
Figure 4.3. Substructure condition rating versus time. the regression analysis, and at the time of the analysis, biennial
NBI inspections had only been occurring for approximately 10
years. Thus, the results may have been influenced by the limited
amount of data available and used.
not be found in the reference. Non-Interstate, RC, and low- The results of the regression analysis were coefficients for a
traffic bridges all have an estimated service life of approxi- third-order polynomial describing the NBI condition rating
mately 80 years, and bridges in the medium-traffic category as a function of bridge age. Coefficients were determined for
have the longest estimated service life (approximately 90 years). the different bridge types and for the deck, superstructure,
Similar to the superstructure ratings, the substructure service and substructure. The equations and a graph of the equations
life for the medium-traffic category is the longest and may be showing the NBI condition rating as a function of time are
due to better maintenance. shown in Table 4.2. Unlike the equations by Bolukbasi et al.
Table 4.2. Rating Prediction Equations and Graphs for Four Categories of Bridges
Steel Interstate Bridges Deck: R = 9 - 0.41141790T + 0.02116563T 2

- 4.0387 × 10 -4T 3
10.00
Deck
9.00 Superstructure: R = 9 - 0.45572206T
7.00 Substructure
Substructure: R = 9 - 0.44818105T
Rating
6.00 + 0.02555900T 2 - 4.9875 × 10 -4T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Concrete Interstate Bridges Deck: R = 9 - 0.36622617T + 0.01659520T 2

- 2.7162 × 10 -4T 3
10.00
Deck
7.00 Substructure
Rating
6.00 + 0.01575857T 2 - 2.6681 × 10 -4T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)

61
Table 4.2. Rating Prediction Equations and Graphs for Four Categories of Bridges (continued)
Steel Non-Interstate Bridges Deck: R = 9 - 0.34979283T + 0.01036093T 2

- 1.1009 × 10 -4T 3
10.00
Deck
7.00 Substructure
Rating
6.00 + 0.01093574T 2 - 1.1953 × 10 -4T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Concrete Non-Interstate Bridges Deck: R = 9 - 0.30199933T + 0.00915111T2

- 9.409 × 10-5T 3
10.00
Deck
8.00 Superstructure + 0.00860726T 2 - 8.815 × 10-5T 3
7.00 Substructure
Rating
6.00 + 0.00961677T 2 - 9.876 × 10-5T 3

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Source: Jiang and Sinha (1989).
(2004), the constant term in the prediction equation is and Figure 4.6 show the predicted deck, superstructure, and
always 9; this assumes that the bridge component was in substructure condition ratings, respectively, versus time. It is
perfect condition when new. typically assumed that the end of service life occurs when the
The equations were also plotted by component (deck, condition rating reaches a value of 3. Figure 4.4 shows that
superstructure, substructure), allowing an investigation into the Interstate bridges typically have a shorter deck service
how other criteria affect the service life of the bridge. Decks life than the non-Interstate bridges. All bridge types have
are believed to be CIP concrete decks. Figure 4.4, Figure 4.5, a similar deterioration rate until approximately 10 years.

10.00 10.00
Steel Interstate Steel Interstate
9.00 9.00
Concrete Interstate Concrete Interstate
8.00 Steel Non-Interstate 8.00 Steel Non-Interstate
7.00 Concrete Non-Interstate 7.00 Concrete Non-Interstate
Rating
Rating
6.00 6.00
5.00 5.00
4.00 4.00
3.00 3.00
2.00 2.00
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Figure 4.4. Deck predicted condition ratings for Figure 4.5. Superstructure predicted condition
different bridge types. ratings for different bridge types.

62
Substructure Ratings which deterioration models were developed specifically for

10.00 Nebraska bridges. The deterioration models were based on
Steel Interstate
9.00
Concrete Interstate
NBI condition ratings for bridge decks, superstructures, and
8.00 Steel Non-Interstate substructures by using data from 1998 to 2010. Factors such
7.00 Concrete Non-Interstate as structure type, deck type, wearing surface, deck protection,
Rating
6.00 average daily traffic (ADT), average daily truck traffic (ADTT),
5.00 and location were considered in the development of the dete-
4.00 rioration models, which were determined using deterministic
3.00 and stochastic methods.
2.00 NBI data were obtained for all bridges in Nebraska from
0 20 40 60 80 100 120
Time (years)
1998 to 2010; only data for state bridges were used in the
analysis as the authors believed that inspections performed
Figure 4.6. Substructure predicted condition ratings by state inspectors have stricter requirements. The determin-
for different bridge types. istic deterioration models developed for state bridges in
Nebraska are shown in Table 4.3. In the second figure in the
table, in which deterioration is related to ADTT, decks sub-
Non-Interstate bridges have a longer period during which jected to more truck traffic appear to have a longer expected
the rating does not change significantly; the concrete bridge life than those subjected to fewer trucks, which is contrary to
rating is higher than the steel bridge rating during this pla- what would be expected. In the third figure, the rating starts
teau period and throughout most of the service life. Con- to increase in District 2 after approximately 60 years, which is
crete bridges are shown to have longer deck service lives than likely a sign that more data were needed to more accurately
steel bridges. The predicted service life for Interstate bridge develop the deterioration model. The last figure shown in
decks is approximately 36 years for steel bridges and 42 years Table 4.3 indicates that the service life of the deck exceeds that
for concrete bridges. For non-Interstate bridges, the service of either the superstructure or substructure, which is also
life increases to approximately 59 years for steel bridges and
contrary to what would be expected.
62 years for concrete bridges. The reference did not provide any
information as to why the deck service life varies between steel
and concrete bridges for both Interstate and non-Interstate 4.5 Comparison of Equations
conditions. from Bolukbasi et al.
Figure 4.5 is very similar to Figure 4.4 with regard to mate- (2004), Jiang and Sinha
rial and highway type. Concrete superstructures have longer (1989), and Hatami and
service lives than steel bridges subjected to the same volume Morcous (2011)
of traffic. Interstate bridges have a shorter service life than non-
4.5.1 Introduction
Interstate bridges. The superstructure service lives predicted
are very similar to those predicted for the deck. The difference The results from Bolukbasi et al. (2004), Jiang and Sinha
in service life between concrete and steel bridges is not dis- (1989), and Hatami and Morcous (2011) are generally simi-
cussed in the reference. lar. The equations are plotted together to provide a comparison
Figure 4.6 shows the predicted substructure condition between resulting equations. Various comparisons are pro-
ratings versus times. This figure is similar to the two previous vided below. Comparisons are based on material type, as well
figures for deck and superstructure service lives. The predicted as highway type and ADTT. Typically the Bolukbasi et al.
substructure service lives are very similar to those predicted for equations have a slower deterioration rate over the service life
the deck and also for the superstructure. This is a surprising of the structure. This may be due to a larger number of struc-
result as it is typically expected that the substructure will last tures being considered and the availability of more inspection
longer than either the deck or the superstructure. The report data. The equations from the Nebraska study are only
did not indicate any specific reasons for the substructure having included in the superstructure ratings for steel bridges; the
predicted service lives similar to the deck and superstructure. results for other bridge and component types were not spe-
cific enough to include elsewhere.
4.4 Hatami and Morcous (2011)
4.5.2 Concrete Superstructure Bridges
A 2011 report by Hatami and Morcous, Developing Deteriora-
tion Models for Nebraska Bridges, presented the results of a Plots of the prediction equations for deck, superstructure,
project performed for the Nebraska Department of Roads in and substructure condition ratings for concrete bridges are

63
Table 4.3. Nebraska Deterioration Models
Deck and Overlay Ratings Original Deck: R = 10.2915 - 0.2531T + 0.0093T 2

- 0.0001T 3
10.00
Original Deck
9.00 Replacement Deck: R = 8.48681 + 0.34139T
8.00 Replacement - 0.05392T 2 + 0.00222T 3 - 0.00003T 4
Deck
7.00 Overlay Overlay: R = 9.6499 - 0.0829T + 0.0009T 2
Rating
6.00 - 0.0002T 3
Low Slump
5.00 Concrete Overlay
Low Slump Concrete Overlay: R = 10.094
4.00
- 0.1902T + 0.0087T 2 - 0.0004T 3
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Deck Ratings based on ADTT ADTT < 100: R = 10.189 - 0.233T + 0.0092T 2
- 0.0002T 3
10.00
Deck w/
9.00 100 < ADTT < 500: R = 10.754 - 0.342T
ADTT<100
8.00 + 0.0127T 2 - 0.0002T 3
Deck w/
7.00 100<ADTT<500
ADTT > 500: R = 10.372 - 0.2311T + 0.0039T 2
Rating
Deck w/
6.00 ADTT>500 - 0.00004T 3
5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Deck Ratings based on Location Districts 1, 3, and 4: R = 9.9984 - 0.1948T

+ 0.0052T 2 - 0.00008T 3
10.00
9.00 District 2: R = 9.9374 - 0.1267T - 0.0015T 2
8.00 + 0.00003T 3
7.00
Districts 5–8: R = 10.252 - 0.2214T + 0.0067T 2
Rating
6.00 - 0.0001T 3
5.00
4.00 Deck Dis. 1, 3, 4
3.00 Deck Dis. 2
Deck Dis. 5-8
2.00
0 20 40 60 80 100 120
Time (years)
Deck and Overlay Ratings Original Deck: R = 10.2915 - 0.2531T + 0.0093T 2

- 0.0001T 3
10.00
9.00 Original Deck Steel Superstructure: R = 10.2731 - 0.1727T
Steel Superstructure + 0.0046T 2 - 0.0001T 3
8.00
Substructure
7.00
Substructure: R = 9.6098 - 0.0657T - 0.0017T 2
Rating
6.00 + 0.00001T 3
5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Source: Hatami and Morcous (2011).

64

10.00 10.00
Bolukbasi - RC Bridges Bolukbasi - RC Bridges
9.00 9.00
8.00 Jiang and Sinha - Concrete 8.00 Jiang and Sinha - Concrete
Interstate Interstate
7.00 7.00 Jiang and Sinha - Concrete
Jiang and Sinha - Concrete
Rating
Rating
6.00 Non-Interstate 6.00 Non-Interstate
5.00 5.00
4.00 4.00
3.00 3.00
2.00 2.00
0 20 40 60 80 100 120 0 20 40 60 80 100 120
(a) (b)
Substructure Ratings
10.00
Bolukbasi - RC Bridges
9.00
Jiang and Sinha - Concrete
8.00
Interstate
7.00 Jiang and Sinha - Concrete
Rating
Non-Interstate
6.00
5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
(c)
Figure 4.7. Comparisons of concrete bridge predicted condition ratings for (a) decks, (b) superstructures,
and (c) substructures.
shown in Figure 4.7. For the deck condition ratings, the Figure 4.8. For the deck condition ratings, the Bolukbasi et al.
Bolukbasi et al. (2004) equation indicates the highest condi- (2004) equation indicates a higher condition rating until an
tion rating until an age of approximately 35 years, after age of approximately 50 years, after which the non-Interstate
which the non-Interstate equation for concrete bridges pro- equation for steel bridges proposed by Jiang and Sinha (1989)
posed by Jiang and Sinha (1989) indicates the highest con- indicates a higher condition rating. The superstructure
dition rating. For the superstructure and substructure, and substructure condition ratings are always higher when
the Bolukbasi et al. equations always indicate the highest con- using the Bolukbasi et al. equations versus either Jiang and
dition rating. The prediction equations provide an estimated Sinha equation. The Hatami and Morcous (2011) equation
service life for the deck, superstructure, and substructure. The
estimated service lives, or the predicted times until a condition
rating of 3 is achieved, are provided in Table 4.4. The Jiang and
Table 4.4. Service Life Comparison:
Sinha equations generally predict a similar service life for all Reinforced Concrete (RC) Bridges
major components of a bridge. The Bolukbasi et al. equation
suggests that the deck has the shortest service life. The super- Service Life (years)
structure and substructure service lives are significantly
Equation Deck Superstructure Substructure
longer.
Bolukbasi et al., RC 54 77 82
Jiang and Sinha, RC 42 41 41
4.5.3 Steel Superstructure Bridges Interstate
Plots of the prediction equations for deck, superstructure, and Jiang and Sinha, RC 63 63 63
non-Interstate
substructure condition ratings for steel bridges are shown in

65

10.00 10
Bolukbasi - Steel Bolukbasi - Steel
9.00 Bridges 9 Bridges
8.00 Jiang and Sinha - 8 Jiang and Sinha - Steel
Steel Interstate Interstate
7.00 7 Jiang and Sinha - Steel
Jiang and Sinha -
Rating
Rating
Non-Interstate
6.00 Steel Non-Interstate 6 NEDOR - Steel Bridges
5.00 5
4.00 4
3.00 3
2.00 2
0 20 40 60 80 100 120 0 20 40 60 80 100 120
(a) (b)
10.00
Bolukbasi - Steel
9.00 Bridges
8.00 Jiang and Sinha -
Steel Interstate
7.00
Jiang and Sinha -
Rating
6.00 Steel Non-Interstate

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
(c)
Figure 4.8. Comparisons of steel bridge predicted condition ratings for (a) decks, (b) superstructures,
[identified as NEDOR (Nebraska Department of Roads) in rating until an age of approximately 25 years. After 25 years,
Figure 4.8] predicts higher condition ratings than both the Bolukbasi et al. equation and the Jiang and Sinha (1989)
Bolukbasi et al. and Jiang and Sinha until approximately concrete equations are very similar. The superstructure and
30 years. Unlike the other equations, the Hatami and substructure condition ratings are always higher when using
Morcous equation does not indicate a period of time when the Bolukbasi et al. equations versus either Jiang and Sinha
the condition rating plateaus. The prediction equations pro- equation, but overall the three indicate similar estimated
vide an estimated service life for the deck, superstructure,
and substructure. The estimated service lives, or the predicted
times until a condition rating of 3 is achieved, are provided in
Table 4.5. The Jiang and Sinha equations generally predict a Table 4.5. Service Life Comparison: Steel Bridges
similar service life for the major components of a bridge. The Service Life (years)
Bolukbasi et al. equation suggests that the deck has the short-
est service life. The superstructure and substructure service Equation Deck Superstructure Substructure
lives are somewhat longer. Bolukbasi et al., steel 53 63 68

Jiang and Sinha, 36 37 36
steel Interstate
4.5.4 Interstate Bridges
Jiang and Sinha, steel 54 56 57
Plots of the prediction equations for deck, superstructure, non-Interstate
and substructure condition ratings for Interstate bridges are Hatami and Morcous, NA 47 NA
shown in Figure 4.9. For the deck condition ratings, the steel
Bolukbasi et al. (2004) equation indicates a higher condition Note: NA = not available.

66

10.00 10.00
Bolukbasi - Interstate Bolukbasi - Interstate
9.00 Bridges 9.00 Bridges
8.00 Jiang and Sinha - 8.00 Jiang and Sinha -
7.00 Steel Interstate 7.00 Steel Interstate
Rating
Rating
Jiang and Sinha - Jiang and Sinha -

6.00 Concrete Interstate 6.00
Concrete Interstate
5.00 5.00
4.00 4.00
3.00 3.00
2.00 2.00
0 20 40 60 80 100 120 0 20 40 60 80 100 120
(a) (b)
10.00
Bolukbasi - Interstate
9.00 Bridges
7.00 Steel Interstate
Rating
Jiang and Sinha -

6.00 Concrete Interstate
5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
(c)
Figure 4.9. Comparisons of Interstate bridge predicted condition ratings for (a) decks, (b) superstructures,
service lives. The prediction equations provide an estimated 4.5.5 Non-Interstate Bridges
service life for the deck, superstructure, and substructure.
The estimated service lives, or the predicted times until a con- The prediction equations for deck, superstructure, and
dition rating of 3 is achieved, are provided in Table 4.6. The substructure condition ratings for non-Interstate bridges
estimated service lives are similar for all components and are shown in Figure 4.10. For the deck condition ratings,
equations. the Bolukbasi et al. (2004) equation indicates a higher rat-
ing until an age of approximately 40 years. After 40 years,
the Jiang and Sinha (1989) concrete equation indicates the
highest condition rating, and after approximately 45 years,
Interstate Bridges the Jiang and Sinha steel equation provides an estima-
ted service life greater than the Bolukbasi et al. equation.
Service Life (years) The superstructure and substructure condition ratings
are always higher when using the Bolukbasi et al. equations
versus either Jiang and Sinha equation. The Bolukbasi
Bolukbasi et al., 41 45 48 et al. equations indicate an overall slower deterioration
Interstate
rate. The prediction equations provide an estimated
Jiang and Sinha, 36 37 36 service life for the deck, superstructure, and substruc-
steel Interstate
ture. The estimated service lives, or the predicted times
Jiang and Sinha, 42 41 41 until a condition rating of 3 is achieved, are provided in
concrete Interstate
Table 4.7. The table shows that the Bolukbasi et al.

67

10.00 10.00
Bolukbasi - Non- Bolukbasi - Non-Interstate
9.00 Interstate Bridges 9.00 Bridges
8.00 Jiang and Sinha - Steel 8.00 Jiang and Sinha - Steel
Non-Interstate Non-Interstate
Rating
Rating
6.00 Concrete Non-Interstate 6.00 Concrete Non-Interstate
5.00 5.00
4.00 4.00
3.00 3.00
2.00 2.00
0 20 40 60 80 100 120 0 20 40 60 80 100 120
(a) (b)
10.00
Bolukbasi - Non-
9.00 Interstate Bridges
8.00 Jiang and Sinha - Steel
Non-Interstate
Rating
6.00 Concrete Non-Interstate
5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
(c)
Figure 4.10. Comparisons of non-Interstate bridge predicted condition ratings for (a) decks, (b) superstructures,
equations predict the shortest deck service life, but they <10,000 for Bolukbasi et al. (2004) and for non-Interstate
also predict the longest superstructure and substructure bridges for Jiang and Sinha (1989) are shown in Fig-
service lives. ure 4.11. For the deck condition ratings, the Bolukbasi
et al. equation for AADT <5,000 indicates the highest
4.5.6 Low- and Medium-AADT Bridges condition rating until an age of approximately 42 years. After
this time, the Jiang and Sinha concrete non-Interstate
The prediction equations for deck, superstructure, and equation indicates the highest condition rating, and after
substructure condition ratings for bridges with AADT approximately 50 years, the Jiang and Sinha steel non-
Interstate equation provides an estimated service life
similar to the Bolukbasi et al. equation for AADT <5,000.
Table 4.7. Service Life Comparison: The Bolukbasi et al. equation for AADT between 5,000
Non-Interstate Bridges
and 10,000 is greater than the Jiang and Sinha equations
Service Life (years)
until approximately 27 years. The superstructure and
substructure condition ratings are always higher when
Equation Deck Superstructure Substructure using the Bolukbasi et al. equations versus the Jiang and
Bolukbasi et al., 52 77 81 Sinha equations. The Bolukbasi et al. equations indicate
non-Interstate an overall slower deterioration rate. The equation for
Jiang and Sinha, 54 56 57 AADT <5,000 is greater until approximately 63 years
steel non-Interstate for the superstructure and approximately 50 years for
Jiang and Sinha, con- 62 62 63 the substructure. The prediction equations provide an
crete non-Interstate
estimated service life for the deck, superstructure, and

68

10 10
Bolukbasi - AADT < Bolukbasi - AADT <
9 5,000 9 5,000
8 Bolukbasi - 5,000 < 8 Bolukbasi - 5,000 <
AADT < 10,000 AADT < 10,000
7 Jiang and Sinha - Steel 7 Jiang and Sinha - Steel
Rating
Rating
6 Non-Interstate Non-Interstate
6
Jiang and Sinha - Jiang and Sinha -
5 Concrete Non-Interstate 5 Concrete Non-Interstate
4 4
3 3
2 2
0 20 40 60 80 100 120 0 20 40 60 80 100 120
(a) (b)
10
Bolukbasi - AADT <
9 5,000
Bolukbasi - 5,000 <
8
AADT < 10,000
7 Jiang and Sinha - Steel
Rating
Non-Interstate
6 Jiang and Sinha -
5 Concrete Non-Interstate
4
3
2
0 20 40 60 80 100 120
Time (years)
(c)
Figure 4.11. Comparisons of low- to medium-AADT bridge predicted condition ratings for (a) decks,
(b) superstructures, and (c) substructures.
substructure. The estimated service lives, or the predicted 4.5.7 High-AADT Bridges
times until a condition rating of 3 is achieved, are provided
in Table 4.8. The table shows that the Bolukbasi et al. equa- The prediction equations for deck, superstructure, and sub-
tions predict the shortest deck service life, but they also structure condition ratings for bridges with AADT >10,000
predict the longest superstructure and substructure ser- for Bolukbasi et al. (2004) and for Interstate bridges for Jiang
vice lives. and Sinha (1989) are shown in Figure 4.12. The Bolukbasi
et al. deck condition rating equation is greater for approxi-
mately 30 years, after which it is very similar to the Jiang
and Sinha concrete Interstate prediction equation. The
Low- to Medium-AADT Bridges superstructure and substructure condition ratings are
always higher when using the Bolukbasi et al. equation
Service Life (years) versus the Jiang and Sinha equations. The Bolukbasi et al.
equations indicate an overall slower deterioration rate. The
prediction equations provide an estimated service life for
Bolukbasi et al., 54 76 80
the deck, superstructure, and substructure. The estimated
AADT <5,000
service lives, or the predicted times until a condition rating
Bolukbasi et al., 5,000 47 81 90
< AADT < 10,000 of 3 is achieved, are provided in Table 4.9. The table shows
Jiang and Sinha, steel 54 56 57
that the Bolukbasi et al. equations predict a deck service life
non-Interstate approximately equal to the service life predicted by the
Jiang and Sinha, 62 62 63 Jiang and Sinha concrete Interstate bridge equation, and
concrete they also predict the longest superstructure and substructure
non-Interstate
service lives.

69

10.00 10.00
Bolukbasi - AADT > Bolukbasi - AADT >
9.00 10,000 9.00 10,000
Steel Interstate Steel Interstate
7.00 7.00
Jiang and Sinha -
Rating
Rating
Jiang and Sinha -
6.00 Concrete Interstate 6.00 Concrete Interstate
5.00 5.00
4.00 4.00
3.00 3.00
2.00 2.00
0 20 40 60 80 100 120 0 20 40 60 80 100 120
(a) (b)
10.00
Bolukbasi - AADT >
9.00 10,000
Steel Interstate
7.00
Jiang and Sinha -
Rating
6.00 Concrete Interstate

5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
(c)
Figure 4.12. Comparisons of high-AADT bridge predicted condition ratings for (a) decks, (b) superstructures,
4.6 Agrawal and Kawaguchi • Abutment wingwall (four options: none, other, wingwall
(2009) exists, and reinforced earth wingwall);
• Abutment bearing (six options: none, steel, polytetra
A 2009 report by Agrawal and Kawaguchi provides regression fluoroethylene [PTFE], multirotational, elastomeric, and
equations relating condition rating (CR) to age for common others);
bridge components in New York State. The following list indi- • Abutment pedestal (all grouped together);
cates the number of options and examples of each component: • Abutment joint (12 options: none, open, finger, sliding
• Abutment backwall (all grouped together); plate, filled elastic material, preformed elastomeric seals,
• Abutment stem (all grouped together); strip seal, sawed and filled, compression, modular, armored,
and other or unknown);
• Pier bearing (six options: none, steel, PTFE, multirota-
Table 4.9. Service Life Comparison: tional, elastomeric, and other or unknown);
High-AADT Bridges • Pier pedestal (five options: none, concrete, masonry, steel,
and timber);
Service Life (years)
• Pier cap top (five options: none, concrete, masonry, steel,
Equation Deck Superstructure Substructure and timber);
Bolukbasi et al., 41 48 49 • Pier cap (five options: none, concrete, masonry, steel, and
AADT >10,000 timber);
Jiang and Sinha, 36 37 36 • Pier stem (all grouped together);
steel Interstate • Pier column (five options: none, concrete, masonry, steel,
Jiang and Sinha, 42 41 41 and timber);
concrete Interstate • Pier footing (all grouped together);

70
• Pier recommendation (five options: none, concrete, 10,175 and 25,457 tons), high (between 16,969 and
masonry, steel, and timber); 40,195 tons), and very high (between 46,375 and
• Pier joint (12 options: none, open, finger, sliding plate, 94,739 tons);
filled elastic material, preformed elastomeric seals, strip • Snow accumulation—There are three snow accumulation
seal, sawed and filled, compression, modular, armored, and categories: low (<171 in.), medium (between 171 and
other or unknown); and 278 in.), and high (between 278 and 458 in.);
• Primary member design type (19 options, such as rolled • Climate groups—The 10 groups are based on climate data
beam, truss, and deck arch). provided by the National Oceanic and Atmospheric
Administration;
The work by Agrawal and Kawaguchi resulted in a com- • Functional class—There are five functional classes ranging
puter program based on synthesized Pontis data that calcu- from Interstate to none; and
lates the deterioration rates of bridge components using • Feature under—This factor has three categories: Interstate
Pontis data. The program contains a cascading algorithm to under, highway under, and water under.
classify bridges based on several factors. These factors are
The factors listed above were used to create a class of bridges
• Element design type—For bearings, for example, the type that have similar characteristics. The number of characteris-
can be one of six choices (none, steel, PTFE, multirota- tics selected allows the deterioration rate to be calculated for a
tional, elastomeric, and others); very narrow or a very broad group of bridges.
• New York State Department of Transportation (NYSDOT) Within a specific component, multiple equations may be
Region—There are 11 regions in New York State; provided for different materials or types of components. As
• Bridge ownership—Various organizations own bridges an example, for abutment bearings, four equations (one
within New York State, including NYSDOT, park authori- each for steel bearings, elastomeric bearings, multirota-
ties or commissions, nonpark authorities or commissions, tional bearings, and PTFE sliding bearings) were provided.
and the New York State Thruway Authority. Bridges are The equations and graphs are shown in Table 4.10. The
also owned locally, privately, by railroads, and by other ratings in New York vary from 1 to 7, with 7 indicating
entities; perfect condition; 5 indicating minor deterioration but still
• Superstructure design type—These include girder and functioning as designed; 3 indicating serious deterioration
floorbeam system, truss, and suspension; or not functioning as designed; and 1 indicating a failed
• Superstructure material type—These include weathering condition. Even-numbered ratings (2, 4, and 6) are used to
steel, timber, and prestressed concrete; provide a middle ground between the odd numbered,
• AADT—AADT is divided into five groups ranging from defined ratings (1, 3, 5, and 7).
no trucks to >5,000 trucks per day; If failure is defined as a condition rating of 3 and the
• Salt usage—Salt usage is divided into four categories: component no longer functioning as intended, then the
low (between 6,893 and 13,492 tons), medium (between (text continues on page 78)
Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data
Abutment Backwall CR = 7 - 0.0564703T + 0.0000667T2
7.00
6.50 Abutment Backwall
6.00
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)

71
Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data (continued)
Abutment Bearings Steel: CR = 7 - 0.0773187T + 0.0002408T 2
7.00 PTFE: CR = 7 - 0.1369652T + 0.0023073T 2

Steel
6.50 - 2.25 × 10 -5T 3
PTFE
6.00
Multirotational
5.50 Multirotational: CR = 7 - 0.1276043T
Elastomeric
5.00 + 0.0020318T 2 - 1.8 × 10 -5T 3
Rating
4.50
4.00 Elastomeric: CR = 7 - 0.0633160T
3.50 + 0.0002109T 2 - 1 × 10 -7T 3
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Abutment Joints Open: CR = 7 - 0.1544542T + 0.0019093T 2

- 1.01 × 10 -5T 3
7.00
Open Compression Compression: CR = 7 - 0.1546255T + 0.0019008T 2
6.50
Modular Armor - 9.3 × 10-6T 3
6.00
Sliding Plate Filled Elastic Modular: CR = 7 - 0.1944402T + 0.0054188T 2 -
5.50
Pref ormed 7.4 × 10 -5T 3
5.00
Rating
4.50 Armor: CR = 7 - 0.1667466T + 0.0022536T 2

4.00 - 1.29 × 10 -5T 3
3.50 Sliding Plate: CR = 7 - 0.1955859T + 0.0043095T 2
3.00 - 3.42 × 10 -5T 3
2.50
Filled Elastic: CR = 7 - 0.1416458T + 0.0016176T 2
2.00
0 20 40 60 80 100 120 - 6.3 × 10-6T 3
Time (years) Preformed: CR = 7 - 0.1563427T + 0.0014834T 2
- 5.0 × 10-6T 3
Abutment Pedestal CR = 7 - 0.0484691T - 0.0000925T 2
7.00
6.50 Abutment Pedestal
6.00
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Abutment Stem CR = 7 - 0.0562065T - 0.0000832T 2
7.00
6.50 Abutment
6.00 Stem
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)

72
Abutment Wingwall CR = 7 - 0.0500728T - 0.0000546T 2 - 6.0 × 10-7T 3
7.00
6.50 Abutment
6.00 Wingwall
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Deck Curb Granite/Stone: CR = 7 - 0.0605424T + 0.0001089T 2

- 1.0 × 10-7T 3
7.00
Granite/Stone
6.50 Steel Plate: CR = 7 - 0.0577393T - 0.0001956T 2
Steel Plate
6.00 - 1.7 × 10-6T 3
Timber
5.50
Concrete
5.00 Timber: CR = 7 - 0.0584921T - 0.0003144T 2
Rating
4.50 - 2.4 × 10-6T 3

4.00
3.50 Concrete: CR = 7 - 0.0507576T - 0.0002625T 2
3.00 - 1.9 × 10-6T 3
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Pier Bearings Steel: CR = 7 - 0.0681319T - 0.0001597T 2 + 3.4

× 10 -6T 3
7.00
6.50 Steel
Multirotational: CR = 7 - 0.0833154T + 0.0008055T 2
6.00 Multirotational - 3.8 × 10 -6T 3
5.50 Elastomeric
5.00 Elastomeric: CR = 7 - 0.0845871T + 0.0008876T 2
Rating
4.50 - 7.3 × 10 -6T 3

4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Pier Cap Concrete: CR = 7 - 0.0575767T + 0.0000583T 2

+ 1.0 × 10-7T 3
7.00
Concrete
6.50 Masonry: CR = 7 - 0.0347071T - 0.0002426T 2 + 1.1
Masonry
6.00 × 10 -6T 3
Steel
5.50
Timber
5.00 Steel: CR = 7 - 0.0172139T - 0.0008876T 2 + 3.8
Rating
4.50 × 10 -6T 3
4.00
3.50 Timber: CR = 7 - 0.0674187T + 0.0001438T 2 + 1.0
3.00 × 10 -6T 3
2.50
2.00
0 20 40 60 80 100 120
Time (years)

73
Pier Cap Top Concrete: CR = 7 - 0.0475800T - 0.0001091T 2 + 1.2

× 10-6T 3
7.00
Concrete
6.50 Masonry: CR = 7 - 0.0094394T - 0.0007153T 2 + 3.8
Masonry
6.00 × 10-6T 3
Steel
5.50
Timber
5.00 Steel: CR = 7 - 0.0131302T - 0.0007820T 2 + 4.9
Rating
4.50 × 10-6T 3
4.00
3.50 Timber: CR = 7 - 0.0467232T + 0.0001051T 2 - 1.3
3.00 × 10-6T 3
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Pier Column Concrete: CR = 7 - 0.0486218T - 0.0001326T 2 + 1.2

× 10-6T 3
7.00
Concrete
6.50 Masonry: CR = 7 - 0.1461181T + 0.0028522T 2 - 2.66
Masonry
6.00 × 10 -5T 3
Steel
5.50
Timber
5.00 Steel: CR = 7 - 0.0594952T + 0.0002300T 2 - 4.0
Rating
4.50 × 10-7T 3
4.00
3.50 Timber: CR = 7 - 0.1077933T + 0.0012051T 2 - 7.9
3.00 × 10-6T 3
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Pier Footing CR = 7 - 0.0361181T - 0.0001836T 2
7.00
6.50 Pier Footing
6.00
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Pier Joint Open: CR = 7 - 0.1746867T + 0.0029733T 2 - 2.24

× 10-5T 3
7.00 Strip Seal: CR = 7 - 0.2222855T + 0.0043429T 2
Open Strip Seal
6.50 - 3.68 × 10-5T 3
Compression Modular
6.00 Compression: CR = 7 - 0.2047452T + 0.0034777T 2
Armor Sliding Plate
5.50 - 2.09 × 10-5T 3
Filled Elastic Preformed
5.00 Modular: CR = 7 - 0.1178004T + 0.0000691T 2 + 1.37
Rating
4.50 × 10-5T 3
4.00 Armor: CR = 7 - 0.1623125T + 0.0012891T 2 - 1.0
3.50 × 10-7T 3
3.00 Sliding Plate: CR = 7 - 0.1581306T + 0.0016926T 2
2.50 - 8.9 × 10-6T 3
2.00 Filled Elastic: CR = 7 - 0.1937046T + 0.0028916T 2
0 20 40 60 80 100 120 - 1.3 × 10-5T 3
Time (years) Preformed: CR = 7 - 0.1725949T + 0.0020362T 2
- 9.6 × 10-6T 3

74
Pier Pedestal Concrete: CR = 7 - 0.0427029T - 0.0003432T 2 + 2.8

× 10-6T 3
7.00
6.50 Concrete
Masonry: CR = 7 - 0.0214166T - 0.0007708T 2 + 5.0
6.00 Masonry × 10-6T 3
5.50
Steel
5.00 Steel: CR = 7 - 0.0294246T - 0.0002940T 2 + 1.5
Rating
4.50 × 10-6T 3
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Pier Design Concrete: CR = 7 - 0.0616063T + 0.0001235T 2 - 1.0

× 10-7T 3
7.00
Concrete
6.50 Masonry: CR = 7 + 0.0189981T - 0.0013498T 2 + 7.5
Masonry
6.00 × 10-6T 3
Steel
5.50
Timber
5.00 Steel: CR = 7 - 0.0335030T - 0.0004089T 2 + 2.4
Rating
4.50 × 10-6T 3
4.00
3.50 Timber: CR = 7 - 0.1156794T + 0.0014818T 2 - 9.8
3.00 × 10-6T 3
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Pier Stem CR = 7 - 0.0445180T - 0.0001482T 2 + 1.1 × 10-6 T 3
7.00
6.50 Pier Stem
6.00
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Primary Members Slab, Box, Box Channel: CR = 7 - 0.0724412T

+ 0.0002255T 2 - 4.0 × 10-7T 3
7.00
Slab, Box, Box Channel
6.50 Tee or I-Beam: CR = 7 - 0.0509168T - 0.0001729T 2
Tee or I-Beam
6.00 + 2.1 × 10-6T 3
Rolled Beam
5.50
Plate Girder
5.00 Rolled Beam: CR = 7 - 0.0573849T + 0.0000603T 2
Rating
4.50 + 1.0 × 10-7T 3

4.00
3.50 Plate Girder: CR = 7 - 0.0533815T + 0.0000618T 2
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)

75
Primary Members Truss: CR = 7 - 0.0962120T + 0.0005460T 2 - 1.6

× 10-6T 3
7.00 Deck Arch: CR = 7 - 0.0608540T - 0.0001644T 2
6.50 Truss
+ 2.5 × 10-6T 3
6.00 Deck Arch
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Primary Members Metal Pipe Arch: CR = 7 - 0.0917752T +

0.0006315T 2 - 1.2 × 10-6T 3
7.00
Metal Pipe Arch
6.50 Frame: CR = 7 - 0.0586090T - 0.0000153T 2 + 9.0
Frame
6.00 × 10-7T 3
Box Culvert
5.50
Pipe Culvert
5.00
Rating
4.50
4.00 Box Culvert: CR = 7 - 0.0662312T + 0.0002877T 2
3.50 - 1.1 × 10-6T 3
3.00
2.50 Pipe Culvert: CR = 7 - 0.0918358T + 0.0005486T 2
2.00 - 1.9 × 10-6T 3
0 20 40 60 80 100 120
Time (years)
Secondary Members Slab, Box, or Box Channel: CR = 7 - 0.0705115T

+ 0.0002846T 2 - 2.0 × 10-7T 3
7.00
6.50 Slab, Box, or Box Channel
Tee or I-Beam: CR = 7 - 0.0371296T - 0.0004970T 2
6.00 Tee/I-Beam
- 4.1 × 10-6T 3
5.50 Rolled Beam
5.00 Plate Girder Rolled Beam: CR = 7 - 0.0536963T + 0.0002090T 2
Rating
4.50 - 1.6 × 10-6T 3

4.00
3.50 Plate Girder: CR = 7 - 0.0403950T - 0.0002383T 2
3.00 + 1.6 × 10-6T 3
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Secondary Members Truss: CR = 7 - 0.0600905T - 0.0001653T 2 + 2.1

× 10-6T 3
7.00
6.50 Truss
Deck Arch: CR = 7 + 0.0225284T - 0.0019546T 2
6.00 Deck Arch
+ 1.38 × 10-5T 3
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)

76
Secondary Members Frame: CR = 7 - 0.0031620T - 0.0007666T 2 + 3.9

× 10-6T 3
7.00
6.50 Frame
6.00
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Sidewalk/Fascia Concrete: CR = 7 - 0.0697598T + 0.0001899T 2 - 4.0

× 10-7T 3
7.00
Concrete
6.50 Steel Plate: CR = 7 - 0.0636279T + 0.0001742T 2
Steel Plate
6.00 + 1.0 × 10-7T 3
Asphalt Concrete
5.50
Steel
5.00 Asphalt Concrete: CR = 7 - 0.1145251T
Rating
4.50 + 0.0015822T 2 - 1.34 × 10-5T 3

4.00
3.50 Steela: CR = 7 - 0.0055077T - 0.0034812T 2 + 4.78
3.00 × 10-5T 3
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Structural Deck CIP Concrete Black Rebar: CR = 7 - 0.0675608T

+ 0.0001411T 2 + 1.0 × 10-7T 3
7.00
CIP Concrete Black Rebar
6.50 Precast Plank: CR = 7 - 0.1188157T + 0.0018646T 2
Precast Plank
6.00 - 2.04 × 10-5T 3
CIP w/ Epoxy Rebar
5.50
CIP w/ Other Prot. or Coating
5.00 CIP with Epoxy Rebar: CR = 7 - 0.0767927T
Rating
4.50 + 0.0007988T 2 - 5.1 × 10-6T 3

4.00
3.50 CIP with Other Protection or Coating: CR = 7
3.00 - 0.0793700T + 0.0005157T 2 - 2.3 × 10-6T 3
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Structural Deck Timber: CR = 7 - 0.1015141T + 0.0010366T 2 - 7.3

× 10-6T 3
7.00
6.50 Timber
Steel Grating: CR = 7 - 0.0971087T + 0.0005147T 2
6.00 Steel Grating - 7 × 10-7T 3
5.50
Steel Plate
5.00 Steel Plate: CR = 7 - 0.1387853T + 0.0032377T 2
Rating
4.50 - 2.53 × 10-5T 3

4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)

77
Superstructure Type Slab: CR = 7 - 0.0550966T - 0.0000107T 2 + 5.0

× 10-7T 3
7.00
Slab
6.50 Stringer/Multi-beam/Girder: CR = 7 - 0.0608104T
Stringer/Multi-beam/Girder
6.00 + 0.0001228T 2 - 2 × 10-7T 3
Girder/F loorbeam System
5.50
Tee Beam
5.00 Girder/Floorbeam System: CR = 7 - 0.0375553T
Rating
Box Beam or Box Girder

4.50 - 0.0003374T 2 + 1.9 × 10-6T 3
4.00
3.50 Tee Beam: CR = 7 - 0.0334694T - 0.0005675T 2
3.00 + 4.1 × 10-6T3
2.50
2.00 Box Beam or Box Girder: CR = 7 - 0.0671339T
0 20 40 60 80 100 120 + 0.0000287T 2 + 1.5 × 10-6T 3
Time (years)
Superstructure Type Frame: CR = 7 - 0.0374148T - 0.0004245T 2 + 3.1

× 10-6T 3
7.00
6.50 Frame
Culvert: CR = 7 - 0.0683836T + 0.0002159T 2 - 1.0
6.00 Culvert × 10-7T 3
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Superstructure Type Thru Truss: CR = 7 - 0.0719036T + 0.0001651T 2
7.00 Deck Arch: CR = 7 - 0.0209106T + 0.0007879T 2

6.50 Thru Truss
+ 5.1 × 10-6T 3
6.00 Deck Arch
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Wearing Surface Integral/Monolithic Portland Cement (PC): CR = 7

- 0.1178904T + 0.0012462T 2 + 7.0 × 10-6T 3
7.00
6.50 Integral/Monolithic PC
Concrete with Membrane: CR = 7 - 0.3488945T
6.00 Concrete w/ Membrane + 0.021168T 2 - 5.196 × 10-4T 3
5.50
Class H Concrete
5.00 Class H Concretea: CR = 7 - 0.0417046T
Rating
4.50 + 0.0022971T 2
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)

78
Wearing Surface PC Overlay: CR = 7 - 0.1517338T + 0.0019529T 2

- 9.7 × 10-6T 3
7.00
6.50 PC Overlay
Asphalt: CR = 7 - 0.1215795T + 0.0008883T 2 - 1.9
6.00 Asphalt × 10-6T 3
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Wearing Surface Wood/Wood Block: CR = 7 - 0.168718T

+ 0.0020122T 2 - 9.2 × 10-6T 3
7.00
6.50 Wood/Wood Block
Open Steel Grate: CR = 7 - 0.0726944T
6.00 Open Steel Grate + 0.0004775T 2 - 3.9 × 10-6T 3
5.50
5.00
Rating
4.50
4.00
3.50
3.00
2.50
2.00
0 20 40 60 80 100 120
Time (years)
Note: CR = condition rating.

a Indicates equations that, when plotted, do not appear correct based on graphs provided in the report.
Source: Agrawal and Kawaguchi (2009).
(continued from page 70) superstructures, and substructures. No distinction between

service life of each component can be estimated. The esti- CIP decks and precast panels was noted in the reference.
mated service lives are shown in Table 4.11 to Table 4.13. Several of the equations are from work completed by others,
The reported service lives were determined by extending but most of the equations either use NBI data for Texas
the graph until a condition rating of 3 was reached. Doing bridges or the expert opinion of Texas bridge engineers. The
so may have resulted in some equations being used outside equations are shown and plotted in Table 4.14. The first set
their intended range of applicability. In addition, the of equations was determined using regression analysis by
bridges used in this analysis were combined into one large the Transportation Systems Center and is a function of both
group. This grouping may result in service lives for one age and ADT.
material being greater than that for another material that NBI data for Texas bridges were used to determine addi-
might be expected to last longer. As an example, in Table 4.12 tional equations relating age and ADT to condition ratings.
timber pier caps are predicted to last longer than concrete Linear, piecewise linear, and nonlinear equations were pro-
pier caps. Performing the analysis on a smaller group of posed; in addition, through a survey, equations based on
bridges may result in the concrete service life being greater expert opinion were determined considering the worst-
than that of the timber. case scenario, the most likely scenario, and the best-case
scenario. Table 4.15 shows the prediction equations for
coastal bridge substructures for different functional clas-
4.7 Stukhart et al. (1991)
sifications, and Table 4.16 shows the prediction equations
In a report to the Texas Department of Transportation for substructures in all regions not considering functional
(DOT), Stukhart et al. (1991) presented numerous equa- classification. The graphs in Table 4.16 and Table 4.17 show
tions predicting the condition rating for bridge decks, that the linear equations suggest service lives (time to reach

79
Table 4.11. Abutment Component Table 4.12. Pier Component

Estimated Service Lives Estimated Service Lives
Component Service Life (years) Component Service Life (years)
Abutment Backwall 78.0 Pier Bearings
Abutment Bearings Steel 61.4
Steel 65.0 Multirotational 98.0
PTFE 51.4 Elastomeric 69.0

Pier Cap
Multirotational 57.0
Concrete 76.0
Elastomeric 88.0
Masonry 84.5
Abutment Joints
Steel 83.4
Open 45.0
Timber 81.6
Compression 46.0
Pier Cap Top
Modular 41.3
Concrete 82.6
Armor 42.4
Masonry 93.0
Sliding plate 66.4
Steel 91.7
Filled elastic 58.0 Timber 84.8
Preformed 37.0 Pier Column
Abutment Pedestal 72.5 Concrete 77.4
Abutment Stem 81.0 Masonry 57.1
Abutment Wingwall 79.0 Steel 98.0
Timber 63.3
Pier Footing 79.0
a condition rating of 3) significantly longer than 100 years, Pier Joints

which is possible, but unlikely. The nonlinear equations are Open 47.7
terminated at the minimum value; beyond this point the Strip seal 34.3
condition rating would appear to increase, which is not
Compression 41.5
possible without maintenance. As only bridges without
maintenance, repair, or rehabilitation were used in the Modular 49.0
analysis, the rating should not increase with increasing age. Armor 33.7
A new structure would have a condition rating of 9; all the Sliding plate 37.2
prediction equations indicate the condition rating to be Filled elastic 42.0
near 8 when new.
Preformed 35.6
Piecewise linear equations were determined for differ-
ent functional classifications for the deck, superstructure, Pier Pedestal
and substructure condition ratings. The coefficients B0, Concrete 76.0

B1, B2, and B3 used in the piecewise linear equations are Masonry 78.3
presented in Table 4.17. The condition rating is described
Steel 91.4
by three linear equations that are applicable during cer-
Overall Pier
tain times of the bridge life; these equations are shown as
Equation 4.1. In the study by Stukhart et al., t1 and t2 are Concrete 75.7
defined as 25 and 45 years, respectively. Several of the Masonry 96.6
graphs are terminated at 45 years as the results of the Steel 78.7
regression analysis indicate that the condition rating would
Timber 66.0
increase, which cannot be true without maintenance being
Pier Stem 81.0
performed.

80
Table 4.13. Superstructure and Deck Estimated Service Lives
Component Service Life (years) Component Service Life (years)
Deck Curb Rolled beam 84.3

Granite or stone 75.6 Plate girder 81.3
Steel plate 75.9 Truss 64.5
Timber 58.3 Deck arch NA
Concrete 66.9 Frame 98.0
Primary Member
Structural Deck
Slab, box, or box channel 67.8
CIP with black rebar 69.9
Tee or I-beam 77.3
Precast plank 51.9
Rolled beam 76.7
CIP with epoxy rebar 87.0
Plate girder 82.9
CIP with other coating 74.5
Truss 56.9
Timber 61.0
Deck arch 65.7
Metal pipe arch 87.0 Steel grating 57.0
Frame 72.8 Steel plate 85.4
Box culvert 79.5 Wearing Surface

Pipe culvert 61.0 Integral or monolithic
Overall Superstructure Portland cement 57.7
Slab 75.4 Concrete with membrane 26.3
Multistringer or beam 76.0
Class H concrete NA
Girder or floorbeam 76.6
Portland cement overlay 53.0
Tee beam 75.6
Asphalt 48.0
Box beam or girder 68.9
Wood or wood block 37.8
Frame 77.4
Open steel grate 68.6
Culvert 76.0
Sidewalk or Fascia
Through truss 65.5
Deck arch 77.9 Concrete 68.0
Secondary Member Steel plate 82.0
Slab, box, or box channel 82.8 Asphalt concrete 59.0
Tee or I-beam 53.2 Steel NA

Note: NA = not available.
CR = superstructures; for most cases, the estimated service lives

were in excess of 150 years. Although this would seem like a
 B0 + B1t if t ≤ t1 good thing, it is known that most bridge decks and super-
 structures will not have a service life of this length. In fact, it
 B0 + B1t1 + B2 (t − t1) if t1 < t ≤ t 2 (4.1)
 is more likely that the service life of a bridge deck is closer to
B0 + B1t1 + B2 (t 2 − t1) + β3 (t − t 2 ) if t > t 2 40 or 50 years than 150 years. Although the estimated service
lives seem extreme, the results are shown to indicate the avail-
A nonlinear regression analysis was performed to deter- able data. The basic equation used to estimate the service life
mine the parameters for the best-fit exponential decay is shown as Equation 4.2; the graphs and parameters are
curve. Parameters were determined for bridge decks and shown in Table 4.18. The estimated service life is approxi-
superstructures based on functional classification using the mately equal to the absolute value of b2 for this set of data.
multiyear data set. The best-fit parameters and equations Looking at the values of b2, the only reasonable values are for
were used to estimate the service life of bridge decks and (text continues on page 86)

81
Table 4.14. Transportation Systems Center Prediction Equations
Deck Ratings CR = 9 - 0.119t - 2.158 × 10-6 (ADTAGE)

10.00
Deck (ADT = 1000)
( ADT ) ( AGE)
Deck (ADT = 5000) ADTAGE =
9.00 Deck (ADT = 10000) 10
Deck (ADT = 15000)
8.00
7.00
Rating
6.00
5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Substructure Ratings CR = 9 - 0.105t - 2.105 × 10-6 ADT

10.00
Substructure (ADT = 1000)
9.00 Substructure (ADT = 10000)
8.00
7.00
Rating
6.00
5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Superstructure Ratings CR = 9 - 0.103t - 1.982 × 10-6 ADT

10.00
Superstructure (ADT = 1000)
9.00 Superstructure (ADT = 10000)
8.00
7.00
Rating
6.00
5.00
4.00
3.00
2.00
0 20 40 60 80 100 120
Time (years)
Source: Stukhart et al. (1991).

82
Table 4.15. Coastal Substructure Condition Rating Prediction Equations
Coastal Substructure Rating-Interstate Highways IH1 = 7.80 - 0.022t

10.00
IH2 = 7.98 - 0.036t + 3.89 × 10-4t2 - 8.00 × 10-8 × t

9.00
× ADT
8.00
ADT assumed as 25,000 vehicles
7.00
Rating
6.00
5.00
4.00
3.00
IH 1 IH 2
2.00
0 20 40 60 80 100 120
Time (years)
Coastal Substructure Ratings-US Highways US1 = 7.81 - 0.017t

10.00
US2 = 7.94 - 0.028t + 2.02 × 10-4t2 - 6.00 × 10-8 ×

9.00
t × ADT
8.00
7.00
Rating
6.00
5.00
4.00
3.00
US 1 US 2
2.00
0 20 40 60 80 100 120
Time (years)
Coastal Substructure Ratings-State Highways SH1 = 8.12 - 0.025t

10.00
SH2 = 8.47 - 0.064t + 7.45 × 10-4 t2 - 2.20 × 10-7

9.00
× t × ADT
8.00
7.00
Rating
6.00
5.00
4.00
3.00
SH 1 SH 2
2.00
0 20 40 60 80 100 120
Time (years)

83
Table 4.15. Coastal Substructure Condition Rating Prediction Equations (continued)
Coastal Substructure Ratings-Farm-Market Highways FM1 = 8.11 - 0.028t

10.00
FM2 = 8.18 - 0.032t + 1.84 × 10-6t3

9.00
Other = 7.93 - 0.034t

8.00
7.00
Rating
6.00
5.00
4.00
3.00
FM 1 FM 2 Other
2.00
0 20 40 60 80 100 120
Time (years)
Note: IH = Interstate highways or principal arterials; US = U.S. highways (non-Interstate) or minor arterials; SH = state highways or minor arterials;
FM = farm-to-market roads or collectors.
Table 4.16. Substructure Condition Rating Prediction Equations by Region (continued)
Coastal Substructure Ratings C1 = 7.97 - 0.024t

10.00
C2 = 8.14 - 0.040t + 3.31 × 10-4t2 - 9.00 × 10-8 × t

9.00
× ADT
8.00
C3 = 8.22 - 0.057t + 1.05 × 10-3t2 - 8.25 × 10-6t3
- 9.00 × 10-8 × t × ADT
7.00
Rating
6.00
5.00
4.00
3.00
C1 C2 C3
2.00
0 20 40 60 80 100 120
Time (years)
East Texas Substructure Ratings ET1 = 8.20 - 0.037t

10.00
ET2 = 8.20 - 0.012t - 1.58 × 10-3t2 + 2.24 × 10-5t3

9.00
8.00
7.00
Rating
6.00
5.00
4.00
3.00
ET1 ET2
2.00
0 20 40 60 80 100 120
Time (years)

84
Inland Texas Substructure Ratings IT1 = 7.93 - 0.015t

10.00
IT2 = 8.05 - 0.028t - 4.40 × 10-4t2 + 3.76 × 10-6t3

9.00
- 6.00 × 10-8 × t × ADT
8.00
7.00
Rating
6.00
5.00
4.00
3.00
IT1 IT2
2.00
0 20 40 60 80 100 120
Time (years)
West Texas Substructure Ratings WT1 = 7.85 - 0.015t

10.00
WT2 = 8.27 - 0.059t - 1.22 × 10-3 t 2 + 9.00 × 10-6 t 3

9.00
- 4.40 × 10-7 × t × ADT
8.00
7.00
Rating
6.00
5.00
4.00
3.00
WT1 WT2
2.00
0 20 40 60 80 100 120
Time (years)
Panhandle Region Substructure Ratings PH1 = 7.72 - 0.015t

9.00
PH2 = 8.56 - 0.109t - 2.52 × 10-3 t 2 + 1.62 × 10-5t 3

8.00 - 3.60 × 10-7 × t × ADT
7.00
6.00
Rating
5.00
4.00
3.00
PH1 PH2
2.00
0 20 40 60 80 100 120
Time (years)

85
All Region Substructure Ratings AR1 = 7.98 - 0.023t

10.00
AR2 = 8.21 - 0.043t - 3.68 × 10-4t2 - 5.00 × 10-8

9.00
× t × ADT
8.00
7.00
Rating
6.00
5.00
4.00
3.00
AR1 AR2
2.00
0 20 40 60 80 100 120
Time (years)
Note: Equations are named according to region (C for coastline, ET for east Texas, and so forth) and are numbered in the order in which they are
presented in Stukhart et al. (1991).
Table 4.17. Piecewise Linear Condition Rating Equations and Coefficients
Deck Ratings
10.00 Mat. B0 B1 B2 B3
9.00 IH RC 8.17 -0.051 0.003 -0.046

IH Other 8.18 -0.025 0.004 -0.063
8.00
SFM RC 8.04 -0.029 -0.016 -0.034

7.00
SFM Other 8.06 -0.012 -0.004 -0.032
Rating
6.00
US RC 8.25 -0.056 0.018 -0.134
5.00 US ST 8.08 -0.016 0.004 -0.040
4.00 IH Reinf. Conc

IH Other
ST/FM Reinf. Conc
3.00 ST/FM Other
US Reinf. Conc
US Other
2.00
0 20 40 60 80 100 120
Time (years)
Superstructure Ratings
10.00 Mat. B0 B1 B2 B3
9.00 IH PS 8.23 -0.029 -0.132 -0.184

IH RC 8.27 -0.036 -0.036 -0.036
8.00
IH ST 8.16 -0.056 0.002 -0.050

7.00
SFM PS 8.33 -0.033 -0.027 0.084
Rating
6.00
SFM RC 8.08 -0.016 -0.026 -0.013
5.00 SFM ST 8.08 -0.035 -0.016 -0.056
IH P/S Conc.
4.00 IH Reinf. Conc US PS 8.34 -0.038 -0.038 0.004
IH Steel
St/FM P/S Conc.
ST/FM Reinf. Conc US RC 8.25 -0.034 -0.002 -0.125
3.00 ST/FM Steel
US P/S Conc.
US Reinf. Conc US ST 8.17 -0.053 -0.016 -0.147
US Steel
2.00
0 20 40 60 80 100 120
Time (years)

86
Table 4.17. Piecewise Linear Condition Rating Equations and Coefficients (continued)
10.00
Mat. B0 B1 B2 B3
IH RC 8.23 -0.039 -0.018 -0.004

9.00
IH ST 8.46 -0.066 -0.066 -0.066

8.00
IH Other 8.30 -0.029 -0.029 -0.029
7.00
SFM RC 8.32 -0.039 -0.039 0.020
Rating
6.00 SFM ST 7.87 -0.028 -0.048 -0.094
5.00 SFM Other 8.23 -0.032 -0.032 -0.032

IH Reinf. Conc.
IH Steel US RC 8.34 -0.046 0.007 -0.118
4.00 IH Other
St/FM Reinf. Conc.
ST/FM Steel US ST 8.08 -0.033 -0.063 -0.063
3.00 ST/FM Other
US Reinf Conc.
US Steel US Other 8.43 -0.040 0.009 -0.109
US Other
2.00
0 20 40 60 80 100 120
Time (years)
Note: IH = Interstate highways or principal arterials; RC = reinforced concrete; SFM = state farm-to-market road; US = U.S. highways (non-Interstate)
or minor arterials; ST = state highways; PS = prestressed; SH = state highways or minor arterials; FM = farm-to-market roads or collectors;
ST/FM = state highways and farm-to-market combined.
(continued from page 80) best-case expected remaining service life based on expert opin-
decks of prestressed concrete bridge on the state and farm-to- ion. The expected remaining service life was based on a given
market highway systems. condition rating: new (9), good (7), fair (5), and poor (3).
t
From these responses, an estimated condition rating deterio-
CR = β1e β2 (4.2) ration rate was determined. As would be expected with any
opinion-based survey, there was significant variation in the
The final method used to develop equations to predict con- responses; in several cases, the standard deviation was
dition ratings for the deck, superstructure, and substructure greater than the mean. The equations and graphs are shown
was a survey of Texas bridge engineers, who were asked to pro- in Table 4.19 for the deck, superstructure, and substructure
vide estimates of the worst-case, the most likely, and the condition ratings.
Table 4.18. Exponential Best-Fit Graphs and Parameters
Deck Ratings
10.00 Material b1 b2
9.00 IH RC 7.997 -198.662

IH Other 8.046 -718.254
8.00
US RC 7.882 -266.081
7.00
US PS 8.229 -255.085
Rating
6.00
US Other 7.980 -1295.727
5.00 SH RC 7.901 -316.403
4.00 SH PS 8.876 -66.876

IH R/F Conc. IH Other
US R/F Conc. US P/S. Conc. SH Other 7.980 -937.262
3.00 US Other State R/F Conc.
State P/S Conc. State Other
FM R/F Conc. FM P/S Conc. FM RC 7.991 -330.937
FM Timber FM Other
2.00
0 20 40 60 80 100 120 FM PS 8.998 -59.482
Time (years)
FM Timber 7.032 -363.467
FM Other 8.046 -718.254

87
Table 4.18. Exponential Best-Fit Graphs and Parameters (continued)
Superstructure Ratings
10.00 Material b1 b2
9.00 IH RC 8.177 -256.504

IH PS 8.269 -247.523
8.00
IH Steel 7.938 -184.626

7.00
US RC 8.155 -302.925
Rating
6.00
US PS 8.356 -204.955
5.00 US Steel 7.630 -371.819
4.00 SH RC 8.186 -332.936

IH R/F Conc. IH P/S Conc.
IH Steel
US P/S Conc.
US R/F Conc.
US Steel
SH PS 8.373 -217.760
3.00
State R/F Conc. State P/S Conc.
State Steel
FM P/S Conc.
FM R/F Conc
FM Steel
SH Steel 7.929 -291.797
2.00
0 20 40 60 80 100 120 FM RC 8.194 -352.926
Time (years)
FM PS 8.187 -410.558
FM Steel 7.957 -289.826
Table 4.19. Expert Opinion Condition Rating Prediction Equations
Deck Ratings Min = 7.560 - 0.145t

10.00
Mean = 7.758 - 0.115t
Max = 7.655 - 0.083t
9.00
8.00
7.00
Rating
6.00
5.00
4.00
3.00
Min Mean Max

2.00
0 20 40 60 80 100 120
Time (years)
Superstructure Ratings Reinforced Concrete:

10.00
RC Min RC Mean RC Max
Min = 7.642 - 0.136t
P/S Min P/S Mean P/S Max
Mean = 7.775 - 0.107t
9.00
Steel Min Steel Mean Steel Max Max = 7.698 - 0.076t
8.00
Prestressed Concrete:
Min = 7.737 - 0.138t
7.00
Mean = 7.707 - 0.099t
Max = 7.633 - 0.076t
Rating
6.00
5.00 Steel:
Min = 7.752 - 0.146t
4.00 Mean = 7.864 - 0.117t
Max = 7.803 - 0.089t
3.00
2.00
0 20 40 60 80 100 120
Time (years)

88
Table 4.19. Expert Opinion Condition Rating Prediction Equations (continued)
Substructure Ratings Reinforced Concrete:

10.00 Min = 7.654 - 0.144t
RC Min RC Mean RC Max
P/S Min P/S Mean P/S Max
Mean = 7.740 - 0.107t
9.00
Steel Min Steel Mean Steel Max Max = 7.701 - 0.081t
Timber Min Timber Mean Timber Max
8.00 Prestressed Concrete:
Min = 7.710 - 0.131t
7.00 Mean = 7.739 - 0.097t
Max = 7.723 - 0.073t
Rating
6.00
Steel:
5.00
Min = 7.881 - 0.177t
Mean = 7.866 - 0.138t
4.00
Max = 7.883 - 0.105t
3.00
Timber:
Min = 7.535 - 0.202t
2.00
0 20 40 60 80 100 120 Mean = 7.846 - 0.174t
Time (years) Max = 7.992 - 0.140t
4.8 Massachusetts DOT was defined as a bridge with an average condition rating of
Items 58, 59, and 60 greater than 5 but less than or equal to 6,
The Massachusetts DOT conducted a study of its bridges to but with none of the individual condition ratings being 4 or
gain a better understanding of the dynamics of how bridges lower. Bridges with a condition rating average greater than 6
age and deteriorate. This knowledge is intended to be used to were considered excellent.
plan strategies for bridge work and to determine required lev- Figure 4.13 shows the average age of a Massachusetts
els of funding. bridge in a given average condition rating. The average age for
The main aspects of this study were to determine the all bridges to reach an average condition rating of 1.0 is
following: greater than that for either steel or concrete and can be attrib-
uted to the age effect from old masonry bridges, the oldest of
• The makeup of the bridge population by age and material
which in Massachusetts is 250 years old.
of construction; By knowing the time it takes a bridge to deteriorate into the
• The average age of bridges for a given average condition
next lower average condition rating, the additional service
rating by material; life that could be obtained by increasing the average condi-
• The probability that a bridge in a given average condition
tion rating can be estimated for a given preservation strategy.
rating will transition to a structurally deficient condition A regression analysis could be used to develop equations
in the following year based on the age and current condi- relating age to condition rating, as done in the previously
tion of the bridge;
• The percentage of bridges in each age group that are in one
of the following categories: structurally deficient, fair, or
satisfactory; and 9
• Equations to predict the growth of bridges in the fair or 8
satisfactory condition categories. 7

Bridge Condition
6
Satisfactory
In undertaking this analysis, the Massachusetts DOT 5
Fair Concrete
defined the bridge condition categories as described here. A 4
Steel
structurally deficient bridge was defined as a bridge with any 3
All
one of the NBI Items 58, 59, or 60 (deck, superstructure, or 2 Structurally Deficient
substructure, respectively) condition ratings less than or 1
equal to 4. A fair bridge was defined as a bridge with an aver- 0
age condition rating of Items 58, 59, and 60 greater than 4 but 0 20 40 60 80 100
Bridge Age
less than or equal to 5, but with none of the individual condi-
tion ratings being 4 or lower. Similarly, a satisfactory bridge Figure 4.13. Massachusetts bridge conditions by age.

89
12.00%
% of Fair Bridges Becoming Structurally

Fair Fair-Conc Fair-Steel
10.00%
8.00%
Deficient
6.00%
4.00%
2.00%
0.00%
100
5
105+
25
30
35
90
95
10
15
20
40
45
50
55
60
65
70
75
80
85
Age Group
Figure 4.14. Probability of all bridges in fair condition becoming
structurally deficient in the following year.
presented studies. The same process could be used to develop transition to structurally deficient the following year (see
additional equations based on ADTT, location, or owner if Figure 4.14). Similarly, Figure 4.15 indicates that approxi-
the required data were available. The average condition rating mately 1.11% of bridges in satisfactory condition transition to
versus age for Massachusetts bridges is similar to those pre- structurally deficient the following year. Figure 4.14 and Fig-
sented above for other states. ure 4.15 also show the transition probabilities for steel bridges
However, the purpose of the Massachusetts DOT study was and concrete bridges; these probabilities could be used if the
not to develop equations to predict the condition rating as a analyst wished to look only at steel or concrete bridges. The
function of time (or age) as the studies presented previously, graphing of these transition probabilities indicates that, except
though if desired, equations could be developed using the for concrete bridges in the fair category, which show some age-
available data. Instead, using the data acquired for the third related influence, age is not as much of a factor as the current
and fifth bullet points above, the number of bridges that condition category of the bridge in determining the transition
become structurally deficient in any given year can be esti- probability. Similar transition probabilities could be devel-
mated from the number of bridges predicted to be in a given oped for different geographic regions or different levels of
condition (i.e., satisfactory or fair). ADT or ADTT.
The analysis results indicated that, for the bridge population To predict the growth of bridges in the fair and satisfactory
as a whole, approximately 4.25% of bridges in fair condition categories, best-fit equations were developed from a regression
5.00%
Satisfactory
% of Satisfactory Bridges Becoming
4.50%
4.00% Satisfactory-Conc
Structurally Deficient
3.50% Satisfactory-Steel
3.00%
2.50%
2.00%
1.50%
1.00%
0.50%
0.00%
100
5
105+
10
30
50
70
90
15
20
25
35
40
45
55
60
65
75
80
85
95
Age Group
Figure 4.15. Probability of all bridges in satisfactory condition
becoming structurally deficient in the following year.

90
Satisfactory Bridges Growth
1660
1640
1620
1600
Count
1580
1560
1540
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Actual 1588 1603 1615 1640 1627 1653 1635 1639
"Predicted" 1588 1605 1615 1622 1628 1632 1636 1639 1642 1645 1647 1649
Figure 4.16. Growth of bridges in satisfactory category.
analysis of the number of bridges that were in those two predicted by the best-fit exponential equation and the best-fit
categories for each year from 2002 through 2009. These equa- straight line equation for further analysis. After applying
tions were used to predict the number of fair and satisfac- the transition probability for fair bridges, the growth in
tory bridges in future years. These equations are graphed in structurally deficient bridges from this category is shown in
Figure 4.16 and Figure 4.17 and show the predicted numbers Figure 4.18 for each of the regression equations, as well as
compared with the actual numbers from 2002 to 2009. from the average.
Due to concerns that the regressed exponential equation A final needs analysis spreadsheet was developed that
for the growth of fair bridges was too aggressive, the Massa- combined the structurally deficient growth predictions and
chusetts DOT decided to use the number of fair bridges that the predictions of the number of projects that could be
would be obtained by averaging the number of fair bridges undertaken for a given amount of funding. The number of
Fair Bridge Growth

1200
1000
800
Count
600
400
200
0
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
"Actual" 496 521 535 567 642 704 716 757
"Predicted Exp" 479 511 546 583 622 664 709 757 808 863 921 983
"Predicted Lin" 476 516 556 597 637 677 717 758 798 838 879 919
Figure 4.17. Growth of bridges in fair category.

91
Growth in SD Bridges from Fair Bridges

100
90
80
70
60
Count
50
40
30
20
10
0
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Linear 57 59 61 63 64 66 68 70 71 73 75
Exponential 57 60 62 64 67 70 73 76 80 83 87
Average 57 59 61 64 66 68 70 73 75 78 81
Figure 4.18. Number of fair bridges becoming structurally deficient (SD).
structurally deficient bridges for a given year was estimated would be prevented from becoming structurally deficient in
by multiplying the transitional probability by the predicted the following year. However, the total number of bridges that
number of fair or satisfactory bridges for that year. had preservation work done were removed from the fair con-
The cost model was calibrated with actual project costs and dition rating population in calculating structurally deficient
considered the costs for a full replacement versus a preserva- bridges for subsequent years.
tion project. Replacement projects assumed the replacement Figure 4.19 and Figure 4.20, respectively, show the effect of
of an already structurally deficient bridge and hence resulted different funding levels on the number of bridges that will be
in a reduction in the number of structurally deficient bridges structurally deficient or in fair condition. The funding levels,
estimated for the following year. It was assumed that preser- shown in the far-left column of Figure 4.19 and Figure 4.20,
vation projects were to be performed on fair bridges and that are normalized to the lowest funding level shown in the top
only a percentage of bridges, based on the transitional prob- row. Figure 4.19 shows the overall bridge program funding
ability times the number of preservation projects undertaken, level and includes both replacement and preservation projects.
SD TRENDS BASED ON CAPITAL AND PRESERVATION SPENDING, 4% INFLATION
700
600
500
400
Count
300
200
100
0
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
$1.00 541 545 559 553 521 494 471 467 473 484 500 526 553 585 620 660
$1.59 541 545 559 553 521 494 463 445 434 424 414 412 407 404 402 402
$1.93 541 545 559 553 521 494 452 421 397 373 348 329 307 287 266 248
YEAR
Figure 4.19. Number of structurally deficient bridges based on various spending levels.

92
FAIR BRIDGE GROWTH TRENDS BY PRESERVATION SPENDING,

4% INFLATION
1200
1000
800
Count
600
400
200
0
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
$1.00 803 811 813 822 829 858 894 936 979 1028 1081 1138
$3.64 803 811 754 707 660 638 623 618 614 619 630 646
$4.48 803 811 735 670 606 567 536 515 497 488 484 487
YEAR
Figure 4.20. Number of fair bridges based on various spending levels.
The second and third rows of Figure 4.19 assume that 1.59 Element condition state for an existing bridge versus that of
and 1.93 times as much money is available for structurally a new bridge, as given in Equation 4.3 (Thompson and
deficient bridges. The 1.53 funding level will trend to a steady Shepard 2000). Current element value is calculated using
state number of structurally deficient bridges. The 1.93 fund- Equation 4.4, and total element value is calculated using
ing level will reduce the number of structurally deficient Equation 4.5. The CoRe elements and associated condition
bridges at a rate that will result in zero structurally deficient states can be found in the AASHTO Guide Manual for Bridge
bridges in 20 years. In Figure 4.20, the spending levels corre- Element Inspection (2011).
late to the preservation spending for each of the total funding
levels in Figure 4.19.
Health Index (HI) =
∑ CEV × 100 (4.3)
As expected, spending more money leads to fewer bridges ∑ TEV
that are structurally deficient or in fair condition and spend-
ing less money leads to more bridges that are either structur- where CEV is current element value and TEV is total element
ally deficient or in fair condition. This graph also indicates value.
the level of funding that would be needed to achieve a given
desired outcome. For example, to achieve a net annual reduc- CEV = (∑ [Quantity in Condition State i × WF (i)]) × FC
tion in the number of structurally deficient bridges for the (4.4)
long term, program funding above the 1.59 level is needed.
In addition to the preceding analysis, the Massachusetts where WF(i) is the condition state weighting factor given in
DOT also developed a utility to rank all the bridges in the Table 4.20 and FC is the failure cost of the element.
state to prioritize which bridges should be worked on first. TEV = Total Element Quantity × FC (4.5)
This ranking methodology is being used to develop the bridge
State Transportation Improvement Program lists. The rank-
ing is a function of three values: the condition loss value, the
Table 4.20. Condition State Weighting Factors
change in health index, and the highway evaluation factor.
Condition loss is simply the difference between a perfect con- Number of
dition rating (9.0) and the current average condition rating Condition
divided by nine and multiplied by 100 to achieve a percentage States State 1 State 2 State 3 State 4 State 5
value. Health index is the change in the bridge’s health index 3 1.00 0.50 0.00
that AASHTOWare Bridge Management (formerly Pontis)
4 1.00 0.67 0.33 0.00
predicts will occur over 15 years, expressed as a percentage
5 1.00 0.75 0.50 0.25 0.00
value. The health index is calculated using the current CoRe

93
The highway evaluation factor is a measure of the func- The ranking factor is then used to sort the bridges to
tionality of the bridge and considers the ADT, detour length, determine each bridge’s overall rank within the Massachu-
functional classification, load-carrying restrictions, and deck setts bridge population and, hence, its priority for work;
geometry deficiencies. The categories within each variable are bridges with the highest ranking factor values are those that
given a value between 1 and 5; the average value for the five require repair or maintenance in the future. The ranking is
variables is determined and then divided by five and multi- not a set order (Bridge 2 can go before Bridge 1) but, in gen-
plied by 100 to achieve a percentage value. The values for con- eral, higher-ranked bridges should be improved before
dition loss (CL), health index (HI), and highway evaluation lower-ranked bridges. The ranking factor, if calculated over
factor (HEF) are then combined using Equation 4.6 to deter- a number of years, may lead to a reasonable estimate of the
mine the final ranking factor for each bridge: amount of deterioration and possible loss of serviceability
Ranking Factor = 0.3CL + 0.4HI + 0.3HEF (4.6) for a bridge.

Chapter 5
Live Load for Calibration
5.1 Development of Live 13 million records) were also removed. The New York data
Load Models for were not considered because they included a considerable
Service Limit States number of extremely heavy vehicles. It was decided that these
data would have a strong effect on the statistical parameters,
5.1.1 Introduction which would cause the remaining states to be unnecessarily
The consideration of limit states, both ultimate (strength) and penalized. Indiana data could not be considered because the
serviceability, requires the knowledge of loads. The objective format was not compatible with the other states. The consid-
of this task is to determine the statistical parameters of live ered database included about 35 million vehicles.
load for the limit states considered in AASHTO LRFD (2012). The obtained WIM data include the following information
For strength limit states, the live load statistics were deter- for each location and each recorded vehicle: number of axles,
mined in NCHRP Project 12-33 and documented in NCHRP spacing between axles, axle loads, gross vehicle weight
Report 368 (Nowak 1999). The emphasis was placed on pre- (GVW), vehicle speed, and exact time of measurement. Sta-
diction of the extreme expected live load effects in the 75-year tistical parameters are determined for the GVW and moment
lifetime of a bridge. The database at that time was a truck sur- caused by the vehicles, including a cumulative distribution
vey carried out by the Ontario Ministry of Transportation in function (CDF); a bias factor (l) that is equal to the mean-to-
Canada. The basic statistical parameters of the maximum nominal ratio (i.e., the ratio of the mean value and the nomi-
75-year live load effect (moment and shear force) were deter- nal, or design, value); and the coefficient of variation (CV), V,
mined by extrapolating the truck survey data. It was assumed which is equal to the ratio of the standard deviation (s) to the
that the survey represented 2 weeks of heavy traffic. The pro- mean (µ).
cedure is described in NCHRP Report 368 (Nowak 1999). The CDFs for the WIM data for each site were plotted on
The serviceability limit states require additional statistical normal probability paper, which is described in Chapter 3,
parameters, not only the maximum values, but also load Section 3.2.1.
spectra (i.e., frequency of occurrence of loads). The maxi-
mum values are needed for shorter time periods, such as a 5.1.2 WIM Database
day, week, month, or year. At present, a considerable amount
of WIM (weigh in motion) truck data is available and the The truck survey includes WIM truck measurements from
research team had access to two sources: NCHRP Project 52 sites obtained from NCHRP Project 12-76 and FHWA.
12-76 data (Sivakumar et al. 2011) and Federal Highway The data obtained from FHWA, which are summarized
Administration (FHWA) files. This chapter provides docu- below, included trucks recorded from special pavement
mentation on the development of the statistical parameters studies (SPSs); each SPS is followed by a number that identi-
of live load for service limit states (SLSs) and fatigue. fies the study’s location (e.g., SPS-1 is Special Pavement Study,
The analysis includes consideration of the WIM database Location 1):
from NCHRP Project 12-76 and FHWA. The obtained data
included over 65 million vehicles. Of that number, about • Arizona (SPS-1)—Data recorded continuously from Janu-
10 million were deleted or filtered because of obvious errors, ary 2008 until December 2008;
leaving about 55 million. Data from New York (about 7.8 mil- • Arizona (SPS-2)—Data recorded continuously from Janu-
lion records) and Indiana other than site SPS-6 (about ary 2008 until December 2008;
94

95
• Arkansas (SPS-2)—Data recorded continuously from Florida

January 2008 until December 2008;
• US-29 (Site 9916)—Data recorded continuously from
• Colorado (SPS-2)—Data recorded continuously from
January 2005 until December 2005 (11 days missing);
• I-95 (Site 9919)—Data recorded continuously from Janu-
• Delaware (SPS-1)—Data recorded continuously from
ary 2005 until December 2005 (16 days missing);
• I-75 (Site 9926)—Data recorded almost continuously from
• Illinois (SPS-6)—Data recorded continuously from Janu-
January 2005 until December 2005 (100 days missing);
ary 2008 until December 2008;
• I-10 (Site 9936)—Data recorded almost continuously from
• Indiana (SPS-6)—Data recorded continuously from July
January 2005 until December 2005 (100 days missing); and
2008 until December 2008; • State Route (Site 9927)—Data recorded almost continuously
• Kansas (SPS-2)—Data recorded continuously from Janu- from January 2004 until December 2004 (5 days missing).
ary 2008 until December 2008;
• Louisiana (SPS-1)—Data recorded continuously from Indiana
• Maine (SPS-5)—Data recorded continuously from Janu- • Site 9511—Data recorded continuously from January 2006
ary 2008 until December 2008; until December 2006;
• Maryland (SPS-5)—Data recorded continuously from • Site 9512—Data recorded continuously from January 2006
January 2008 until December 2008; until December 2006;
• Minnesota (SPS-5)—Data recorded continuously from • Site 9532—Data recorded continuously from January 2006
January 2008 until December 2008; until December 2006;
• New Mexico (SPS-1)—Data recorded continuously from • Site 9534—Data recorded continuously from January 2006
May 2008 until December 2008; until December 2006; and
• New Mexico (SPS-5)—Data recorded continuously from • Site 9552—Data recorded continuously from January 2006
May 2008 until December 2008; until December 2006.
• Pennsylvania (SPS-6)—Data recorded continuously from
January 2008 until December 2008; Mississippi
• Tennessee (SPS-6)—Data recorded continuously from • I-10 (Site 3015)—Data recorded almost continuously from
January 2008 until December 2008; January 2006 until December 2006 (28 days missing);
• Virginia (SPS-1)—Data recorded continuously from Janu- • I-55 (Site 2606)—Data recorded almost continuously from
ary 2008 until December 2008; and January 2006 until December 2006 (16 days missing);
• Wisconsin (SPS-1)—Data recorded continuously from • I-55 (Site 4506)—Data recorded almost continuously
January 2008 until December 2008. from March 2006 until December 2006 (39 days missing);
• US-49 (Site 6104)—Data recorded almost continuously from
Data obtained from NCHRP projects are also summarized January 2006 until December 2006 (5 days missing); and
here, and include trucks recorded from • US-61 (Site 7900)—Data recorded almost continuously
from January 2006 until December 2006 (49 days missing).
California
• Lodi (Site 003)—Data recorded continuously from June
New York
2006 until March 2007; • I-95 Northbound (Site 0199)—Data recorded continu-
• Antelope Eastbound (Site 003)—Data recorded almost ously from March 2006 until December 2006;
continuously from April 2006 until March 2007 (107 days • I-95 Southbound (Site 0199)—Data recorded continu-
missing); ously from July 2006 until November 2006;
• Antelope Westbound (Site 003)—Data recorded almost • I-495 Westbound (Site 0580)—Data recorded continu-
continuously from April 2006 until March 2007 (109 days ously from January 2006 until December 2006;
missing); • I-495 Eastbound (Site 0580)—Data recorded continuously
• LA 710 Southbound (Site 059)—Data recorded continu- from January 2006 until December 2006;
ously from April 2006 until March 2007; • Highway 12 (Site 2680)—Data recorded continuously
• LA 710 Northbound (Site 060)—Data recorded almost from January 2005 until December 2005;
continuously from April 2006 until March 2007 (32 days • I-84 Eastbound and Westbound (Site 8280)—Data recorded
missing); and continuously from January 2006 until December 2006;
• Bowman (Site 072)—Data recorded almost continuously • I-84 Eastbound and Westbound (Site 8382)—Data recorded
from April 2006 until February 2007 (139 days missing). continuously from January 2005 until December 2005;

96
• I-81 Northbound and Southbound (Site 9121)—Data • Record in which the length of the truck varied from the
recorded continuously from January 2005 until December sum of the axle spacings by more than 1 ft, based on
2005; and NCHRP 12-76;
• Highway 17 Eastbound and Westbound (Site 9631)—Data • Record that had a GVW less than a threshold; at various
recorded continuously from February 2006 until Decem- times the threshold was 10 or 12 kips;
ber 2006. • Record in which the steering axle was less than 6 kips,
based on NCHRP 12-76;
5.1.3 WIM Data Filtering • Record in which the sum of the axle spacing lengths was
less than 7 ft, based on Pelphrey et al. (2008);
The WIM data both from NCHRP Project 12-76 and FHWA • Class of the vehicle according to FHWA, from Class 3 to 14,
include vehicle records that appear to be incorrect. There are to filter out cars, motorcycles, and so on; and
various reasons for questioning the data (e.g., GVW is too • Speed ranges from 10 to 100 mph, based on NCHRP 12-76.
low, unrealistic geometry). The data were filtered to eliminate
questionable vehicles by using the following criteria: The filtering process is illustrated in the flowchart in Fig-
ure 5.1. Because a heavy vehicle meeting all the conditional
• Weight per axle less than 2 kips or greater than 70 kips, filters involving GVW would pass the filters, the research
based on NCHRP 12-76; team reviewed exceptionally heavy vehicles to check if their
• Record in which the first axle spacing was less than 5 ft, configuration resembled permit vehicles, such as cranes and
based on NCHRP 12-76; garbage trucks. The data were divided into two sets. The first set
• Record in which any axle spacing was less than 3.4 ft, based contained regular truck traffic. These data were used for the live
on NCHRP 12-76; load model for SLSs. The remaining set of data included permit
• Record in which GVW varied from the sum of the axle vehicles and illegally overloaded vehicles, which occurred rela-
weights by more than 10%, based on NCHRP 12-76; tively infrequently. The latter data were used along with the
Figure 5.1. Flowchart of the filtering process.

97
Standard Normal Variable
GVW [kips]
Figure 5.2. CDF of GVW FHWA and Ontario data.
regular truck traffic for live load for SLS II. The GVW criteria of Figure 5.2 represents the CDF of the GVW of trucks from
20 kips in Step 3 is a traditional, albeit arbitrary, cutoff used in FHWA sites plotted on probability paper. Data collected from
virtually all previous fatigue studies to reduce the calculation 14 sites represent 1 year of traffic, data from the Indiana site
effort by not considering light traffic, which does not contribute represent 6 months of traffic, and data from the New Mexico
significantly to cumulative damage. sites represent 8 months of traffic. The maximum truck GVW
The CDFs of GVWs were plotted on probability paper; was 220 kips. Mean values ranged from 20 to 65 kips.
examples are shown in Figure 5.2 to Figure 5.5. The live load Figure 5.3 to Figure 5.5 represent CDFs of the GVWs for
model based on the Ontario truck survey data that were used Ontario and the following states: Oregon and Florida (Fig-
in calibration for strength limit states is also shown. The rela- ure 5.3), Indiana and Mississippi (Figure 5.4), and California
tive position of the Ontario curve is a result of the intentional and New York (Figure 5.5) (i.e., the NCHRP 12-76 data). The
selection of seemingly heavy vehicles, albeit based solely on corresponding traffic data from these figures are given in
the appearance of the vehicles. Table 5.1.
NCHRP Data - Oregon NCHRP Data - Florida

GVW [kips] GVW [kips]
Figure 5.3. CDFs of GVW for Oregon, Florida, and Ontario.

98
NCHRP Data - Indiana NCHRP Data - Mississippi

Figure 5.4. CDFs of GVW for Indiana, Mississippi, and Ontario.
NCHRP Data - California NCHRP Data - New York

Figure 5.5. CDFs of GVW for California, New York, and Ontario.
Table 5.1. Summary of State Sites and Their Traffic Data

for Figures 5.3 to 5.5
No. of No. of Months Maximum Mean Value

Figure State Sites of Data GVW (kips) Range (kips)
Figure 5.3 Oregon 4 4 200 43–52

Florida 5 12 250 20–50
Figure 5.4 Indiana 5 12 250 25–57
Mississippi 5 12 260 38–57
Figure 5.5 California 2 8.7 250 40–50
1 7
New York 7 12 380 35–50

99
As an initial observation, the data shown in Figure 5.2 to Table 5.2. WIM Locations and Number
Figure 5.5 are generally consistent for the majority of the sites of Recorded Vehicles
(consistent refers to the similarity of the general shape of the
curves, i.e., the CDFs). Exceptions are the following heavily No. of Days Total No. of Lane
Site in Data Truck Records ADTT
loaded sites from New York:
Arizona (SPS-1) 365 35,572 97
• Site 9121 on I-81 by Whitney Point; Arizona (SPS-2) 365 1,430,461 3,919
• Site 8382 on I-84 by Port Jervis; Arkansas (SPS-2) 365 1,675,349 4,590
• Site 8280 on I-84 by Fishkill; and
Colorado (SPS-2) 365 343,603 941
• Site 0580 on I-495 in Queens in New York City.
Delaware (SPS-1) 365 201,677 553
Because these sites were so exceptional, it was decided not Illinois (SPS-6) 365 854,075 2,340
to include the New York WIM data in developing a national, Indiana (SPS-6) 214 185,267 508
notional SLS live load. In addition, several sites for which the
Kansas (SPS-2) 365 477,922 1,309
recording format differed or had considerably less than one
tier of data were eliminated from consideration. A summary Louisiana (SPS-1) 365 85,702 235
of the remaining 32 sites and filtered data, including the WIM Maine (SPS-5) 365 183,576 503
locations, number of records, and average daily truck traffic Maryland (SPS-5) 365 164,389 450
(ADTT), is shown in Table 5.2. Approximately 35 million
Minnesota (SPS-5) 365 55,572 152
records are represented by these sites.
A copy of the raw WIM data and of the filtered WIM data New Mexico (SPS-1) 245 117,102 321
is available at http://www.trb.org/Main/Blurbs/170201.aspx. New Mexico (SPS-5) 245 608,280 1,667
A sample of the filtered WIM data is included in Appendix F. Pennsylvania (SPS-6) 365 1,495,741 4,098
The CDFs of GVWs and moment are plotted as separate
Tennessee (SPS-6) 365 1,622,320 4,445
curves for each location. The legend for all CDFs is shown in
Virginia (SPS-1) 365 259,190 710
Figure 5.6.
Wisconsin (SPS-1) 365 226,943 622
California Antelope EB 258 837,667 2,192a
5.2 Initial Data Analysis
California Antelope WB 256 943,147 2,258a
5.2.1 Gross Vehicle Weight
California Bowman 134 651,090 2,018a
The CDFs for the GVWs from the remaining FHWA and California LA-710 NB 333 4,092,484 6,380a
NCHRP sites are plotted on probability paper in Figure 5.7.
California LA-710 SB 365 4,661,287 8,366a
Each of the 32 curves represents a different location. The result-
ing curves indicate that the distribution of GVW is not normal. California Lodi 304 3,298,499 5,186a
Irregularity of the CDFs is a result of different types of vehicles Florida I-10 354 1,641,480 2,207a
(such as long and short, fully loaded and empty, or loaded by Florida I-95 349 2,112,518 2,558a
volume only) in the WIM data. For the considered locations, Florida US-29 354 389,164 606a
the mean GVWs are between 25 and 65 kips. The upper tails of
Mississippi I-10 337 1,965,022 2,967a
the CDF curves show a similar trend, but there is a considerable
spread of the maximum values, from 150 to over 250 kips. Mississippi I-55UI 268 1,232,223 2,054a
Mississippi I-55R 349 1,333,268 1,790a
5.2.2 Moments from WIM Data Mississippi US-49 359 1,225,138 1,475a
Mississippi US-61 319 159,299 254a
The distribution of simple-span moments due to WIM trucks
was obtained by calculating the maximum bending moment Total 35,856,898
for each vehicle in the database. Each vehicle was run over Note: EB = eastbound; WB = westbound; NB = northbound; SB = southbound.
a NCHRP data are for multilane cases; the lane with maximum ADTT is listed.
influence lines to determine the maximum moment by using
a specially developed computer program. The calculations
were carried out for spans from 30 to 200 ft. For easier inter-
pretation and comparison of results, the calculated WIM data
moments were then divided by the corresponding HL-93
moment. Normalizing the data to a common reference makes

100
FHWA Data NCHRP Data category that fits the detected configuration. Heavy vehicles are
assumed to be permit vehicles or illegally loaded vehicles.
The initial study indicated that the removal of a very small
number of the heaviest vehicles drastically changed the upper
tail of the CDF of moments and shears. It was decided to explore
this by investigating the number of vehicles that exceeded an
upper value of 1.35 times HL-93, which corresponds to the max-
imum bias ratio obtained from the Ontario measurements.
The results of the analysis for sites from New York and Mis-
sissippi were plotted on probability paper and are shown in
Figures 5.14 to 5.16. It can be observed that, as expected, the
very upper tail of the distribution changed drastically by
removing only a very small percentage of vehicles.
For example, for 90-ft spans at New York Site 8382 (Fig-
ure 5.15), the bias changes from about 2.35 to about 1.65—but
only when considering the six largest moment ratios (corre-
sponding to the six heaviest trucks, including the 391-kip vehi-
Figure 5.6. Legend for all graphs. cle shown in Figure 5.13) out of the 1.55 million data records
remaining after application of the additional filter to remove
the data easier to interpret. HL-93 was a convenient reference moments less than 15% of the corresponding HL-93 moment.
and ties this work to the original strength limit state calibra- Even for the WIM sites that demonstrated very extreme tails,
tion and associated published information. these extreme trucks constituted only the upper 0.01% to
The CDFs for the ratio of the WIM truck moment and 0.22% of the truck population. For most of the locations
HL-93 moment are plotted on normal probability paper in reviewed, the percentage was lower (see Table 5.3). The heavi-
est loads may have an important impact on calibration of the
Figure 5.8 to Figure 5.12; the shape of the CDF curves is simi-
ultimate or strength limit states; however, in the case of SLSs,
lar to that of GVW. The mean WIM moments were between
the upper tail of the CDF of the live load is not important, as it
0.2 and 0.4 of the HL-93 moments for all span lengths consid-
is the main body of the CDF that affects SLS performance.
ered. The probability of a WIM moment exceeding 0.4 to 0.5
Therefore, for SLS calibration, it was decided to ignore the
of the HL-93 moment was about 0.15. The maximum values
upper tip of the CDF of live load.
of the WIM moment were between 1.0 and 1.4 of HL-93
moment in most cases.
The obtained results served as the basis for determining 5.2.4 Multiple Presence Analysis
the statistical parameters of live load needed for the reliability
Multiple presence was investigated by a correlation analysis
analysis of the serviceability limit states.
of the WIM data sets. The objective of the correlation analysis
was to select two trucks that were simultaneously positioned
5.2.3 Filtering of Presumed Illegal Overloads on the bridge as shown in Figure 5.17 and that satisfied the
and Special Permit Loads following requirements:
The goal of this analysis was to observe the change in the very • Both trucks had the same number of axles.
top tail of the distribution after removing the heaviest vehicles • GVWs of the trucks were within ±5%.
from the database. These extremely heavy vehicles seemed to be • All corresponding spacings between axles were within ±10%.
either permit vehicles that should be included in the design pro-
cess (as some states do) or vehicles reviewed for permit issuance The maximum load effect is often caused by the simultane-
by using the Strength II limit state load combination; otherwise, ous presence of two or more trucks on a bridge. The statistical
they are illegal overloads. An example of the heaviest truck in the parameters of these effects are influenced by the degree of
WIM data is presented in Figure 5.13. This truck was recorded correlation. In calibration for the strength limit states, certain
at Site 8382 near Port Jervis, New York. The total length of the probabilities of occurrence of correlated trucks were assumed
truck was 100.6 ft. The GVW was 391.4 kips. The position of the on the basis of engineering judgment applied to limited obser-
12 axles, their weight, and the vehicle’s length suggest that it vations of the presence of multiple trucks of unknown weight.
should be categorized as a permit vehicle. WIM equipment cap- The available WIM data allowed for verification of these
tures each vehicle, including permit vehicles, as a string of axles, assumptions.
and an FHWA designation is given based on the best FHWA (text continues on page 108)

101
Figure 5.7. CDFs of GVWs.

102
Figure 5.8. CDFs of WIM moment and HL-93 moment ratio, span 5 30 ft.

103

104

105

106
Figure 5.12. CDFs of WIM moment and HL-93 moment ratio,

span 5 200 ft.

107
Figure 5.13. Configuration of extremely loaded truck.
New York 0580 Span 90 ft

Truck Moment / HL93 Moment Truck Moment / HL93 Moment
Figure 5.14. Data removal from New York Sites 0580 and 2680.

Figure 5.15. Data removal from New York Sites 8280 and 8382.

108
New York 9121 Span 90 ft Mississippi - I10 Span 90 ft

Figure 5.16. Data removal from New York Site 9121 and Mississippi I-10 locations.
(continued from page 100) The selected trucks were plotted on probability paper and
A special program was developed to filter the data by using compared with all recorded vehicles. The GVW of both cor-
the time of a record and the speed of the truck to find related trucks were added together and divided by two to
instances when either of the events shown in Figure 5.17 obtain the average GVW. (Note that the correlation criteria
occurred involving similar trucks. The filter resulted in select- ensure that the average is similar to the two selected trucks
ing the observed cases of two trucks with a headway distance in each pair.) The comparison of the mean correlated GVW
less than 200 ft in either the same lane or two adjacent lanes. of the trucks recorded in adjacent lanes with the GVW of the
whole population from Florida and New York is shown in
Figure 5.19.
Two Trucks: Side by Side
The analysis of the degree of correlation was performed for Two Trucks: One After the Other
Site 9936 in Florida along I-10 and Site 8382 in New York Filtering the data resulted in the selection of 8,380 fully cor-
with 1,654,004 and 1,594,674 site-specific total records, related trucks in one lane in Florida and 9,868 fully correlated
respectively. Filtering the data resulted in the selection of trucks in one lane in New York. Histograms of these trucks
2,518 fully correlated trucks in adjacent lanes in Florida and are shown in Figure 5.20. The comparison of the mean cor-
3,748 fully correlated trucks in adjacent lanes in New York. related GVW of the trucks recorded in one lane with the
Histograms of the GVWs of these fully correlated side-by- GVW of the whole data set from Florida and New York is
side trucks are shown in Figure 5.18. shown in Figure 5.21.
Table 5.3. Removal of Heaviest Vehicles (90-ft Span)
No. of Trucks No. of Trucks No. of Removed

Figure State Site Before Filtering After Filtering Removed Trucks Trucks (%)
Figure 5.14 New York 0580 2,474,407 2,468,952 5,455 0.22

Figure 5.14 New York 2680 89,286 89,250 36 0.04
Figure 5.15 New York 8280 1,717,972 1,717,428 544 0.03
Figure 5.15 New York 8382 1,551,454 1,550,914 540 0.03
Figure 5.16 New York 9121 1,235,963 1,235,886 77 0.01
Figure 5.16 Mississippi I-10 2,103,302 2,103,300 2 0.00

109
T1 T1
Headway Distance max 200 ft Headway Distance max 200 ft
T2 T2
Figure 5.17. Two cases of the simultaneous presence of

two trucks with headway distance less than 200 ft.
Florida I-10 New York Site 8382
(a) (b)
Figure 5.18. Histograms of trucks side by side (a) on Florida I-10 and (b) at New York
Site 8382.
(a) (b)
Figure 5.19. Comparison of mean GVW and GVW of the whole population for
(a) Florida and (b) New York.

110
(a) (b)
Figure 5.20. Histogram of trucks one after another (a) on Florida I-10 and
(b) at New York Site 8382.
Implications for Specification Development correlated to the design loading of two lanes of HL-93 with a
The study of multiple presence based on WIM data indicated load factor of 1.75 and a multiple presence factor of 1.0. (The
that, for SLSs, the vehicles representing the extreme tails of the multiple presence factor for a single-lane loading is 1.20 to
CDF need not be considered as being simultaneously present account for the occasional truck that creates more force effect
in multiple lanes. The implication is that only a single-lane than the family of configurations used to develop the HL-93
live load model needs to be considered on the load side (Q) load configuration.)
of limit state functions. The resistance side (R) of limit state
functions should represent the requirements of the applica-
5.2.5 Project Guidelines Regarding Live load
ble design requirement, even if that is a multiple-lane loading
situation. The following guidelines are based on live load bias factors
The issue of multiple load lanes was considered in the and CVs determined from the preliminary analysis of WIM
development of HL-93 for AASHTO LRFD strength limit measurements and previous work by the research team
states, and the conclusion was that extreme truck load does (Nowak 1999):
not occur simultaneously with another fully correlated extreme
truck, but was considered to occur simultaneously with a • The use of dynamic load as 10% of live load, with CV = 80%,
truck about 15% to 20% lighter. This two-lane loading was is recommended.
(a) (b)
Figure 5.21. Comparison of mean GVW and GVW of the whole population for
(a) Florida and (b) New York.

111
• Generally use a single loaded lane (no multiple loaded In further analysis, five ADTTs were considered: 250, 1,000,
lanes). 2,500, 5,000, and 10,000. The calculations were performed
• The national load (i.e., notional load) should not try to separately for each ADTT. To determine the mean maximum
encompass all WIM records. Some of the extremely heavy moments corresponding to the considered time periods, the
vehicles are permit loads and some are illegal overloads. A vertical coordinates were found first.
relatively small number of loads were excluded for most of Starting with ADTT = 250, the vertical coordinate of the
the SLS studies, but they were included for the overload mean maximum 1-day moment z is given by Equation 5.2:
limit state.
• It is likely that different probabilities of exceedance will be z = − Φ −1 (1 250) = 2.65 (5.2)
used for various limit states based on consequences.
• Some jurisdictions may need exceptions based on their because the number of trucks per 1 day is 250.
legal loads and extent of enforcement. The mean maximum 2-week moment z is given by Equa-
• The basic HL-93 load model, scaled by calibrated load fac- tion 5.3:
tors, is appropriate for SLS.
z = − Φ −1 (1 3500) = 3.44 (5.3)
With these recommendations, the evaluation of numerical
live load models continued. The processes used and results because the number of trucks per 2 weeks is (250 trucks)
obtained are summarized here. Further details and extensive (14 days) = 3,500 trucks.
graphical presentations are contained in Rakoczy (2011). Finally, the mean maximum 100-year moment z is given by
Equation 5.4:
5.3 Statistical Parameters z = − Φ −1 (1 9,125,000) = 5.18 (5.4)
for Service Limit States
Other than Fatigue because the number of trucks per 100 years is (250 trucks)
5.3.1 Maximum Moments for Different (365 days)(100 years) = 9,125,000 trucks.
Time Periods Similarly, for ADTT = 1,000, the vertical coordinate of the
mean maximum 1-day moment z is given by Equation 5.5:
The maximum moment is a random variable. It depends on
the period of time, ADTT, and distribution of traffic (e.g., CDF z = − Φ −1 (1 1000) = 3.09 (5.5)
of WIM moments). For a given CDF of WIM moments [F(x)],
period of time (T), and ADTT, the mean value of the maxi- because the number of trucks per 1 day is 1,000.
mum moment can be determined as follows. The total number The mean maximum 2-week moment z is given by Equa-
of vehicles (N) expected during the considered time period T tion 5.6:
(in days) is T × ADTT. The expected or mean value of the max-
imum moment for time T [Mmax(T)] is equal to the moment z = − Φ −1 (1 14,000) = 3.8 (5.6)
corresponding to probability {1 - F[1/N(T)]}, where F(x) is
the CDF of WIM moments, which is F-1[1 - 1/N(T)], where because the number of trucks per 2 weeks is (1,000 trucks)
F-1 is the inverse of CDF. (14 days) = 14,000 trucks.
The objective is to determine the mean maximum moment Finally, the mean maximum 100-year moment z is given by
for different time periods (i.e., 1 day, 2 weeks, 1 month, 2 months, Equation 5.7:
6 months, 1 year, 5 years, 50 years, 75 years, and 100 years). The
number of recorded vehicles for each location is given in z = − Φ −1 (1 36,500,000) = 5.67 (5.7)
Table 5.2. The data were collected over different time periods,
in most cases about 1 year, but the number of vehicles varies because the number of trucks per 100 years is (1,000 trucks)
because ADTT varies. Each CDF in Figure 5.8 to Figure 5.12 (365 days)(100 years) = 36,500,000 trucks.
includes the number of data points equal to the corresponding Values of z for the considered ADTTs and time periods
number of vehicles (N). For each CDF, the vertical coordinate from 1 day to 100 years are summarized in Table 5.4.
of the maximum moment (Zmax) is given by Equation 5.1: For example, for the WIM moments in Figure 5.11 (span =
120 ft), the vertical coordinates corresponding to different
z max = − Φ −1 (1 N ) (5.1) time periods are shown in Figure 5.22 for ADTT = 1,000.
There were 32 WIM locations and, therefore, 32 curves rep-
where -F-1 is the inverse standard normal distribution func- resenting CDFs of WIM moments in each of Figures 5.8 to
tion. For example, if N = 1,000,000, then Zmax = 4.75. 5.12. The mean maximum moment can be obtained directly

112
Table 5.4. Vertical Coordinates for the Mean randomly; rather, the selection was based on the availability
Maximum Moment of WIM stations with truck data and the credibility of the
measured data (truck records). If the considered WIM loca-
ADTT
tions are biased (i.e., nonrepresentative), then the processed
Time Period 250 1,000 2,500 5,000 10,000 database can underestimate or overestimate the statistical
parameters of the national live load. Therefore, for the pur-
1 Day 2.65 3.09 3.35 3.54 3.72
pose of further reliability analysis, it is conservatively assumed
2 Weeks 3.44 3.08 4.02 4.18 4.33
that the calculated mean maximum live load is increased by
1 Month 3.65 4.00 4.20 4.35 4.50 1.5 standard deviations. The probability of exceeding this
2 Months 3.82 4.15 4.35 4.50 4.65 value (mean plus 1.5 standard deviations) is about 5%, so that
6 Months 4.09 4.39 4.59 4.73 4.87
it will be exceeded by 5% of 32 WIM locations (i.e., in one or
two WIM locations).
1 Year 4.24 4.55 4.73 4.87 5.01
As the upper parts of the CDFs are almost straight lines, the
5 Years 4.59 4.87 5.05 5.18 5.31 fitting by normal distributions is justified. The mean values
50 Years 5.05 5.31 5.47 5.60 5.72 can be read directly from the graph as the intersection of CDFs
75 Years 5.13 5.38 5.55 5.67 5.78 (represented by straight lines) and the horizontal axis at zero
100 Years 5.18 5.44 5.60 5.72 5.83
on the vertical scale. This process is depicted in Figure 5.24.
The visual comparison of how the actual CDF fits a straight
line is much better than any curve-fitting formula because the
from the graph by reading the moment ratio (horizontal axis) research team was mostly interested in only some parts of the
corresponding to the vertical coordinate representing the con- CDF. Different curves can have different slopes, which are
sidered time period. For example, from Figure 5.22, the mean reflected in the standard deviations.
maximum 1-day moment ratio for Florida US-29 is 0.95, and Calculations were carried out for all considered cases of
the mean maximum 1-year moment ratio is 1.39. Values for lon- ADTT and span length. The results, which were extrapolated
ger time periods were projected or interpolated as appropriate. to 100 years and span length of 300 ft, are summarized in
For each ADTT and span length, there are 32 values of the Table 5.5 to Table 5.9. Statistical parameters were calculated
mean maximum 1-day moment, 32 values of the mean maxi- for a variety of ADTTs (500, 1,000, 2,500, 5,000, and 10,000);
mum 2-week moment, and so on. For an easier review and however, the AASHTO LRFD is based on 5,000 (consistent
comparison, CDFs of these 32 values obtained from Fig- with strength limit states). Live load data for values of ADTT
ure 5.22 were plotted on normal probability paper and are other than 5,000 were tabulated so owners can repeat the cali-
shown in Figure 5.23. There is one CDF for 1-day values, one bration process with other data. For a given bridge, use of a
for 2 weeks, and so on. These are CDFs of extreme variables, as lower ADTT should lead to a higher reliability index.
each of the 32 values is the maximum moment for a WIM loca- Bias factors vary depending on ADTT for shorter time
tion. The obtained CDFs are almost parallel; in particular, this periods; however, for longer time periods, the bias factor is
applies to the upper part. Because of regularity, it is easier to about 1.4.
determine the statistical parameters. Each data point repre-
sents the mean of the maximum value for one of 32 WIM loca-
5.3.3 Reactions
tions, which means that the CDFs in Figure 5.23 are extreme
value distributions rather than hypothetical curves. Tables of statistics for reactions of simply supported spans
were developed for the same spans, time periods, and ADTTs
as presented for bending moments by using a methodology
5.3.2 Statistical Parameters of Live Load
analogous to the one presented in Section 5.3.2. The results
It was assumed that the 32 WIM locations considered are rep- are shown in Table 5.10 to Table 5.14. Graphical representa-
resentative for the truck traffic in the United States. The statisti- tions are presented in Rakoczy (2011).
cal parameters (the mean maximum and CV of the maximum
live load) were determined for each WIM location. The CDFs
5.3.4 Axle Loads
of the mean maximum values were plotted on probability
paper. This is an extreme value distribution. The mean of these Statistical parameters for various time periods and ADTTs
mean maximum values can be considered as the mean maxi- are developed using a methodology analogous to that pre-
mum national live load. The standard deviation of the mean sented in Section 5.3.2 applied to axle loads instead of
maximum values can be determined from the graphs (slope of moments. The results are presented in Table 5.15.
the CDF). However, the WIM locations were not selected (text continues on page 121)

113
Figure 5.22. Vertical coordinates for different time periods for

ADTT 5 1,000 and span 5 120 ft.

114
Figure 5.23. CDFs of mean maximum moment ratios for ADTT 5 1,000 and span
length 5 120 ft.

115
Figure 5.24. Determination of mean values at 1.5 s.

116
Table 5.5. Statistical Parameters of Live Load Moments for ADTT 250, l 5   1.5s
Span
30 ft 60 ft 90 ft 120 ft 200 ft 300 ft

Time
Period l µ CV l µ CV l µ CV l µ CV l µ CV l µ CV
1 Day 0.92 0.65 0.28 0.82 0.64 0.23 0.80 0.66 0.17 0.79 0.65 0.15 0.71 0.56 0.18 0.61 0.48 0.18
2 Weeks 1.06 0.80 0.21 1.05 0.80 0.16 1.01 0.80 0.18 1.02 0.80 0.16 0.93 0.73 0.16 0.84 0.67 0.16
1 Month 1.12 0.85 0.21 1.09 0.85 0.19 1.08 0.85 0.18 1.08 0.85 0.17 1.01 0.78 0.19 0.90 0.73 0.16
2 Months 1.14 0.90 0.18 1.15 0.91 0.17 1.14 0.90 0.18 1.14 0.90 0.17 1.05 0.85 0.15 0.95 0.77 0.15
6 Months 1.19 0.95 0.17 1.23 0.96 0.19 1.20 0.97 0.15 1.19 0.98 0.14 1.12 0.91 0.15 1.04 0.85 0.15
1 Year 1.23 1.00 0.15 1.27 0.98 0.19 1.24 1.00 0.16 1.22 1.04 0.12 1.15 0.94 0.15 1.08 0.88 0.15
5 Years 1.31 1.07 0.15 1.35 1.09 0.16 1.31 1.13 0.11 1.31 1.14 0.10 1.25 1.02 0.15 1.18 0.97 0.15
50 Years 1.37 1.17 0.11 1.39 1.16 0.13 1.39 1.25 0.07 1.37 1.19 0.10 1.32 1.06 0.16 1.25 1.02 0.15
75 Years 1.38 1.20 0.10 1.40 1.19 0.12 1.41 1.27 0.07 1.39 1.21 0.10 1.34 1.08 0.16 1.27 1.04 0.15
100 Years 1.39 1.22 0.09 1.43 1.21 0.12 1.42 1.28 0.07 1.41 1.22 0.10 1.35 1.09 0.16 1.29 1.05 0.15
Table 5.6. Statistical Parameters of Live Load Moments for ADTT 1,000, l 5   1.5s
Span
30 ft 60 ft 90 ft 120 ft 200 ft 300 ft

Time
1 Day 0.99 0.72 0.28 0.89 0.71 0.20 0.90 0.72 0.17 0.89 0.71 0.17 0.81 0.63 0.19 0.71 0.55 0.19
2 Weeks 1.14 0.87 0.21 1.13 0.90 0.16 1.13 0.89 0.18 1.14 0.91 0.16 1.06 0.85 0.16 0.97 0.77 0.16
1 Month 1.18 0.95 0.16 1.19 0.95 0.16 1.19 0.95 0.17 1.19 0.96 0.16 1.11 0.91 0.14 1.01 0.83 0.14
2 Months 1.23 0.99 0.16 1.26 0.99 0.18 1.26 1.00 0.17 1.23 1.03 0.13 1.16 0.96 0.14 1.07 0.89 0.14
6 Months 1.27 1.04 0.14 1.31 1.05 0.16 1.30 1.10 0.12 1.27 1.09 0.11 1.22 0.99 0.15 1.15 0.93 0.15
1 Year 1.33 1.07 0.16 1.34 1.08 0.16 1.32 1.15 0.10 1.31 1.14 0.10 1.25 1.01 0.16 1.18 0.95 0.16
5 Years 1.37 1.11 0.15 1.37 1.14 0.13 1.36 1.21 0.08 1.35 1.17 0.10 1.30 1.06 0.15 1.24 1.01 0.15
50 Years 1.38 1.24 0.07 1.42 1.21 0.12 1.41 1.26 0.08 1.41 1.21 0.11 1.35 1.11 0.14 1.28 1.05 0.14
75 Years 1.40 1.26 0.07 1.42 1.23 0.11 1.42 1.28 0.07 1.41 1.23 0.10 1.36 1.13 0.13 1.29 1.07 0.13
100 Years 1.40 1.27 0.07 1.44 1.24 0.11 1.43 1.29 0.07 1.43 1.24 0.10 1.37 1.14 0.13 1.30 1.09 0.13

117
Span
30 ft 60 ft 90 ft 120 ft 200 ft 300 ft

Time
1 Day 1.03 0.80 0.19 0.97 0.79 0.18 0.97 0.77 0.17 0.98 0.78 0.17 0.90 0.70 0.19 0.80 0.62 0.19
2 Weeks 1.20 0.93 0.19 1.20 0.96 0.17 1.20 0.96 0.17 1.20 0.97 0.15 1.12 0.92 0.14 1.02 0.84 0.14
1 Month 1.23 0.99 0.16 1.25 0.99 0.17 1.26 1.00 0.17 1.22 1.04 0.12 1.16 0.95 0.15 1.09 0.89 0.15
2 Months 1.28 1.04 0.15 1.31 1.04 0.17 1.29 1.11 0.11 1.27 1.12 0.09 1.21 0.98 0.15 1.12 0.91 0.15
6 Months 1.31 1.07 0.15 1.34 1.07 0.17 1.32 1.15 0.10 1.31 1.14 0.10 1.25 1.01 0.16 1.18 0.95 0.16
1 Year 1.34 1.11 0.14 1.35 1.11 0.14 1.36 1.19 0.09 1.34 1.17 0.09 1.28 1.04 0.15 1.21 0.98 0.15
5 Years 1.36 1.15 0.12 1.39 1.18 0.12 1.39 1.24 0.08 1.38 1.20 0.10 1.33 1.07 0.16 1.26 1.01 0.16
50 Years 1.40 1.25 0.08 1.42 1.22 0.11 1.43 1.29 0.07 1.43 1.23 0.11 1.37 1.11 0.15 1.29 1.05 0.15
75 Years 1.40 1.26 0.07 1.43 1.24 0.10 1.43 1.30 0.07 1.44 1.24 0.10 1.37 1.13 0.14 1.29 1.06 0.14
100 Years 1.40 1.27 0.07 1.44 1.25 0.10 1.44 1.31 0.07 1.44 1.25 0.10 1.39 1.14 0.14 1.32 1.09 0.14
Span
30 ft 60 ft 90 ft 120 ft 200 ft 300 ft

Time
1 Day 1.08 0.85 0.18 1.02 0.82 0.17 1.03 0.82 0.17 1.03 0.82 0.17 0.95 0.75 0.17 0.84 0.67 0.17
2 Weeks 1.24 0.98 0.17 1.26 1.00 0.17 1.24 1.00 0.16 1.24 1.04 0.13 1.16 0.96 0.14 1.06 0.88 0.14
1 Month 1.28 1.04 0.15 1.32 1.03 0.18 1.30 1.12 0.11 1.26 1.11 0.09 1.20 0.99 0.14 1.13 0.93 0.14
2 Months 1.31 1.07 0.15 1.34 1.07 0.17 1.32 1.15 0.10 1.31 1.14 0.10 1.23 1.02 0.14 1.16 0.96 0.14
6 Months 1.34 1.11 0.14 1.35 1.11 0.14 1.34 1.19 0.08 1.32 1.17 0.09 1.28 1.04 0.15 1.23 1.00 0.15
1 Year 1.35 1.14 0.12 1.38 1.14 0.14 1.38 1.21 0.09 1.36 1.19 0.09 1.31 1.07 0.15 1.25 1.02 0.15
5 Years 1.39 1.16 0.13 1.40 1.19 0.12 1.40 1.25 0.08 1.41 1.21 0.11 1.34 1.10 0.15 1.28 1.05 0.15
50 Years 1.41 1.21 0.11 1.44 1.24 0.10 1.44 1.27 0.09 1.46 1.23 0.12 1.39 1.13 0.15 1.30 1.06 0.15
75 Years 1.42 1.22 0.11 1.45 1.25 0.10 1.45 1.29 0.08 1.46 1.25 0.11 1.40 1.14 0.15 1.31 1.07 0.15
100 Years 1.42 1.23 0.11 1.45 1.26 0.10 1.47 1.30 0.08 1.47 1.26 0.11 1.40 1.15 0.15 1.33 1.08 0.15

118
Span
30 ft 60 ft 90 ft 120 ft 200 ft 300 ft

Time
1 Day 1.17 0.88 0.22 1.09 0.89 0.16 1.11 0.87 0.18 1.13 0.87 0.20 1.02 0.81 0.17 0.91 0.75 0.17
2 Weeks 1.29 1.02 0.18 1.31 1.04 0.17 1.29 1.11 0.11 1.27 1.12 0.09 1.22 0.98 0.16 1.16 0.93 0.16
1 Month 1.32 1.06 0.16 1.34 1.08 0.16 1.32 1.15 0.10 1.29 1.14 0.09 1.25 1.01 0.16 1.20 0.97 0.16
2 Months 1.35 1.09 0.16 1.35 1.11 0.14 1.35 1.18 0.09 1.32 1.17 0.09 1.28 1.04 0.15 1.23 1.00 0.15
6 Months 1.35 1.12 0.13 1.37 1.14 0.13 1.37 1.20 0.09 1.34 1.19 0.08 1.30 1.06 0.15 1.25 1.02 0.15
1 Year 1.37 1.17 0.11 1.39 1.16 0.13 1.39 1.24 0.08 1.38 1.20 0.10 1.32 1.08 0.15 1.27 1.04 0.15
5 Years 1.39 1.24 0.08 1.41 1.21 0.11 1.42 1.27 0.08 1.42 1.22 0.11 1.37 1.11 0.15 1.30 1.06 0.15
50 Years 1.40 1.28 0.06 1.45 1.24 0.11 1.45 1.30 0.08 1.46 1.25 0.11 1.40 1.14 0.15 1.31 1.07 0.15
75 Years 1.41 1.29 0.06 1.46 1.26 0.10 1.47 1.32 0.08 1.47 1.26 0.11 1.40 1.16 0.14 1.32 1.09 0.14
100 Years 1.42 1.30 0.06 1.47 1.27 0.10 1.49 1.33 0.08 1.48 1.27 0.11 1.42 1.17 0.14 1.33 1.10 0.14
Table 5.10. Statistical Parameters of Live Load Reactions for ADTT 250, l 5   1.5s
Span
30 ft 60 ft 90 ft 120 ft 200 ft 300 ft
Time µ µ µ µ µ µ
Period 1.5s µ CV 1.5s µ CV 1.5s µ CV 1.5s µ CV 1.5s µ CV 1.5s µ CV
1 Day 1.02 0.85 0.13 0.88 0.74 0.12 0.88 0.74 0.12 0.86 0.72 0.13 0.73 0.61 0.13 0.57 0.48 0.13
2 Weeks 1.22 1.02 0.13 1.08 0.91 0.12 1.11 0.94 0.12 1.08 0.90 0.13 0.97 0.80 0.14 0.82 0.68 0.14
1 Month 1.28 1.07 0.13 1.14 0.96 0.13 1.17 0.99 0.12 1.15 0.97 0.12 1.06 0.88 0.14 0.93 0.77 0.14
2 Months 1.32 1.11 0.13 1.19 1.01 0.12 1.22 1.04 0.12 1.20 1.02 0.12 1.12 0.92 0.14 0.98 0.81 0.14
6 Months 1.37 1.16 0.12 1.27 1.07 0.12 1.32 1.11 0.13 1.30 1.10 0.12 1.18 0.97 0.14 1.08 0.89 0.14
1 Year 1.41 1.20 0.12 1.31 1.10 0.13 1.37 1.14 0.13 1.35 1.12 0.13 1.22 1.01 0.14 1.12 0.93 0.14
5 Years 1.49 1.26 0.12 1.38 1.15 0.13 1.46 1.22 0.13 1.44 1.20 0.13 1.35 1.11 0.14 1.24 1.02 0.14
50 Years 1.54 1.30 0.12 1.49 1.23 0.14 1.52 1.28 0.13 1.52 1.28 0.13 1.45 1.18 0.15 1.36 1.11 0.15
75 Years 1.55 1.31 0.12 1.50 1.24 0.14 1.55 1.29 0.13 1.55 1.29 0.13 1.46 1.19 0.15 1.37 1.12 0.15
100 Years 1.56 1.32 0.12 1.50 1.25 0.14 1.55 1.30 0.13 1.55 1.30 0.13 1.47 1.20 0.15 1.38 1.12 0.15

119
Table 5.11. Statistical Parameters of Live Load Reactions for ADTT 1,000, l 5   1.5s
Span
30 ft 60 ft 90 ft 120 ft 200 ft 300 ft
1 Day 1.14 0.94 0.14 0.95 0.80 0.13 0.94 0.80 0.11 0.91 0.79 0.10 0.84 0.70 0.13 0.74 0.62 0.13
2 Weeks 1.31 1.10 0.13 1.17 0.99 0.12 1.19 1.02 0.11 1.19 1.02 0.11 1.09 0.91 0.13 0.97 0.81 0.13
1 Month 1.35 1.15 0.12 1.23 1.03 0.13 1.26 1.08 0.11 1.25 1.07 0.11 1.17 0.97 0.13 1.06 0.88 0.13
2 Months 1.38 1.18 0.11 1.26 1.08 0.11 1.31 1.11 0.12 1.31 1.11 0.12 1.22 1.01 0.14 1.11 0.92 0.14
6 Months 1.42 1.22 0.11 1.29 1.11 0.11 1.38 1.15 0.13 1.37 1.16 0.12 1.28 1.05 0.14 1.18 0.97 0.14
1 Year 1.45 1.25 0.11 1.32 1.14 0.11 1.40 1.19 0.12 1.40 1.19 0.12 1.32 1.09 0.14 1.21 1.00 0.14
5 Years 1.50 1.29 0.11 1.40 1.20 0.11 1.49 1.26 0.12 1.50 1.26 0.13 1.38 1.14 0.14 1.28 1.06 0.14
50 Years 1.56 1.33 0.11 1.46 1.25 0.11 1.56 1.30 0.13 1.57 1.30 0.14 1.47 1.20 0.15 1.35 1.10 0.15
75 Years 1.57 1.34 0.11 1.47 1.26 0.11 1.57 1.31 0.13 1.58 1.31 0.14 1.48 1.21 0.15 1.36 1.11 0.15
100 Years 1.57 1.35 0.11 1.48 1.27 0.11 1.57 1.32 0.13 1.59 1.32 0.14 1.49 1.22 0.15 1.36 1.12 0.15
Span
30 ft 60 ft 90 ft 120 ft 200 ft 300 ft
1 Day 1.18 1.00 0.12 1.02 0.88 0.10 1.07 0.90 0.12 1.04 0.89 0.11 0.93 0.78 0.13 0.79 0.66 0.13
2 Weeks 1.35 1.14 0.12 1.23 1.05 0.11 1.29 1.09 0.12 1.29 1.09 0.12 1.19 0.99 0.13 1.06 0.89 0.13
1 Month 1.38 1.17 0.12 1.26 1.08 0.11 1.35 1.14 0.12 1.34 1.13 0.12 1.23 1.02 0.14 1.12 0.93 0.14
2 Months 1.41 1.20 0.12 1.29 1.11 0.11 1.40 1.17 0.13 1.38 1.17 0.12 1.29 1.06 0.14 1.17 0.96 0.14
6 Months 1.47 1.24 0.12 1.34 1.14 0.11 1.44 1.20 0.13 1.44 1.20 0.13 1.33 1.09 0.15 1.22 1.00 0.15
1 Year 1.49 1.25 0.13 1.36 1.16 0.11 1.47 1.23 0.13 1.48 1.24 0.13 1.38 1.12 0.15 1.25 1.02 0.15
5 Years 1.55 1.29 0.13 1.44 1.21 0.12 1.55 1.29 0.13 1.54 1.28 0.13 1.43 1.17 0.15 1.31 1.08 0.15
50 Years 1.59 1.33 0.13 1.53 1.27 0.13 1.58 1.32 0.13 1.59 1.32 0.14 1.50 1.21 0.16 1.38 1.11 0.16
75 Years 1.60 1.34 0.13 1.54 1.28 0.13 1.59 1.33 0.13 1.60 1.33 0.14 1.51 1.22 0.16 1.39 1.12 0.16
100 Years 1.60 1.35 0.13 1.54 1.29 0.13 1.59 1.34 0.13 1.61 1.34 0.14 1.51 1.23 0.16 1.40 1.13 0.16

120
Span
30 ft 60 ft 90 ft 120 ft 200 ft 300 ft
1 Day 1.25 1.05 0.12 1.09 0.94 0.11 1.14 0.96 0.13 1.12 0.94 0.13 1.02 0.84 0.14 0.90 0.74 0.14
2 Weeks 1.42 1.19 0.13 1.30 1.10 0.12 1.36 1.13 0.13 1.36 1.13 0.13 1.26 1.03 0.15 1.13 0.93 0.15
1 Month 1.46 1.22 0.13 1.34 1.13 0.12 1.39 1.16 0.13 1.40 1.17 0.13 1.30 1.06 0.15 1.18 0.96 0.15
2 Months 1.48 1.24 0.13 1.36 1.15 0.12 1.43 1.20 0.13 1.44 1.20 0.13 1.33 1.09 0.15 1.21 0.99 0.15
6 Months 1.51 1.27 0.13 1.39 1.18 0.12 1.47 1.23 0.13 1.48 1.24 0.13 1.39 1.13 0.15 1.27 1.03 0.15
1 Year 1.54 1.28 0.13 1.41 1.20 0.12 1.50 1.26 0.13 1.51 1.27 0.13 1.41 1.15 0.15 1.29 1.06 0.15
5 Years 1.58 1.32 0.13 1.48 1.25 0.12 1.54 1.30 0.12 1.56 1.30 0.13 1.46 1.19 0.15 1.34 1.09 0.15
50 Years 1.62 1.36 0.13 1.53 1.29 0.12 1.59 1.35 0.12 1.61 1.35 0.13 1.52 1.23 0.15 1.40 1.14 0.15
75 Years 1.63 1.37 0.12 1.54 1.30 0.12 1.60 1.36 0.12 1.62 1.36 0.13 1.53 1.24 0.15 1.41 1.15 0.15
100 Years 1.63 1.38 0.12 1.55 1.31 0.12 1.61 1.37 0.12 1.62 1.37 0.13 1.53 1.25 0.15 1.42 1.15 0.15
Span
30 ft 60 ft 90 ft 120 ft 200 ft 300 ft
1 Day 1.31 1.10 0.13 1.20 1.00 0.13 1.23 1.03 0.13 1.21 1.01 0.13 1.11 0.91 0.14 0.98 0.81 0.14
2 Weeks 1.45 1.21 0.13 1.35 1.12 0.13 1.40 1.17 0.13 1.41 1.18 0.13 1.31 1.07 0.15 1.19 0.97 0.15
1 Month 1.48 1.24 0.13 1.39 1.16 0.13 1.43 1.20 0.13 1.45 1.21 0.13 1.36 1.10 0.15 1.24 1.00 0.15
2 Months 1.50 1.26 0.13 1.42 1.19 0.13 1.46 1.23 0.12 1.48 1.24 0.13 1.39 1.13 0.15 1.27 1.03 0.15
6 Months 1.52 1.28 0.13 1.45 1.21 0.13 1.48 1.25 0.12 1.52 1.26 0.13 1.41 1.15 0.15 1.31 1.07 0.15
1 Year 1.55 1.29 0.13 1.46 1.22 0.13 1.51 1.28 0.12 1.54 1.28 0.13 1.44 1.17 0.15 1.33 1.08 0.15
5 Years 1.60 1.34 0.13 1.50 1.26 0.13 1.55 1.31 0.12 1.59 1.33 0.13 1.49 1.22 0.15 1.37 1.12 0.15
50 Years 1.64 1.37 0.13 1.56 1.30 0.13 1.62 1.36 0.13 1.62 1.35 0.13 1.54 1.25 0.15 1.43 1.16 0.15
75 Years 1.65 1.38 0.13 1.57 1.31 0.13 1.63 1.37 0.12 1.63 1.36 0.13 1.55 1.26 0.15 1.44 1.17 0.15
100 Years 1.66 1.39 0.13 1.57 1.32 0.13 1.63 1.38 0.12 1.64 1.37 0.13 1.55 1.27 0.15 1.45 1.18 0.15

121
Table 5.15. Statistical Parameters for Axle Loads, l 5   1.5s
ADTT
250 1,000 2,500 5,000 10,000

Time
Period l CV (%) l CV (%) l CV (%) l CV (%) l CV (%)
1 Day 0.91 0.17 1.00 0.17 1.07 0.16 1.11 0.16 1.15 0.16
2 Weeks 1.09 0.16 1.17 0.16 1.24 0.15 1.29 0.15 1.32 0.15
1 Month 1.14 0.16 1.23 0.15 1.28 0.15 1.32 0.14 1.36 0.14
2 Months 1.18 0.15 1.27 0.15 1.32 0.14 1.36 0.14 1.38 0.14
6 Months 1.24 0.15 1.32 0.14 1.37 0.14 1.40 0.14 1.42 0.13
1 Year 1.30 0.14 1.37 0.14 1.41 0.13 1.42 0.13 1.45 0.13
5 Years 1.38 0.14 1.43 0.13 1.46 0.13 1.47 0.13 1.49 0.13
50 Years 1.45 0.13 1.48 0.13 1.50 0.13 1.51 0.13 1.53 0.12
75 Years 1.45 0.13 1.48 0.12 1.50 0.12 1.51 0.12 1.53 0.12
100 Years 1.46 0.13 1.49 0.12 1.51 0.12 1.52 0.12 1.53 0.12
(continued from page 112) failure. However, knowledge about the real fatigue stress caused
5.4 Development of Statistical by current truck traffic, which was based on research done in
Parameters of Fatigue Load the 1980s, was limited and outdated.
The current AASHTO LRFD (2012) has two fatigue limit
5.4.1 Objective states. Fatigue Limit State I is related to infinite load-induced
Fatigue is one of the major causes of distress in steel highway fatigue life. The fatigue load in this limit state reflects the load
bridges. Cracking or rupture of components and connections levels found to be representative of the maximum stress range
calls for costly repairs or replacements. The durability of affected of the truck population for infinite fatigue life design. Fatigue
structures can be enhanced by applying reliability theory to this Limit State II is related to finite load-induced fatigue life. The
limit state. The limit state of fatigue is reached when accumu- fatigue load in this limit state is intended to reflect a load level
lated load spectra exceed the fatigue resistance of material. A found to be representative of the effective stress range of the
rational approach to the evaluation of existing bridges and truck population with respect to the induced number of load
design for new bridges requires knowledge of the load-carrying cycles and their cumulative damage effects on the bridge
capacity and accumulated loads, as shown on Figure 5.25. A components. Only Fatigue I applies to fatigue of concrete and
considerable effort was directed toward tests of materials under the considered types of reinforcement.
cyclic loading to establish the so-called S-N curves, where S is The focus of this section is to develop statistical models of
the applied stress, and N is the number of load applications to fatigue load based on the WIM truck survey data. The fatigue
load is intended to be used in calibration of the design provi-
sions in the AASHTO LRFD (2012). The WIM measurements
provide an unbiased data set. The 15 WIM sites provided by
FHWA are considered as representative for the United States
for this analysis. Only sites with one full year of constant
reading were used for fatigue analysis.
Three cases are considered: midspan moment for a simply
supported bridge, moment at the interior support of a two-
span continuous bridge, and moment at 0.4 of the span length
of a continuous bridge. The surveyed vehicles were run over
influence lines as traffic streams to determine the number and
magnitude of moment cycles for a wide range of span lengths
for each case. The fatigue load time history was then devel-
Figure 5.25. Fatigue failure on S-N curve. oped for the bending moment. The Fatigue II (finite life) load

122
was calculated as an equivalent moment by using the linear is very important to find the largest load cycles, because they
damage rule first proposed by Palmgren (1924) and later pop- cause the major fatigue damage. Experimental studies showed
ularized by Miner (1945) as the Palmgren–Miner rule. The that there is a linear relationship between the magnitude of
Fatigue I (infinite life) load for each location was determined load cycle and fatigue damage. S-N curves for fatigue load
by finding the highest 0.01% of the load cycles and using the tests show a log-log relationship between the cycle amplitude
smallest of them as the fatigue load for the considered loca- and the number of cycles to failure. This relationship is reflected
tion. The obtained results combined with fatigue resistance in the Palmgren–Miner formula for equivalent load, shown as
models served as the basis for the development of calibrated Equations 5.13 and 5.14.
criteria for SLS in the AASHTO LRFD. Recent WIM data provide much more complex and more
accurate information about measured trucks. The WIM data
5.4.2 WIM Data Used for Fatigue Calculation include not only axle loads and spacing between axles, but
also truck speed and time of measurement with an accuracy
To be consistent with research done by Fisher (1977), in addi- of 1 s. Using these data, the team simulated truck traffic on a
tion to the two filters used for live load, a third filter was used bridge for a 1-year period, and the time history of the bend-
to remove light trucks with GVW under 20 kips because light ing moment was recorded. This allowed calculation of the
vehicles cause relatively low fatigue damage. A summary of load effect due to the presence of multiple trucks. Calcula-
the data used for fatigue analysis, including WIM locations, tions were carried out for span lengths from 30 to 200 ft. The
number of records, and ADTT, is shown in Table 5.16. considered continuous bridges had two equal-length spans.
Examples of moment time histories for a single truck passage
5.4.3 Truck Traffic Simulation and Calculation are shown in Figure 5.26, Figure 5.27, and Figure 5.28.
of Bending Moment Time History
Live load on bridges is caused mainly by moving trucks. Longer 5.4.4 Rainflow Cycle-Counting Method
bridges often experience more than one vehicle in one span at The development of fatigue load models requires a collection
the same time. Multiple vehicles in one span produce a larger of the actual load time histories. The collected time histories
load effect than a single truck. For fatigue load calculations, it must be processed to obtain a usable form. In general, load
histories may be considered as either narrow-band or wide-
band processes, as shown in Figure 5.29. Narrow-band time
Table 5.16. WIM Locations and Number of Vehicles histories are characterized by an approximately constant period.
Used for Fatigue Analysis Wide-band time histories are characterized by a variable fre-
quency and random amplitude. For fatigue calculations, the
Total No. stress range is determined (i.e., the difference between peak
No. of Days of Truck Single-Lane
Site in Data Records ADTT and valley).
Bending moment histories due to truck passages are wide
Arizona (SPS-1) 365 26,501 97
band. The cycles are irregular with variable frequencies and
Arizona (SPS-2) 365 1,391,098 3,919 amplitudes. Wide-band histories do not allow for simple
Arkansas (SPS-2) 365 1,642,334 4,590 cycle counting. The Palmgren–Miner rule is applicable only
Colorado (SPS-2) 365 326,017 941 when the individual events are isolated, (i.e., narrow-band
time histories). Different counting procedures have been pro-
Delaware (SPS-1) 365 175,889 553
posed and used, all of which were studied and compared to
Illinois (SPS-6) 365 821,809 2,340
select the most efficient approach for this study. Only two
Kansas (SPS-2) 365 456,881 1,309 counting algorithms seemed to provide accurate results: rain-
Louisiana (SPS-1) 365 70,831 235 flow and range pair (Dowling 1972). Rainflow counting was
Maine (SPS-5) 365 172,333 503
used in this study.
A rainflow cycle-counting procedure was proposed for the
Maryland (SPS-5) 365 124,474 450
first time by Matsuishi and Endo in 1968. This method counts
Minnesota (SPS-5) 365 47,794 152 the number of full reversal cycles, as well as partial cycles, and
Pennsylvania (SPS-6) 365 1,458,818 4,098 their range amplitude for a given load time history. A full
Tennessee (SPS-6) 365 1,583,151 4,445 reversal cycle occurs when the cycle range goes up to its peak
and back to the starting position. A partial cycle goes in only
Virginia (SPS-1) 365 237,804 710
one direction, from the valley to the peak or from the peak to
Wisconsin (SPS-1) 365 209,239 622
the valley.

123
Figure 5.26. Bending moment time history for a single truck

passage on simple-supported bridges at middle of the span.

passage on continuous bridges at middle support.

124

passage on continuous bridges at 0.4 of the span length.
The summary of the steps in rainflow cycle counting are as 7. Assign a magnitude to each half-cycle equal to the stress
follows: difference between its start and termination (Table 5.17).
8. Pair up half-cycles of identical magnitude to count the
1. Reduce the time history to a sequence of (tensile) peaks number of complete cycles (Table 5.18). Typically, there
and (compressive) troughs. are some residual half-cycles (Downing and Socie 1982).
2. Imagine that the time history is a template for a rigid sheet
(pagoda roof). The moment time histories obtained from the truck traffic
3. Turn the sheet clockwise 90° (earliest time to the top). simulation for each WIM site, span length, and case were pro-
4. Each tensile peak is imagined as a source of water that cessed using the rainflow counting method. Total number of
“drips” down the pagoda. cycles was divided by number of trucks in the database to get
5. Count the number of half-cycles by looking for termina- an average number of load cycles per truck passage. The
tions in the flow occurring when results for the simple-span case are summarized in Table 5.19,
• It reaches the end of the time history (Figure 5.30, for the negative moment over the support in continuous
Path 3-4-end or Path 4-5-7-9-11-end); spans in Table 5.20, and for positive moment at 0.4 of the
• It merges with a flow that started at an earlier tensile span length in continuous bridges in Table 5.21. For simply
peak; or supported bridges, the number of cycles at the midspan was
• It flows opposite a tensile peak of greater magnitude 2 to 2.5 cycles per truck passage for short spans; this value
(Figure 5.30, Path 5-6, 6-6′, 8-8′, or 10-10′). dropped linearly to 1 cycle for a span length of about 100 ft.
6. Repeat Step 5 for compressive troughs. Similarly, for continuous bridges at 0.4 of the span length, the
Figure 5.29. Wide-band versus narrow-band history.

125
Table 5.17. Half-Cycles After Rainflow

Counting
Positive Direction Negative Direction
Range Amplitude Range Amplitude
1–2 2 2–3 3
3–4–end 4 4–5–7–9–11–end 6
5–6 1 6–6′ 1
7–8 1 8–8′ 1
9–10 4 10–10′ 4
11–12 5 12–13 4
13–14 2 — —
Note: For range values, see Figure 5.30; — = no further negative
direction values.
Table 5.18. Load Cycles

After Rainflow Counting
Amplitude No. of Cycles
1 2
2 1
Figure 5.30. Rainflow counting diagram.
3 0.5
4 2
5 0.5
6 0.5
Table 5.19. Total Number of Load Cycles and Average Number of Load Cycles per Truck Passage
for Simply Supported Bridges at the Midspan
No. of Cycles No. of Cycles per Truck

No. of
Site Vehicles 30 ft 60 ft 90 ft 120 ft 200 ft 30 ft 60 ft 90 ft 120 ft 200 ft
Arizona 26,501 59,427.5 36,397 27,321 26,505 26,501 2.24 1.37 1.03 1.00 1.00
(SPS-1)
Arizona 1,391,098 3,667,719.5 2,632,482.5 1,650,818.0 1,407,468.0 1,397,629.5 2.64 1.89 1.19 1.01 1.00
(SPS-2)
Arkansas 1,642,334 4,216,668.5 3,108,866.5 1,983,249.5 1,667,856.0 1,640,182.5 2.57 1.89 1.21 1.02 1.00
(SPS-2)
Colorado 326,017 824,366.5 591,565.5 377,138.0 328,271.0 327,680.5 2.53 1.81 1.16 1.01 1.01
(SPS-2)
Delaware 175,889 391,173.0 272,989.0 184,061.0 176,696.5 175,664.5 2.22 1.55 1.05 1.00 1.00
(SPS-1)
Illinois 821,809 2,104,493.5 1,552,007.5 990,256.0 831,086.0 823,435.0 2.56 1.89 1.20 1.01 1.00
(SPS-6)
Kansas 456,881 1,182,596.0 839,726.0 542,967.5 460,973.5 459,671.5 2.59 1.84 1.19 1.01 1.01
(SPS-2)
Louisiana 70,831 162,679.5 113,121.5 74,619.5 70,947.0 70,838.0 2.30 1.60 1.05 1.00 1.00
(SPS-1)
Maine 172,333 417,837.5 294,010.5 185,121.0 173,174.0 172,727.0 2.42 1.71 1.07 1.00 1.00
(SPS-5)

126
for Simply Supported Bridges at the Midspan (continued)

No. of
Maryland 124,474 271,233.5 186,120.0 129,968.0 124,930.5 124,482.0 2.18 1.50 1.04 1.00 1.00
(SPS-5)
Minnesota 47,794 96,065.0 68,750.0 48,829.0 47,798.0 47,752.0 2.01 1.44 1.02 1.00 1.00
(SPS-5)
Pennsylvania 1,458,818 3,669,978.0 2,667,443.0 1,676,101.0 1,477,196.0 1,459,284.0 2.52 1.83 1.15 1.01 1.00
(SPS-6)
Tennessee 1,583,151 3,492,829.0 2,816,652.0 1,673,936.0 1,600,563.0 1,583,300.0 2.21 1.78 1.06 1.01 1.00
(SPS-6)
Virginia 237,804 563,467.5 416,252.5 260,806.0 239,251.0 238,315.0 2.37 1.75 1.10 1.01 1.00
(SPS-1)
Wisconsin 209,239 483,546.0 366,955.0 225,109.0 210,644.0 210,164.5 2.31 1.75 1.08 1.01 1.00
(SPS-1)
for Continuous Bridges at the Middle Support

No. of
Arizona 26,501 65,563.5 64,115.5 69,703.5 65,402 58,905 2.47 2.42 2.63 2.47 2.22
(SPS-1)
Arizona 1,391,098 4,584,915.0 4,804,207.0 4,971,600.0 4,220,277.5 3,423,766.0 3.30 3.45 3.57 3.03 2.46
(SPS-2)
Arkansas 1,642,334 5,437,711.0 5,654,802.0 5,774,335.5 4,949,930.5 3,902,161.0 3.31 3.44 3.52 3.01 2.38
(SPS-2)
Colorado 326,017 1,020,374.5 989,200.0 1,100,728.5 983,802.0 767,937.0 3.13 3.03 3.38 3.02 2.36
(SPS-2)
Delaware 175,889 543,754.5 502,112.5 527,143.0 484,787.5 419,294.5 3.09 2.85 3.00 2.76 2.38
(SPS-1)
Illinois 821,809 2,716,902.0 2,768,327.0 2,836,337.0 2,489,643.5 1,987,891.5 3.31 3.37 3.45 3.03 2.42
(SPS-6)
Kansas 456,881 1,505,890.5 1,507,880.5 1,608,769.0 1,387,383.0 1,116,965.5 3.30 3.30 3.52 3.04 2.44
(SPS-2)
Louisiana 70,831 217,990.0 199,088.0 215,738.0 200,995.5 166,450.5 3.08 2.81 3.05 2.84 2.35
(SPS-1)
Maine 172,333 518,377.5 502,246.5 558,181.0 508,993.0 383,351.5 3.01 2.91 3.24 2.95 2.22
(SPS-5)
Maryland 124,474 397,197.5 346,614.5 376,056.5 342,106.5 290,348.0 3.19 2.78 3.02 2.75 2.33
(SPS-5)
Minnesota 47,794 135,741.0 131,289.0 139,940.0 123,124.0 107,837.0 2.84 2.75 2.93 2.58 2.26
(SPS-5)
Pennsylvania 1,458,818 3,896,713.0 3,604,125.0 4,019,137.0 3,955,368.0 3,174,582.0 2.67 2.47 2.76 2.71 2.18
(SPS-6)
Tennessee 1,583,151 4,298,789.0 3,889,255.0 4,468,069.0 4,346,233.0 3,427,878.0 2.72 2.46 2.82 2.75 2.17
(SPS-6)
Virginia 237,804 743,162.0 716,559.5 770,125.5 700,670.5 561,742.5 3.13 3.01 3.24 2.95 2.36
(SPS-1)
Wisconsin 209,239 646,250.5 633,403.0 657,828.5 608,381.0 492,283.5 3.09 3.03 3.14 2.91 2.35
(SPS-1)

127
for Continuous Bridges at 0.4 of the Span Length

No. of
Arizona 26,501 68,688.0 39,328.0 29,363.0 27,695.0 26,509.0 2.59 1.48 1.11 1.05 1.00
(SPS-1)
Arizona 1,391,098 4,032,130.0 2,699,800.5 2,281,797.0 2,017,321.5 1,767,920.0 2.90 1.94 1.64 1.45 1.27
(SPS-2)
Arkansas 1,642,334 5,610,372.0 4,069,843.0 3,532,308.0 3,132,234.0 2,872,888.0 3.42 2.48 2.15 1.91 1.75
(SPS-2)
Colorado 326,017 885,651.0 617,440.5 458,136.5 410,761.5 385,205.5 2.72 1.89 1.41 1.26 1.18
(SPS-2)
Delaware 175,889 410,830.0 293,946.0 223,028.5 210,104.0 199,350.0 2.34 1.67 1.27 1.19 1.13
(SPS-1)
Illinois 821,809 2,304,196.0 1,579,655.0 1,313,036.5 1,118,188.0 1,037,709.5 2.80 1.92 1.60 1.36 1.26
(SPS-6)
Kansas 456,881 1,292,694.0 872,400.0 702,959.5 616,645.0 554,203.0 2.83 1.91 1.54 1.35 1.21
(SPS-2)
Louisiana 70,831 171,703.5 120,584.5 91,168.0 85,553.5 80,458.0 2.42 1.70 1.29 1.21 1.14
(SPS-1)
Maine 172,333 433,793.5 313,517.0 231,617.0 204,775.5 190,443.0 2.52 1.82 1.34 1.19 1.11
(SPS-5)
Maryland 124,474 279,856.5 200,955.5 155,882.5 143,168.0 138,347.5 2.25 1.61 1.25 1.15 1.11
(SPS-5)
Minnesota 47,794 123,298.0 70,383.5 59,891.5 52,727.5 48,541.5 2.58 1.47 1.25 1.10 1.02
(SPS-5)
Pennsylvania 1,458,818 3,992,907.0 2,784,565.0 2,243,835.5 1,943,551.0 1,756,756.0 2.74 1.91 1.54 1.33 1.20
(SPS-6)
Tennessee 1,583,151 4,590,126.0 2,929,061.5 2,273,958.5 1,888,805.5 1,651,117.5 2.90 1.85 1.44 1.19 1.04
(SPS-6)
Virginia 237,804 599,977.0 434,778.0 338,100.0 299,309.0 278,883.5 2.52 1.83 1.42 1.26 1.17
(SPS-1)
Wisconsin 209,239 516,843.0 376,098.5 298,936.5 267,981.0 246,176.0 2.47 1.80 1.43 1.28 1.18
(SPS-1)
number of cycles per truck was 2.3 to 3.5 for short spans, applying a cumulative damage rule. Many rules have been
which dropped to 1 to 1.5 cycles for a span length of about proposed. According to the Palmgren–Miner rule, which
100 ft. The results for negative moment over the support in seems to provide a reasonable means of accounting for ran-
continuous bridges were 2.5 to 3.5 for short spans and about dom variable loading, fatigue damage due to variable ampli-
2.5 for longer spans. More load cycles for short spans is caused tude loading is expressed by Equation 5.8:
by groups of axles rather than whole trucks due to relatively
n1 n n n n
short spans compared with the vehicle length. + 2 + 3 +…+ n = ∑ i =1 (5.8)
N1 N2 N3 Nn Ni
5.4.5 Fatigue Damage Accumulation
where ni/Ni is the incremental damage that results from the
and Equivalent Fatigue Load
stress range cycles with magnitude Si that occurs ni times
Because bridge structures are subjected to loads of different (Figure 5.31), and Ni is the number of cycles to failure with a
magnitude and frequency occurring at different times, the constant amplitude equal to Si (Figure 5.32). Failure occurs
load can be considered as a randomly varying amplitude when the sum of the incremental damage equals or exceeds 1.
load. The effect of such a loading can be accounted for by The tests of welded details (Fisher et al. 1983; Schilling et al.

128
S3 Substituting Equation 5.11 into Equation 5.10 yields
ni = pi N T (5.11)
Stress range, S
pi N T pi Se− n
S1 ∑ AS −n
=∑
ASi− n
=1 (5.12)
i
S2
or Sen = ∑ pi Sin
Time Se = n ∑ pi Sin (5.13)

n1 n2 n3
The exponent n for most structural metal details is about 3.
Figure 5.31. Number of cycles ni for stress range S i. Equation 5.13 is often referred to as a root mean cube of the
stress distribution. The equivalent stress is a convenient con-
1977) and Barsom’s crack growth studies (Rolfe and Barsom cept to be used for comparison of stress histograms obtained
1977) showed a good correlation with the Palmgren–Miner using the rainflow counting method.
rule assumptions. Because fatigue crack nucleation and further propagation
Schilling et al. (1977) showed that the Palmgren–Miner occur mostly at tensile stress conditions that are related to
rule can be used to develop an equivalent constant amplitude bending moment, it is convenient to use the bending moment
cyclic loading that produces the same fatigue damage as a formulation instead of the stress formulation for an equiva-
variable amplitude load for the same number of load cycles. lent load. The bending moment formulation of Equation 5.13
This theory is based on the exponential model of the stress is given by Equation 5.14:
range–life relationship as given by Equation 5.9 (Fisher 1977):
Meq = n ∑ pi M in (5.14)
−n
N = AS (5.9)
where
where Meq = equivalent moment cycle load;
N = number of cycles to failure; Mi = incremental moment cycle; and
S = nominal stress range; pi = probability of occurrence of Mi.
A = a constant for a given detail; and
Calculation of the equivalent moment requires the prob-
n = slope constant.
ability of occurrence for each incremental moment Mi. The
The concept of fatigue design based on stress range alone was corresponding probability distribution functions (PDFs) of
adopted by AASHO in 1974 (Fisher et al. 1970, 1974). Equa- the moment cycles for each site were calculated for spans
tion 5.10 is obtained by substituting Equation 5.9 into Equa- from 30 to 200 ft. As an example, the PDFs for moments cor-
tion 5.8: responding to the FHWA WIM data from Arkansas (SPS-1)
are shown in Figure 5.33. The area under the curve represent-
ni
∑ AS −n
=1 (5.10) ing the PDF for each span length is equal to 1.
The equivalent moment was calculated from moment
i
cycles obtained using rainflow counting. The equivalent load
was calculated for all considered WIM sites, a wide range of
span lengths between 30 and 200 ft, and three bridge configu-
rations. Next, the calculated equivalent moments were divided
Stress range, S
by moment due to the AASHTO LRFD fatigue truck. Results

S3 are summarized in Table 5.22 to Table 5.24. The results show
S1 that the moment ratio is smaller for short spans.
S2
5.4.6 Fatigue Limit State II:

N3 N1 N2 Fatigue Damage Ratio
Number of load cycles to failure, N
Finite fatigue life depends on the number of load cycles dur-
Figure 5.32. Number of load cycles to failure Ni for ing the service life and their magnitude. According to the
stress range S i. AASHTO LRFD (2012) provisions, the number of load cycles

129
30 ft. To compare fatigue damage due to design fatigue load and

actual fatigue load, it is convenient to remove the resistance part
from limit state Equations 5.16, 5.17, and 5.18 [AASHTO LRFD
(2012, Equations 6.6.1.2.5-1 and 6.6.1.2.5-2)]:
γ (∆ f ) ≤ (∆ F )n (5.16)
A
(∆ F )n = 3 (5.17)
N
A
γ (∆ f ) ≤ 3 (5.18)
N
where
g = load factor;
Figure 5.33. PDFs of WIM moments for data from D f = force effect (i.e., live load stress range due to the pas-
Arkansas (SPS-1). sage of a fatigue truck);
A = resistance constant that depends on the class of the
during bridge service life (N) is calculated using Equa- structural detail; and
tion 6.6.1.2.5-3, shown here as Equation 5.15: N = number of load cycles during the service life calcu-
lated according to Equation 5.15.
N = (365)(75)n (ADTT )SL (5.15)
Stress due to truck passage is calculated according to Equa-
where (ADTT)SL is a single lane of ADTT, and n is the number tion 5.19:
of load cycles per truck taken from Table 5.25 [AASHTO ∆f = M S (5.19)
LRFD (2012, Table 6.6.1.2.5-2)].
The magnitude of load cycles is calculated as a stress due to where S is section modulus and M is moment due to truck
the HL-93 fatigue truck with the second axle spacing equal to passage.
Table 5.22. Equivalent Moments for Simply Supported Bridges at the Midspan
Equivalent Moment/HL-93
Equivalent Moment (kip-ft) Fatigue Moment
No. of
Arizona (SPS-1) 26,501 151.63 426.18 889.67 1,362.17 2,593.90 0.62 0.78 0.82 0.84 0.85
Arizona (SPS-2) 1,391,098 145.82 357.81 790.59 1,316.93 2,601.01 0.60 0.66 0.73 0.81 0.85
Arkansas (SPS-2) 1,642,334 146.25 354.83 770.41 1,290.54 2,554.63 0.60 0.65 0.71 0.79 0.83
Colorado (SPS-2) 326,017 132.61 325.49 713.45 1,173.08 2,311.31 0.54 0.60 0.66 0.72 0.75
Delaware (SPS-1) 175,889 155.16 400.92 831.01 1,270.55 2,424.36 0.64 0.74 0.77 0.78 0.79
Illinois (SPS-6) 821,809 146.48 354.91 762.76 1,279.33 2,532.79 0.60 0.65 0.70 0.79 0.83
Kansas (SPS-2) 456,881 141.00 355.18 767.58 1,277.57 2,524.67 0.58 0.65 0.71 0.79 0.82
Louisiana (SPS-1) 70,831 142.42 363.30 775.00 1,202.37 2,318.98 0.58 0.67 0.72 0.74 0.76
Maine (SPS-5) 172,333 129.72 328.38 707.39 1,126.24 2,206.36 0.53 0.60 0.65 0.69 0.72
Maryland (SPS-5) 124,474 132.44 335.88 675.87 1,033.81 1,982.63 0.54 0.62 0.62 0.64 0.65
Minnesota (SPS-5) 47,794 142.39 353.48 731.81 1,138.96 2,219.99 0.58 0.65 0.68 0.70 0.72
Pennsylvania (SPS-6) 1,458,818 151.46 363.23 777.74 1,259.78 2,468.70 0.62 0.67 0.72 0.78 0.81
Tennessee (SPS-6) 1,583,151 153.14 351.05 772.72 1,227.46 2,417.64 0.63 0.65 0.71 0.76 0.79
Virginia (SPS-1) 237,804 140.35 344.56 749.93 1,202.76 2,356.27 0.58 0.63 0.69 0.74 0.77
Wisconsin (SPS-1) 209,239 142.47 360.19 772.69 1,213.03 2,349.64 0.58 0.66 0.71 0.75 0.77

130
Table 5.23. Equivalent Moments for Continuous Bridges at the Middle Support
No. of
Arizona (SPS-1) 26,501 -96.91 -212.79 -314.01 -483.56 -960.50 0.53 0.59 0.59 0.63 0.72
Arizona (SPS-2) 1,391,098 -89.91 -221.06 -296.65 -454.32 -975.62 0.49 0.61 0.56 0.60 0.73
Arkansas (SPS-2) 1,642,334 -87.98 -219.14 -294.68 -450.44 -998.99 0.48 0.61 0.56 0.59 0.74
Colorado (SPS-2) 326,017 -82.94 -203.52 -268.50 -407.76 -844.78 0.45 0.56 0.51 0.54 0.63
Delaware (SPS-1) 175,889 -90.38 -214.99 -299.91 -451.62 -896.29 0.49 0.60 0.57 0.59 0.67
Illinois (SPS-6) 821,809 -87.55 -219.79 -295.45 -444.61 -964.62 0.48 0.61 0.56 0.58 0.72
Kansas (SPS-2) 456,881 -85.97 -216.73 -290.84 -439.49 -916.36 0.47 0.60 0.55 0.58 0.68
Louisiana (SPS-1) 70,831 -86.45 -205.76 -280.85 -423.51 -858.73 0.47 0.57 0.53 0.56 0.64
Maine (SPS-5) 172,333 -79.39 -198.30 -262.39 -393.39 -825.92 0.43 0.55 0.50 0.52 0.62
Maryland (SPS-5) 124,474 -79.35 -192.49 -263.24 -403.19 -814.86 0.43 0.53 0.50 0.53 0.61
Minnesota (SPS-5) 47,794 -79.86 -201.32 -270.79 -405.61 -814.03 0.43 0.56 0.51 0.53 0.61
Pennsylvania (SPS-6) 1,458,818 -90.89 -235.11 -310.77 -449.53 -974.43 0.49 0.65 0.59 0.59 0.73
Tennessee (SPS-6) 1,583,151 -87.39 -231.37 -300.99 -436.22 -961.13 0.48 0.64 0.57 0.57 0.72
Virginia (SPS-1) 237,804 -84.56 -208.61 -278.84 -418.94 -868.36 0.46 0.58 0.53 0.55 0.65
Wisconsin (SPS-1) 209,239 -83.68 -206.92 -285.18 -422.87 -860.95 0.45 0.57 0.54 0.56 0.64
Table 5.24. Equivalent Moments for Continuous Bridges at 0.4 of the Span Length
No. of
Arizona (SPS-1) 26,501 134.25 413.51 838.11 1,291.21 2,503.65 0.55 0.71 0.76 0.80 0.83
Arizona (SPS-2) 1,391,098 133.46 349.66 663.89 1,096.11 2,282.14 0.54 0.60 0.60 0.68 0.75
Arkansas (SPS-2) 1,642,334 122.64 272.45 540.68 899.92 1,881.91 0.50 0.47 0.49 0.55 0.62
Colorado (SPS-2) 326,017 121.69 317.11 634.34 1,032.34 2,101.21 0.49 0.54 0.58 0.64 0.69
Delaware (SPS-1) 175,889 144.84 386.50 743.25 1,143.10 2,230.39 0.59 0.66 0.68 0.70 0.74
Illinois (SPS-6) 821,809 135.47 345.78 657.19 1,091.43 2,222.58 0.55 0.59 0.60 0.67 0.73
Kansas (SPS-2) 456,881 129.86 342.26 665.25 1,095.22 2,272.29 0.53 0.58 0.61 0.68 0.75
Louisiana (SPS-1) 70,831 131.42 353.55 691.10 1,076.33 2,130.97 0.53 0.60 0.63 0.66 0.70
Maine (SPS-5) 172,333 121.26 312.87 618.65 1,008.24 2,050.51 0.49 0.53 0.56 0.62 0.68
Maryland (SPS-5) 124,474 126.68 339.42 654.83 1,023.41 1,994.36 0.52 0.58 0.60 0.63 0.66
Minnesota (SPS-5) 47,794 120.71 344.34 655.63 1,054.97 2,132.90 0.49 0.59 0.60 0.65 0.70
Pennsylvania (SPS-6) 1,458,818 135.74 352.48 668.88 1,087.55 2,204.49 0.55 0.60 0.61 0.67 0.73
Tennessee (SPS-6) 1,583,151 128.44 339.17 665.81 1,104.94 2,275.40 0.52 0.58 0.61 0.68 0.75
Virginia (SPS-1) 237,804 130.01 334.69 649.10 1,055.60 2,142.89 0.53 0.57 0.59 0.65 0.71
Wisconsin (SPS-1) 209,239 133.10 349.47 666.43 1,061.17 2,138.88 0.54 0.60 0.61 0.65 0.71

131
Table 5.25. Number of Cycles per Truck The fatigue damage ratio was calculated according to the
Passage (n) for AASHTO Fatigue Design current AASHTO LRFD provisions for each WIM site, span
length, and case. Results are summarized in Table 5.27 to
n
Table 5.29 in the column labeled Fatigue Damage Ratio (cur-
Span Span rent). The fatigue damage ratio is smaller for shorter spans.
Length Length The difference between short and longer spans is due to dif-
Longitudinal Members >40 ft <
– 40 ft
ferent code provisions for short spans with a given number of
Simple-span girders 1.0 2.0 load cycles per truck passage (see Table 5.25). For short spans,
Continuous girders Near interior support 1.5 2.0 a truck causes more load cycles than for longer spans. How-
Elsewhere 1.0 2.0 ever, it is balanced by a smaller moment ratio (equivalent
moment/HL-93 fatigue truck moment) for short spans. If the
number of load cycles due to a truck passage were equal for
To calculate the ratio of fatigue damage caused by the all spans, as shown in Table 5.26, then the resulting fatigue
actual fatigue load and design fatigue load, the load factor has damage ratio would be more uniform.
to be removed from Equation 5.18. From Equations 5.18 and The fatigue damage ratio for the proposed fatigue design
5.19, it is possible to calculate the ratio of fatigue damage due was calculated for each WIM site, span length, and case. The
to the actual load and fatigue damage due to design load by results are summarized in Table 5.27 to Table 5.29 in the col-
using Equation 5.20: umn labeled Fatigue Damage Ratio (proposed). For simply
supported bridges at midspan and continuous bridges at 0.4
of the span length, the results are very uniform for all span
Boundary of Actual Boundary of Code
lengths. At the middle support of continuous bridges, the
Fatigue Damage Fatigue Damage
difference between short and longer spans is reduced by
M eq about 10%.
A M 3 A
S = 3 S = Because fatigue resistance depends on structural detail and
S NR S N
material characteristics but not on span length, the variation in
where fatigue load due to span length produces a variation in reli-
where
M = moment due to ability indices. The design parameters proposed in Section 5.4.9
Meq = equivalent moment
fatigue design truck; eliminate this problem.
from Miner’s Rule;
A = resistance constant; A = resistance constant;
and and 5.4.7 Fatigue Limit State I: Maximum
NR = actual number of N = number of cycles Moment Range Ratio
cycles. (from Equation 5.15).
Fatigue Limit State I is related to an infinite load-induced
A S A S fatigue life. The fatigue load in this limit state reflects the load
1= 3 p 1= 3 p levels found to be representative of the maximum stress range
NR M eq N M
of the truck population for an infinite fatigue life design
(AASHTO LRFD 2012). In other words, if the majority of stress
cycles are below a threshold magnitude [(D F)TH], then failure
A S A S will require so many load cycles that the considered detail will
3 p =3 p
NR M eq N M have an infinite fatigue life. (D F)TH is a boundary between the
finite and infinite fatigue life, as shown in Figure 5.34.
A NR SM eq
λ= 3 p p
N A SM
Table 5.26. Number of Cycles per
N M eq Truck Passage (n) for Proposed
λ= 3 R p (5.20)
N M Fatigue Design
l is the ratio of the fatigue damage due to the actual fatigue Longitudinal Members n
load to the fatigue damage due to the design fatigue load. Simple-span girders 1.0
Because resistance was removed from Equation 5.20, the
Continuous girders Near interior support 1.5
fatigue damage ratio is the same regardless of the bridge com-
Elsewhere 1.0
ponent or detail class.

132
Table 5.27. Fatigue Damage Ratios for Simply Supported Bridges at the Midspan
Fatigue Damage Ratio (current) Fatigue Damage Ratio (proposed)

No. of
Arizona (SPS-1) 26,501 0.65 0.87 0.83 0.84 0.85 0.81 0.87 0.83 0.84 0.85
Arizona (SPS-2) 1,391,098 0.66 0.81 0.77 0.81 0.85 0.83 0.81 0.77 0.81 0.85
Arkansas (SPS-2) 1,642,334 0.65 0.81 0.76 0.80 0.83 0.82 0.81 0.76 0.80 0.83
Colorado (SPS-2) 326,017 0.59 0.73 0.69 0.72 0.76 0.74 0.73 0.69 0.72 0.76
Delaware (SPS-1) 175,889 0.66 0.85 0.78 0.78 0.79 0.83 0.85 0.78 0.78 0.79
Illinois (SPS-6) 821,809 0.65 0.81 0.75 0.79 0.83 0.82 0.81 0.75 0.79 0.83
Kansas (SPS-2) 456,881 0.63 0.80 0.75 0.79 0.83 0.79 0.80 0.75 0.79 0.83
Louisiana (SPS-1) 70,831 0.61 0.78 0.73 0.74 0.76 0.77 0.78 0.73 0.74 0.76
Maine (SPS-5) 172,333 0.57 0.72 0.67 0.69 0.72 0.71 0.72 0.67 0.69 0.72
Maryland (SPS-5) 124,474 0.56 0.71 0.63 0.64 0.65 0.70 0.71 0.63 0.64 0.65
Minnesota (SPS-5) 47,794 0.58 0.73 0.68 0.70 0.72 0.74 0.73 0.68 0.70 0.72
Pennsylvania (SPS-6) 1,458,818 0.67 0.82 0.75 0.78 0.81 0.84 0.82 0.75 0.78 0.81
Tennessee (SPS-6) 1,583,151 0.65 0.78 0.73 0.76 0.79 0.82 0.78 0.73 0.76 0.79
Virginia (SPS-1) 237,804 0.61 0.76 0.71 0.74 0.77 0.77 0.76 0.71 0.74 0.77
Wisconsin (SPS-1) 209,239 0.61 0.80 0.73 0.75 0.77 0.77 0.80 0.73 0.75 0.77
Table 5.28. Fatigue Damage Ratios for Continuous Bridges at the Middle Support

No. of
Arizona (SPS-1) 26,501 0.57 0.69 0.71 0.75 0.82 0.62 0.69 0.71 0.75 0.82
Arizona (SPS-2) 1,391,098 0.58 0.81 0.75 0.75 0.86 0.64 0.81 0.75 0.75 0.86
Arkansas (SPS-2) 1,642,334 0.57 0.80 0.74 0.75 0.87 0.62 0.80 0.74 0.75 0.87
Colorado (SPS-2) 326,017 0.52 0.71 0.66 0.68 0.73 0.58 0.71 0.66 0.68 0.73
Delaware (SPS-1) 175,889 0.57 0.74 0.71 0.73 0.78 0.63 0.74 0.71 0.73 0.78
Illinois (SPS-6) 821,809 0.56 0.80 0.74 0.74 0.84 0.62 0.80 0.74 0.74 0.84
Kansas (SPS-2) 456,881 0.55 0.78 0.73 0.73 0.80 0.61 0.78 0.73 0.73 0.80
Louisiana (SPS-1) 70,831 0.54 0.70 0.67 0.69 0.74 0.60 0.70 0.67 0.69 0.74
Maine (SPS-5) 172,333 0.49 0.69 0.64 0.65 0.70 0.54 0.69 0.64 0.65 0.70
Maryland (SPS-5) 124,474 0.50 0.66 0.63 0.65 0.70 0.55 0.66 0.63 0.65 0.70
Minnesota (SPS-5) 47,794 0.49 0.68 0.64 0.64 0.70 0.54 0.68 0.64 0.64 0.70
Tennessee (SPS-6) 1,583,151 0.53 0.76 0.70 0.70 0.81 0.58 0.76 0.70 0.70 0.81
Virginia (SPS-1) 237,804 0.53 0.73 0.68 0.69 0.75 0.59 0.73 0.68 0.69 0.75
Wisconsin (SPS-1) 209,239 0.53 0.73 0.69 0.69 0.75 0.58 0.73 0.69 0.69 0.75

133
Table 5.29. Fatigue Damage Ratios for Continuous Bridges at 0.4 of the Span Length

No. of
Arizona (SPS-1) 26,501 0.60 0.81 0.79 0.81 0.83 0.75 0.81 0.79 0.81 0.83
Arizona (SPS-2) 1,391,098 0.61 0.75 0.71 0.76 0.82 0.77 0.75 0.71 0.76 0.82
Arkansas (SPS-2) 1,642,334 0.60 0.63 0.64 0.69 0.75 0.75 0.63 0.64 0.69 0.75
Colorado (SPS-2) 326,017 0.55 0.67 0.65 0.69 0.73 0.69 0.67 0.65 0.69 0.73
Delaware (SPS-1) 175,889 0.62 0.78 0.73 0.75 0.77 0.78 0.78 0.73 0.75 0.77
Illinois (SPS-6) 821,809 0.62 0.73 0.70 0.75 0.79 0.78 0.73 0.70 0.75 0.79
Kansas (SPS-2) 456,881 0.59 0.73 0.70 0.75 0.80 0.75 0.73 0.70 0.75 0.80
Louisiana (SPS-1) 70,831 0.57 0.72 0.68 0.71 0.73 0.72 0.72 0.68 0.71 0.73
Maine (SPS-5) 172,333 0.53 0.65 0.62 0.66 0.70 0.67 0.65 0.62 0.66 0.70
Maryland (SPS-5) 124,474 0.54 0.68 0.64 0.66 0.68 0.67 0.68 0.64 0.66 0.68
Minnesota (SPS-5) 47,794 0.53 0.67 0.64 0.67 0.71 0.67 0.67 0.64 0.67 0.71
Tennessee (SPS-6) 1,583,151 0.59 0.71 0.68 0.72 0.76 0.74 0.71 0.68 0.72 0.76
Virginia (SPS-1) 237,804 0.57 0.70 0.66 0.70 0.75 0.72 0.70 0.66 0.70 0.75
Wisconsin (SPS-1) 209,239 0.58 0.73 0.68 0.71 0.75 0.73 0.73 0.68 0.71 0.75
Fatigue Limit State I refers to the stress value that has The calculations were performed for the considered loca-
1/10,000 probability of being exceeded. It is assumed that the tions, cases, and span lengths. The obtained values of moment
distribution of stress has the same CDF shape as that of the were divided by the corresponding AASHTO fatigue truck
corresponding moments. Thus, the fatigue load analysis is moment. The results are summarized in Table 5.30 to Table 5.32.
performed using the developed CDFs for moments for vari-
ous considered sites, cases, and spans from 30 to 200 ft. The 5.4.8 Statistical Parameters
moment corresponding to the upper 0.01% is determined as of Fatigue Live Load
a percentile corresponding to the probability of 0.9999, or 3.8
The objective was to determine the statistical parameters of
on the vertical axis in Figure 5.35. This moment represents
fatigue load that can be considered as representative for the
the maximum stress range corresponding to an unlimited
national load. The statistical parameters will be different for
fatigue life. For example, for the WIM data from Arkansas
the maximum and equivalent fatigue load specified for
(SPS-1), the moment for span of 120 ft corresponding to the
Fatigue Limit States I and II, respectively. The ratios of the
upper 0.01% is 2,505.5 kip-ft (Figure 5.35).
1/10,000 moment to the HL-93 fatigue moment were plotted
on normal probability paper and are shown in Figure 5.36 to
Figure 5.38, and the proposed fatigue damage ratios are
shown in Figure 5.39 to Figure 5.41. Each point on the graphs
Stress range (log scale)
represents one of 15 sites considered.

To determine the statistical parameters from the graphs, a
Resistance, R
straight line was fitted for each distribution. A straight line cor-
Finite Fatigue Life responds to a normal distribution on the normal probability
(F)TH paper. The intersection of the straight line with the horizontal
axis is at the mean value. The standard deviation is determined
Infinite Fatigue Life from the slope of the straight line. The statistical parameters of
fatigue load (i.e., mean, µ, and CV), based on data from 15 con-
Number of load cycles to failure (log scale) sidered sites, were calculated as the ratio of standard deviation
Figure 5.34. The threshold stress (DF)TH on an (s) and the mean and are listed in Table 5.33 and Table 5.34.
S-N curve. (text continues on page 142)

134
Figure 5.35. Moment corresponding to the upper 0.01%,

span 5 120 ft.
Table 5.30. Maximum Moment Range for Simply Supported Bridges at the Midspan
1/10,000 Moment/HL-93
1/10,000 Moment Cycle Fatigue Moment
No. of
Arizona (SPS-1) 26,501 424 1,003 1,761 2,754 5,640 1.74 1.84 1.63 1.70 1.84
Arizona (SPS-2) 1,391,098 308 765 1,416 2,246 4,711 1.26 1.41 1.31 1.38 1.54
Arkansas (SPS-2) 1,642,334 352 860 1,526 2,460 5,066 1.44 1.58 1.41 1.52 1.65
Colorado (SPS-2) 326,017 336 814 1,497 2,409 4,854 1.38 1.50 1.38 1.48 1.58
Delaware (SPS-1) 175,889 454 1,257 2,302 3,212 5,735 1.86 2.31 2.12 1.98 1.87
Illinois (SPS-6) 821,809 350 844 1,480 2,408 5,033 1.43 1.55 1.37 1.48 1.64
Kansas (SPS-2) 456,881 411 1,018 1,989 3,112 6,083 1.69 1.87 1.84 1.92 1.99
Louisiana (SPS-1) 70,831 460 1,237 2,126 3,332 6,616 1.89 2.27 1.96 2.05 2.16
Maine (SPS-5) 172,333 397 964 1,722 2,726 5,549 1.63 1.77 1.59 1.68 1.81
Maryland (SPS-5) 124,474 412 1,038 1,802 2,599 5,061 1.69 1.91 1.66 1.60 1.65
Minnesota (SPS-5) 47,794 392 1,111 2,220 3,316 6,225 1.61 2.04 2.05 2.04 2.03
Pennsylvania (SPS-6) 1,458,818 402 1,003 1,730 2,623 5,291 1.65 1.84 1.60 1.62 1.73
Tennessee (SPS-6) 1,583,151 419 1,020 1,652 2,387 4,906 1.72 1.88 1.52 1.47 1.60
Virginia (SPS-1) 237,804 369 946 1,709 2,562 5,055 1.51 1.74 1.58 1.58 1.65
Wisconsin (SPS-1) 209,239 393 968 1,712 2,717 5,396 1.61 1.78 1.58 1.67 1.76

135
Table 5.31. Maximum Moment Range for Continuous Bridges at the Middle Support
1/10,000 Moment/HL-93 Fatigue

1/10,000 Moment Cycle Moment
No. of
Arizona (SPS-1) 26,501 -266 -701 -1,026 -1,608 -3,089 1.45 1.95 1.94 2.11 2.30
Arizona (SPS-2) 1,391,098 -211 -549 -968 -1,526 -3,019 1.15 1.52 1.83 2.00 2.25
Arkansas (SPS-2) 1,642,334 -213 -643 -995 -1,522 -3,187 1.16 1.78 1.88 2.00 2.38
Colorado (SPS-2) 326,017 -231 -579 -877 -1,312 -2,813 1.25 1.61 1.66 1.72 2.10
Delaware (SPS-1) 175,889 -248 -650 -1,173 -1,643 -3,303 1.35 1.80 2.21 2.16 2.46
Illinois (SPS-6) 821,809 -207 -640 -1,005 -1,506 -3,093 1.13 1.78 1.90 1.98 2.31
Kansas (SPS-2) 456,881 -294 -755 -1,015 -1,469 -2,937 1.60 2.10 1.92 1.93 2.19
Louisiana (SPS-1) 70,831 -278 -815 -1,128 -1,539 -3,255 1.51 2.26 2.13 2.02 2.43
Maine (SPS-5) 172,333 -251 -694 -970 -1,418 -2,967 1.37 1.93 1.83 1.86 2.21
Maryland (SPS-5) 124,474 -240 -592 -1,049 -1,564 -3,281 1.31 1.64 1.98 2.05 2.45
Minnesota (SPS-5) 47,794 -292 -695 -1,034 -1,487 -2,753 1.59 1.93 1.95 1.95 2.05
Pennsylvania (SPS-6) 1,458,818 -245 -638 -1,067 -1,588 -3,131 1.33 1.77 2.01 2.09 2.33
Tennessee (SPS-6) 1,583,151 -222 -628 -1,025 -1,559 -2,977 1.21 1.74 1.93 2.05 2.22
Virginia (SPS-1) 237,804 -223 -603 -973 -1,477 -3,010 1.21 1.67 1.84 1.94 2.24
Wisconsin (SPS-1) 209,239 -250 -671 -953 -1,394 -2,892 1.36 1.86 1.80 1.83 2.16
Table 5.32. Maximum Moment Range for Continuous Bridges at 0.4 of the Span Length
1/10,000 Moment/HL-93
1/10,000 Moment Cycle Fatigue Moment
No. of
Arizona (SPS-1) 26,501 399 976 1,764 2,769 5,542 1.62 1.67 1.61 1.71 1.83
Arizona (SPS-2) 1,391,098 293 761 1,431 2,228 4,636 1.19 1.30 1.30 1.37 1.53
Arkansas (SPS-2) 1,642,334 338 849 1,527 2,416 4,914 1.37 1.45 1.39 1.49 1.62
Colorado (SPS-2) 326,017 319 805 1,528 2,428 4,857 1.30 1.38 1.39 1.50 1.60
Delaware (SPS-1) 175,889 439 1,279 2,243 3,141 5,635 1.78 2.19 2.04 1.94 1.86
Illinois (SPS-6) 821,809 334 814 1,508 2,399 4,893 1.36 1.39 1.37 1.48 1.61
Kansas (SPS-2) 456,881 394 1,049 1,983 3,088 5,988 1.60 1.79 1.81 1.90 1.98
Louisiana (SPS-1) 70,831 458 1,126 2,174 3,349 6,486 1.86 1.92 1.98 2.06 2.14
Maine (SPS-5) 172,333 377 937 1,811 2,768 5,525 1.53 1.60 1.65 1.71 1.82
Maryland (SPS-5) 124,474 406 1,036 1,817 2,618 4,941 1.65 1.77 1.65 1.61 1.63
Minnesota (SPS-5) 47,794 382 1,142 2,134 3,223 6,065 1.55 1.95 1.94 1.99 2.00
Pennsylvania (SPS-6) 1,458,818 395 1,020 1,726 2,608 5,243 1.61 1.74 1.57 1.61 1.73
Tennessee (SPS-6) 1,583,151 416 1,012 1,636 2,379 4,868 1.69 1.73 1.49 1.47 1.61
Virginia (SPS-1) 237,804 356 955 1,704 2,509 4,947 1.45 1.63 1.55 1.55 1.63
Wisconsin (SPS-1) 209,239 375 958 1,705 2,662 5,326 1.53 1.64 1.55 1.64 1.76

136
Figure 5.36. Maximum moment range ratio (Fatigue Limit State I)

for simple-supported bridges at the midspan.

137
Ratio: “1/10000 Moment” / HL-93 Fatigue Moment

for continuous bridges at the middle support.

138
Ratio: “1/10000 Moment” / HL-93 Fatigue

for continuous bridges at 0.4 of the span length.

139
Figure 5.39. Fatigue damage ratio for proposed change (Fatigue

Limit State II) for simple-supported bridges at the midspan.

140

Limit State II) for continuous bridges at the middle support.

141

Limit State II) for continuous bridges at 0.4 of the span length.

142
Table 5.33. Maximum Moment Range Ratio (continued from page 133)
for Fatigue Limit State I It is assumed that the considered 15 WIM locations are rep-
resentative of the truck traffic in the United States. For the pur-
Bridge Type Span (ft) Mean Mean  1.5s CV
pose of further reliability analysis, it is recommended to assume
Simple-supported 30 1.60 1.90 0.13 that the mean fatigue load is equal to the mean for the 15 WIM
midspan
60 1.83 2.24 0.15 locations plus 1.5 standard deviations (1.5 s). The probability
of exceeding this value is about 5%, and as Figure 5.42 shows,
90 1.60 1.96 0.15
95% of sites in the United States are below this value. The
120 1.64 1.88 0.10
moment ratios corresponding to the mean plus 1.5 standard
200 1.70 2.15 0.18 deviations are also listed in Table 5.33 and Table 5.34.
Continuous middle 30 1.35 1.61 0.13 The statistical parameters were calculated for all consid-
support ered cases and span length.
60 1.81 2.13 0.12
90 1.92 2.18 0.09
120 1.97 2.17 0.07 5.4.9 Recommendations
200 2.27 2.47 0.06 Use of the proposed number of cycles of stress range per
Continuous 0.4 of 30 1.54 1.86 0.14 truck shown in Table 5.26 resulted in the relatively tightly clus-
the span length tered moment range ratios shown in Table 5.33 and Table 5.34
60 1.67 2.06 0.16
for the Fatigue II and Fatigue I limit states, respectively. As
90 1.60 1.92 0.13
with other live load recommendations in this report, the val-
120 1.65 1.97 0.13 ues to be considered for calibration are the moment ratios at
200 1.72 2.11 0.15 the “mean plus 1.5 standard deviations” and the CVs. For sim-
plicity, the recommended values for the calibration of the
fatigue limit states are further simplified into single values
Table 5.34. Proposed Fatigue Damage Ratio independent of span length as follows:
for Fatigue Limit State II
• For Fatigue I, use stress ranges (loads) based on 2.0 HL-93
Bridge Type Span (ft) Mean Mean  1.5s CV and a CV = 0.12.
• For Fatigue II, use stress ranges (loads) based on 0.80
Simple-supported 30 0.79 0.87 0.07
midspan HL-93 and a CV = 0.07.
60 0.78 0.86 0.06
90 0.73 0.81 0.07 The corresponding load factors are determined from
120 0.76 0.84 0.07 Monte Carlo simulation using the statistics of resistance
200 0.78 0.86 0.07 based on past laboratory testing, as summarized in Keating
Continuous middle 30 0.59 0.65 0.07
and Fisher (1986). The development of the load factors for
support steel and concrete components and details is explained in
60 0.74 0.82 0.07
Chapter 6.
90 0.69 0.77 0.07
120 0.71 0.78 0.06
Probability Density Function
200 0.79 0.87 0.07
Continuous 0.4 of 30 0.73 0.81 0.07
the span length
60 0.72 0.80 0.07
Probability
95% of the
90 0.68 0.75 0.07 population
120 0.72 0.79 0.06
200 0.76 0.84 0.07
Mean Mean + 1.5σ

Fatigue Load Ratio
Figure 5.42. Probability density function of the

national fatigue load.

143
5.5 Development of Overload directed onto this road, which undoubtedly accounted for the
(Service II) Parameters relatively large number of times the HL-93 was exceeded for
the various percentages indicated. The total number of times
WIM data also forms the basis for estimating how often a the various ratios of HL-93 were exceeded, excluding Florida
given design moment (or shear) is exceeded. Table 5.35 shows Route 29, is shown in the Table 5.35, as well as the average
the number of times the live load moment exceeded 100%, number per site. Most of the data were collected for a year, so
110%, 120%, and 130% of HL-93 for the 32 WIM sites. One that the lowest row in the table indicates the average number
of the sites, Florida Route 29, clearly has a unique traffic pat- of times each of the criteria was exceeded on an average site
tern. The Florida Department of Transportation explained during a year. This information was used to assess the signifi-
that truck traffic from several other highways was being cance of the Service II limit states in Chapter 6.
Table 5.35. Number of Times WIM Moments Exceeded Factored HL-93 Loadings
Moment
Ratio Truck/HL-93 >

–1.1 Ratio Truck/HL-93 >
–1.3
Site 30 ft 60 ft 90 ft 120 ft 200 ft 30 ft 60 ft 90 ft 120 ft 200 ft 30 ft 60 ft 90 ft 120 ft 200 ft
Arizona 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
(SPS-1)
Arizona 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0
(SPS-2)
Arkansas 2 7 3 0 0 0 3 0 0 0 0 0 0 0 0
(SPS-2)
Colorado 0 2 5 4 0 0 0 2 0 0 0 0 0 0 0
(SPS-2)
Delaware 36 33 22 11 0 10 22 10 1 0 1 11 1 0 0
(SPS-1)
Illinois 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0
(SPS-6)
Indiana 3 11 11 10 2 2 4 5 4 0 0 0 1 0 0
(SPS-6)
Kansas 16 33 35 31 2 7 16 17 7 0 6 7 6 0 0
(SPS-2)
Louisiana 44 6 12 14 7 26 6 7 7 0 6 6 5 4 0
(SPS-1)
Maine 4 4 5 2 0 0 4 2 0 0 0 2 0 0 0
(SPS-5)
Maryland 5 6 2 2 0 0 1 1 0 0 0 1 0 0 0
(SPS-5)
Minnesota 7 5 6 5 0 4 2 2 1 0 2 1 1 0 0
(SPS-5)
New Mexico 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0
(SPS-1)
New Mexico 3 1 1 2 0 2 0 0 0 0 0 0 0 0 0
(SPS-5)
Pennsylvania 32 22 17 14 1 13 17 13 1 0 3 13 2 0 0
(SPS-6)
Tennessee 53 4 4 0 0 5 1 0 0 0 1 0 0 0 0
(SPS-6)

144
Table 5.35. Number of Times WIM Moments Exceeded Factored HL-93 Loadings (continued)
Moment
Ratio Truck/HL-93 >

–1.3
Site 30 ft 60 ft 90 ft 120 ft 200 ft 30 ft 60 ft 90 ft 120 ft 200 ft 30 ft 60 ft 90 ft 120 ft 200 ft
Virginia 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0
(SPS-1)
Wisconsin 1 0 3 3 1 0 0 1 1 0 0 0 0 0 0
(SPS-1)
California 0 1 0 0 5 0 0 0 0 0 0 0 0 0 0
Antelope
EB
California 0 5 4 13 28 0 0 0 1 9 0 0 0 0 1
Antelope
WB
California 0 0 0 1 1 0 0 0 0 1 0 0 0 0 0
Bowman
California 1 31 50 51 15 0 6 24 19 0 0 0 4 1 0
LA-710 NB
California 1 17 45 48 14 0 3 18 19 0 0 0 1 1 0
LA-710 SB
California 0 4 16 46 140 0 0 1 2 32 0 0 0 0 2
Lodi
Florida 79 40 46 75 37 22 16 14 17 5 10 5 4 5 2
I-10
Florida 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
I-95
Florida 653 495 322 245 106 360 266 174 119 51 177 160 82 59 21
US-29
Mississippi 24 22 31 33 22 7 2 10 19 2 2 2 2 2 1
I-10
Mississippi 0 0 0 1 2 0 0 0 0 0 0 0 0 0 0
I-55UI
Mississippi 19 30 48 58 32 7 8 16 21 19 2 3 5 8 9
I-55R
Mississippi 0 0 2 1 0 0 0 0 0 0 0 0 0 0 0
US-49
Mississippi 0 0 1 2 1 0 0 1 1 0 0 0 0 0 0
US-61
Total 331 285 373 430 310 105 111 144 121 68 33 51 32 21 15
(without
Florida
US-29)
Average per 10.7 9.2 12.0 13.9 10.0 3.4 3.6 4.6 3.9 2.2 1.1 1.6 1.0 0.7 0.5
site per
year

Chapter 6
Calibration Results
6.1 Foundation Deformations, estimated settlement, is only one of many load factors. Further-
Service I: Lifetime more, the primary moments due to the dead and live loads
are much larger than the additional (secondary) moments
The geotechnical limit states for serviceability of a bridge struc- due to settlement. Because of these considerations and based
ture relate to foundation deformations. Within the context of on a consideration of the reversible and irreversible SLSs
foundation deformation, the geotechnical limit states can be for bridge superstructures described earlier in this report,
broadly categorized into vertical and horizontal deformations a target reliability index (bT) in the range of 0.50 to 1.00 for
for any foundation type (e.g., spread footings, driven piles, the calibration of load factor gSE for foundation deformation
drilled shafts, micropiles). Table 6.1 summarizes the various in the Service I limit state is used.
relevant articles in AASHTO LRFD (2012) that address vertical
(settlement) and horizontal deformations for various types of
structural foundations. 6.1.2 Calculation Models
This section describes procedures that can be used for
Evaluation of an SLS involves consideration of the deformation
calibrating service limit states (SLSs) to evaluate the effect of
aspects of the structure or members of the structure. The load
vertical or horizontal deformations of all structural foundation
deformation characteristics of the structure or its member
types such as footings, drilled shafts, and driven piles. The
are important to understand because resistance must now
procedure is demonstrated by using the case of immediate
be quantified as a function of the deformation. This section
settlements of spread footings, and the effect of foundation
first discusses the extension of the AASHTO LRFD frame-
deformations on bridge superstructures is discussed in the
work to incorporate the load deformation behavior, after
context of construction stages.
which a calibration framework for SLSs for foundation defor-
mations is presented. The proposed step-by-step procedure
6.1.1 Target Reliability Index for calibration is described in Section 6.1.2.5, which leads to
a load factor for deformations based on the target reliability
For strength limit states, reliability index values in the range of index that was discussed in Section 6.1.1. The proposed pro-
3.09 to 3.54 are used. Strength (or ultimate) limit states pertain cedure is demonstrated by an example for immediate settle-
to structural safety and the loss of load-carrying capacity. ments of spread footings by using various analytical methods
In contrast, SLSs are user-defined limiting conditions that affect in Section 6.1.3.
the function of the structure under expected service conditions.
Violation of SLSs occurs at loads much smaller than those for
strength limit states. As there is no danger of collapse if an 6.1.2.1 Incorporation of Load Effect
SLS is violated, a smaller value of target reliability index may Deformation (Q-d) Characteristics
be used for SLSs. In the case of foundation deformation, such in the AASHTO LRFD Framework
as settlement, the structural load effect is manifested in terms
The basic AASHTO LRFD framework in terms of distributions
of increased moments and potential cracking. The load effect
of load effects and resistances is shown in Figure 6.1, where
due to settlement relative to the load effect due to dead and live
loads would generally be small because in the Service I limit Q = load;
state the load factor gSE, which represents the uncertainty in Qmean = mean load;
145

146
Table 6.1. Summary of AASHTO LRFD (2012) Articles for Estimation of Vertical and Horizontal Deformation
of Structural Foundations
AASHTO LRFD Article Comment
10.6.2.4 Article 10.6.2.4 presents methods to estimate the settlement of spread footings. Settlement
Settlement Analyses for Spread Footings analysis is based on the elastic and semiempirical Hough (1959) method for immediate
settlement and the one-dimensional consolidation method for long-term settlement.
10.7.2.3 The procedures in these articles (10.7.2.3, 10.8.2.2, and 10.9.2.3) refer to settlement analysis
Settlement (related to driven pile groups) for an equivalent spread footing; see AASHTO LRFD (2012, Figure 10.7.2.3.1-1).
10.8.2.2
Settlement (related to drilled shaft groups)
10.9.2.3
Settlement (related to micropile groups)
10.7.2.4 Lateral analysis based on the p-y method and strain wedge method is included in AASHTO
Horizontal Pile Foundation Movement LRFD (2012) for estimating horizontal (lateral) deformations of deep foundations.
10.8.2.4
Horizontal Movement of Shaft and Shaft Groups
10.9.2.4
Horizontal Micropile Foundation Movement
Note: Section 11 (Abutments, Piers and Walls) Article 11.6.2 refers to the various articles noted in the left column of this table. Thus, the articles noted in this table also
apply to fill retaining walls and their foundations.
g = load factor; the evaluation of SLS, the basic AASHTO LRFD framework
R = resistance; shown in Figure 6.1 needs to be modified to include load effect
Rmean = mean resistance; deformation, or Q-d behavior. The Q-d behavior can be consid-
ϕ = resistance factor; ered as another dimension of the basic AASHTO LRFD frame-
Qn = nominal load; work, as shown in Figure 6.2, where
lQ = bias factor for load;
d = deformation;
f = frequency;
dS = deformation at nominal load effect (Qn);
Rn = nominal resistance; and
lR = bias factor for resistance.
Details of the AASHTO LRFD framework can be found in
Nowak and Collins (2013). Strength limit states were evaluated Qmean Rmean
by using this framework. Because determination of defor Qn Rn
f(R, Q)
mation is a necessary part of the evaluation of serviceability, for λQ λR
γQn φRn
Qmean Rmean
Qn Rn
f(R, Q)
Q R
Q, R
Qn Rn δS S
δF F
δN N
Mean curve
PDF of curve
Q, R δ
Figure 6.1. Basic AASHTO LRFD framework Figure 6.2. Incorporation of Q-d mechanism
for load effects and resistances. into the basic AASHTO LRFD framework.

147
Load Effect, Q
QN N
Nominal Resistance Level

QF
Strength Limit State

QS
SLS
Deformation, δ
O
δS δF δN
Figure 6.3. Significant points of interest on the mean Q-d curve.
dF = deformation at factored load effect [QF = g(Qn)]; and • The load effect deformation (Q-d) curves shown in Fig-
dN = deformation at load corresponding to nominal resis- ure 6.2 and Figure 6.3 represent the measured mean curves
tance (Rn). based on field measurements.
• Field measurements have upper and lower bounds with
Although Q-d curves can have many shapes, for illustration
purposes, a strain-hardening curve is shown in Figure 6.2. For respect to the mean of the measured data. These bounds are
discussion purposes, the mean Q-d curve is shown, and the shown schematically in Figure 6.4 and also in Figure 6.2
spread of the Q-d data about the mean curve is represented sche- and Figure 6.3 through a PDF. Although PDFs for nor-
matically by a probability distribution function (PDF) that is mal distributions are shown, the spread of the data about
discussed later in this chapter. The various relevant load effect the mean may be represented by normal or nonnormal
and deformation quantities shown in the Q-d space in Figure 6.2 distributions, as appropriate. In general, the spread of
are shown in the regular first quadrant of the two-dimensional the data around the mean curve increases with increasing
plot in Figure 6.3. Note that the nominal resistance is equated deformations.
to a load effect that would correspond to this resistance. • Many theoretical methods are available to predict load effect
Figure 6.2 combines different aspects of material behavior deformation behavior. The theoretical models may predict a
that cover both load effects and resistances. It is important to stiffer or softer material response compared with the actual
understand the interrelationships among the various param- response. For the purpose of discussion, a softer material
eters displayed on the curves. To that end the following points behavior is shown in Figure 6.5. Because the bias factor is
are made: defined as the ratio of measured mean to predicted values,
Q
Q
Measured
C
Upper Bound
C
Mean
Measured Mean
Lower Bound
Theoretical
D
Prediction
δ δ
Figure 6.4. Significant points of interest Figure 6.5. Relationship of measured

on the Q-d curve. mean with theoretical prediction.

148
the bias factor for deformations (ld) will vary over the full modeling the uncertainty in predicted deformations (dP) are
range of the Q-d curve. available, the same is not true for tolerable deformations (dT).
Some attempts have been made (e.g., Zhang and Ng 2005) to
evaluate the distribution of tolerable deformations, but from
6.1.2.2 Consideration of Bias Factor
a geotechnical viewpoint, it may not be possible to obtain a
in Calibration of SLS
PDF for tolerable deformation that is applicable to the vari-
A varying bias factor along the Q-d curve, although a reality, ous structural SLSs discussed in other sections of this report.
can be cumbersome to handle in the calibration process. This is largely because it is virtually impossible to identify a
However, the problem is made easier by realizing that for consistent tolerable deformation across all elements of a
calibration of SLS the load effects between Points O and S, as structure. Many variables can affect the value of tolerable
shown in Figure 6.3, are of primary interest. Point S represents deformation for a given element. To bypass these difficulties,
the service load effects, and the deformation corresponding to a single deterministic value of dT is often used for comparison
this point is of primary interest. As the bias factor will generally against the potential spread of data for dP. In practical terms,
increase with increasing deformations, the value of the bias a bridge engineer often assumes a deterministic tolerable
factor at Point S will be the maximum between Points O and S. deformation that would limit deformations according to the
Thus, use of the bias factor at Point S will be conservative. type of bridge structure being designed. In this case, the con-
In this context, the bias factor at Point S is most relevant ventional calibration processes, such as the Monte Carlo pro-
and, at a minimum, field data under full service loads are of cedure, would not be necessary as there would be a PDF for
importance in geotechnical SLS calibrations. The data most load (Q), but a deterministic value for resistance (R). To use
particularly needed for SLS evaluations are the full range of Monte Carlo in this situation an arbitrarily small value of
incremental loads and deformations measured on in-service standard deviation or coefficient of variation (CV) would
structures from the beginning of construction of the first have to be used. Although theoretically possible, this process
element (e.g., the foundation) to the completion of the road- could lead to spurious results. Thus, an alternative approach
way and beyond. Such data will help in the evaluation of the to calibration of SLSs for geotechnical features is necessary.
variability in predicted deformations for structural, as well as When a deterministic value for dT is used, then by using
geotechnical, features. At present, these types of data are not Figure 6.1 as the basis, the resistance PDF is reduced to a single
routinely available; however, programs such as the Federal value while the load effect PDF can be used to represent the
Highway Administration’s (FHWA) Long-Term Bridge Perfor- predicted deformations. This modified treatment for defor-
mance Program may offer a good avenue to collect such data. mations is shown in Figure 6.6. In this approach, the probabil-
ity of exceedance (Pe) for the predicted deformations to exceed
the tolerable deformation is given by the area of the overlap
6.1.2.3 Application of Q-d Curves in the
of the two curves (the shaded zone shown in Figure 6.6). As
LRFD Framework
the goal is to prevent serviceability-related problems, Pe can
The AASHTO LRFD calibration of the strength limit state
was performed by using the general concepts in Figure 6.1.
This approach presumes that statistical data are available to δP δT
quantify the spread of the load effects and resistances. In the f(R, Q)
context of deformations, tolerable deformations (dT) can be
considered as resistances, and predicted deformations (dP) can
be considered as loads. Thus, a limit state function (g) can be
written as shown by Equation 6.1:
g = δT − δ P (6.1) Probability of
Exceedance, Pe
Once the deformations are expressed in the form of a limit

state, probabilistic calibration processes similar to those used Q, R
for the strength limit state can be used. For strength limit Figure 6.6. Relationship of deterministic value
states, Monte Carlo analysis is often used for calibrations. of tolerable deformation (dT) and a PDF for
One of the assumptions of the Monte Carlo procedure is that predicted deformation (dP). Q = load effect;
PDFs for both the load (Q) and resistance (R) are available. R = resistance; dP = predicted deformations
However, for geotechnical SLS calibration, there are practical (load effect); and dT = deterministic value of
limitations to this approach. Although the statistical data for tolerable deformation (resistance).

149
be selected on the basis of the acceptable value of the target designer will perform the necessary tests and, as appropriate,
reliability index (bT). The ratio dT/dP can be thought of as a load implement the necessary mitigation measures to minimize
factor for deformations for a given Pe, corresponding to bT. the inevitable effects of corrosion or degradation on the foun-
The PDF for the predicted deformations shown in Figure 6.6 dation and wall elements and the structures these elements
is obtained from the data at Point S shown in Figure 6.2 and support. The most common approach is to estimate the rate
Figure 6.3. This is where the concept of the Q-d curve fits into of corrosion or degradation over the design life of the struc-
the framework to calibrate the SLS on the basis of on defor- ture and provide additional sectional or strength properties
mations. Thus, any model that can predict a Q-d curve can be (or both) that will permit the structure to perform within its
used in the conventional AASHTO LRFD framework as long strength and serviceability requirements. For example, metal
as the data at Point S corresponding to SLS load effects are elements are often provided additional section based on the
available through field measurements. The effect of material anticipated loss of metal over the design life of the structure.
brittleness (or ductility) can now be introduced in the AASHTO Concrete deterioration due to sulfate attack is often mitigated
LRFD framework through the use of an appropriate Q-d model. by use of an appropriate type of cement.
Examples of Q-d models are stress–strain curves, vertical load-
settlement curves for foundations, p-y (lateral load–lateral
6.1.2.5 Determination of Load Factor
displacement) curves for laterally loaded piles, shear force–
for Deformations
shear strain curves, moment–curvature curves, and so forth.
The proposed framework can incorporate any Q-d model and The concept presented in Figure 6.6 assumes that the designer
is therefore a general framework that is applicable to structural has unique (fixed) values of tolerable deformation (dT) and
or geotechnical aspects. predicted deformation (dP). However, these values are func-
tions of many parameters for a given element and the mode
of deformation being evaluated. Thus, it is more practical to
6.1.2.4 Deterioration of Foundation
express the load factor for deformation as a function of the
and Wall Elements
value of dP. The load factor is more conveniently determined
Most, if not all, foundation elements are buried in geomaterials. by using an alternative form of the concept, as shown in
This is also true for most earth-retaining structures. Thus, the Figure 6.7, in which the cumulative distribution function
long-term performance of the foundation and wall elements (CDF) is used instead of the PDF. In this concept it is more
can be affected by the corrosion or degradation potential of the convenient to use the data based on the inverse of the bias
geomaterials. The term corrosion applies to metal components, factor because the predicted (calculated) deformation is
and degradation applies to nonmetal components such as poly- plotted on the x-axis. The format shown in Figure 6.7 is used
meric soil reinforcements in mechanically stabilized earth as follows:
(MSE) walls.
If the geomaterials have significant corrosion or degradation 1. Obtain data for predicted (dP) and measured (dM) deforma-
potential, then the sectional properties of the foundation and tions for the deformation mode of interest (e.g., immediate
wall elements will deteriorate by reduction in the section or settlement of spread footings). Recognize that the value of
loss of strength, or both. The AASHTO LRFD Specifications dM can be considered as resistance and equivalent to the
clearly recognizes this mode of deterioration and provides tolerable settlement (dT).
definitive guidelines. For example, Articles 10.7.5 and 10.9.5 of 2. Modify the data to be expressed in terms of the ratio dP/dT.
Section 10 (Foundations) provide guidelines for evaluation of In geotechnical literature (e.g., Tan and Duncan 1991) this
corrosion and deterioration of driven piles and micropiles, ratio is often referred to as accuracy. Label this ratio as X.
respectively. Similarly, Section 11 (Abutments, Piers, and Walls) X is a random variable that can now be modeled by an
provides guidance in Article 11.8.7 for nongravity cantilevered appropriate PDF. Develop the appropriate statistics, and
walls, Article 11.9.7 for anchored walls, and Articles 11.10.2.3.3 select a suitable distribution function. Express the data in
and 11.10.6.4 for MSE walls. Supplementary guidance can be terms of a CDF.
found in Elias et al. (2009) and Fishman and Withiam (2011). 3. As shown in Figure 6.7, plot a family of CDF curves for a
The AASHTO, Elias et al., and Fishman and Withiam docu- range of values of tolerable deformation (e.g., dT1 > dT2 > dT3)
ments cross reference various publications that discuss the that permits the determination of values of the probability
corrosion or degradation potential of geomaterials. of exceedance (Pe) for a range of dP. The CDFs are generated
In general, the AASHTO articles and other documents cited by multiplying the CDF for accuracy (i.e., X = dP/dT) or by
above provide guidance for testing frequencies and protocols selected values of tolerable deformations (dT1, dT2, dT3).
to evaluate the corrosion or degradation potential of various The plot shown in Figure 6.7 is referred to as a probability
geomaterials. It is assumed that the foundation and wall exceedance chart (PEC).

150
Probability of Exceedance, Pe
δ T1
δ T2
δ T3
δT
PeT
Deformation Load Factor, γ = δ T /δ P

δP
Predicted (Calculated) Deformation, δP

Figure 6.7. PEC for evaluation of load factor for a target probability
of exceedance (PeT) at the applicable SLS combination.
4. Select the design value of probability of exceedance (PeT) For SLS calibration for foundation deformation, the limit
corresponding to the target reliability index (beT), and state g expressed as a ratio is more appropriate, as given by
determine the value of dT for a given value of dP, as shown Equation 6.3:
in Figure 6.7.
5. Compute the value of the deformation load factor (g = dT/dP), g = δ P δT (6.3)
as shown in Figure 6.7.
The benefit of this approach is that once the designer 6.1.3.2 Step 2: Identify and Select Representative
computes (predicts) a deformation for any given deformation Structural Types and Design Cases
mechanism, then the designer simply multiplies the computed In general, the vertical and lateral deformations for all structural
value by the deformation load factor corresponding to that foundation types (e.g., footings, drilled shafts, and driven piles)
value of deformation and uses the factored value for evaluation can be calibrated using the process described in this example.
at the applicable SLS load combination. This concept is valid For the purpose of demonstration of the calibration process,
whether structural or geotechnical deformation mechanisms immediate vertical settlement of spread footings is used as a
are evaluated. This concept is demonstrated in the next sec- design case.
tion, in which immediate settlements for spread footings are
evaluated.
6.1.3.3 Step 3: Determine Load and Resistance
Parameters for the Selected Design Cases
6.1.3 Calibration Results
The load and resistance parameters for the selected design
The proposed procedure for calibration described in Sec- case of immediate vertical settlement of spread footings are as
tion 6.1.2.5 is demonstrated by an example for immediate follows. Load is predicted (or calculated) immediate vertical
settlement of spread footings. The calibration results are settlement (dP) and resistance is tolerable (or limiting or mea-
presented in a step-by-step format that is generally consistent sured) immediate vertical settlement (dT).
with other similar results presented in this report.
6.1.3.4 Step 4: Develop Statistical Models
6.1.3.1 Step 1: Formulate the Limit State Function for Load and Resistance
and Identify Basic Variables Table 6.2 shows a data set for spread footings based on vertical
In the context of deformations, tolerable deformations (dT) can settlements of footings measured at 20 footings for 10 instru-
be considered as resistances, and predicted deformations (dP) mented bridges in the northeastern United States (Gifford
can be considered as loads. Thus, a limit state function (g) can et al. 1987). The bridges included five simple-span and five
be given by Equation 6.2 (first introduced as Equation 6.1): continuous-beam structures. Each of the footing designations
in Table 6.2 represents a footing supporting a single sub
g = δT − δ P (6.2) structure unit (abutment or pier). Four of the instrumented

151
Table 6.2. Data for Measured Settlement (dM ) and Calculated Settlement ( dP)
Shown in Figure 6.8
Calculated Settlement (in.)

Measured
Settlement Schmertmann D’Appolonia Peck and Burland and
Site (in.) et al. (1978) Hough (1959) et al. (1968) Bazaraa (1969) Burbridge (1984)
S1 0.35 0.79 0.75 0.65 0.29 0.30

S2 0.67 1.85 0.94 0.39 0.16 0.12
S3 0.94 0.86 1.21 0.30 0.19 0.13
S4 0.76 0.46 1.46 0.58 0.36 0.39
S5 0.61 0.30 0.98 0.38 0.42 0.57
S6 0.42 0.52 0.61 0.50 0.17 0.34
S7 0.61 0.18 0.40 0.19 0.30 0.19
S8 0.28 0.30 0.60 0.26 0.16 0.14
S9 0.26 0.18 0.53 0.20 0.16 0.11
S10 0.29 0.29 0.40 0.23 0.16 0.09
S11 0.25 0.36 0.47 0.29 0.16 0.06
S14 0.46 0.41 1.27 0.57 0.50 0.40
S15 0.34 1.57 1.46 0.74 1.36 1.61
S16 0.23 0.26 0.74 0.39 0.17 0.17
S17 0.44 0.40 0.82 0.46 0.28 0.23
S20 0.64 1.21 0.33 0.10 0.07 0.65
S21 0.46 0.29 1.05 0.49 0.21 0.54
S22 0.66 0.54 0.84 0.56 0.52 0.31
S23 0.61 1.02 1.39 0.61 0.34 0.64
S24 0.28 0.64 0.99 0.59 0.33 0.44
Note: Gifford et al. (1987), the source for the table, note that data for Footings S12, S13, and S18 were not included because
construction problems at these sites resulted in disturbance of the subgrade soils, and short-term settlement was increased.
Data for Footing S19 appear to be anomalous and have been excluded in this table and Figure 6.8.
bridges were single-span structures. Two two-span and three Figure 6.8 shows a plot of the data in Table 6.2 and the
four-span bridges were also monitored, in addition to a single spread of the data about the 1:1 diagonal line, which defines the
five-span structure. Nine of the structures were designed case for which the predicted and measured values are equal.
to carry highway traffic, and one four-span bridge carried Such a plot provides a visual frame of reference to judge the
railroad traffic across an Interstate highway. Additional infor- accuracy of the prediction method. If the data points align
mation on the instrumentation and data collection at the closely with the 1:1 diagonal line, then the predictions based
10 bridges can be found in Gifford et al. (1987). Similar and on the analytical method being evaluated are close to the
more extensive databases are available for spread footings measured values and are more accurate than the case for which
(e.g., Sargand et al. 1999; Sargand and Masada 2006; Akbas the data points do not align closely with the 1:1 diagonal line.
and Kulhawy 2009; Samtani et al. 2010) and other foundation In the geotechnical literature (e.g., Tan and Duncan 1991),
types, such as driven piles and drilled shafts. Similar data- accuracy is defined as the mean value of the ratio of the pre-
bases are also available for lateral load behavior. However, for dicted (calculated) to the measured settlements. Table 6.3
this report, the calibration concepts for SLS evaluations are shows the values of accuracy (denoted by X, where X = dP/dM)
demonstrated by use of the limited data set for spread footings for each footing based on the data in Table 6.2.
shown in Table 6.2. Although spread footings are used as an As noted in Step 3 of the calibration process, the value
example, all the concepts discussed here are applicable to other of dM can be considered as the resistance and equivalent to
foundation types and deformation patterns. the tolerable settlement (dT). The accuracy (i.e., X = dP/dM

152
Figure 6.8. Comparison of measured and calculated (predicted) settlements based on

service load data in Table 6.2.
Table 6.3. Accuracy ( X = dP/dM ) Values Based on Data Shown in Table 6.2
Schmertmann D’Appolonia Peck and Burland and

Site et al. (1978) Hough (1959) et al. (1968) Bazaraa (1969) Burbridge (1984)
S1 2.257 2.143 1.857 0.829 0.857

S2 2.761 1.403 0.582 0.239 0.179
S3 0.915 1.287 0.319 0.202 0.138
S4 0.605 1.921 0.763 0.474 0.513
S5 0.492 1.607 0.623 0.689 0.934
S6 1.238 1.452 1.190 0.405 0.810
S7 0.295 0.656 0.311 0.492 0.311
S8 1.071 2.143 0.929 0.571 0.500
S9 0.692 2.038 0.769 0.615 0.423
S10 1.000 1.379 0.793 0.552 0.310
S11 1.440 1.880 1.160 0.640 0.240
S14 0.891 2.761 1.239 1.087 0.870
S15 4.618 4.294 2.176 4.000 4.735
S16 1.130 3.217 1.696 0.739 0.739
S17 0.909 1.864 1.045 0.636 0.523
S20 1.891 1.641 0.766 0.328 0.844
S21 0.630 1.826 1.217 1.130 0.674
S22 0.818 2.106 0.924 0.515 0.970
S23 1.672 1.623 0.967 0.541 0.721
S24 2.286 2.179 1.286 0.893 1.286

153
[or dP/dT]) is a random variable that can now be modeled by discussed above. As Figure 6.9b to Figure 6.13b show, the data
an appropriate PDF. To develop an appropriate PDF, an eval- points based on Table 6.3 do not plot on the straight line, which
uation of the data spread around the mean value is needed. confirms the observation of nonnormal distributions made
This evaluation involves statistical analysis and the develop- on the basis of the histograms in Figure 6.9a to Figure 6.13a.
ment of histograms. By using procedures described in Allen et al. (2005), a log-
Table 6.4 presents the arithmetic mean (µ) and standard normal distribution was used to evaluate the nonnormal
deviation (s) values for various methods. AASHTO LRFD data. As seen in Figure 6.9b to Figure 6.13b, the lognormal
recommends the use of Hough’s (1959) method, which has the distribution fits the data better than the normal distribution.
smallest CV, for calculating immediate settlement. However, The lognormal distribution, which is valid between values
the Hough method is conservative by a factor of approximately of 0 and +∞, is used in these figures because (1) immediate
two (see mean value in Table 6.4), which leads to an unnecessary settlement cannot have negative values, and (2) lognormal
use of deep foundations instead of spread footings. FHWA PDFs have been used in the past for nonnormal distributions
(Samtani and Nowatzki 2006; Samtani et al. 2010) recommends during calibration of the strength limit state for geotechnical,
the use of the method proposed by Schmertmann et al. (1978) as well as structural, features in the AASHTO LRFD frame-
because it is a rational method that considers not only the work. For SLS, a PDF with an upper bound and lower bound
applied stress and its associated strain influence distribution (e.g., beta distribution) instead of open tail(s) (e.g., normal
with depth for various footing shapes, but also the elastic prop- or lognormal distribution) may be more appropriate because
erties of the foundation soils, even if they are layered. the conditions represented by an open-tail PDF are not physi-
Even though FHWA and AASHTO recommend the cally possible when one considers foundation deformations.
Schmertmann et al. (1978) and Hough (1959) methods, respec- As noted, the lognormal PDF is used here to be consistent
tively, all the methods noted in Table 6.2 to Table 6.4 were with the PDFs that have been used in LRFD calibration pro-
evaluated as part of the calibration process because some agen- cesses to date. Guidance for the selection of an appropriate
cies may use one of the remaining three methods as a result PDF and development of the distribution parameters shown
of past successful local practice. in Table 6.5 is provided in Nowak and Collins (2013) or other
As noted earlier, accuracy (X = dP/dM) is a random variable similar books that deal with probabilistic methods.
that can be modeled by an appropriate PDF. The data for X in Values of the lognormal mean and lognormal standard devi-
Table 6.3 were used to develop histograms. ation are needed to use the lognormal PDF. These values can be
The histograms of the data for X taken from Columns 2 to 6 obtained by using correlations with the mean and standard
of Table 6.3 are shown in Figure 6.9a to Figure 6.13a, respectively. deviation values for normal distribution or calculated directly
None of the histograms resembles a classic bell shape character- from the natural logarithm (ln) of the values of the data points.
istic of normally distributed data. Thus, nonnormal distribu- Table 6.5 presents the values for correlated mean (µLNC) and
tions would be more appropriate in these cases. To evaluate the correlated standard deviation (sLNC). Table 6.6 shows the log-
deviation of the data from a classic normal PDF, the data for the normal of accuracy values of data in Table 6.3, and Table 6.7
value of X in Table 6.3 were plotted against the standard normal presents the values for arithmetic mean (µLNA) and arithmetic
variable (z) to generate CDFs, as shown in Figure 6.9b to Fig- standard deviation (sLNA) based on the ln(X) values in Table 6.6.
ure 6.13b. See Allen et al. (2005, Chapter 5) for a definition of z The correlated and arithmetic values of the mean (µLNC and
and procedures to develop the lower graphs (b) in Figures 6.9– µLNA, respectively) and standard deviation (sLNC and sLNA,
6.13. The beneficial attributes of this probability plot are (text continues on page 160)
Table 6.4. Statistics of Accuracy ( X) Values Based on Data Shown in Table 6.3

Statistic et al. (1978) Hough (1959) et al. (1968) Bazaraa (1969) Burbridge (1984)
Count 20 20 20 20 20
Minimum 0.295 0.656 0.311 0.202 0.138
Maximum 4.618 4.294 2.176 4.000 4.735
µ 1.381 1.971 1.031 0.779 0.829
s 1.006 0.769 0.476 0.796 0.968
CV 0.729 0.390 0.462 1.022 1.168
Note: CV = s/µ.

154
7
Frequency of Occurrence
0
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Accuracy, X (Predicted/Measured)
(a)
Accuracy Data for Schmertmann et al. (1978) - Full Data Fit

4.0
3.5
3.0
2.5
Standard Normal Variable, z
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Data Points Predicted LN Fitted from Normal Statistics Predicted Normal Distribution
(b)
Figure 6.9. Schmertmann et al. (1978) method: (a) histograms

for accuracy ( X) and (b) plot of standard normal variable (z) as
a function of X.

155
6
0
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
(a)
Accuracy Data for Hough (1959) - Full Data Fit

4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
(b)
Figure 6.10. Hough (1959) method: (a) histograms for accuracy ( X)

and (b) plot of standard normal variable (z) as a function of X.

156
10
9
8
7
6
5
4
3
2
1
0
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
(a)
Accuracy Data for D'Appolonia et al. (1968) - Full Data Fit

4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
(b)
Figure 6.11. D’Appolonia et al. (1968) method: (a) histograms

for accuracy (X) and (b) plot of standard normal variable (z) as
a function of X.

157
12
10
0
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
(a)
Accuracy Data for Peck and Bazaraa (1969) - Full Data Fit
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
(b)
Figure 6.12. Peck and Bazaraa (1969) method: (a) histograms

a function of X.

158
12
10
0
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
(a)
Accuracy Data for Burland and Burbridge (1984) - Full Data Fit
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
(b)
Figure 6.13. Burland and Burbridge (1984) method: (a) histograms

a function of X.

159
Table 6.5. Correlated Statistics of Accuracy ( X) for Lognormal PDFs

µLNC 0.1095 0.6076 -0.0665 -0.6078 -0.6177

sLNC 0.6528 0.3766 0.4398 0.8459 0.9274
Note: The µLNC and sLNC values for lognormal distribution were calculated from the normal (arithmetic) mean and
standard deviation values in Table 6.4, respectively, by using the following equations based on idealized normal
and lognormal PDFs: µLNC = ln(µ) - 0.50(sLNC)2; and sLNC = [ln{(s/µ)2 + 1}]0.5.
Table 6.6. Lognormal of Accuracy Values [ln(X)] Based on Data

Shown in Table 6.3

Site et al. (1978) Hough (1959) et al. (1968) Bazaraa (1969) Burbridge (1984)
S1 0.8141 0.7621 0.6190 -0.1881 -0.1542

S2 1.0157 0.3386 -0.5411 -1.4321 -1.7198
S3 -0.0889 0.2525 -1.1421 -1.5989 -1.9783
S4 -0.5021 0.6529 -0.2703 -0.7472 -0.6672
S5 -0.7097 0.4741 -0.4733 -0.3732 -0.0678
S6 0.2136 0.3732 0.1744 -0.9045 -0.2113
S7 -1.2205 -0.4220 -1.1664 -0.7097 -1.1664
S8 0.0690 0.7621 -0.0741 -0.5596 -0.6931
S9 -0.3677 0.7122 -0.2624 -0.4855 -0.8602
S10 0.0000 0.3216 -0.2318 -0.5947 -1.1701
S11 0.3646 0.6313 0.1484 -0.4463 -1.4271
S14 -0.1151 1.0155 0.2144 0.0834 -0.1398
S15 1.5299 1.4572 0.7777 1.3863 1.5550
S16 0.1226 1.1686 0.5281 -0.3023 -0.3023
S17 -0.0953 0.6225 0.0445 -0.4520 -0.6487
S20 0.6369 0.4951 -0.2671 -1.1144 -0.1699
S21 -0.4613 0.6022 0.1967 0.1226 -0.3947
S22 -0.2007 0.7448 -0.0788 -0.6633 -0.0308
S23 0.5141 0.4842 -0.0333 -0.6144 -0.3267
S24 0.8267 0.7787 0.2513 -0.1133 0.2513
Table 6.7. Statistics of ln( X) Values Based on Data Shown in Table 6.6

Count 20 20 20 20 20
Minimum -1.2205 -0.4220 -1.1664 -1.5989 -1.9783
Maximum 1.5299 1.4572 0.7777 1.3863 1.5550
µLNA 0.1173 0.6114 -0.0793 -0.4854 -0.5161
sLNA 0.6479 0.3807 0.5029 0.6226 0.7731
Note: µLNA = arithmetic mean of ln(X) values; sLNA = arithmetic standard deviation of ln(X) values.

160
(continued from page 153)

respectively) for lognormal distributions are not equal. This
is because the correlated values were based on derivations for
an idealized lognormal distribution and not a sample dis-
tribution from actual data, which may not necessarily fit an
idealized lognormal distribution. In contrast, the arithmetic
values were obtained by taking the arithmetic mean and stan- C B
D F
dard deviation directly from the ln(X) value of each data point
noted in Columns 2 to 6 of Table 6.3.
It is important to use the appropriate values of mean and A E
standard deviation based on the syntax for a lognormal dis-
tribution function used by a particular computational program.
For example, if one is using the @RISK program by Palisade
Corporation, then the RISKLOGNORM function in that pro-
gram is based on arithmetic values (µ and s) of the normal Figure 6.15. PEC for the Schmertmann et al. (1978)
method.
distribution. In contrast, the Microsoft Excel LOGNORMDIST
(or LOGNORM.DIST) function uses the arithmetic mean
(µLNA) and standard deviation (sLNA) values of ln(X). Use of
improper values of mean and standard deviation can lead to by drawing line AB, finding the intersection of the line with the
drastically different results. This issue is of critical importance curve for 1 in., drawing line BC, and reading the value from the
because calibration in this report, as mentioned earlier, is based ordinate of the PEC in Figure 6.15. Four additional curves for
on Microsoft Excel. settlements of 1.5, 2, 2.5, and 3 in. are shown in Figure 6.15.
Figure 6.14 shows the CDFs for accuracy (X) for various Using the procedure demonstrated for the example above
analytical methods based on the LOGNORM.DIST function (see dashed arrows in Figure 6.15), if the predicted (calculated)
in the 2010 version of Microsoft Excel using the µLNA and sLNA value is 0.85 in., then the probability of the measured value
values noted in Table 6.7. These CDFs can now be used to being greater than 1.5, 2, 2.5, and 3 in. is approximately 14%,
develop the PEC discussed in Section 6.1.2 for various analyti- 6%, 3%, and 2%, respectively.
cal methods. A load factor for settlement (gSE) can be determined using
Figure 6.15 shows the PEC for the method by Schmertmann the procedure in Section 6.1.2.5. For example, assume the
et al. (1978). The probability of exceedance corresponding to predicted settlement is 1 in. To determine the value of gSE for
a given predicted settlement can now be readily determined. a 25% target probability of exceedance (PeT), draw a horizontal
For example, assume that the geotechnical engineer has pre- line from Point D on the ordinate corresponding to a value of
dicted a settlement of 0.85 in. The probability of exceedance 25%. Next, draw a vertical line from Point E on the abscissa
of 1 in. in this case is approximately 32%. This can be found corresponding to a value of 1 in. Locate the point of inter
section, F, which lies between the curves for 1 in. and 1.5 in.
Interpolating between the two curves leads to a value of approx-
imately 1.35 in. Based on the definition of gSE noted above, the
value of gSE is equal to 1.35 in./1.0 in., or 1.35.
PECs for other analytical methods noted in Figure 6.14
are given in Figure 6.16 to Figure 6.19. Those PECs can be
used in a similar manner as demonstrated for the PEC for the
Schmertmann et al. (1978) method.
A PEC chart is essentially a representation of the CDF of
accuracy, or X. Similar charts are referred to as probabilistic
design charts by Das and Sivakugan (2007) and Sivakugan and
Johnson (2002, 2004) and artificial neural network charts
by Shahin et al. (2002) and Musso and Provenzano (2003).
Although not specifically in chart format, similar concepts
are presented in Tan and Duncan (1991) and Duncan (2000).
Figure 6.14. CDFs for various analytical methods The specific format of PEC that is developed and used here is
for estimation of immediate settlement of spread amenable to correlation to the AASHTO LRFD–based concept
footings. of target reliability index, as explained in Step 5.

161
Figure 6.16. PEC for the Hough (1959) method. Figure 6.19. PEC for the Burland and Burbridge
(1984) method.
6.1.3.5 Step 5: Develop Reliability

Analysis Procedure
The estimation of load factor for settlement (gSE) in terms of
probability of exceedance (Pe) was demonstrated in the previ-
ous step. In the AASHTO LRFD framework, calibrations are
expressed in terms of a reliability index (b). b can be expressed
in terms of Pe of a predicted value by using Equation 6.4,
which applies to normally distributed data. As observed
from Step 4, lognormal distributions have been used. Fur-
thermore, the CV values noted in Table 6.4 are rather large.
For a normal random variable, the relationship between b
and Pe depends only on CV (i.e., one parameter), but for a
Figure 6.17. PEC for the D’Appolonia et al. lognormal distribution, it depends on the mean and stan-
(1968) method. dard deviation or the mean and CV (i.e., two parameters).
Therefore, strictly speaking, b should be based on a lognor-
mal function. However, for b <2.0 there is not a significant
practical difference in the Pe values for data that are normally
or lognormally distributed for the wide range of CVs noted
in Table 6.4. An assumption of a normal distribution is gen-
erally conservative in the sense that for a given b it gives a
larger Pe compared with a lognormal distribution. Further-
more, conventionally the normal distribution has been
assumed for strength limit states in AASHTO LRFD (as well
as other international codes), which has b values larger than
2.0. The key consideration is that the type of distribution is
not as important as being consistent and not mixing different
distributions while comparing b values. When these various
issues are taken into account, the Microsoft Excel formula
that assumes normally distributed data is considered to be
acceptable for SLS calibrations.
Figure 6.18. PEC for the Peck and Bazaraa (1969)
Table 6.8 and Figure 6.20 were generated by using Equa-
method.
tion 6.4:
β = NORMSINV (1 − Pe ) (6.4)

162
Table 6.8. Values of b and Corresponding Pe bridge should be performed using a reliability index of 0.50.
Based on Normally Distributed Data What is the value of gSE and the tolerable settlement that the
bridge designer should use?
b Pe (%) b Pe (%) b Pe (%) b Pe (%)
The load factor (gSE) is a function of the probability of
2.00 2.28 1.50 6.68 1.00 15.87 0.50 30.85 exceedance (Pe) of the foundation deformation under consid-
1.95 2.56 1.45 7.35 0.95 17.11 0.45 32.64 eration, which in this example is the immediate settlement of
spread footings calculated by using the analytical method
1.90 2.87 1.40 8.08 0.90 18.41 0.40 34.46
of Schmertmann et al. (1978). By using either Equation 6.4 or
1.85 3.22 1.35 8.85 0.85 19.77 0.35 36.32
Table 6.8, a value of Pe ≈ 0.3085 (or 30.85%) is obtained for
1.80 3.59 1.30 9.68 0.80 21.19 0.30 38.21 b = 0.50.
1.75 4.01 1.25 10.56 0.75 22.66 0.25 40.13 Equation 6.5 is the formula used in Microsoft Excel to deter-
1.70 4.46 1.20 11.51 0.70 24.20 0.20 42.07 mine a value of accuracy (X) in terms of Pe, the mean value
(µLNA), and the standard deviation (sLNA) of the lognormal dis-
1.65 4.95 1.15 12.51 0.65 25.78 0.15 44.04
tribution function as computed in Step 4. The value of X repre-
1.60 5.48 1.10 13.57 0.60 27.43 0.10 46.02 sents the probability of the accuracy value (dP/dT) being less
1.55 6.06 1.05 14.69 0.55 29.12 0.05 48.01 than a specified value.
0.00 50.00
Note: Linear interpolation may be used as an approximation for Pe = LOGNORMDIST (X, µ LNA , σ LNA ) (6.5)
intermediate values.
From Table 6.7, for the Schmertmann et al. (1978) method,

µLNA = 0.1173 and sLNA = 0.6479. The goal is to determine the
The correlation between b and Pe can now be used to value of X that gives Pe = 0.3085. Thus, for this example, the
rephrase the earlier discussion with respect to Figure 6.15. In expression for Pe can be written as shown by Equation 6.6:
that discussion, as an example, it was assumed that the geo-
technical engineer has predicted a settlement of 0.85 in. From Pe = LOGNORMDIST(X, 0.1173, 0.6479) = 0.3085 or 30.85%
Figure 6.15, it was determined that the probability of exceed- (6.6)
ance of 1, 1.5, 2, 2.5, and 3 in. was approximately 32%, 14%,
6%, 3%, and 2%, respectively. Using Table 6.8 (or Figure 6.20
Using Goal Seek in Microsoft Excel, X (i.e., dP/dT) ≈ 0.813.
or Equation 6.3), the results can now be expressed in terms of
Note that in the 2010 version of Microsoft Excel, another
reliability index values. Thus, it can be stated that if the pre-
function, LOGNORM.DIST, can also be used. In this case,
dicted settlement is 0.85 in., then the assumption of tolerable
the same result (X ≈ 0.813) is obtained by using the follow-
settlement values of 1, 1.5, 2, 2.5, and 3 in. means a reliability
ing syntax and using the Goal Seek function to determine X
index of approximately 0.45, 1.10, 1.55, 1.90, and >2.00,
(TRUE indicates the use of a CDF):
respectively.
In the example, the geotechnical engineer has predicted
Pe = LOGNORM.DIST(X, 0.1173, 0.6479, TRUE) = 0.3085.
settlement dP = 0.85 in. by using the Schmertmann et al. (1978)
method. The owner has specified that the SLS design for the
In the context of the AASHTO LRFD framework, the load
factor (gSE) is the reciprocal of X. Thus, for immediate settle-
Reliability Index, ment of spread footings based on the method of Schmertmann
0.0 0.5 1.0 1.5 2.0 et al. (1978), gSE = 1/0.813 ≈ 1.23.
1% 1 in 100
As per the AASHTO LRFD framework, the load factor is

rounded up to the nearest 0.05, and thus gSE = 1.25 should
be used.
Therefore, in the bridge design example, the bridge designer
10% 1 in 10
should use a settlement value of (gSE)(dP) = (1.25)(0.85 in.) =
1.06 in. to assess the effect of settlement on the structure.
This value can also be obtained using the graphic technique
explained earlier with respect to Figure 6.15. The example
100% 1 in 1
demonstrated with respect to Figure 6.15 also assumed a toler-
Figure 6.20. Relationship between b and Pe for able settlement of 0.85 in., and it was found that a settlement
the case of a single load and single resistance. of 1 in. would imply a 32% probability of exceedance. These

163
Table 6.9. Computed Values of gSE for Various Methods to Estimate Immediate
Settlement of Spread Footings on Cohesionless Soils
Reliability Schmertmann D’Appolonia Peck and Burland and

Index (b) et al. (1978) Hough (1959) et al. (1968) Bazaraa (1969) Burbridge (1984)
0.00 0.89 0.54 1.08 1.62 1.68

0.50 1.23 0.66 1.39 2.22 2.47
1.00 1.70 0.79 1.79 3.03 3.63
1.50 2.35 0.96 2.30 4.13 5.34
2.00 3.25 1.16 2.96 5.64 7.86
2.50 4.49 1.41 3.81 7.71 11.58
3.00 6.21 1.70 4.89 10.52 17.04
3.50 8.59 2.06 6.29 14.36 25.08
values are close to the value of 1.06 in. for a 30.85% probability 6.1.3.6 Step 6: Review Results and Selection
of exceedance obtained here. Given that the load factor is of Load Factor for Settlement
rounded to the nearest 0.05, the result from the graphic tech-
Figure 6.21 shows a plot of gSE versus b based on the data shown
nique is sufficiently accurate.
in Table 6.10. The current practice based on AASHTO LRFD
The procedure demonstrated in the above example can
(2012) is as follows:
be used to develop values of gSE for any desired b by using the
lognormal distribution of X for the method of Schmertmann
et al. (1978). A similar approach can be used for other analytical 1. Use the Hough (1959) method to estimate immediate
methods and distributions. settlements.
Table 6.9 presents the values of gSE results for the various 2. Use gSE = 1.0.
analytical methods shown in Figure 6.8 and Table 6.2. Obvi-
ously, gSE values less than 1.0 should not be allowed to prevent The data in Table 6.10 and the graph in Figure 6.21 imply
the risk of bridges being underdesigned. Furthermore, the that b ≈ 1.65 corresponds to the current practice noted above.
values of gSE should be rounded to the nearest 0.05, because b ≈ 1.65 is based on the data set in Table 6.2. If additional data
not doing so implies a level of confidence that is not justified were included, or if a different regional data set were to be
by the available data. Table 6.10 presents values of gSE that are used, then the value of b may be different. However, based on
bounded by 1.0 and rounded to the nearest 0.05. a review of state practices performed as part of Samtani and
Table 6.10. Proposed Values of gSE for Various Methods to Estimate Immediate
Settlement of Spread Footings on Cohesionless Soils
Reliability Schmertmann D’Appolonia Peck and Burland and

Index (b) et al. (1978) Hough (1959) et al. (1968) Bazaraa (1969) Burbridge (1984)
0.00 1.00 1.00 1.10 1.60 1.70

0.50 1.25 1.00 1.40 2.20 2.45
1.00 1.70 1.00 1.80 3.05 3.65
1.50 2.35 1.00 2.30 4.15 5.35
2.00 3.25 1.15 2.95 5.65 7.85
2.50 4.50 1.40 3.80 7.70 11.60
3.00 6.20 1.70 4.90 10.50 17.05
3.50 8.60 2.05 6.30 14.35 25.10
Note: The values of gSE have been rounded to the nearest 0.05 and limited to 1.00 or larger.

164
Figure 6.21. Evaluation of gSE based on current and target

reliability indices.
Nowatzki (2006) and Samtani et al. (2010), it is anticipated 6.1.3.7 Step 7: Select gSE for Service I Limit State
that, based on its inherent conservatism, the value of b is
As demonstrated in Steps 5 and 6, the gSE value can be deter-
anticipated to be large and greater than 1.0 for the Hough
mined for any reliability index (b) for various analytical
(1959) method and gSE = 1.0. The majority of the data points
methods. Use of the format shown in Figure 6.21 will lead to
for the Hough method plot below gSE = 1.0, which suggests a
better regional practices in the sense that owners desiring to
significant conservatism in the Hough method. This is con-
calibrate their local practices can readily see the implication
sistent with the earlier observation that the Hough method
of a certain method on the selection and cost of a foundation
is conservative (overpredicts) by a factor of approximately
system. This is because the chart in Figure 6.21 shows the
two (see Table 6.4), which leads to an unnecessary use of deep
reliability of various methods and permits selection of an
foundations instead of spread footings. Based on a consider-
appropriate method that would lead to selection of a proper
ation of reversible and irreversible SLSs for bridge super-
foundation system for a given set of b and gSE values (i.e., not
structures, as shown earlier, a target reliability index (bT) in
using a deep foundation system when a spread foundation
the range of 0.50 to 1.00 for calibration of load factor gSE for
would be feasible). The agency that calibrates a value of gSE
foundation deformation in the Service I limit state is accept-
based on a locally accepted analytical method must ensure
able. Settlement is clearly an irreversible limit state with
that the chosen value of gSE is consistent with the serviceability
respect to the foundation elements, but it may be reversible
of the substructure and superstructure design, as discussed
through intervention with respect to the superstructure. This
in Step 6.
type of logic would lead to consideration of 0.50 as the bT for
calibration of immediate settlements under spread footings
on cohesionless soils.
6.1.4 Meaning and Use of gSE
In Figure 6.21, the horizontal bold dashed line corresponds
to b = 0.50 for SLS evaluation. The boxes around the markers The meaning and use of gSE must be understood in the specific
for various methods represent the spread of predicted values for context of structural implications within the AASHTO LRFD
the five methods evaluated here. For b = 0.50, if gSE = 1.25 is framework. In particular, the value of gSE is used to assess
adopted, then it would encompass three of the five methods. structural implications, such as the generation of additional
The value of gSE = 1.25 includes the Schmertmann et al. (1978) (secondary) moments within a given span because of settle-
method, which is currently recommended by Samtani and ment of one of the support elements and effect on the riding
Nowatzki (2006) and Samtani et al. (2010) and is commonly surface, and conceivably even appearance and roadway damage
used in U.S. practice. Based on these observations, gSE = 1.25 is issues. If taken literally, the value of gSE = 1.25 in the example
recommended. could be interpreted to mean that the settlement predicted (dP)

165
by the analytical method of Schmertmann et al. (1978) needs calculation of creep effects could be beneficial in some cases.
to be increased by 25% to limit the probability of exceedance Use of the construction point concept (see next section) would
(Pe) of the tolerable settlement (dT) to less than 30.85%, cor- also mitigate the settlement moments. Nevertheless, Schopen’s
responding to a target reliability index (bT) of 0.50. However, results suggest that the use of permissible angular distortions
this literal interpretation is not entirely correct because the approaching those currently allowed by AASHTO LRFD
value of gSE (1.25 in this case) is just one of many load factors requires careful consideration of the particular bridge and its
in the Service I limit state load combinations within the overall design objectives.
AASHTO LRFD framework. As the predicted (estimated or calculated) settlement (dP) is
In addition to the SLS, settlements need to be considered at based on the Service I load combination and the load factor
applicable strength limit states, because although settlements (gSE) is used to modify the Service I load combination, the
can cause serviceability problems, they can also have a signifi- use of gSE can potentially lead to a circular reference in the
cant effect on moments in continuous superstructures that can bridge design process that may require significant iterations.
result in increased member sizes. Settlement is handled more The following procedure is recommended to avoid a circular
explicitly in AASHTO LRFD (in which it is listed among the reference:
loads in Table 3.4.1-1, Load Combinations and Load Factors)
than it was in the Standard Specifications for Highway Bridges. 1. Assume zero settlement and determine the service load by
It appears in four of the five strength load combinations and using the Service I load combination. When settlement is
three of the four service load combinations. This emphasis may assumed to be zero, the value of gSE is irrelevant.
appear to be a departure from past practice, as exemplified 2. Determine dP based on the method that has been calibrated
by AASHTO’s Standard Specifications for Highway Bridges, in using the procedure described here.
that settlement does not appear in those load combinations. 3. Multiply dP by gSE = 1.25 to determine the tolerable (lim-
But settlement is mentioned in Article 3.3.2.1 of Standard iting) settlement (dT) that should be incorporated into
Specifications (2002), which states “If differential settlement bridge design through use of the d-0 angular distortion
is anticipated in a structure, consideration should be given and construction point concept described below. The use
to stresses resulting from this settlement.” The parent article of the d-0 angular distortion and construction point con-
(3.3, Dead Load) implies that settlement effects should be cept incorporates the span lengths, differential settlements
considered wherever dead load appears in the allowable stress between support elements, and the various stages of con-
design or load factor design (LFD) load combinations. struction into the bridge design process.
The additional moments due to the effect of settlement are
very dependent on the stiffness of the bridge, as well as the
6.1.5 Effect of Foundation Deformations
angular rotation (i.e., differential settlement normalized with
on Bridge Superstructures
respect to span length, as discussed in Chapter 2 in the section
on tolerable vertical deformation criteria). A limited study Uneven displacements of bridge abutments and pier founda-
(Schopen 2010) of several two- and three-span steel and pre- tions often lead to costly maintenance and repair measures
stressed concrete continuous bridges selected from the NCHRP associated with the structural distress that the bridge super-
Project 12-78 database showed that allowing the full angular structure and substructure might experience. The bridge
distortion suggested in Table 2.2 could result in an increase superstructure and substructure displacements can be due to
in the factored Strength I moments on the general order of as a variety of reasons, including foundation deformations. The
little as 10% for the more flexible units considered to more foundation deformations need to be evaluated in the context
than double the moment from only the factored dead and live of span lengths and various construction steps to understand
load moments for the stiffer units. These order of magnitude their effect on the bridge superstructures. These aspects and
estimates are based on elastic analysis without consideration of the concepts of angular distortion and construction points are
creep, which could significantly reduce the moments, especially discussed in this section. The application of load factors due
for relatively stiff concrete bridges. For example, a W 36 × 194 to deformation (e.g., gSE) is also presented.
rolled beam with a 10 × 17⁄8 -in. bottom cover plate composite For all bridges, stiffness should be appropriate to the con-
with a 96 × 7 3⁄4-in. deck is presented in Brockenbrough and sidered limit state. Similarly, the effects of continuity with the
Merritt (2011). The computed moments of inertia for the basic substructure should be considered. In assessing the structural
beam and short-term composite and long-term composite implications of foundation deformations of concrete bridges,
sections were in the approximate ratio 1:2:3. This indicates that the determination of the stiffness of the bridge components
consideration of construction sequence, an appropriate choice should consider the effects of cracking, creep, and other
of section properties, and possibly even a time-dependent inelastic responses.

166
6.1.5.1
d-0 Concept for Vertical Deformations differential settlement on structures is roughly proportional
(Settlements) to the angular distortion (A), which is a normalized measure
of differential settlement that includes the distance over which
Because of the inherent variability of geomaterials, the verti-
the differential settlement occurs. Angular distortion is defined
cal deformations at the support elements of a given bridge as the difference in settlement between two points (Dd) divided
(i.e., piers and abutments) will generally be different. This is by the distance between the two points (L), as shown in Fig-
true regardless of whether deep foundations or spread footings ure 6.22. Angular distortion is a dimensionless quantity that is
are used. Figure 6.22 shows the hypothetical case of a four-span expressed as an angle in radians. Theoretically, the ratio Dd/L
bridge structure with five support elements (two abutments represents the tangent of the angle of distortion, but for small
and three piers) for which the calculated settlement (d) at each values of the tangent, the angles are also very small. Thus, the
support is different (the figure assumes rigid substructure tangents (i.e., A) are shown as angles in Figure 6.22. For bridge
units between the foundations and bridge superstructure). structures, the two points used to evaluate the differential
Differential settlements induce bending moments and shear settlement are commonly selected as the distance between
in the bridge superstructure when spans are continuous over adjacent support elements (see Figure 6.22).
supports and potentially cause structural damage. To a lesser Although all analytical methods for estimating settlements
extent, they can also cause damage to a simple-span bridge. have a certain degree of uncertainty, the uncertainty of the
However, the major concern with simple-span bridges is the calculated differential settlement is larger than the uncertainty
quality of the riding surface and aesthetics. Due to a lack of of the calculated total settlement at each of the two support
continuity over the supports, the changes in slope of the riding elements used to calculate the differential settlement (e.g.,
surface near the supports of a simple-span bridge induced by between an abutment and a pier or between two adjacent piers).
differential settlements may be more severe than those in a For example, if one support element settles less than the
continuous-span bridge. amount calculated, and the other support element settles
Depending on factors such as the type of superstructure, the amount calculated, the actual differential settlement
the connections between the superstructure and substructure will be larger than the difference between the two values of
units, and the span lengths and widths, the magnitudes of calculated settlement at the support elements. On the basis of
differential settlement that can cause damage to the bridge these considerations and guidance in Samtani and Nowatzki
structure can vary significantly. For example, the damage (2006) and Barker et al. (1991), the following limit state criteria
to the bridge structure due to a differential settlement of are suggested to estimate a realistic value of differential settle-
2 in. over a 50-ft span is likely to be more severe than the ment and angular distortion:
same amount of differential settlement over a 150-ft span.
Various studies, including Grant et al. (1974) and Skempton • The actual settlement of any support element could be as
and MacDonald (1956), have determined that the severity of large as the value calculated by using a given method.
Abutment 1 Pier 1 Pier 2 Pier 3 Abutment 2
L1 L2 L3 L4
Span 1 Span 2 Span 3 Span 4
A1 P1 P2 P3 A2
A3 A4
A1 A2
Differential
Support Element Settlement Span Settlement Angular Distortion
Abutment 1 A1 1 | A1 – P1| A1 = (| A1 – P1|)/L1
Pier 1 P1 2 | P1 – P2| A2 = (| P1 – P2|)/L2
Pier 2 P2 3 | P2 – P3| A3 = (| P2 – P3|)/L3
Pier 3 P3 4 | P3 – A2| A4 = (| P3 – A2|)/L4
Abutment 2 A2
Figure 6.22. Concept of settlement and angular distortion

in bridges.

167
Table 6.11. Estimation of Design damage to the bridge superstructure, the concept shown in
Differential Settlements and Design Figure 6.23 is more important for continuous-span structures
Angular Distortions for Hypothetical than single-span structures because of the ability of the latter
Case Shown in Figure 6.22 to permit larger movements at support elements.
Design Differential Design Angular
Span Settlement Distortion 6.1.5.2 Construction Point Concept
1 dP1 = dP1 (assume dA1 = 0) A1 = dP1/L1
Most designers analyze foundation deformations as if a
2 dP1 = dP1 (assume dP2 = 0) A2 = dP1/L2 weightless bridge structure is instantaneously set in place and
3 dP3 = dP3 (assume dP2 = 0) A3 = dP3/L3 all the loads are applied at the same time. In reality, loads are
4 dA2 = dA2 (assume dP3 = 0) A4 = dA2/L4 applied gradually as construction proceeds. Consequently,
settlements also occur gradually as construction proceeds.
Several critical construction points or stages during construc-
tion should be evaluated separately by the designer. Figure 6.3
• At the same time, the actual settlement of the adjacent sup- shows the critical construction stages and their associated load-
port element could be zero instead of the value calculated displacement behavior. The baseline format of Figure 6.24 is
by using the same given method. the same as that of Figure 6.3 except that the figure considers
vertical load and vertical displacement (i.e., settlement). For-
The concept outlined above is referred to as the d-0 concept, mulation of settlements in the manner shown in Figure 6.24
with a value of d representing full calculated settlement at one would permit an assessment of settlements up to that point
support of a span and a value of 0 representing zero settle- that can affect the bridge superstructure. For example, the
ment at an adjacent support. Use of the d-0 approach would settlements that occur before placement of the superstructure
result in an estimated maximum possible differential settle- may not be relevant to the design of the superstructure. Thus,
ment between two adjacent supports equal to the larger of the the settlements between application of loads X and Z are the
two total settlements calculated at either end of any span. Thus, most relevant.
with respect to Figure 6.22, where dA1 < dP1 > dP2 < dP3 < dA2 The percentage of settlement between the placement of
represents the relative magnitudes of the total settlement at beams and the end of construction is generally in the range of
each support point, the differential settlements and angular 25% to 75%, depending on the type of superstructure and
distortion for design are evaluated as shown in Table 6.11. the construction sequence. With respect to the example of the
The values in Table 6.11 represent the maximum values for four-span bridge and the angular distortions in Table 6.11, the
each span according to the criteria above and should be used use of the construction point concept would result in smaller
for design. The hypothetical settlement profile assumed for angular distortions to be considered in the structural design.
computation of the design angular distortion for each span is This will be true of any bridge evaluation. Using Figure 6.22
represented by the dashed lines in Figure 6.23. It should not as a reference, Figure 6.25 compares the profiles of the calcu-
be confused with the calculated total settlement profile, which lated settlements (solid lines), hypothetical maximum angular
is represented by the solid lines. From the viewpoint of the distortions (dashed lines), and the range of actual angular
L1 L2 L3 L4
A1 P1 P2 P3 A2
DA1 DA3
DA2 DA4
Legend:
Calculated settlement profile (refer to Figure 6.22)
Hypothetical settlement profile assumed for computation of maximum
angular distortion
Figure 6.23. Estimation of maximum angular distortion

in bridges.

168
Z Wearing Surface
Y
Superstructure
X
Substructure
W Foundation (shallow or deep)
(a)
(b)
Legend:
W load after foundation construction δW displacement under load W
X load after pier column and wall construction δX displacement under load X
Y load after superstructure construction δY displacement under load Y
Z load after wearing surface construction δZ displacement under load Z
S service load (or limit) state (SLS)
F factored load (strength limit state)
Figure 6.24. Construction point concept for a bridge pier: (a) identification
of critical construction points and (b) conceptual load-displacement pattern
for a given foundation.
distortions (hatched-pattern zones) based on the construction should be evaluated in terms of angular distortions using the
point concept. The angular distortions shown in Figure 6.25 construction point concept. While using the construction
should be compared with the limit state criteria for angular point concept, it is important to recognize that the various
distortions provided in Article 10.5.2.2 of AASHTO LRFD quantities are being measured at discrete construction stages,
(2012) and Table 6.2 here. and the associated settlements are considered to be immediate.
On the basis of the discussions above, it is recommended However, the evaluation of total settlement and the maximum
that the limit state of vertical deformations (i.e., settlements) (design) angular distortion must also account for long-term
L1 L2 L3 L4
A1 P1 P2 P3 A2
Legend:
Calculated settlement profile (refer to Figure 6.22)
Hypothetical settlement profile assumed for computation of maximum
angular distortion
Range of relevant angular distortions using construction point concept
Figure 6.25. Angular distortion in bridges based on construction

point concept.

169
settlements. For example, significant long-term settlements and lateral movements due to sliding or lateral deformations
may occur if foundations are founded on saturated clay depos- of deep foundations can be estimated by structural specialists
its or if a layer of saturated clay falls within the zone of stress using input from geotechnical specialists. The limiting hori-
influence below the foundation, even though the foundation zontal movements are strongly dependent on the type of
itself is founded on competent soil. In such cases, long-term superstructure and the connection with substructure and are
settlements will continue under the total construction load therefore project specific.
(Z), as shown by the dashed line in Figure 6.24. Continued
settlements during the service life of the structure will tend to
6.1.6 Practical AASHTO LRFD Application
reduce the vertical clearance under the bridge, which may
of the Load Factor for Deformation
cause problems when large vehicles pass below the bridge
Using the Construction Point Concept
superstructure. As a result, the geotechnical specialist must
estimate and report to the structural specialist the magnitude Article 10.5.2.2 of AASHTO LRFD (2012) addresses the topic
of the long-term settlement that will occur during the design of tolerable movements and movement criteria. This section
life of the bridge. A key point in evaluating settlements at critical is intended to provide additional guidance to incorporate the
construction points is the close coordination required between concept of the load factor for deformation and the construc-
the structural and geotechnical specialists. tion point concept into LRFD Article 10.5.2.2. The following
steps should be followed to estimate a practical value of angular
distortion of the superstructure on the basis of foundation
6.1.5.3 Foundations Proportioned
settlement; a similar approach can be applied and is recom-
for Equal Settlement
mended for evaluation of horizontal movement and rotation
Often geotechnical and structural specialists will try to proof foundations.
portion foundations for equal settlement. In this case, the
argument is made that there will be no differential settlement.
6.1.6.1 Vertical Deformations (Settlement)
Although this concept may work for a building structure
because the footprint is localized, it is a fallacy to assume zero 1. Compute total foundation settlement at each support ele-
differential settlement for a long linear highway structure such ment by using an owner-approved method for the assumed
as a bridge or a wall due to the inevitable variation of the foundation type (e.g., spread footings, driven piles, drilled
properties of geomaterials along the length of the structure. shafts) as follows:
Furthermore, as noted earlier, the prediction of settlements a. Determine dta, the total foundation settlement using all
from any given method is uncertain in itself. Thus for high- applicable loads in the Service I load combination.
way structures, even when the foundations are proportioned b. Determine dtp, the total foundation settlement before
for equal settlement, it is advisable to evaluate differential construction of bridge superstructure. This settlement
settlement assuming that the settlement of any support ele- would generally be as a result of all applicable sub-
ment could be as large as the value calculated by using a given structure loads computed in accordance with a Service I
method, while at the same time, the settlement of the adjacent load combination.
support element could be zero. c. Determine dtr, relevant total settlement: dtr = dta - dtp.
2. At a given support element assume that the actual relevant
settlement could be as large as the value calculated by the
6.1.5.4 Horizontal Deformations
chosen method. At the same time, assume that the settle-
Horizontal deformations generally occur due to sliding or ment of the adjacent support element could be zero instead
rotation (or both) of the foundation. Horizontal deformations of the relevant settlement value calculated by the same
cause more severe and widespread problems than do equal chosen method. Thus, differential settlement (dd) within a
magnitudes of vertical movement (Moulton et al. 1985). The given bridge span is equal to the larger of the relevant settle-
most common location of horizontal deformations is at ment at each of two supports of a bridge span. Compute
the abutments, which are subject to lateral earth pressure. angular distortion (Ad) as the ratio of dd to span length (Ls),
Horizontal movements can also occur at the piers as a result where Ad is measured in radians.
of lateral loads and moments at the top of the substructure The discussion with respect to Table 6.11 and Figure 6.23
unit. The estimation of the magnitudes of horizontal move- in Section 6.1.5 is applicable to this step.
ments should take into account the movements associated 3. Compute modified angular distortion (Adm) by multiplying
with lateral squeeze, as discussed in Samtani and Nowatzki Ad from Step 2 with the gSE values for settlement using the
(2006) and Samtani et al. (2010). Lateral movements due to approach discussed in Section 6.1.4 based on the analytical
lateral squeeze can be estimated by geotechnical specialists, method used for computing the total settlement value.

170
4. Compare the Adm value with owner-specified angular dis- walls to abutment walls. The angular distortion values along
tortion criteria. If owner-specified criteria are not available, a retaining wall can be used to select an appropriate wall type
then use 0.008 radians for the case of simple spans and (e.g., MSE walls can tolerate larger angular distortions than
0.004 radians for the case of continuous spans as the limit- cast-in-place walls).
ing angular distortions. This was discussed in Chapter 2 in
the section on tolerable vertical deformation criteria.
Other angular distortion limits may be appropriate after 6.1.6.4 General Comments
consideration of The following guidance from AASHTO LRFD Article 10.5.2.2
• Cost of mitigation through larger foundations, realign- should be followed while implementing the recommenda-
ment, or surcharge; tions made in previous sections:
• Rideability;
• Vertical clearance; Foundation movement criteria shall be consistent with the
• Tolerable limits of deformation of other structures function and type of structure, anticipated service life, and
associated with the bridge (e.g., approach slabs, wing- consequences of unacceptable movements on structure perfor-
walls, pavement structures, drainage grades, utilities on mance. Foundation movement shall include vertical, horizontal,
the bridge); and rotational movements. The tolerable movement criteria
• Roadway drainage; shall be established by either empirical procedures or structural
• Aesthetics; and analyses, or by consideration of both.
• Safety. Foundation settlement shall be investigated using all
applicable loads in the Service I load combination specified in
5. Evaluate the structural ramifications of the computed
Table 3.4.1-1. Transient loads may be omitted from settlement
angular distortions that are within acceptable limits as per
analyses for foundations bearing on or in cohesive soil deposits
Step 4. Modify foundation design as appropriate based on that are subject to time-dependent consolidation settlements.
structural ramifications. All applicable service limit state load combinations in
6. The above procedure should also be used for cases in Table 3.4.1-1 shall be used for evaluating horizontal move-
which the foundations of various support elements are ment and rotation of foundations.
proportioned for equal total settlement, because the pre-
diction of settlements from any given method is in itself Additional guidance is provided below:
uncertain.
• All foundation deformation evaluations should be based
6.1.6.2 Lateral Deformations on the geomaterial information obtained in accordance
with Article 10.4 of AASHTO LRFD (2012).
Using procedures similar to settlement evaluation specified • The bridge engineer should add deformations from the sub-
for vertical deformation in the previous subsection, lateral structure (elements between foundation and superstructure)
(horizontal) movement at the foundation level can also be as appropriate in evaluation of angular distortions at the
evaluated. Horizontal movement criteria should be estab- deck elevation.
lished at the top of the foundation based on the tolerance of • Although the angular distortion is generally applied in the
the structure to lateral movement, with consideration of the longitudinal direction of a bridge, similar analyses should
column length and stiffness. be performed in the transverse direction based on consid-
The above guidance should take into account the following eration of bridge width and stiffness.
guidance from Article C10.5.2.2 of AASHTO LRFD (2012):
Rotation movements should be evaluated at the top of the 6.1.7 Proposed AASHTO LRFD Provisions
substructure unit in plan location and at the deck elevation.
Tolerance of the superstructure to lateral movement will In AASHTO LRFD (2012), Article 10.5.2 (Service Limit
depend on bridge seat or joint widths, bearing type(s), struc- States) in Section 10 (Foundations) is the primary article that
ture type, and load distribution effects. provides guidance for SLS design for bridge foundations in
terms of tolerable movements. Article 10.5.2 is referenced in
other articles, as indicated in Table 6.12.
6.1.6.3 Walls
The changes based on geotechnical considerations are pri-
The procedure for computing angular distortions can also be marily needed in Article 10.5.2. As Article 10.5.2 references
applied for evaluating angular distortions along and trans- Article 3.4, changes are also needed in that article. These
verse to retaining walls, as well as the junction of the approach changes are provided in Chapter 7.

171
Table 6.12. Summary of Relevant Articles in bridge decks exposed to deicing salts and substructures exposed
AASHTO LRFD for Foundation Deformations to water. Class 1 is used for less corrosive conditions and could
be thought of as an upper bound in regard to crack width for
Article Title Relates to
appearance and corrosion. Previous research indicates that
10.6.2.2 Tolerable Movements Spread footings there appears to be little or no correlation between crack width
10.6.2.5 Overall Stability Spread footings and corrosion. However, the different classes of exposure con-
ditions have been so defined in the design specifications in
10.7.2.2 Tolerable Movements Driven piles
order to provide flexibility in the application of these provi-
10.7.2.4 Horizontal Pile Foundation Driven piles
sions to meet the needs of the bridge owner.
Movement
The load factors for dead load (DL) and live load (LL) spec-
C10.7.2.5 Commentary to Settlement Driven piles
ified for the Service I load combination are as follows: DL load
Due to Downdrag
factor = 1.0 and LL load factor = 1.0.
10.8.2.1 Service Limit State Drilled shafts
When designing reinforced concrete bridge decks using the
10.8.2.2.1 General Drilled shafts conventional design method, most designers follow a similar
10.8.2.3 Horizontal Movements of Drilled shafts approach in selecting the deck thickness and reinforcement.
Shafts and Shaft Groups The thickness is typically selected as the minimum acceptable
10.9.2.2 Tolerable Movements Micropiles thickness, often based on the owner’s standards. The choice of
10.9.2.4 Horizontal Micropile Micropiles main reinforcement bar diameter is typically limited to No. 5
Foundation Movement and No. 6 bars, and the designer does not switch to No. 6 bars
C11.10.11 Commentary to MSE MSE walls unless No. 5 bars result in bar spacing less than the minimum
Abutments spacing allowed. This practice limits the number of possible
14.5.2.1 Number of Joints Joints and bearings variations and allows the development of a deck database that
Note: Article 10.5.2 and its subarticles are frequently referenced in the can be used in the calibration.
articles noted in the left-hand column and their corresponding commentary For decks designed using the empirical method, not deter-
portion. In this table, the article number is based on the first occurrence of
mined on the basis of a calculated design load, the reinforce-
the reference to Article 10.5.2.
ment does not change with the change in girder spacing,
which results in varying crack resistance. As the statistical
parameters for both the load effect and the resistance are
6.2 Cracking of Reinforced required to perform the calibration, a meaningful calibration
Concrete Components, of decks designed using the empirical design method could
Service I Limit State: not be performed.
Annual Probability For other components, including prestressed decks, design-
ers may select different member dimensions, resulting in dif-
Traditionally, reinforced concrete components are designed to
ferent reinforcement areas. Even for the same reinforcement
satisfy the requirements of the strength limit state, after which
area, the designer may use bars or strands of different diam-
they are checked for the Service I limit state load combination
eters and spacing and, consequently, obtain different crack
to ensure that the crack width under service conditions does
resistance and a different reliability index for each possible
not exceed a certain value. However, the specifications provi-
variation. The variation in the cracking behavior of the same
sions are written in a form emphasizing reinforcement details
component with the change in the selected reinforcement
(i.e., limiting bar spacing rather than crack width). Satisfying
prohibits the performance of a meaningful calibration for
the Service I limit state for crack control through the distribu-
such components.
tion of reinforcement may require a reduction in the reinforce-
Due to the reasons indicated above, the calibration for the
ment spacing. This may require the use of smaller bar diameters
Service I limit state for crack control through the distribution of
or, if the smallest allowed bar diameters are already being used,
reinforcement was limited to reinforced concrete decks designed
an increase in the number of reinforcement bars leading to an
increase in the reinforcement area. using the conventional design method. The decks are assumed
Two exposure classifications exist in AASHTO LRFD: Class 1 to be supported on parallel longitudinal girders.
exposure condition and Class 2 exposure condition. Class 1
relates to an estimated maximum crack width of 0.017 in.,
6.2.1 Live Load Model
and Class 2 relates to an estimated maximum crack width of
0.01275 in. Class 2 is typically used for situations in which the Reinforced concrete decks designed using the conventional
concrete is subjected to severe corrosion conditions, such as method have been designed for the heavy axles of the design

172
truck. This practice required developing the statistical param- 6.2.2.2 Statistical Parameters of Variables
eters of the axle loads of the trucks in the weigh-in-motion Included in the Design
(WIM) data. The statistical parameters for the axle loads are
Several variables affect the resistance of prestressed compo-
presented in Chapter 5. Statistical parameters corresponding
nents. Table 6.13 shows a list of variables that were considered
to a 1-year return period were assumed in the reliability analy-
to be random variables during the performance of the reliability
sis. Average daily truck traffic (ADTT) of 1,000, 2,500, 5,000,
analyses. These variables represent a summary of the infor-
and 10,000 were considered; however, an ADTT of 5,000 was
mation based on research studies by Siriaksorn and Naaman
used as the basis for the calibration.
(1980) and Nowak et al. (2008).
6.2.2 Target Reliability Index

6.2.2.3 Database of Reinforced Concrete Decks
6.2.2.1 Limit State Function
A database consisting of 15 reinforced concrete decks designed
For the control of cracking of reinforced concrete through the using the conventional method of deck design was developed.
distribution of reinforcement, the limiting criteria are the cal- As typical in deck design, No. 5 bars were used unless they
culated crack widths, assumed to be 0.017 and 0.01275 in. for resulted in bar spacing less than 5 in., the minimum spacing
Class 1 and Class 2, respectively. Due to the lack of clear con many jurisdictions allow in deck design. If No. 5 bars resulted in
sequences for violating the limiting crack width, there was no a bar spacing less than 5 in., No. 6 bars were used. No maximum
basis to change the nature or the limiting values of the limit bar spacing was considered in the design to ensure that all
state function (i.e., the crack width criteria). The work was decks produced a calculated crack width equal to the maxi-
based on maintaining the current crack width values and mum allowed crack width allowed by the specifications. The
calibrating the limit state to produce a uniform reliability designs were not checked for other limit states because the
index similar to the average reliability index produced by the purpose was to calibrate the Service I limit state. The design
current designs. of the 15 decks was repeated twice, once assuming Class 1
Table 6.13. Summary of Statistical Information for Variables Used

in the Calibration of Service I Limit State for Crack Control
Variable Distribution Mean CV Source
As Normal 0.9As 0.015 Siriaksorn and Naaman (1980)

b Normal bn 0.04 Siriaksorn and Naaman (1980)
CEc Normal 33.6 0.1217 Siriaksorn and Naaman (1980)
d Normal 0.99dn 0.04 Nowak et al. (2008)
dc Normal dcn 0.04 Nowak et al. (2008)
Es Normal Esn 0.024 Siriaksorn and Naaman (1980)
f c′ Lognormal ′
3,000 psi:1.31 f cn 3,000:0.17 Siriaksorn and Naaman (1980)
Ec = CEc ′
3,500 psi:1.27 f cn 3,500:0.16
g 1.5
c fc′ ′
4,000 psi:1.24 f cn 4,000:0.15
′
4,500 psi:1.21 f cn 4,500:0.14
′
5,000 psi:1.19 f cn 5,000:0.135
fy Lognormal 1.13fyn 0.03 Nowak et al. (2008)
h Normal hn 1/(6.4µ) Siriaksorn and Naaman (1980)
gc Normal 150 0.03 Siriaksorn and Naaman (1980)
As = area of steel rebar (in. );
2
b = width of equivalent transverse strip of concrete deck (in.);

CEc = constant parameter for concrete elasticity modulus;
d = effective depth of concrete section (in.);
dc = bottom cover measured from center of lowest bar (in.);
Es = modulus of elasticity of steel reinforcement (psi);
f′c = specified compressive strength of concrete (psi);
fy = yield strength of steel reinforcement (psi);
h = deck thickness (in.); and
gc = unit weight of concrete (lb/ft3).

173
Table 6.14. Summary Information of 15 Bridge the statistical parameters for live load were taken from Sec-
Decks Designed Using AASHTO LRFD Conventional tion 5.3.4. The reliability indices for various ADTTs and expo-
Deck Design Method sure conditions for the 15 decks are summarized in Table 6.15.
Due to the difference in positive and negative moment (bottom
Deck Group No. Girder Spacing (ft) Deck Thickness (in.)
and top) reinforcement of the deck, the reliability index was
1 6 7 calculated separately for the positive and negative moment
7.5 reinforcement.
8
Even though the design for Class 2 resulted in more rein-
forcement than for Class 1 exposure conditions, the reliability
2 8 7.5
index for Class 2 is lower than that for Class 1 due to the more
8 stringent limiting criteria (narrower crack width).
8.5 Current practices rarely result in the deck positive moment
3 10 8 reinforcement being controlled by the Service I limit state
8.5
due to the small bottom concrete cover. When Strength I limit
state is considered, more positive moment reinforcement is
9
typically required than by Service I. The additional reinforce-
9.5 ment results in reliability indices for the positive moment
4 12 8 region higher than those shown in Table 6.15.
8.5 For the negative moment region, the design is often con-
9
trolled by the Service I limit state. Thus, the reliability indices
shown for the negative moment region in Table 6.15 are
9.5
considered representative of the reliability indices that would
10 be calculated when all limit states, including Strength I, are
considered in the design.
Therefore, it is recommended that the target reliability index
exposure conditions and a second time assuming Class 2
be based on the reliability index for the negative moment
exposure conditions.
region. Because the Class 2 case is the more common case for
Table 6.14 presents the summary information of the
decks, the reliability index for Class 2 was used as the basis
15 designed bridge decks.
for selecting the target reliability index. The reliability index
for Class 1 was assumed to represent a relaxation of the base
6.2.2.4 Selection of Target Reliability Index
requirements. The case of ADTT = 5,000 was also considered
Monte Carlo simulation was used to obtain the statistical as the base case on which the reliability analysis was performed.
parameters of resistance (or capacity) and dead load, and Table 6.16 shows the inherent reliability indices for the negative
Table 6.15. Summary of Reliability Indices for Concrete Decks Designed

According to AASHTO LRFD (2012)
Positive Moment Region Negative Moment Region
Reliability Index Reliability Index Reliability Index Reliability Index

ADTT (Class 1) (Class 2) (Class 1) (Class 2)
1,000 2.44 1.54 2.37 1.77

2,500 1.95 1.07 1.79 1.27
5,000 1.66 0.85 1.61 1.05
10,000 1.39 0.33 1.02 0.5
Average 1.86 0.95 1.70 1.15
Maximum 2.44 1.54 2.37 1.77
Minimum 1.39 0.33 1.02 0.50
Standard deviation 0.45 0.50 0.56 0.53
CV 24% 53% 33% 46%

174
Table 6.16. Reliability Indices of Existing information includes the probability distribution and statisti-
Bridges Based on 1-Year Return Period cal parameters for axle loads presented in Section 5.3.4 in
Chapter 5 and for other variables affecting the resistance pre-
Current Practice Current Practice
sented in Section 6.2.2.2.
ADTT (Class 1, Negative) (Class 2, Negative)
1,000 2.37 1.77

2,500 1.79 1.27 Analysis Procedure
5,000 1.61 1.05
The statistical information of all the required variables was
10,000 1.02 0.50
used to determine the statistical parameters of the resistance
by using Monte Carlo simulation.
For each deck, Monte Carlo simulation was performed for
moment region of decks designed for the current AASHTO each random variable associated with the calculation of the
LRFD. The selected target reliability indices are 1.6 and 1.0 for resistance and dead load. One thousand simulations were per-
Class 1 and Class 2, respectively, based on ADTT = 5,000. formed. For each random variable 1,000 values were generated
independently on the basis of the statistics and distribution
of that random variable. For each simulation, the dead load
6.2.3 Calibration Result and the resistance were calculated using one of the 1,000 sets
The basic steps of the calibration process are shown below as of values of the random variable (i.e., the nth simulation used
they relate to the Service I calibration. the nth value of each random variable, where n varied from
1 to 1,000). This process resulted in 1,000 values of the dead
load and the resistance. The mean and standard deviation of
6.2.3.1 Step 1: Formulate Limit State Function the dead load and the resistance were then calculated based
and Identify Basic Variables on the 1,000 simulations.
The limit state function considered was the limit on the esti-
mated crack width. In the absence of information suggesting 6.2.3.6 Step 6: Calculate Reliability Indices for
that the current criteria (based on a crack width of 0.017 and Current Design Code and Current Practice
0.01275 in. for Class 1 and Class 2, respectively) are not ade-
Using the statistics of the dead load and the resistance, calcu-
quate, the current crack widths were maintained as the limiting
lated from the Monte Carlo simulation as described above, and
criteria. A discussion of crack width equations in the literature
the statistics of the live load as derived from the WIM data, as
is included in Appendix C.
described in Chapter 5, the reliability index (b) was calculated
for each deck by using Equation 6.7:
6.2.3.2 Step 2: Identify and Select Representative
Structural Types and Design Cases µ R − µQ
β= (6.7)
σ 2R + σ Q2
The database of decks used in this study is described in Sec-
tion 6.2.2.3.
where
µR = mean value of the resistance;
6.2.3.3 Step 3: Determine Load and Resistance µQ = mean value of the applied loads;
Parameters for Selected Design Cases sR = standard deviation of the resistance; and
The variables include the dimension of the cross section and the sQ = standard deviation of the applied loads.
material properties. The statistical information includes the The calculated reliability indices of the decks in the data-
probability distribution and statistical parameters such as base are shown in Table 6.15 for both positive and negative
mean (µ) and standard deviation (s). moment reinforcement and for Class 1 and Class 2 exposure
conditions.
for Load and Resistance 6.2.3.7 Step 7: Review Results and
Select Target Reliability Index
The variables affecting the load and resistance were identified.
These include live load; resistance, including the dimensions The initial target reliability index (bT) was determined as shown
of the cross section; and the material properties. The statistical in Table 6.16.

175
Figure 6.26. Reliability indices of various Figure 6.28. Reliability indices of various
bridge decks designed using a 1.0 live bridge decks designed using a 1.0 live
load factor over a 1-year return period load factor over a 1-year return period
(ADTT  5,000) for positive moment (ADTT  5,000) for positive moment
region, Class 1 exposure. region, Class 2 exposure.
6.2.3.8 Step 8: Select Potential Load and reliability index are 1.66 and 1.61 for positive and negative
Resistance Factors for Service I, moment regions, respectively.
Crack Control Through the Distribution For a Class 2 exposure condition (maximum crack width of
of Reinforcement 0.01275 in.), Figure 6.28 and Figure 6.29 present the reliability
indices for the bridge decks in the database designed using a
The load factors for dead loads and live loads for the Service I live load factor of 1.0 over a 1-year return period for an ADTT
limit state in the AASHTO LRFD (2012) are 1.0. The existing of 5,000. As indicated in Table 6.15, the average values of the
specifications do not explicitly include a resistance factor for reliability index are 0.85 and 1.05 for positive and negative
the distribution of the control of cracking through the distri- moment regions, respectively.
bution of reinforcement. This omission results in an implied As discussed above, for positive moment (bottom) reinforce-
resistance factor of 1.0. The load and resistance factors were ment, Strength I limit state requirements typically result in
maintained for the initial reliability index calculations. more reinforcement than needed to satisfy Service I require-
For a Class 1 exposure condition (maximum crack width of ments, and the reliability index for cracking at the bottom
0.017 in.), Figure 6.26 and Figure 6.27 present the reliability will be higher than shown in Figure 6.26 and Figure 6.28. This
indices for the bridge decks in the database designed using a difference in required reinforcement resulted in the recommen-
live load factor of 1.0 over a 1-year return period for an ADTT dation that the reliability index should be based on the negative
of 5,000. As indicated in Table 6.15, the average values of the moment (top) reinforcement.
Figure 6.27. Reliability indices of various Figure 6.29. Reliability indices of various
bridge decks designed using a 1.0 live bridge decks designed using a 1.0 live
load factor over a 1-year return period load factor over a 1-year return period
(ADTT  5,000) for negative moment (ADTT  5,000) for negative moment
region, Class 1 exposure. region, Class 2 exposure.

176
6.2.3.9 Step 9: Calculate Reliability Indices 6.3 Live Load Deflections,

As shown in Figure 6.27 and Figure 6.29, the reliability index
Service I: Annual Probability
associated with cracking at the top of the deck appears to be 6.3.1 Proposed Resistance Criteria
very uniform across the range of girder spacings considered.
It was concluded that there was no need to redesign the decks The background and state of the art of SLS for live load deflec-
for different load or resistance factors to improve the unifor- tion was reviewed in Chapter 2. There is considerable dis-
mity of the results. With this conclusion, the reliability indices agreement in the literature as to whether live load deflection
are the same as shown in Table 6.15 and Table 6.16 and in alone is an effective measure of dynamic response or a principal
Figure 6.27 and Figure 6.29. contributor to deck deterioration. Human beings are more sen-
sitive to acceleration than displacement per se, especially when
stationary on a bridge. The combined frequency–displacement
6.2.3.10 Summary and Recommendations criteria in the Canadian Highway Bridge Design Code (CHBDC)
for Service I Limit State, Crack Control (2006), based on a comparison of computed values to Figure 2.1,
Through the Distribution of Reinforcement appear to have more aspects of response accounted for than
The following conclusions are based on the reported reliabil- merely comparing a computed live load deflection to span
ity analyses: length divided by a constant (L/N criterion), which is the case
in AASHTO LRFD.
• Assessment of current practice led to recommended target A direct comparison to Canadian practice (CHBDC 2006)
reliability indices of 1.6 for the base case (Class 1 exposure) requires consideration of the magnitude of the design live load,
and 1.0 for the enhanced requirements (i.e., smaller maxi- dynamic load allowance, load factors, and analysis assumptions.
mum crack width) for Class 2 exposure conditions. These For consideration of vibrations, the CHBDC uses 90% of one
values correspond to an ADTT of 5,000. CL-625 truck loading without a dynamic load allowance. The
• The current requirements in the specifications produce CL-625 truck has six axles totaling about 140 kips and a wheel
uniform reliability across the range of girder spacings con- base of about 58 ft. AASHTO LRFD uses the larger of the design
sidered, so there is no need to change the load or the resistance truck without the uniform load or 25% of the design truck with
factors. the uniform load. The dynamic load allowance is included
for this purpose; the current load factor is 1.0. A comparison
of one lane of CHBDC and AASHTO LRFD loadings for
6.2.4 Proposed AASHTO LRFD Revisions
this limit state is shown in Figure 6.30. Note that the cur-
As indicated above, no revisions to applicable AASHTO LRFD rent provisions of AASHTO LRFD consider live load in each
provisions related to control of cracking by distributed design lane.
reinforcement in reinforced concrete components are war- To determine whether the L/N criterion captures the dynamic
ranted by the results of this research. response criteria in Figure 2.1 sufficiently to be “deemed to
Mss-100 MAX: (HL-93 Ratios)

Ratios of: Centerline Moment for Simple Span
Ratio: (1.00*0.90*CL-625)/ (1.00*AASHTO
1.40
LRFD Deflection Limit Loading)
1.20
1.00
0.80
0.60
0.40
0.20
CL-625
0.00
0 50 100 150 200 250 300
Base Span Length (ft)
Figure 6.30. Comparison of CHBDC and AASHTO LRFD service load moments.

177
satisfy” a need for minimum global stiffness, the live load deflec- all applicable requirements, it is reasonable to suspect they are
tion for 41 bridges chosen from the NCHRP Project 12-78 data- overdesigned with respect to just the live load deflection require-
base (Mlynarski et al. 2011) was calculated using VIRTIS and ment. For calibration purposes, the considered database should
plotted against frequency, as shown in Figure 6.31. The fre- satisfy the limit state under investigation but not necessarily
quency was calculated using the method of Barth and Wu other limit states. To accomplish this, a set of simply supported
(2007) presented in Section 2.3.1.2 in Chapter 2, with l = 1.0. bridges was designed to satisfy all applicable limit states by
The 41 bridges contained simple-span and continuous using the load and resistance factor design of the Pennsylvania
bridges and steel and prestressed concrete bridges, which sat- Department of Transportation’s girder design program and
isfy all applicable requirements, not just live load deflection. then forced to satisfy only the AASHTO LRFD L/800 criteria.
The general trend of results was similar to the criteria curves If the designs forced to satisfy L/800 satisfied any other criteria,
in the figure and fell mostly between the curves for bridges with that outcome was unintentional. The considered span lengths
sidewalks and frequent pedestrian use and bridges with side- were 60, 90, 120, 160, 200, and 300 ft, and girder spacings were
walks and occasional pedestrian usage. It is reasonable to 9 and 12 ft. Deflections and frequencies were calculated, plotted,
assume that most, if not all, of the 41 sample bridges were and compared with the CHBDC criteria. The results are shown
essentially highway bridges without sidewalks. As shown in in Figure 6.32. When the girders were forced to meet the L/800
Figure 6.31, concrete bridges tend to be stiffer than steel bridges, criteria, the frequencies for each pair of spacings for a given
but some of the concrete sample bridges exhibited responses span were so similar that the solid circles in the figure cannot
relatively close to the CHBDC acceptance curves. be distinguished.
For this purpose, the comparisons were made between the For calibration purposes, the CHBDC curves were treated
provisions of AASHTO LRFD and acceptable CHBDC response as deterministic (i.e., the bias was assumed to be 1.0, and the CV
as indicated by the criteria curves. As the 41 bridges satisfied was assumed to be 0.0). This is analogous to the geotechnical
Deflection Limitations for Highway Bridge Superstructure Vibration

1000
Existing Spread Box Girders
Existing Adjacent Box Girders
500
Existing I-Girders
Existing Steel Girders UNACCEPTABLE
200
without sidewalks
100
with sidewalks,

50
with sidewalks,
20.0
10.0
5.0
ACCEPTABLE
2.0
1.0
0 1 2 3 4 5 6 7 8 9 10
Figure 6.31. Comparison of CHBDC requirements and various bridges

satisfying all AASHTO LRFD design requirements.

178
Deflection Limitations for Highway Bridge Superstructure Vibration
1000
Simulated Steel Bridges Designed to Satisfy
AASHTO LRFD Deflection Limits Only
500
Simulated Steel Bridges Designed to Satisfy
AASHTO LRFD Specifications UNACCEPTABLE
200
without sidewalks
100
with sidewalks,
50
with sidewalks,
20.0
10.0
5.0
ACCEPTABLE
2.0
1.0
0 1 2 3 4 5 6 7 8 9 10
Figure 6.32. Comparison of CHBDC requirements and various steel bridges

satisfying all AASHTO LRFD design requirements and similar bridges satisfying
only L/800 criteria.
calibration using a deterministic value of tolerable deformation, Substituting these variables into the general limit state
as presented in Section 6.1.2.3 and illustrated in Figure 6.6. function yields Equation 6.9:
γ LL ∆ LL ≤ ∆ limit (6.9)
6.3.2 Calibration Results
6.3.2.1 Formulate Limit State Function
6.3.2.2 Select Structural Types and Design Cases
The live load deflection limit state function is merely the sum
of the factored loads and must be less than or equal to the All structural types and materials were considered for this
factored resistance, as shown by Equation 6.8: limit state.
∑ γ Q ≤ ϕR
i i (6.8) 6.3.2.3 Determine Load and Resistance Parameters
i
for Selected Design Cases
The basic load effect for live load deflection is obviously As discussed above, the load currently used for deflection
the deflection due to live load. The resistance was taken as the calculations in AASHTO LRFD was maintained with a bias of
appropriate deflection limit from Figure 2.1, as discussed in 1.35 and a CV of 0.12 per Section 5.5.2, which is representa-
Chapter 2. tive of the 1-year live load results for ADTT = 5,000 shown in

179
Table 5.8. The resistance was taken from the CHBDC curves, 6.3.2.9 Calculate Reliability Indices
which were treated as deterministic (i.e., the bias was assumed
A Monte Carlo simulation was again performed for the live
to be 1.0, and the CV was assumed to be 0.0).
load deflection limit state. With the live load deflection limit
considered invariant, all cases yield a reliability index of
6.3.2.4 Develop Statistical Models for Loads, Load about 1.0 using a live load load factor of 1.5 and a resistance
Combinations, and Resistance Variables factor of 1.0.
Uncertainties of Load
From Tables 5.6 to 5.9, the uncertainties of live load moment 6.3.3 Potential LRFD Revisions
were taken as approximately 0.12 for the CV and 1.35 for the
bias. It was assumed that the uncertainty of deflections is the 6.3.3.1 Theoretical Conclusions
same as the uncertainty of moments. The live load deflection limit state provisions could be modified
from those of the AASHTO LRFD to satisfy frequency, percep-
Uncertainties of Resistance tion, and deflection by adopting the CHBDC provisions. A rec-
The live load deflection limits in Figure 2.1 were taken as ommendation to use user comfort was made by Roeder et al.
invariant with a bias of unity and a CV of zero. Thus resistance (2002). The provisions of Article 2.5.2.6 could be revised to
was set equal to the curves shown in Figure 2.1 for the purpose include Figure 2.1 and remove the L/N criterion for steel, alu-
of calibration. minum, and concrete vehicular bridges. If this were done, the
live load deflection limit state should be as defined in AASHTO
6.3.2.5 Develop Reliability Analysis Procedure LRFD Table 3.4.1-1 as Service V, with the load factor for live
load given as 1.50. Dead load would also be needed in the load
As discussed in Chapter 3, Monte Carlo simulation using MS combination for use in calculating frequency. Descriptive text
Excel formed the basis of the reliability analysis procedure for and commentary would be needed in Article 3.4.1.
the live load deflection limit state.
6.3.3.2 Practical Assessment of Results

6.3.2.6 Calculate Reliability Indices for Current
Design Code or Current Practice The use of the CHBDC criteria appears to be a more realistic
approach in that it incorporates both deflection and frequency
Using the current AASHTO LRFD Service I limit state with load and compares them with a set of human factors response
factors equal to unity, the probability of failure is almost 100%. curves. The current (and historic) AASHTO requirement
Thus, load and resistance factors other than unity must be includes only stiffness and compares the result to a criterion,
considered to achieve reliability indices comparable to the other L/N, whose background is reviewed in Article 2.3.1.2. The
SLSs considered. survey of owners (Barker and Barth 2007) indicated that the
range of applications of the current criteria could produce
6.3.2.7 Review Results and Select Target results that differ more than mere consideration of the vari-
Reliability Index ability of loads would indicate.
Despite the obvious limitations of the current AASHTO
Using the SLS reliability indices developed above and a reli- LRFD criteria, Figure 6.31 and Figure 6.32 indicate that the
ability analysis that showed current practice for most fatigue two criteria are somewhat similar. A review of the 12 steel bridge
limit states yields a reliability index of 1.0, a target reliability designs that satisfied all the relevant AASHTO LRFD require-
index (bT) of about 1.0 was chosen for the calibration of the live ments showed that they would not be changed if the deflection
load deflection limit state to be consistent with other reversible criteria were based on a load factor of 1.40, which resulted from
SLSs. Thus, the proposed specifications maintain the historic the calibration described above. Stated more simply, other
level of reliability. criteria still controlled those 12 bridges.
6.3.2.8 Select Potential Load and 6.3.3.3 Recommendations

Resistance Factors
Based on this research, there does not appear to be a compel-
Through trial-and-error testing, the live load load factor for ling need to change the current AASHTO LRFD provision for
the deflection limit state was selected as 1.50, along with a live load response (i.e., load deflection) in Article 2.5.6.2.
resistance factor of unity. Such an approach is basically “deemed to satisfy.” However,

180
if AASHTO chooses to adopt the more complete approach of rideability?” This subject was discussed with the AASHTO Sub-
combining frequency, displacement, and perception, a pos- committee on Bridges and Structures Steel Structures Technical
sible set of revisions to accomplish that change are proposed in Committee. Several states are represented on that committee,
Chapter 7. Other considerations could include basing the deter- but the meeting was open to researchers and practitioners.
mination of deflection on the fatigue truck of Article 3.6.1.4.1, No one present offered any evidence, either documented or
as its longer wheel base is more representative of actual traffic. anecdotal, that there has been any significant issue with perma-
The fatigue truck for orthotropic decks could also be used, nent deformations. The research team is aware of some anec-
but the substitution of tandem axles for the 32K single axle is dotal discussion of permanent set in stringers in the floor
not apt to be significant except for very short spans. systems of some long-span bridges, but that has not been docu-
mented. In summary, the current provisions appear to be serv-
ing well but just how well has not been quantified statistically.
6.4 Overload, Service II: Question 3, “What reliability index is provided in current
Annual Probability designs for which the overload provision controlled?” is dis-
cussed in Section 6.4.3. As detailed below, it is possible to
6.4.1 Basis of Limit State
provide some insight into how often the factored live load is
The basis for this limit state was presented in Section 2.3.1.5. exceeded, which bears on Question 4, “How often can this
Several questions regarding the criteria for control of perma- criteria be exceeded without creating a significant permanent
nent deformations arose: deflection of the structure?”
1. Was the LFD requirement a good target?

6.4.2 Load Model
2. What is the current level of experience with these pro
visions? That is, has there been a significant issue with per- The application of WIM data to the development of load
manent set in girder bridges that affect appearance or models for the SLSs was discussed in Chapter 5. The location
rideability? of WIM sites, the number of days of measurements, and the
3. What reliability index is provided in current designs for number of trucks after filtering are shown in Figure 5.1. The
which the overload provision controlled? indicated number of filtered truck records includes the trucks
4. How often can these criteria be exceeded without creating thought to be permit trucks, as discussed in Section 5.2.3.
a significant permanent deflection of the structure? Thus the trucks used in this study include the permit trucks,
5. Is another choice of load factor, other than the 1.30 used which were filtered for consideration of other limit states.
in the current Service II load combination, equally valid? Table 6.17 shows the number of times bending moments
6. Should this requirement be applied to multilane loading? from the trucks in the database exceeded 1.0, 1.1, 1.2, and
7. If it is used as a single-lane criterion, should the multiple 1.3 times the HL-93 moment for simple spans of 30, 60, 90,
presence factor (MPF) of 1.2 currently used for single- 120, and 200 ft for each WIM site scaled to 1 year of data. Note
lane loading in the AASHTO LRFD be applicable to this that results for the WIM site on Florida US-29 look much
condition? different from the other sites. It was determined that trucks
from other parallel routes were being diverted to Route 29,
The LFD criteria resulted from an assessment of experience creating an unusual situation. This site was excluded from the
at the AASHO Road Test (1962); these criteria are summarized discussion below.
in Figure 2.3. The six data points shown, three for composite Figure 6.33 shows the average exceedance for the 31 remain-
structures and three for noncomposite structures, comprise ing WIM sites by HL-93 ratio for each span length considered.
the full data set known to the research team. Most of the per- Figure 6.34 shows the same information by span length for
manent set occurred during the early repetitions of load, as each HL-93 ratio considered. The reduction in the rate of
would be expected. Because the weight of the test trucks did exceedance with increasing HL-93 ratio is clearly evident. These
not vary during most of the circuits of the test track, the per- data do not show how much a given HL-93 ratio was exceeded
manent set accumulated slowly after the period of initial load in terms of stress, but this limit state has historically been based
cycles. This behavior was expected. on infrequently exceeding the criteria. The rate of exceedance
Two of the questions to be explored in regard to the control of 1.3 HL-93, the current criteria, is seen to be quite small.
of permanent deformations noted above were Question 1, A more meaningful assessment of the exceedance rate is
“Was the LFD requirement a good target?” and Question 2, presented in Table 6.18, Figure 6.35, and Figure 6.36. In this
“What is the current level of experience with these provi- case, the exceedance data have been scaled to an assumed ADTT
sions? That is, has there been a significant issue with perma- of 2,500 at each site assuming that the distribution of trucks
nent deformations in girder bridges that affect appearance or (text continues on page 186)

Table 6.17. Bending Moment Exceedances per Year
Ratio Truck/HL-93 ≥1.0 Ratio Truck/HL-93 ≥1.1 Ratio Truck/HL-93 ≥1.2 Ratio Truck/HL-93 ≥1.3
Site 30 ft 60 ft 90 ft 120 ft 200 ft 30 ft 60 ft 90 ft 120 ft 200 ft 30 ft 60 ft 90 ft 120 ft 200 ft 30 ft 60 ft 90 ft 120 ft 200 ft
Arizona 4 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
(SPS-1)
Arizona 0 2 6 5 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0
(SPS-2)
Arkansas 14 10 17 10 0 2 7 3 0 0 0 3 0 0 0 0 0 0 0 0
(SPS-2)
Colorado 0 5 6 6 2 0 2 5 4 0 0 0 2 0 0 0 0 0 0 0
(SPS-2)
Delaware 140 48 33 27 1 36 33 22 11 0 10 22 10 1 0 1 11 1 0 0
(SPS-1)
Illinois 1 3 4 4 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0
(SPS-6)
Indiana 27 32 24 19 14 5 19 19 17 3 3 7 9 7 0 0 0 2 0 0
(SPS-6)
Kansas 42 47 80 96 10 16 33 35 31 2 7 16 17 7 0 6 7 6 0 0
(SPS-2)
Louisiana 76 16 25 30 13 44 6 12 14 7 26 6 7 7 0 6 6 5 4 0
(SPS-1)
Maine 6 7 8 7 1 4 4 5 2 0 0 4 2 0 0 0 2 0 0 0
(SPS-5)
Maryland 25 8 8 2 1 5 6 2 2 0 0 1 1 0 0 0 1 0 0 0
(SPS-5)
Minnesota 9 8 18 19 2 7 5 6 5 0 4 2 2 1 0 2 1 1 0 0
(SPS-5)
New Mexico 1 1 1 3 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0
(SPS-1)
New Mexico 12 7 7 9 4 4 1 1 3 0 3 0 0 0 0 0 0 0 0 0
(SPS-5)
Pennsylvania 155 45 22 21 1 32 22 17 14 1 13 17 13 1 0 3 13 2 0 0
(SPS-6)
Tennessee 2,085 29 8 7 0 53 4 4 0 0 5 1 0 0 0 1 0 0 0 0
(SPS-6)
Virginia 7 10 1 2 1 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0
(SPS-1)
181
182
Table 6.17. Bending Moment Exceedances per Year (continued)
Wisconsin 6 3 5 4 2 1 0 3 3 1 0 0 1 1 0 0 0 0 0 0
(SPS-1)
California 0 13 25 31 25 0 1 0 0 7 0 0 0 0 0 0 0 0 0 0
Antelope
EB
California 0 30 71 100 84 0 7 6 19 40 0 0 0 1 13 0 0 0 0 1
Antelope
WB
California 0 3 3 8 16 0 0 0 3 3 0 0 0 0 3 0 0 0 0 0
Bowman
California 10 99 150 153 85 1 34 55 56 16 0 7 26 21 0 0 0 4 1 0
LA-710 NB
California 3 62 105 111 54 1 17 45 48 14 0 3 18 19 0 0 0 1 1 0
LA-710 SB
California 0 110 137 281 417 0 5 19 55 168 0 0 1 2 38 0 0 0 0 2
Lodi
Florida I-10 279 141 159 264 152 81 41 47 77 38 23 16 14 18 5 10 5 4 5 2
Florida I-95 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Mississippi 41 48 53 53 44 26 24 34 36 24 8 2 11 21 2 2 2 2 2 1
I-10
Mississippi 0 4 5 11 8 0 0 0 1 3 0 0 0 0 0 0 0 0 0 0
I-55UI
Mississippi 142 100 255 349 89 20 31 50 61 33 7 8 17 22 20 2 3 5 8 9
I-55R
Mississippi 0 3 11 13 7 0 0 2 1 0 0 0 0 0 0 0 0 0 0 0
US-49
Mississippi 0 1 5 8 6 0 0 1 2 1 0 0 1 1 0 0 0 0 0 0
US-61
Florida 1,291 995 651 496 204 673 510 332 253 109 371 274 179 123 53 183 165 85 61 22
US-29
Annual 99.6 28.9 40.4 53.4 33.6 11.0 9.8 12.8 15.1 11.7 3.5 3.7 4.9 4.2 2.6 1.1 1.7 1.1 0.7 0.5
Average
Note: EB = eastbound; WB = westbound; NB = northbound; SB = southbound.
183
120
>= 1.0HL93
>= 1.1HL93
100
>= 1.2HL93
80
Annual Average
>= 1.3HL93
60
40
20
0
30 ft 60 ft 90 ft 120 ft 200 ft
Span (ft.)
Figure 6.33. Annual average exceedances versus span.
120
30 ft
60 ft
100
90 ft
80 120 ft
Annual Average
200 ft
60
40
20
0
>= 1.0HL93 >= 1.1HL93 >= 1.2HL93 >= 1.3HL93
Ratio Truck/HL93
Figure 6.34. Annual average exceedances versus ratio truck/HL-93.

184
Table 6.18. Events per Year Scaled to ADTT  2,500
Arizona (SPS-1) 103 0 0 26 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Arizona (SPS-2) 0 1 4 3 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0
Arkansas (SPS-2) 8 5 9 5 0 1 4 2 0 0 0 2 0 0 0 0 0 0 0 0
Colorado (SPS-2) 0 13 16 16 5 0 5 13 11 0 0 0 5 0 0 0 0 0 0 0
Delaware (SPS-1) 633 217 149 122 5 163 149 100 50 0 45 100 45 5 0 5 50 5 0 0
Illinois (SPS-6) 1 3 4 4 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0
Indiana (SPS-6) 79 94 69 54 39 15 54 54 49 10 10 20 25 20 0 0 0 5 0 0
Kansas (SPS-2) 80 90 153 183 19 31 63 67 59 4 13 31 32 13 0 11 13 11 0 0
Louisiana (SPS-1) 808 170 266 319 138 468 64 128 149 74 277 64 74 74 0 64 64 53 43 0
Maine (SPS-5) 30 35 40 35 5 20 20 25 10 0 0 20 10 0 0 0 10 0 0 0
Maryland (SPS-5) 139 44 44 11 6 28 33 11 11 0 0 6 6 0 0 0 6 0 0 0
Minnesota 148 131 296 312 33 115 82 99 82 0 66 33 33 16 0 33 16 16 0 0
(SPS-5)
New Mexico 8 8 8 16 0 0 8 8 8 0 0 0 0 0 0 0 0 0 0 0
(SPS-1)
New Mexico 12 8 8 9 5 5 2 2 3 0 3 0 0 0 0 0 0 0 0 0
(SPS-5)
Pennsylvania 95 27 13 13 1 20 13 10 9 1 8 10 8 1 0 2 8 1 0 0
(SPS-6)
Tennessee 1,173 16 4 4 0 30 2 2 0 0 3 1 0 0 0 1 0 0 0 0
(SPS-6)
Virginia (SPS-1) 25 35 4 7 4 0 0 4 4 0 0 0 0 0 0 0 0 0 0 0

Table 6.18. Events per Year Scaled to ADTT  2,500 (continued)

Wisconsin 24 12 20 16 8 4 0 12 12 4 0 0 4 4 0 0 0 0 0 0
(SPS-1)
California 0 10 20 24 20 0 1 0 0 5 0 0 0 0 0 0 0 0 0 0
Antelope EB
California 0 20 48 68 57 0 5 4 13 27 0 0 0 1 9 0 0 0 0 1
Antelope WB
California 0 1 1 4 8 0 0 0 1 1 0 0 0 0 1 0 0 0 0 0
Bowman
California 2 20 31 31 17 0 7 11 11 3 0 1 5 4 0 0 0 1 0 0
LA-710 NB
California 1 12 21 22 11 0 3 9 9 3 0 1 4 4 0 0 0 0 0 0
LA-710 SB
California Lodi 0 25 32 65 96 0 1 4 13 39 0 0 0 1 9 0 0 0 0 1
Florida I-10 151 76 86 142 82 44 22 26 42 21 12 9 8 9 3 6 3 2 3 1
Florida I-95 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Mississippi I-10 0 2 3 6 4 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0
Mississippi I-55UI 0 2 3 6 4 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0
Mississippi I-55R 93 66 167 229 58 13 21 33 40 22 5 5 11 14 13 1 2 3 5 6
Mississippi US-49 0 2 8 10 5 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0
Mississippi US-61 0 6 23 40 29 0 0 6 11 6 0 0 6 6 0 0 0 0 0 0
Florida US-29 2,922 2,252 1,473 1,122 462 1,524 1,155 751 572 247 840 621 406 278 119 413 373 191 138 49
Annual Average 117.0 37.8 50.6 58.7 21.7 32.0 18.4 20.8 19.8 7.5 14.3 9.7 9.1 5.8 1.2 4.0 5.6 3.2 1.7 0.3
185
186
120
>= 1.0HL93
100 >= 1.1HL93
>= 1.2HL93
80
Annual Average
>= 1.3HL93
60
40
20
0
30 ft 60 ft 90 ft 120 ft 200 ft
Span (ft.)
Figure 6.35. Annual average events scaled to ADTT  2,500 versus span.
(continued from page 180) load factor higher than the current value of 1.30 except pos-
is the same (i.e., the data are scalable). The average rate of sibly for locations with extraordinary levels of truck traffic.
exceedance in Table 6.18 is higher than in Table 6.17 because For example, for a location with an ADTT of 7,500, over
many of the WIM sites were on roads with ADTTs less than a 100-year service life the average exceedance would be on
2,500. Nevertheless, the rate at which 1.3 HL-93 was exceeded the order of 1,000 events, although one of the sites shown in
remains quite low. The values in Table 6.18 can be scaled for Table 6.18 could see about 10 times that number. However,
locations with an ADTT other than 2,500 with the same slightly over half the sites recorded no events when the moments
assumption of scalability. for the spans indicated in Table 6.17 exceeded 1.3 HL-93 dur-
Question 5 asked, “Is another choice of load factor, other ing the recording period. Given the lack of field evidence of a
than the 1.30 used in the current Service II load combination, significant number of bridges with a permanent deformation
equally valid?” Given the relatively low number of exceedances due to overloads, it was not possible to establish an ADTT
in Table 6.18, it is difficult to rationalize the need for a national criterion at which the load factor should be increased.
120
30 ft
60 ft
100
90 ft
80 120 ft
Annual Average
200 ft
60
40
20
0
>= 1.0HL93 >= 1.1HL93 >= 1.2HL93 >= 1.3HL93
Ratio Truck/HL93
Figure 6.36. Annual average events scaled to ADTT  2,500 versus ratio truck/HL-93.

187
Judgment and experience are still necessary. This issue is also 1.3 LL +1.0 (DC + DW) ≤ 0.80 Fy for noncomposite sections
clouded by the issues in Questions 6 and 7, “Should this require- (6.12)
ment be applied to multilane loading?” and “If it is used as a
single-lane criterion, should the MPF of 1.2 currently used These are the current Service II limit state functions for
for single-lane loading in the AASHTO LRFD be applicable to investigation.
this condition?”
The issue of number of loaded lanes is discussed in Sec- 6.4.3.2 Select Structural Types and Design Cases
tion 5.2.4. For the WIM sites where data were recorded in
The Service II limit state is currently intended only for steel
two lanes, and given the definition of correlated events in that
superstructures and governs only for composite and compact
discussion, it was shown that the number of events of multiple
sections in the positive moment region. For these regions of
lanes loaded with correlated trucks was quite small, and the
composite and compact sections, the Service II limit state
histograms of gross vehicle weight showed that the number of
often governs the design over the Strength I limit state. Non-
events of two heavy trucks was even smaller. It was concluded
composite sections are not typically compact.
that multiple lanes of heavy trucks need not be considered for
Thus, the structure types being considered are positive
the SLSs. Thus, it was concluded that in most cases, design for
moment regions of steel girder superstructures that were
control of permanent distortions need not be based on multiple
modeled as simple spans, and the design case is a composite
lanes of overload (i.e., 1.3 HL-93). In the calibration process girder, which is the governing case.
described in Section 6.4.3, a single-lane loading with no MPF
was used on the load side of the limit state function.
To summarize, based on a review of the WIM data 6.4.3.3 Determine Load and Resistance Parameters
• There is little basis for lowering the current Service II load As discussed in the previous subsection, the Service II limit
factor. state can be investigated by concentrating on simple spans of
• Site-specific consideration of sites with unusually high composite steel girders. A set of 41 simple-span composite
volumes of heavy trucks is warranted. steel girder bridges was extracted from Mlynarski et al. (2011).
• Design for a single-lane loading is justified by this study. The flexural resistances of the interior girders of these bridges
• Elimination of the single-lane MPF of 1.20 for Service II is were used to study the Service II limit states. The documenta-
justified by this study. tion of the 41 bridges is given in Appendix F.
It was established in Chapter 5 that although the Service II
limit state is evaluated assuming multiple lanes loaded, the
6.4.3 Calibration Procedure and Results
WIM study suggests that the Service II live load does not
6.4.3.1 Formulate Limit State Function occur often enough to warrant design for multiple lanes.
The Service II limit state function requires that the sum of
the factored loads must be less than or equal to the factored 6.4.3.4 Develop Statistical Models for Loads, Load
resistance, as shown by Equation 6.10: Combinations, and Resistance Variables
Uncertainties of Load
∑ γ Q ≤ ϕR
i i (6.10)
The uncertainties of the various components of dead load
i
were investigated previously with the results documented in
The two basic loads for Service II are dead load (DC + DW) Kulicki et al. (2007) and reproduced in Table 3.2.
and live load (LL); DC is the load factor for structural com- From Table 5.5 to Table 5.9, the uncertainties of live load are
ponents and attachments, and DW is the load factor for wear- taken as approximately 0.12 for the CV and 1.35 for the bias.
ing surfaces and utilities. Currently, the resistance is taken as
0.95 Fy for composite sections and 0.80 Fy for noncomposite Uncertainties of Resistance
sections with resistance factors of unity for both. The limited The uncertainties of the flexural resistance of composite steel
basis for these criteria was presented in Section 2.3.1.5. Sub- girders have also been investigated and similarly documented
stituting these variables and the current Service II load factors in Kulicki et al. (2007) and reproduced in Table 3.1.
of AASHTO LRFD Table 3.4.1-1 yields Equation 6.11 and
Equation 6.12: 6.4.3.5 Develop Reliability Analysis Procedure
1.3 LL +1.0 (DC + DW) ≤ 0.95 Fy for composite sections Monte Carlo simulation using Microsoft Excel formed the
(6.11) basis of the reliability analysis procedure for the Service II

188
limit states. Use of the Monte Carlo analysis is presented in 6.4.3.9 Calibrate Reliability Indices
Section 3.2.3.
Monte Carlo simulations were again performed for the
41 bridges by using a single lane of AASHTO LRFD live load
6.4.3.6 Calculate Reliability Indices for Current (which represents the load today as suggested by the WIM
Design Code or Current Practice studies) and dead load times the AASHTO LRFD load factors
compared with the flexural resistance consistent with multiple
The Service II limit state was first introduced with LFD in the
lanes of the AASHTO LRFD live load, as most of today’s new
AASHTO Standard Specifications for Highway Bridges (2002)
bridges will be designed. This assumes that design for multiple
as “overload” provisions. In the development of the AASHTO
lanes of live load will continue.
LRFD, a simple calibration was made in an attempt to yield
An average reliability index of 1.8 with a CV of 0.09 resulted
similar member proportions as with the Standard Specifications;
from using the current AASHTO LRFD load and resistance
this attempt is discussed in Section 2.3.1.5.
factors. Thus, the reliability is comparable to the inherent
The calculation of the inherent reliability indices for cur- reliability of current practice, but with much more unifor-
rent design practice was based on LFD overload provisions. mity and with a low CV compared with the original overload
Monte Carlo simulations were performed for the 41 bridges provisions.
by using a single lane of AASHTO LFD live load times the over-
load load factor of 5/3 compared with the flexural resistance
consistent with multiple lanes of the LFD live load. This is the 6.4.4 Proposed AASHTO LRFD Revisions
requirement for which most of today’s in-service bridges were The Service II limit state provisions do not require any
designed. The results are summarized in Table 6.19. In addi- modification from those of the AASHTO LRFD. Thus, the
tion, similar simulations were made assuming that the load Service II limit state will continue as defined in AASHTO LRFD
was multiple lanes of AASHTO LRFD live load, as the original Table 3.4.1-1.
conceivers of the limit state assumed traffic to be. Given the limited background on this limit state, the
These results suggest that the current inherent reliability possibility of reducing the demand was discussed informally
index associated with the Service II limit state is on average with three of the AASHTO Technical Committees: T-5, T-10,
about 2.0. The originally assumed multiple lanes of loading and T-14, although in the case of T-14 fewer than half the
suggest lower reliability but, as discussed, this loading is not members were present. Most of the members of these three
very probable. committees expressed reservations about decreasing the cur-
rent design requirement, citing the increasing numbers of
6.4.3.7 Review Results and Select Target trucks on the roads and the continual pressure to increase
Reliability Index legal loads. At the time of this work, an increase of about 20%
in legal gross vehicle weight is under discussion.
Using the values in Table 6.19, a target reliability index (bT)
of about 2.0 was chosen for the calibration of the Service II
limit state. 6.5 Tension in Prestressed
Concrete Beams,
Service III Limit State:
6.4.3.8 Select Potential Load and Annual Probability
Resistance Factors
Traditionally, prestressed concrete beams are proportioned
As an initial trial, the Service II load factors of the AASHTO for the SLS such that the concrete tensile and compressive
LRFD (1.3 for live load and 1.0 for dead load) were selected, stresses immediately after transfer and at the final stage are
along with resistance factors of unity. within certain stress limits defined in the specifications.
Under the current AASHTO LRFD (2012), two SLS load
combinations are used to calculate the stresses in prestressed
concrete components: the Service I and Service III load com-
Table 6.19. Inherent
Reliability Indices binations. The two service load combinations are described
as follows:
Live Load b CV
• Service I—Load combination relating to the normal opera-
Single lane (reality) 1.8 0.32
tional use of the bridge with a 55 mph wind and all loads
Multiple lane (assumed) 1.6 0.92
taken at their nominal values. Service I is also related to

189
deflection control in buried metal structures, tunnel liner frequency of the crack opening being sufficiently small that
plate, and thermoplastic pipe; to control crack width in adverse strand fatigue problems at crack locations are not
reinforced concrete structures; and for transverse analysis produced.
relating to tension in concrete segmental girders. This load
combination should also be used for the investigation of
slope stability. 6.5.1 History of Major Relevant Design
• Service III—Load combination for longitudinal analysis Provisions and Revisions to
relating to tension in prestressed concrete superstructures AASHTO LRFD
with the objective of crack control and to the principal
6.5.1.1 Load Factor for Live Load in Service III
tension in the webs of segmental concrete girders.
Load Combination
The load factors for DL and LL specified for the two load During the early stages of the development of AASHTO LRFD
combinations are as follows. in the early 1990s, only the Service I load combination was
considered for calculating all stresses in prestressed concrete
• Service I: DL load factor = 1.0, and LL load factor = 1.0. components. The load factor for live load was 1.0, which is the
• Service III: DL load factor = 1.0, and LL load factor = 0.8. same load factor used for service loads under the AASHTO
Standard Specifications for Highway Bridges, the predecessor
On the basis of the definition of the two limit states, the to AASHTO LRFD.
Service I limit state is used for calculating all service stresses in The design live load specified in AASHTO LRFD produces
the superstructure and substructure components at all stages, higher unfactored, undistributed load effects than that specified
with the exception that the Service III limit state is used to in the AASHTO Standard Specifications. The girder distribution
calculate the tensile stresses in the superstructure components factors, particularly for interior girders, for many typical girder
under full service loads and the principal tension in webs of systems in AASHTO LRFD are lower than those in Standard
segmental concrete. Specifications, thus reducing the difference between the unfac-
Stresses immediately after transfer are independent of the tored distributed load effects in the two specifications. Even
live loads. At the final stage, typically the design is controlled with the smaller distribution factor, the unfactored distributed
by the tensile stress in the concrete and not by the compressive load effects from AASHTO LRFD were higher for most girder
stresses on the opposite side of the girders. Thus, the calibration systems. Using the same load factor for SLS (1.0) resulted in
for prestressed concrete superstructures was performed for higher design-factored load effects for the AASHTO LRFD
the Service III limit state, and no calibration was performed designs than for those designed to the AASHTO Standard Speci-
for the Service I limit state. fications requirements. The results from the trial designs con-
In addition to being designed for the SLS, all prestressed ducted during the development of AASHTO LRFD indicated
concrete components are checked for the strength limit state. a larger number of strands than required by AASHTO Stan-
For typical precast prestressed superstructure beams (e.g., dard Specifications. This finding would suggest that designs
I-shapes, bulb-T shapes, and adjacent and spread box beams), performed under AASHTO Standard Specifications resulted
the controlling case of the design is usually the SLS. in underdesigned components that should have shown signs
The SLS stresses are calculated assuming an uncracked of cracking. In the absence of widespread cracking, the load
section. The concrete is assumed to be subjected to tensile factor for live load was decreased to 0.8, and the Service III load
stresses. However, due to the relatively low load factors used combination was created and was specified for tension in pre-
for the SLSs, it is highly probable that the structure is subjected stressed concrete components. This resulted in a similar num-
to heavy trucks that produce live load effects higher than those ber of strands for the designs conducted using both AASHTO
produced by the design-factored service loads. When a heavy Standard Specifications and AASHTO LRFD.
truck causes the tensile stress in the concrete to exceed the
modulus of rupture, the concrete is expected to crack. Once
6.5.1.2 Method of Calculating Prestressing Losses
the load passes, the prestressing force will cause the crack to
close, and it will remain closed as long as the concrete at the AASHTO LRFD (2012) includes three methods for determining
crack location remains under compression. However, if a the time-dependent prestressing losses. These three methods
truck heavy enough to cause the concrete stress calculated are as follows:
on the basis of the uncracked section to be tensile, the crack
will reopen. 1. Approximate method—This method is termed approxi-
Successful past performance of prestressed concrete com- mate estimate of time-dependent losses and is the least
ponents suggests that past design requirements result in a detailed method. It requires limited calculations to estimate

190
the time-dependent losses. Before 2005, the specifications the lower load factor used for the Service III load combination
included a simpler approximate method termed approxi- of AASHTO LRFD. Some of the work presented in the follow-
mate lump-sum estimate of time-dependent losses. The ing subsections was intended to investigate the effect of dif-
lump-sum method allowed selecting a value for the time- ferent loss methods and different design specifications on the
dependent losses from a table. The value varied according reliability index for the Service III load combination.
to the type of girders and the type and grade of prestressing
steel. Some concrete compressive strength requirements
6.5.2 Live Load Model
were allowed to use this method.
2. Refined estimates of time-dependent losses—This method is Traditionally, prestressed concrete components have been
more detailed than the approximate method. More details designed for the number of traffic lanes, including MPFs, which
on this method are presented below. produced the highest load effects. This was assumed to continue
3. Time-step method—This method is highly detailed and is in the future, and all sections designed as part of this study used
based on tracking the changes in the material properties this approach.
with time. The loss calculations are based on the time of the However, as indicated in Section 5.2.4, the presence of heavy
application of loads and the material properties at the time loads in adjacent traffic lanes simultaneously is not likely. Thus,
of the load application. This method is required to be used the load side of the limit state function in the reliability analysis
in the design of posttensioned segmental bridges. It may also was calculated assuming the live load existed in only one lane,
be used for other types of bridges; however, due to the level and no MPF was included. The design truck, tandem, and
of effort required, it is typically limited to segmental bridges. uniform lane load specified in AASHTO LRFD were used unless
otherwise noted. The live load distribution factors specified in
Throughout the remainder of this section, unless explicitly AASHTO LRFD were used in distributing the design loads. The
indicated otherwise, time-dependent losses are calculated dynamic load allowance (10%) used in the original calibration
using the techniques outlined in Refined Estimates of Time- of the strength limit state in AASHTO LRFD was applied to
Dependent Losses in AASHTO LRFD. the load side.
Originally, the method of calculating prestressing force The return period considered in the calibration of the Ser-
losses in AASHTO LRFD (the “pre-2005” method) was the vice III limit state was 1 year. This return period was selected
same method used in AASHTO Standard Specifications. A new because the live load statistics were developed based on 1 year
method of loss calculations (the “post-2005” method) first of reliable WIM data from various WIM sites. Furthermore,
appeared in the 2005 Interim to the third edition of AASHTO as only three of the 32 WIM sites had an ADTT larger than
LRFD. The post-2005 method is thought to produce a more 5,000, and only one of the 32 WIM sites had an ADTT larger
accurate estimate of the losses. The post-2005 method has than 8,000, an ADTT of 5,000 was used for the bulk of the
new equations for calculating the time-dependent prestressing calibration. The bias and CV of live load were taken as shown
losses, and it also introduced the concept of “elastic gain.” After in Table 5.5 to Table 5.9.
the initial prestressing loss at transfer, when load components
that produce tensile stresses in the concrete at the strand
locations are applied to the girder, the strands are subjected to 6.5.3 Methods of Analysis for Study Bridges
an additional tensile strain equal to the strain in the surround- Unless explicitly indicated otherwise, the methods of analysis
ing concrete due to the application of the loads. This results in used in designing and analyzing the study bridges throughout
an increase in the force in the strands. The increase in the force Section 6.5.4 and Section 6.5.5 are listed below.
in the strands was termed “elastic gain,” and the post-2005 pre- For bridges designed or analyzed using the post-2005 pre-
stressing loss method allows including elastic gain to be used to stressing loss method,
offset some of the losses.
When the elastic gain was considered, the post-2005 pre- • The time-dependent prestressing loss method used is the
stress loss method produced lower prestressing force losses method designated in AASHTO LRFD (2012) as the Refined
than the earlier method. The reduction in prestressing losses Estimates of Time-Dependent Losses.
resulted in fewer strands than what was required under • The section properties used in the analysis were based on
AASHTO Standard Specifications and under earlier editions the gross section of the concrete.
of AASHTO LRFD. This change raised some concern as some • The calculations of prestressing losses considered the effects
practitioners and researchers thought that the higher pre- of elastic gain as allowed by the current design provisions.
stressing losses calculated using the pre-2005 loss method
compensated for the lower live load effects caused by the lower Regardless of the method of design used in designing a
design live load used in AASHTO Standard Specifications or girder, the stresses in the girder used as part of the reliability

191
index calculations were determined by analyzing the girder produced uniform reliability. The bulk of the calibration
using the above assumptions. was performed using a crack width of 0.016 in. The differen-
For bridges designed using the pre-2005 prestressing loss tiation between different environments was accounted for
method, through the use of different reliability indices in association
with the same crack width.
• The time-dependent prestressing loss method used is the
method designated in the pre-2005 AASHTO LRFD editions For each girder, the design was performed according to
as the Refined Estimates of Time-Dependent Losses. certain stress limits, as is conventionally done, and the girder
• The section properties used in the analysis were based on the section and number of strands were determined. The reli-
gross section of the concrete. ability index was determined for each of the three limit state
• The calculations neglected the effects of elastic gain. functions described above by using the same girder design
(i.e., the same girder section and same number of strands).
Each of the limit state functions requires a different level
6.5.4 Target Reliability Index of loading before the criteria are violated. The frequency at
In the development of AASHTO LRFD, the target reliability which any of the three limit states will be violated and the
index for the strength limit states was 3.5. The limit state was corresponding reliability index depend on the level of loading
assumed to be violated when the applied load effects exceeded required to cause the limit state to be violated. For a specific
the resistance, which was in turn assumed to be equal to the cross section with a specific prestressing area and force, reach-
design-factored load. Failure under the strength limit state is ing the decompression limit state requires less applied load
well defined as it relates to a certain criterion related to the than reaching a specified tensile stress, which in turn requires
properties of the materials used (such as steel yield stress less load than that required to reach a specific crack width.
or concrete compressive strength) or to a behavior criterion Requiring a higher load to violate a specific limit state means
that, if violated, may lead to instability of the component (such that the section resistance is higher, which would cause the
as local or global buckling). Due to the lack of clear conse- curve representing the resistance in Figure 6.1 to be shifted to
quences for violating the limiting stress specified for the con- the right. This results in a higher reliability index. Table 6.20
crete in a prestressed concrete component, selecting the limit shows the required load and the corresponding reliability index
for the three limit states relative to each other.
state function required investigating different alternatives.
With the target reliability index dependent on the definition
of the limit state, selecting the target reliability index required
6.5.4.1 Limit State Functions Investigated investigating all three criteria and selecting the one that pro-
vided more uniform reliability across a wide range of bridge
The following three limit state functions were investigated:
geometrical characteristics.
• Decompression limit state—This limit state assumes that
failure occurs when the stress in the concrete on the tension 6.5.4.2 Statistical Parameters of Variables
face calculated on the basis of the uncracked section under Included in the Design
the combined effect of factored dead load and live load ceases
Several variables affect the resistance of prestressed compo-
to be in compression.
nents. Table 6.21 shows a list of variables considered to be
• Stress limit state—This limit state assumes that failure
occurs when the tensile stress in the concrete on the ten-
sion face calculated on the basis of the uncracked section Table 6.20. Relation Between Limiting Criteria
under the combined effect of factored dead load and live and Reliability Index for a Given Girder
load exceeds a certain tensile stress limit calculated on
Live Load
the basis of the uncracked section properties regardless of Required to Frequency
whether the section has previously been cracked. Stress Violate the of Exceeding
limits of ft = 0.0948 fc′, ft = 0.19 fc′, and ft = 0.25 fc′ were Limiting the Limiting Reliability
Limiting Criterion Criterion Criterion Index
initially considered in the reliability analysis; ft = 0.19 fc′
was used for the final calibration. Decompression Lowest Highest Lowest
• Crack width limit state—This limit state assumes that fail- Maximum allowable Middle Middle Middle
ure occurs when a previously formed crack in the concrete tensile stress limit
opens, and the crack width reaches a certain prespecified Maximum allowable Highest Lowest Highest
crack width. Crack widths of 0.008, 0.012, and 0.016 in. crack width limit
state
were initially considered in the reliability analysis, but none

192
Table 6.21. Random Variables and the Value of Their Statistical Parameters
Variable Distribution Mean () CV () Remarks
As Normal 0.9Asn 0.015 Siriaksorn and Naaman (1980)

Aps Normal 1.01176Apsn 0.0125 Siriaksorn and Naaman (1980)
b, b0, b1, bw Normal bn 0.04 Siriaksorn and Naaman (1980)
CEc Normal 33.6 0.1217 Siriaksorn and Naaman (1980); nominal = 33
CEc = Ec/(g 1.5
c z fc′ )
Cfci Normal 0.6445 0.073 nominal = 0.8; Cfci = fci /f c′

dp, ds Normal dpn, dsn 0.04 Siriaksorn and Naaman (1980)
e1 Normal e0n 0.04 Siriaksorn and Naaman (1980)
Eps Normal 1.011Epsn 0.01 Siriaksorn and Naaman (1980); Epsn = 29,000 ksi
Es Normal Esn 0.024 Siriaksorn and Naaman (1980)
f c′ Lognormal 1.11f c′n 0.11 Nowak et al. (2008)
fpu Lognormal 1.03fpun 0.015 Nowak et al. (2008); fpun = 270 ksi
fsi Normal 0.97fsin 0.08 Developed from Gross and Burns (2000)
fy Lognormal 1.13fyn 0.03 Nowak et al. (2008)
h, hf, hf1, hf2 Normal hn, hfn, hf1n, hf2n 0.025 Siriaksorn and Naaman (1980)
l Normal ln 11/(32µ) Siriaksorn and Naaman (1980)
gc Normal gcn = 150 0.03 Siriaksorn and Naaman (1980)
Dfs Normal 1.05Dfsn 0.10 Developed from Gross and Burns (2000) and Tadros et al. (2003)
S0 Normal S0n 0.03 Siriaksorn and Naaman (1980)
Note: Subscript n refers to nominal values.
Notations:
As = area of nonprestressing steel (in.2);
Aps = area of prestressing steel in tension zone (in.2);
b = prestressed beam top flange width (in.);
b0 = deck width transformed to beam material (in.);
b1 = prestressed beam bottom flange width (in.);
bw = web thickness (in.);
c = depth of neutral axis from extreme compression fiber (in.);
Cfci = fci /f′c;
dp = distance from extreme compression fiber to centroid of prestressing steel (in.);
ds = distance from extreme compression fiber to centroid of nonprestressing steel (in.);
e1 = eccentricity of prestressing force with respect to centroid of the section at midspan (in.);
Eps = modulus of elasticity of prestressing steel (psi);
Es = modulus of elasticity of nonprestressing steel (psi);
f′c = specified compressive strength of concrete (psi);
fpu = specified tensile strength of prestressing steel (psi);
fsi = initial stress in prestressing steel (psi);
fy = yield strength of nonprestressing steel (psi);
h = girder depth (in.);
hf = deck thickness (in.);
hf1 = top flange thickness (in.);
hf2 = bottom flange thickness (in.);
l = clear span length of beam members (ft);
gc = unit weight of concrete (lb/ft3);
Dfs = prestress losses (psi); and
S0 = sum of reinforcing element circumferences (in.).

193
random during the performance of the reliability analyses. Table 6.22. Summary of Reliability Indices for
These variables represent a summary of the information from Existing I- and Bulb-T Girder Bridges with One Lane
research studies by Siriaksorn and Naaman (1980) and Nowak Loaded and Return Period of 1 Year
et al. (2008).
ADTT
Performance Level 1,000 2,500 5,000 10,000

6.5.4.3 Database of Existing Bridges
Decompression 0.95 0.85 0.74 0.61
A database of existing prestressed concrete girder bridges was Maximum ft = 0.0948 fc′ 1.15 1.01 0.94 0.82
extracted from the database of bridges used in the NCHRP tensile
stress limit ft = 0.19 fc′ 1.24 1.14 1.05 0.95
12-78 project (Mlynarski et al. 2011). The database used in
this study included 30 I- and bulb-T girder bridges, 31 adja- ft = 0.25 fc′ 1.40 1.27 1.19 1.07
cent box girder bridges, and 36 spread box girder bridges. Maximum 0.008 2.29 2.21 1.99 1.85
The geometric characteristics of the bridges are included in crack
0.012 2.65 2.60 2.37 2.22
Appendix F. width (in.)
0.016 3.06 2.89 2.69 2.56
Depending on the environmental exposure conditions, both
AASHTO Standard Specifications and AASHTO LRFD allow
designing conventional prestressed components for a maxi-
mum concrete tensile stress of ft = 0.0948 fc′ or ft = 0.19 fc′ girder spacing of 6, 8, 10, and 12 ft. This database was analyzed
for severe corrosion conditions or no worse than moderate to determine the effect of the change in the method of estimat-
corrosion conditions, respectively. When either specification is
ing prestressing losses (pre-2005 and post-2005 methods)
applied without owner’s exceptions, most bridges are designed
and the design environment (severe corrosive conditions and
for ft = 0.19 fc′, with a small number of bridges in coastal areas
normal or not worse than moderate corrosion conditions). The
designed for ft = 0.0948 fc′. The stress limit for which each
two environmental conditions were signified by the maximum
bridge in the database was designed was unknown. As the per-
concrete tensile stress limit (ft = 0.0948 fc′ or ft = 0.19 fc′)
centage of bridges designed for severe corrosive conditions is
used in the design. The four cases of design considered were
small, it was assumed that most bridges in the database were
as follows:
likely to have been designed for the higher limit.
The construction dates of the bridges considered suggest that
Case 1: AASHTO LRFD with maximum concrete tensile stress
they were all designed using the prestressing loss provisions
of ft = 0.0948 fc′ and pre-2005 prestress loss method;
method that existed in both the AASHTO Standard Specifica-
tions and the pre-2005 AASHTO LRFD.
The database of existing bridges was used to estimate the of ft = 0.0948 fc′ and post-2005 prestress loss method;
reliability index inherent in the existing bridge system and Case 3: AASHTO LRFD with maximum concrete tensile stress
used this as the starting point for the calibration. of ft = 0.19 fc′ and pre-2005 prestress loss method; and
of ft = 0.19 fc′ and post-2005 prestress loss method.
6.5.4.4 Estimated Reliability Index
of Existing Bridges Table 6.23 and Table 6.24, respectively, show the span length
and girder spacing along with the calculated reliability indices
Table 6.22 summarizes the average reliability indices for the for I-girder bridges designed for maximum concrete tensile
existing I- and bulb-T girder bridges database. For example, stress ft = 0.0948 fc′ (Case 1 and Case 2) and ft = 0.19 fc′
the average reliability indices at decompression level, maxi- (Case 3 and Case 4) for ADTT = 5,000.
mum allowable tensile stress limit under service loads of In performing the design, the cases using the post-2005
ft = 0.19 fc′, and maximum allowable crack width limit of prestress loss method (Case 2 and Case 4) were designed using
0.016 in. are 0.74, 1.05, and 2.69, respectively, for an ADTT the smallest possible AASHTO girder size. To facilitate the com-
of 5,000 and a return period of 1 year. parisons, when possible, Case 1 and Case 3 were then designed
using the same AASHTO section used for Case 2 and Case 4,
6.5.4.5 Database of Simulated Bridges respectively. For the cases for which the section used for Case 2
or Case 4 was too small to be used for the corresponding
A database of simulated simple-span bridges was designed Case 1 or Case 3, no design is shown in Table 6.23 for Case 1 or
using AASHTO I-girder sections for four cases. The simu- in Table 6.24 for Case 3. For the 140-ft span bridges with 12-ft
lated bridges have span lengths of 30, 60, 80, 100, and 140 ft and girder spacing, no AASHTO I-girder section was sufficient.

194
Table 6.23. Summary of Reliability Indices of Simulated Bridges Designed Using AASHTO Girders
with ADTT 5 5,000 and ft 5 0.0948 fc′
Case 1 Case 2
Designed Using Pre-2005 Loss Method Designed Using Post-2005 Loss Method
Span Spacing Maximum Maximum Maximum Maximum

Case Section Type Length (ft) (ft) Decompression Tensile Crack Decompression Tensile Crack
1 AASHTO I 30 6 1.05 1.49 2.92 1.03 1.51 2.55

2 AASHTO I 30 8 0.90 0.94 2.41 0.93 1.00 2.32
3 AASHTO I 30 10 1.16 1.68 2.87 1.28 1.67 2.82
4 AASHTO I 30 12 1.28 1.67 2.91 0.63 0.97 2.29
Average for 30-ft span 1.10 1.45 2.78 0.97 1.29 2.50
5 AASHTO II 60 6 0.66 1.01 3.35 0.23 0.61 2.47
6 AASHTO II 60 8 — — — 0.73 1.04 2.42
7 AASHTO III 60 10 1.22 1.62 3.01 0.43 0.76 1.97
8 AASHTO III 60 12 1.57 1.96 3.68 0.73 0.99 2.51
9 AASHTO III 80 6 1.35 1.66 4.1 0.61 0.92 3.07
10 AASHTO III 80 8 1.8 2.14 5.23 0.82 1.13 3.64
11 AASHTO III 80 10 — — — 0.90 1.19 2.93
12 AASHTO IV 80 12 2.2 2.49 5.11 0.83 1.17 3.32
13 AASHTO III 100 6 — — — 1.45 1.85 3.51
14 AASHTO IV 100 8 1.86 2.00 3.86 1.33 1.43 3.44
15 AASHTO IV 100 10 — — — 1.33 1.65 3.37
16 AASHTO V 100 12 1.68 1.99 4.08 0.93 1.24 3.33
17 AASHTO IV 120 6 — — — 1.32 1.76 3.81
18 AASHTO V 120 8 1.54 2.05 3.65 0.92 1.4 3.14
19 AASHTO V 120 10 — — — 0.95 1.46 3.02
20 AASHTO VI 120 12 1.82 2.26 3.88 0.9 1.35 3.38
21 AASHTO VI 140 6 1.48 1.99 3.91 0.86 1.36 2.32
22 AASHTO VI 140 8 — — — 0.99 1.47 2.79
23 AASHTO VI 140 10 — — — 1.05 1.53 3.22
24 na 140 12 — — — — — —
Average for all spans 1.44 1.80 3.66 0.92 1.28 2.94
Note: — = there is no design; na = not applicable because no AASHTO I-girder was sufficient.

195
Table 6.24. Summary of Reliability Indices of Simulated Bridges Designed Using AASHTO Girders
with ADTT 5 5,000 and ft 5 0.19 fc′
Case 3 Case 4
Designed Using Pre-2005 Loss Method Designed Using Post-2005 Loss Method
Span Spacing Maximum Maximum Maximum Maximum

Case Section Type Length (ft) (ft) Decompression Tensile Crack Decompression Tensile Crack
1 AASHTO I 30 6 1.00 1.55 2.39 0.97 1.55 2.46

2 AASHTO I 30 8 0.94 0.92 2.35 0.91 1.00 2.16
3 AASHTO I 30 10 1.29 1.66 2.91 1.18 1.66 2.79
4 AASHTO I 30 12 1.30 1.72 3.02 1.26 1.70 2.91
5 AASHTO II 60 6 0.74 1.13 3.11 0.18 0.58 2.41
6 AASHTO II 60 8 1.04 1.39 2.82 0.28 0.66 1.91
7 AASHTO III 60 10 0.42 0.79 2.05 0.42 0.78 2.07
8 AASHTO III 60 12 0.66 1.00 2.5 0.68 0.96 2.53
9 AASHTO III 80 6 0.56 0.97 3.13 0.13 0.51 2.53
10 AASHTO III 80 8 1.06 1.46 3.43 0.42 0.78 3.2
11 AASHTO III 80 10 1.58 1.84 3.65 0.37 0.65 2.72
12 AASHTO IV 80 12 0.83 1.15 3.72 0.51 0.87 3.11
13 AASHTO III 100 6 — — — 0.82 1.23 3.44
14 AASHTO IV 100 8 1.31 1.42 3.60 0.69 0.76 2.76
15 AASHTO IV 100 10 1.80 1.98 3.67 0.75 1.04 3.12
16 AASHTO V 100 12 1.08 1.37 3.43 0.40 0.72 2.55
17 AASHTO IV 120 6 1.53 1.98 3.71 0.70 1.28 3.10
18 AASHTO V 120 8 0.90 1.30 3.31 0.46 0.85 2.55
19 AASHTO V 120 10 1.25 1.65 3.35 0.26 0.78 2.68
20 AASHTO VI 120 12 1.19 1.66 3.37 0.47 0.91 2.69
21 AASHTO VI 140 6 0.84 1.41 3.23 0.28 0.82 2.41
22 AASHTO VI 140 8 1.22 1.68 3.30 0.53 0.98 3.04
23 AASHTO VI 140 10 — — — 0.62 1.08 2.46
24 na 140 12 — — — — — —
Average for all spans 1.07 1.43 3.15 0.58 0.96 2.68
Note: — = there is no design; na = not applicable because no AASHTO I-girder was sufficient.

196
Bridges designed for Case 1 and Case 3 are also thought to two columns showing the reliability indices of the simulated
be similar to those designed using AASHTO Standard Speci- bridges are for cases for which the pre-2005 prestressing loss
fications for the two environmental conditions. The reliability method was used, as these are thought to better represent the
indices calculated for Case 1 and Case 3 represent the inherent bridges currently on the system.
reliability of bridges currently on the system, as most of them For example, the reliability index at the decompression
were designed before 2005. Case 2 and Case 4 generally repre- performance level for existing bridges, simulated bridges
sent the inherent reliability of newer bridges designed using designed for severe environments, and simulated bridges
the 2005 and later versions of AASHTO LRFD for severe and designed for normal environments was around 0.74, 1.44, and
normal environmental conditions, respectively. 1.07, respectively (see Table 6.22 to Table 6.25). Consequently,
Comparing Case 1 with Case 2 and Case 3 with Case 4 shows a target reliability index of 1.2 and 1.0 was selected for the
the effect of changing the prestressing loss method. decompression performance level for bridges designed for
Using the post-2005 prestress loss method resulted in a severe environments and bridges designed for normal envi-
smaller number of strands than the pre-2005 loss method. ronments, respectively. A reliability index of 1.0 means that
As shown in Table 6.23 and Table 6.24, the lower number of 15 of 100 bridges will probably have the bottom of the girder
strands resulted in lower reliability indices for bridges designed decompress in any given year.
using the post-2005 prestress loss method.
As shown in Table 6.23 and Table 6.24, regardless of the
6.5.5 Calibration Result
loss method and/or the limit state used, the reliability indices
for each case varied significantly. This variation in values The basic steps of the calibration process are shown below as
suggested the need to calibrate the limit state to develop a they relate to the Service III calibration.
combination of load and resistance factors to produce a more
uniform reliability index across the range of different span
6.5.5.1 Step 1: Formulate Limit State Function
lengths and girder spacings.
and Identify Basic Variables
The three limit state functions that were investigated are listed
6.5.4.6 Selection of Target Reliability Index
in Section 6.5.4.1. The limit state function is formulated by
The target reliability indices were selected on the basis of deriving an expression for the resistance prediction equation.
the calculated average values of the reliability levels of exist- For the decompression and tensile stress limits, the stress in
ing bridges and previous practice, with some consideration the concrete is calculated as it is usually done for the design of
given to experiences from other codes (Eurocode and ISO 2394 prestressed concrete components. For the crack width limit
document). A return period of 1 year and an ADTT equal to state, Appendix D presents a detailed derivation of the resis-
5,000 were used. tance prediction equation for a typical prestressed concrete
Table 6.25 shows the target reliability indices selected in bridge girder. The derived equation considers uncracked and
this study, as well as the reliability indices for the existing and cracked section behavior in a general format by including
simulated bridge databases. Note that the environmental the crack width equation. In lieu of setting the stress to zero,
condition for existing bridges was not known and that the the resistance for the decompression limit state can also be
Table 6.25. Reliability Indices (b) for Existing and Simulated Bridges with Return Period of 1 Year
and ADTT 5 5,000
Average b for
Simulated Bridges Simulated Bridges

Designed for Designed for Proposed Target b for
ft 5 0.0948 f c′ ft 5 0.19 f c′
Existing Bridges and Pre-2005 and Pre-2005 Bridges in Severe Bridges in Normal
Performance Level in NCHRP 12-78 Loss Method Loss Method Environment Environment
Decompression 0.74 1.44 1.07 1.20 1.00

Maximum allowable tensile 1.05 1.80 1.43 1.50 1.25
stress of ft = 0.19 fc′
Maximum allowable crack 2.69 3.68 3.15 3.30 3.10
width of 0.016 in.

197
derived by setting the crack width to zero in the general equa- generated independently on the basis of the statistics and dis-
tion for crack width. tribution of that random variable. For each simulation, the
The majority of the equations for the prediction of the dead load and the resistance were calculated using one of
maximum crack width are given in terms of the stress in the the 1,000 sets of values of each random variable, resulting in
steel. Various maximum crack width prediction equations 1,000 values of the dead load and the resistance. The mean and
were evaluated using test data available in the literature. Appen- standard deviation of the dead load and the resistance were
dix C presents a comparison and evaluation of maximum then calculated on the basis of the 1,000 simulations.
crack width prediction equations for prestressed concrete
members.
6.5.5.6 Step 6: Calculate Reliability Indices for
Current Design Code and Current Practice
6.5.5.2 Step 2: Identify and Select Representative
Using the statistics of the dead load and the resistance calcu-
Structural Types and Design Cases
lated from Monte Carlo simulation (as described above) and
Various design cases for span lengths ranging from 30 to 140 ft the statistics of the live load as derived from the WIM data
were designed, as shown in Section 6.5.4.5. For a maximum (as described in Section 5), the reliability index was calculated
crack width limit state, a crack width of 0.016 in. was consid- for each girder.
ered. For the maximum allowable stress limit state, the stress The reliability index (b) was calculated using Equation 6.13:
considered is as stated in the discussion included in the fol-
lowing sections. µ R − µQ
β= (6.13)
σ 2R + σ Q2
6.5.5.3 Step 3: Determine Load and Resistance
Parameters for Selected Design Cases where
µR = mean value of resistance;
The variables included the dimension of the cross section and
µQ = mean value of applied loads;
the material properties. The statistical information included the
sR = standard deviation of resistance; and
probability distribution and statistical parameters, such as
sQ = standard deviation of applied loads.
mean (µ) and standard deviation (s).
The calculated reliability indices of existing and simulated
bridges are shown in Table 6.22 to Table 6.24.
for Load and Resistance
6.5.5.7 Step 7: Review Results and Select Target
The variables affecting the load and resistance were identified.
Reliability Index
These variables included live load; those affecting resistance,
such as the dimensions of the cross section; and the material The initial target reliability index (bT) was determined as shown
properties. The statistical information included the probability in Table 6.25.
distribution and statistical parameters for live load presented
in Section 5.3.2 and for other variables affecting the resis-
6.5.5.8 Step 8: Select Potential Load and
tance presented in Section 6.5.4.2.
Resistance Factors for Service III
For all steps, the resistance factor was assumed to be the same
as in the current AASHTO LRFD (2012) (i.e., equal to 1.0).
Analysis Procedure
The Service III limit state resistance is affected by the ten-
The statistical information of all the required variables was sile stress limit used in the design. Therefore, in addition to
used to determine the statistical parameters of the resistance trying different load factors, different stress limits for the
by using Monte Carlo simulation. Monte Carlo simulation is design were also investigated. Maximum concrete design ten-
useful in generating a large number of random cases that sile stresses of ft = 0.0948 fc′, ft = 0.19 fc′, and ft = 0.25 fc′
are used in defining the mean and standard deviation of the were considered. In addition, the simulated bridge database
resistance. used in determining the target resistance factor was expanded
For each girder, Monte Carlo simulation was performed to allow longer spans.
for each random variable associated with calculation of the Because there were three concrete tensile stress limits, Step 8
resistance and dead load. One thousand simulations were is divided into three repetitions designated 8a, 8b, and 8c. For
performed. For each random variable, 1,000 values were this step, the range of span lengths was increased to 220 ft.

198
Step 8a: Select Potential Load and Resistance Factors of the bridges are shown in Table 6.26. The average reliability
for Service III: Bridges Designed for Maximum Concrete index for the decompression limit state, maximum allowable
Tensile Stress of ft = 0.0948 fc′ tensile stress limit state, and maximum allowable crack width
The calibration for a selected bridge database (shown in limit state were 0.97, 1.31, and 3.06, respectively. As the reli-
Table 6.26) was performed assuming an ADTT of 5,000 and ability indices were lower than the target reliability indices
a maximum concrete design tensile stress of ft = 0.0948 fc′. and were not uniform across different spans, modifications
to the load factor were applied in the next step in an attempt
1. Calculate the reliability level of designs according to to achieve higher, and more uniform, reliability indices.
AASHTO LRFD (2012) (Figure 6.37 to Figure 6.39). 2. Redesign the bridges with a live load factor of 1.0.
Figures 6.37 to 6.39 show the reliability indices for the
bridges designed using AASHTO type girders according to In this step, the bridges were redesigned using a live load
AASHTO LRFD (2012), including a load factor of 0.8 for the factor of 1.0, and the dead load and resistance factors were
Service III limit state, and assuming a maximum concrete kept the same. Table 6.27 shows the design geometric charac-
tensile stress of ft = 0.0948 fc′. The geometric characteristics teristics of the redesigned bridges.
Table 6.26. Summary Information of Bridges Designed

with gLL 5 0.8 and ft 5 0.0948 fc′
Span Girder Aps

Case Section Type Length (ft) Spacing (ft) (in.2) No. of Strands
1 AASHTO I 30 6 1.224 8
2 AASHTO I 30 8 1.530 10
3 AASHTO I 30 10 1.836 12
4 AASHTO I 30 12 2.142 14
5 AASHTO II 60 6 2.448 16
6 AASHTO II 60 8 3.366 22
7 AASHTO III 60 10 3.060 20
8 AASHTO III 60 12 3.672 24
9 AASHTO III 80 6 3.672 24
10 AASHTO III 80 8 4.590 30
11 AASHTO III 80 10 5.508 36
12 AASHTO IV 80 12 5.202 34
13 AASHTO III 100 6 6.120 40
14 AASHTO IV 100 8 6.426 42
15 AASHTO IV 100 10 7.344 48
16 AASHTO V 100 12 7.038 46
17 AASHTO IV 120 6 7.956 52
18 AASHTO V 120 8 7.956 52
19 AASHTO V 120 10 9.180 60
20 AASHTO VI 120 12 8.874 58
21 AASHTO VI 140 6 8.262 54
22 AASHTO VI 140 8 9.792 64
23 AASHTO VI 140 10 11.322 74
24 AASHTO VI 140 12 — —
25 FIB-96 160 6 5.508 36

199
Table 6.26. Summary Information of Bridges Designed

with gLL 5 0.8 and ft 5 0.0948 fc′ (continued)
Span Girder Aps

26 FIB-96 160 8 6.426 42

27 FIB-96 160 10 7.344 48
28 FIB-96 160 12 — —
29 FIB-96 180 6 7.344 48
30 Mod. BT-72 180 9 16.218 106
31 Mod. AASHTO VI 180 9 15.912 104
32 Mod. AASHTO VI 200 9 20.502 134
33 Mod. NEBT-2200 200 9 16.830 110
34 Mod. W95PTMG 200 9 16.830 110
35 Mod. NEBT-2200 220 9 20.808 136
Note: — = a practical solution was not found.
Figure 6.40 to Figure 6.42 show the reliability indices for more uniform than for the case of using a live load factor of
the redesigned bridges using a live load factor of 1.0. The 0.8, particularly for the decompression and maximum tensile
average reliability indices for the decompression limit state, stress limit states. Consequently, a live load factor of 1.0 was
the maximum allowable tensile stress limit state, and the proposed if the tensile stress was limited to ft = 0.0948 fc′.
maximum allowable crack width limit state were 1.33, 1.70,
and 3.32, respectively. The reliability level of bridges became Step 8b: Select Potential Load and Resistance Factors
for Service III: Bridges Designed for Maximum Concrete
Tensile Stress of ft = 0.19 fc′
The work described under Step 8a was repeated, except the
Reliability Index
girders were redesigned assuming a maximum concrete tensile

stress of ft = 0.19 fc′.
1. Calculate the reliability level of designs according to

AASHTO LRFD (2012) with a maximum concrete tensile
stress for design of ft = 0.19 fc′ (Figure 6.43 to Figure 6.45).
Span Length (ft.) 2. Redesign the bridges with a live load factor of 1.0.
Figure 6.37. Reliability indices for bridges
at decompression limit state (ADTT  5,000, Figure 6.46 to Figure 6.48 show the reliability indices for the
gLL  0.8, and ft  0.0948 fc′ ). redesigned bridges using a live load factor of 1.0 and ft = 0.19
Reliability Index
Reliability Index
Span Length (ft.) Span Length (ft.)
Figure 6.38. Reliability indices for bridges Figure 6.39. Reliability indices for bridges at
at maximum allowable tensile stress limit state maximum allowable crack width limit state
(ADTT  5,000, gLL  0.8, and ft  0.0948 fc′ ). (ADTT  5,000, gLL  0.8, and ft  0.0948 fc′ ).

200
Table 6.27. Summary Information of Bridges Designed with

gLL 5 1.0 and ft 5 0.0948 fc′
Span Girder Aps

1 AASHTO I 30 6 1.224 8
2 AASHTO I 30 8 1.530 10
3 AASHTO I 30 10 1.836 12
4 AASHTO I 30 12 2.142 14
5 AASHTO II 60 6 3.06 20
6 AASHTO II 60 8 3.978 26
7 AASHTO III 60 10 3.366 22
8 AASHTO III 60 12 4.284 28
9 AASHTO III 80 6 4.284 28
10 AASHTO III 80 8 5.202 34
11 AASHTO III 80 10 6.120 40
12 AASHTO IV 80 12 5.814 38
13 AASHTO III 100 6 7.038 46
14 AASHTO IV 100 8 7.038 46
15 AASHTO IV 100 10 8.262 54
16 AASHTO V 100 12 7.650 50
17 AASHTO IV 120 6 8.874 58
18 AASHTO V 120 8 8.874 58
19 AASHTO V 120 10 10.404 68
20 AASHTO VI 120 12 9.792 64
21 AASHTO VI 140 6 8.874 58
22 AASHTO VI 140 8 10.710 70
23 AASHTO VI 140 10 — —
24 AASHTO VI 140 12 — —
25 FIB-96 160 6 5.814 38
26 FIB-96 160 8 7.344 48
27 FIB-96 160 10 7.956 52
28 FIB-96 160 12 — —
29 FIB-96 180 6 7.956 52
30 Mod. BT-72 180 9 17.442 114
31 Mod. AASHTO VI 180 9 17.442 114
32 Mod. AASHTO VI 200 9 22.032 144
33 Mod. NEBT-2200 200 9 18.360 120
34 Mod. W95PTMG 200 9 18.360 120
35 Mod. NEBT-2200 220 9 22.338 146
Note: — = a practical solution was not found.

201
Reliability Index
Reliability Index
Figure 6.40. Reliability indices for bridges Figure 6.44. Reliability indices for bridges
at decompression limit state (ADTT  5,000, at maximum allowable tensile stress limit state
gLL  1.0, and ft  0.0948 fc′ ). (ADTT  5,000, gLL  0.8, and ft  0.19 fc′ ).
Reliability Index
Span Length (ft.) Reliability Index Span Length (ft.)

at maximum allowable tensile stress limit state at maximum allowable crack width limit state
(ADTT  5,000, gLL  1.0, and ft  0.0948 fc′ ). (ADTT  5,000, gLL  0.8, and ft  0.19 fc′ ).
Reliability Index
Reliability Index

at maximum allowable crack width limit state at decompression limit state (ADTT  5,000,
(ADTT  5,000, gLL  1.0, and ft  0.0948 fc′ ). gLL  1.0, and ft  0.19 fc′ ).
Reliability Index
Reliability Index
Span Length (ft.)

Span Length (ft.)
Figure 6.47. Reliability indices for bridges at
at decompression limit state (ADTT  5,000,
maximum tensile stress limit state (ADTT  5,000,
gLL  0.8, and ft  0.19 fc′ ).
gLL  1.0, and ft  0.19 fc′ ).

202
Reliability Index
Reliability Index
Figure 6.48. Reliability indices for bridges at Figure 6.50. Reliability indices for bridges
maximum crack width limit state (ADTT  5,000, at maximum allowable tensile stress limit state
gLL  1.0, and ft  0.19 fc′ ). (ADTT  5,000, gLL  0.8, and ft  0.25 fc′ ).
fc′. Similar to the case of bridges designed for a maximum

concrete tensile stress of ft = 0.0948 fc′, the reliability level of
Reliability Index
bridges became more uniform than the case of using a live load
factor of 0.8, particularly for the decompression and maximum
tensile stress limit states. Consequently, a live load factor of
1.0 was proposed if the maximum tensile stress was limited to
0.19 fc′.
Step 8c: Select Potential Load and Resistance Factors Span Length (ft.)
for Service III: Bridges Designed for Maximum Concrete
Tensile Stress of ft = 0.25 fc′
at maximum allowable crack width limit state
The work described under Step 8a and Step 8b was repeated, (ADTT  5,000, gLL  0.8, and ft  0.25 fc′ ).
except the girders were redesigned assuming a maximum
concrete tensile stress of ft = 0.25 fc′.
1. Calculate the reliability level of designs according to

Reliability Index
AASHTO LRFD (2010) with a maximum concrete tensile

stress for design of ft = 0.25 fc′ (Figure 6.49 to Figure 6.51).
2. Redesign the bridges with a live load factor of 1.0.
Figure 6.52 to Figure 6.54 show the reliability indices for

the redesigned bridges using a live load factor of 1.0 and ft =
Span Length (ft.)
0.25 fc′. Similar to the case of bridges designed for maxi-
mum concrete tensile stresses of ft = 0.0948 fc′ and Figure 6.52. Reliability indices for bridges
at decompression limit state (ADTT  5,000,
gLL  1.0, and ft  0.25 fc′ ).
Reliability Index
ft = 0.16 fc′, the reliability level of bridges became more uni-

form than the case of using a live load factor of 0.8, particu-
larly for the decompression and maximum tensile stress limit
states. Consequently, a live load factor of 1.0 was proposed
if the maximum tensile stress was limited to ft = 0.25 fc′.
Span Length (ft.) 6.5.5.9 Step 9: Calculate Reliability Indices
Figure 6.49. Reliability indices for bridges The reliability indices were calculated for three cases, as
at decompression limit state (ADTT  5,000, shown in Step 8. In Step 9, the calculated values were reviewed
gLL  0.8, and ft  0.25 fc′ ). to determine whether they were close to the target reliability

203
index and whether they were uniform across the range of

spans considered. If they were not, the load factors, resistance
factors, and/or the concrete tensile stress limit used for design
Reliability Index
were changed, and Step 8 was repeated. The limit state function
to be used as the basis for the calibration was also determined
in Step 9.
6.5.5.10 Summary of Target Reliability Indices for

Span Length (ft.) Different Design and Performance Levels
Figure 6.53. Reliability indices for bridges at Summaries of the average reliability indices calculated for the
maximum tensile stress limit state (ADTT  5,000, different cases are given in Table 6.28 to Table 6.30. Regardless
gLL  1.0, and ft  0.25 fc′ ).
of the maximum tensile stress limits used in the design, the
limiting criterion for the maximum tensile stress when deter-
mining the reliability index was taken as ft = 0.19 fc′.
As indicated earlier, the calibration of the specifications
were based on an ADTT of 5,000. For this ADTT, the reli-
Reliability Index
ability indices obtained assuming the bridges were designed

for maximum stress limits of ft = 0.0948 fc′ and ft = 0.19 fc′
(see the bold outlined cells below in Table 6.28 and Table 6.29,
respectively) are very close to the target reliability indices shown
in Table 6.25.
Span Length (ft.)

6.5.5.11 Effect of Proposed Changes on Design
Figure 6.54. Reliability indices for bridges at
maximum crack width limit state (ADTT  5,000, To investigate the effect of the proposed change in the load
gLL  1.0, and ft  0.25 fc′ ). factor, the number of strands required for different design cases
Table 6.28. Summary of Reliability Indices for Simulated Bridges Designed for ft 5 0.0948 fc′
Live Load Factor 5 0.8 Live Load Factor 5 1.0
Maximum Tensile Crack Width Maximum Tensile Crack Width

ADTT Decompression Stress Limit (in.) Decompression Stress Limit (in.)
1,000 1.05 1.41 3.16 1.42 1.79 3.36

2,500 1.01 1.35 3.11 1.38 1.75 3.33
5,000 0.97 1.31 3.06 1.33 1.70 3.32
10,000 0.94 1.30 3.00 1.32 1.66 3.28

1,000 0.84 1.27 2.92 1.11 1.53 3.25

2,500 0.70 1.15 2.87 1.04 1.46 3.17
5,000 0.68 1.10 2.82 1.00 1.41 3.14
10,000 0.64 1.07 2.78 0.98 1.34 3.11

204

1,000 0.20 0.55 2.83 0.93 1.29 3.03

2,500 0.08 0.49 2.77 0.89 1.27 2.95
5,000 0.06 0.44 2.72 0.85 1.23 2.92
10,000 0.02 0.41 2.66 0.82 1.20 2.88
was compared (see Table 6.31). The comparison indicated The decompression limit state showed the highest level of
that when a live load factor of 0.8 was used in both cases, the uniformity and is recommended to be used as the basis for
post-2005 prestress loss method resulted in a smaller number the reliability analysis (i.e., the determination of the load
of strands than when the pre-2005 prestress loss method was and resistance factors and associated design criteria).
used. It also indicated that when the post-2005 loss method 3. It is recommended that the reliability indices correspond-
was used with a load factor of 1.0, the required number of ing to an ADTT of 5,000 be used as the basis for the cali-
strands was similar to that required when a load factor of bration. The reliability index is not highly sensitive to
0.8 was used in conjunction with the pre-2005 prestress loss changes in the ADTT, so there is no need to use different
method (i.e., the designs were similar between the pre-2005 load factors for ADTTs up to 10,000.
and post-2005 methods). 4. With satisfactory past performance of prestressed beams,
the target reliability index is selected to be similar to the
average inherent reliability index of the bridges on the
6.5.5.12 Summary and Recommendations system. There is no scientific reason to support targeting a
for Service III Limit State different (either higher or lower) reliability index.
For typical I-girders designed using the post-2005 prestress 5. The recommended target reliability index for the decom-
loss method and the assumptions listed in Section 6.5.3, and pression limit state is 1.0 for bridges designed for no worse
comparing the target reliability indices shown in Table 6.25 than moderate corrosion conditions and 1.2 for bridges
and the calculated reliability indices for different design crite- designed for severe corrosion conditions. The reliability
ria, load factors, and design live load as shown in Table 6.28 index, which was based on the study of the WIM data, is
to Table 6.30 and Figure 6.37 to Figure 6.54, the following determined assuming live load exists in a single lane and
conclusions were drawn and summarized: without applying the MPF. This would appear on the
“load side” of the limit state function.
1. For a specific girder of known cross section and specific 6. Based on the reliability indices calculated for different design
number and arrangement of prestressing strands, the reli- and load scenarios, to achieve the target reliability index, it is
ability index varies on the basis of the following: recommended that the following parameters be used for
• The design maximum concrete tensile stress [maximum designing for the Service III limit state:
tensile stresses of ft = 0.0948 fc′ and ft = 0.19 fc′ are • Live load factor of 1.0;
currently shown in AASHTO LRFD (2012) and are pro- • Maximum concrete tensile stress of ft = 0.0948 f c′ and
posed to remain the same]; ft = 0.19 fc′ for bridges in severe corrosion conditions
• The limit state function [i.e., decompression, tensile stress and for bridges in no worse than moderate corrosion
of a certain value (assumed to be ft = 0.19 fc′ in the conditions, respectively; and
work shown above), or a crack width of a certain value • Girders to be designed following conventional design
(assumed to be 0.016 in.)]; and methods and assuming that live loads exist in single lane
• ADTT. or multiple lanes, whichever produces higher load effects.
The effect of different factors can be deduced from The appropriate MPF applies.
Table 6.28 to Table 6.30. These design parameters would appear on the resistance
2. The target reliability index can be achieved uniformly side of the limit state function during calibration.
across various span lengths by using the load factor devel- 7. The results of the calibration demonstrated that girders
oped by following the proposed calibration procedure. designed using the conventional design methods and the
The level of uniformity varies with the limiting criteria. controlling number of loaded traffic lanes produce uniform

Table 6.31. Comparison of Number of Strands Required for Different Design Assumptions
ft 5 0.0948 f c′ , ft 5 0.0948 f c′ , ft 5 0.0948 f c′ , ft 5 0.19 f c′ , ft 5 0.19 f c′ , ft 5 0.19 f c′ ,

Section Span Girder gLL 5 0.8, gLL 5 0.8, gLL 5 1.0, gLL 5 0.8, gLL 5 0.8, gLL 5 1.0,
Case Type Length (ft) Spacing (ft) Pre-2005 Losses Post-2005 Losses Post-2005 Losses Pre-2005 Losses Post-2005 Losses Post-2005 Losses
1 AASHTO I 30 6 8 8 8 8 8 8
2 AASHTO I 30 8 10 10 10 10 10 10
3 AASHTO I 30 10 12 12 12 12 12 12
4 AASHTO I 30 12 14 14 14 14 14 14
5 AASHTO II 60 6 20 16 20 18 16 16
6 AASHTO II 60 8 — 22 26 24 20 22
7 AASHTO III 60 10 22 20 22 20 20 20
8 AASHTO III 60 12 28 24 28 24 24 24
9 AASHTO III 80 6 28 24 28 24 22 24
10 AASHTO III 80 8 38 30 34 32 28 30
11 AASHTO III 80 10 — 36 40 42 32 38
12 AASHTO IV 80 12 40 34 38 34 32 34
13 AASHTO III 100 6 — 40 46 — 38 42
14 AASHTO IV 100 8 50 42 46 44 38 42
15 AASHTO IV 100 10 — 48 54 56 44 50
16 AASHTO V 100 12 56 46 50 48 42 46
17 AASHTO IV 120 6 — 52 58 58 48 52
18 AASHTO V 120 8 62 52 58 54 48 52
19 AASHTO V 120 10 — 60 68 68 54 60
20 AASHTO VI 120 12 74 58 64 64 54 58
21 AASHTO VI 140 6 62 54 58 54 48 52
22 AASHTO VI 140 8 — 64 70 68 58 64
23 AASHTO VI 140 10 — 74 — — 68 74
24 na 140 12 — — — — — —
Note: — = a practical solution was not found; na = not applicable.
205
206
reliability approximately equal to the target reliability index

provided that the load factor is based on a reliability index
calculated using the decompression criteria and assuming
Reliability Index
one lane of traffic.
6.5.6 Results for Adjacent Box Beams,

Spread Box Beams, and American
Segmental Box Institute Box Beams
Work similar to that described above for I-beams was per-
formed for adjacent box beams, spread box beams, and
American Segmental Box Institute (ASBI) box beams. The Span Length (ft.)
details of the work are shown in Appendix D. The final results Figure 6.57. Spread box beams, reliability
assuming the decompression limit state, ADTT of 5,000, indices for bridges at decompression limit state
return period of 1 year, and a load factor of 1.0 for live load (ADTT  5,000, gLL  1.0, and ft  0.0948 fc′ ).
are shown in Figure 6.55 to Figure 6.60. Table 6.32 shows the
average reliability indices represented graphically in Figure 6.55
to Figure 6.60.
Reliability Index
Reliability Index
Span Length (ft.)
Figure 6.58. Spread box beams, reliability

indices for bridges at decompression limit state
Span Length (ft.) (ADTT  5,000, gLL  1.0, and ft  0.19 fc′ ).
Figure 6.55. Adjacent box beams, reliability
(ADTT  5,000, gLL  1.0, and ft  0.0948 fc′ ).
Reliability Index
Reliability Index
Span Length (ft.)
Figure 6.59. ASBI box beams, reliability indices

for bridges at decompression limit state
Span Length (ft.) (ADTT  5,000, gLL  1.0, and ft  0.0948 fc′ ).
Figure 6.56. Adjacent box beams, reliability

(ADTT  5,000, gLL  1.0, and ft  0.19 fc′ ).

207
Losses. AASHTO LRFD requires the time-dependent losses

for segmental bridges to be determined using detailed time-
step methods. The 2005 revisions to Refined Estimates of
Time-Dependent Losses did not affect the time-dependent
Reliability Index
prestressing loss calculations for segmental bridges. Historically,

segmental bridges have been designed using gross section
properties, not transformed section properties, and the effects
of elastic gain have been neglected. If approved by the owner,
the time-step method may also be used to design prestressed
concrete components other than segmental bridges. However,
the level of effort required to perform time-step analysis typi-
Span Length (ft.) cally precludes this method for nonsegmental construction.
Figure 6.60. ASBI box beams, reliability indices The proposed increase in the load factor for live load for the
for bridges at decompression limit state Service III limit state from 0.8 to 1.0 is based on comparing
(ADTT  5,000, gLL  1.0, and ft  0.19 fc′ ). sections designed using the AASHTO LRFD pre-2005 pro-
visions and the post-2005 provisions without making any
exceptions to the specifications requirements and assuming
that the Refined Estimates of Time-Dependent Losses method
The results shown in Figure 6.55 to Figure 6.60 indicate
in the AASHTO LRFD was used for calculating the time-
that the reliability indices for each type of girder are reason-
ably uniform across the range of spans considered. With the dependent losses.
exception of the adjacent box beams, the average reliability The development of the method termed Approximate Esti-
indices for other section types were very close to each other mate of Time-Dependent Losses in AASHTO LRFD was based
and to the target reliability index. For adjacent box beams, on producing prestress losses similar to those produced by the
the average reliability index was slightly higher. However, Refined Estimates of Time-Dependent Losses method. Thus,
the difference did not warrant incorporating measures to the change in the load factor should also be applied to the
reduce the resistance of the beams, such as revising the distri- former method.
bution factor equations or using lower load factors for adjacent Because the changes in the prestress loss methods in 2005 did
box beams. not affect the time-step method, the increase in the load factor
should not be applied to sections designed using the time-step
method. These sections have to satisfy the following conditions
6.5.7 Sections Designed Using to continue using the 0.8 load factor for live load:
Other Methods of Determining
Prestressing Time-Dependent • Time-dependent losses are determined using the time-step
Losses or Section Properties method.
• Gross sections properties are used for the calculations.
As indicated in Section 6.5.3, the calibration of Service III
• The calculations of the force in the prestressing steel neglects
limit states assumed that the sections were designed using
the AASHTO LRFD Refined Estimates of Time-Dependent the effects of the elastic gain.

Table 6.32. Average Reliability Indices AASHTO LRFD (2012) Article 5.9.4.2.2 (Tension Stresses,
for Different Types of Girders
which discusses stresses in fully prestressed components at SLS
Maximum Tensile Stress
after losses) contains the design stress limits that are affected by
Used in Design (ksi) the calibration of the Service III limit state. Due to the lack of
changes to the design stress limits, no revisions to this section
Type of Section ft  0.0948 f c′ ft  0.19 f c′ are required.
I- and bulb-T girders 1.33 1.00 With respect to the calibration of the limit state for tension
Adjacent box beams 1.85 1.31 in prestressed concrete presented above, the only required
revisions to the specifications are those in Article 3.4.1, which
Spread box beams 1.45 1.01
should specify the load factor for live load as 0.8 or 1.0 depend-
ASBI box beams 1.41 1.00
ing on the design procedure used.

208
6.6 Fatigue Limit States: Table 6.33. Current Fatigue

Lifetime Load Factors
6.6.1 Steel Members Fatigue Limit State Live Load Load Factor
6.6.1.1 Formulate Limit State Function Fatigue I 1.5

Fatigue II 0.75
Two limit states for load-induced fatigue of steel details are
defined in AASHTO LRFD Article 3.4.1: Fatigue I, related
to infinite load-induced fatigue life; and Fatigue II, related to
finite load-induced fatigue life. range of the truck population for infinite fatigue life design.
For load-induced fatigue considerations, according to The factor was chosen on the assumption that the maximum
AASHTO LRFD Article 6.6.1.2.2, each steel detail should stress range in the random variable spectrum is twice the effec-
satisfy Equation 6.14: tive stress range caused by the Fatigue II load combination.
The load factor for the Fatigue II load combination reflects
γ (∆f ) ≤ (∆F)n (6.14)
load levels found to be representative of the effective stress
range of the truck population with respect to a small number
where
of stress range cycles and to their cumulative effects in steel
g = load factor;
elements, components, and connections for finite fatigue life
Df = force effect (live load stress range due to the passage
design.
of the fatigue load); and
The resistance factors for the fatigue limit states (f) are
(DF)n = nominal fatigue resistance.
inherently taken as unity and hence do not appear in Equa-
This general limit state function is used for the calibration of tion 6.14.
the fatigue limit states.
The fatigue load of AASHTO LRFD Article 3.7.1.4 and the
fatigue live load load factors of AASHTO LRFD Table 3.4.1-1
are based on extensive research of structural steel highway Components and details susceptible to load-induced
bridges. The fatigue load is the AASHTO LRFD design truck fatigue cracking have been grouped into eight groups, called
[the HS20-44 truck of AASHTO Standard Specifications detail categories, by fatigue resistance. AASHTO LRFD
(2002)], but with a fixed rear-axle spacing of 30 ft. The live Table 6.6.1.2.3-1 illustrates many common details found in
load load factors for the fatigue limit state load combinations steel bridge construction and identifies potential crack sites
are summarized in Table 6.33. for each detail. Figure 6.61 shows the current AASHTO LRFD
The load factor for the Fatigue I load combination reflects fatigue design curves with the eight detail categories ranging
load levels found to be representative of the maximum stress from A to E′.
Source: AASHTO LRFD (2012).
Figure 6.61. AASHTO fatigue design curves: stress range

versus number of cycles. Used with permission of the American
Association of State Highway and Transportation Officials.

209
6.6.1.3 Determine Load and Resistance Parameters 6.6.1.4 Develop Statistical Models

for Selected Design Cases for Loads and Resistances
A comprehensive database containing constant and variable Load Uncertainties
amplitude fatigue test results for various welded steel bridge On the basis of the analysis of WIM data discussed in Chap-
detail types was developed by Keating and Fisher (1986). This ter 5, it is suggested that the current load factor of 1.5 for the
database includes the test data from various NCHRP test pro- Fatigue I limit state be increased to 2.0 to account for current
grams and other available data. These data are presented in and projected truck loads. Similarly, it is proposed that the
Appendix F. load factor of 0.75 for the Fatigue II limit state be increased
The fatigue data include the detail type of each specimen, to 0.80. The mean values and CVs from Chapter 5 are shown
the minimum and maximum stress values, and the number of in Table 6.34.
cycles observed until fatigue failure was evident. From these
data, the stress range was taken as the only significant param- Resistance Uncertainties
eter in the determination of the fatigue life, and a relationship Fatigue Damage Parameter
between the stress range and number of cycles to failure was To properly calibrate the fatigue limit states of the AASHTO
developed for the combined fatigue data (Keating and Fisher LRFD, it was necessary to determine the statistical param-
1986). The regression analysis performed on the stress range eters of the fatigue test data used in the bridge fatigue resis-
versus cycle relation showed that this relation was log-log in tance model. These parameters include the bias and the CV
nature. The curves of the data plotted in log form are charac- of the fatigue test data. As previously described, the fatigue
terized by Equation 6.15: data are commonly presented in terms of the stress range and
number of cycles to failure, or S-N curves in log-log space.
log N = log A − Blog Sr (6.15) The use of this relationship with the given constant ampli-
tude fatigue test data, however, causes difficulty in accurately
determining the statistical parameters. The available data
or in exponential form as shown in Equation 6.16 were not sufficiently distributed along the S-N curves in log-
log space for a regression analysis; the data were often gath-
N = ASr− B (6.16) ered over a small increment of stress ranges and were limited
in number. Any number of regression lines could have been
where used to describe this relationship between the stress range
N = number of cycles to failure; and fatigue life. To better analyze the fatigue data, a different
Sr = constant amplitude stress range (ksi); relationship between the number of cycles and stress range
log A = log-N-axis intercept of S-N curve (a constant taken was developed.
from AASHTO LRFD Table 6.6.1.2.5-1 for the vari- The test data were arranged to couple the number of cycles
ous detail categories); and and stress range in the form of an effective stress range for
B = slope of the curve. each test specimen. The effective stress range as presented in
Article 6.6.2.2 of AASHTO LRFD (2010) was taken as the cube
The combined fatigue data for each detail type were root of the sum of the cubes of the measured stress ranges, as
placed in the eight detail categories on the basis of the fatigue seen in Equation 6.17. The effective stress range is an accepted
performance of the details as specified by AASHTO LRFD means to compare variable amplitude fatigue data with con-
Article 6.6.1.2.3. Fatigue design curves were then determined stant amplitude fatigue test data.
for each of the fatigue categories. The design curves represent
allowable stress range values that are based on a 98% confi-
(Sr )eff = (Σγ i Sri3 )
13
(6.17)
dence limit or lower bound of fatigue resistance. Thus, for a
particular detail type, most of the fatigue data fall above the
design curve, and the test data should not deviate significantly
from the curve. The slopes of all the design curves were deter-
Table 6.34. Load
mined to be very close to a constant value of –3.0, as shown
Uncertainties
through the use of regression analysis (Keating and Fisher
1986). Thus, a constant slope of -3.0 was imposed on the Limit State Mean CV
equations in the regression analysis. Figure 6.61 showed the
Fatigue I 2.0 0.12
current AASHTO LRFD fatigue design curves with the eight
Fatigue II 0.8 0.07
detail categories.

210
where general, the majority of the lower portion of each curve was
(Sr)eff = effective constant amplitude stress range; selected for each detail category. The lower tail of the data was
gi = percentage of cycles at a particular stress range; and selected because it was the portion of the curve that fit the nor-
Sri = constant amplitude stress range for a group of cycles mal distribution (i.e., it was the straight portion of the normal
(ksi). probability plot). Moreover, the lower portion of the fatigue
data represented the range of values within which fatigue crack-
The formula describing the parameter used for the test data
ing was expected to occur when analyzed for the fatigue limit
follows the form of Equation 6.17; however, this equation is state load combinations using the Monte Carlo simulation
applied to each of the test specimens. Thus, the percentage approach, which is discussed in more detail below. Failure
term is equated to a value of one and is subsequently multi- occurs when load exceeds resistance; thus, the higher portions
plied by the number of cycles (N) to yield Equation 6.18: of the fatigue data sets represented fatigue resistance data that
were very unlikely to be exceeded by the fatigue loads used
Sfi = (N p Sri3 )
13
(6.18) within this study and were therefore considered insignificant.
Different approaches for selecting the cutoff values for each
category were investigated to determine the sensitivity of the
where Sfi is a fatigue damage parameter.
resulting reliability indices. It was determined that the relative
The fatigue damage parameter is taken as a normally distrib-
differences of the results determined from the different tech-
uted random variable in order to determine the bias and CV of
niques were negligible. Other techniques used to determine
the fatigue resistance for each of the detail categories. The data the cutoff values included the use of constant cutoff values for
were fitted to many of the typical distributions commonly all the detail categories and having different analysts manually
used, and it was determined that the normal distribution best insert best-fit lines. Table 6.35 shows the resulting cutoff values
characterized the nature of the fatigue data. The bias is a ratio for the standard normal variable. Figure 6.62 and Figure 6.63
of the mean value of the test data to the nominal value described show the normal probability plots of the full fatigue data set
in the specifications. The calculation of the nominal, mean, and the truncated data for categories C and C′, respectively.
and CV values are described in the following subsections. Determining the statistical parameters of the data was
relatively straightforward once the data for each detail cate-
Probability Paper to Determine Statistical Parameters gory were filtered and fitted with a line of best fit by using
The collection of the fatigue data in terms of the new fatigue Microsoft Excel software. The mean value of the stress param-
parameter for each detail category was statistically analyzed eter is simply the intersection of the best-fit line with the
using normal probability paper, as the data best fit the normal horizontal axis. The standard deviation of the data is taken
distribution. The use of normal probability paper is explained as the inverse of the slope of the best-fit line. More simply
in Chapter 3. stated, the standard deviation is the change in the horizontal
The fatigue data for each detail category were filtered to coordinates divided by the change in the vertical coordinates.
include the data that most accurately reflected the fatigue CV is the ratio of the standard deviation to the mean of the data.
behavior of each category. In other words, the data were trun- The resulting statistical parameters are given in Table 6.35. The
cated based on the nature of the curve within each normal probability plots of the fatigue data and corresponding trun-
probability plot to include the pertinent fatigue data. In cated data for all detail categories can be seen in Appendix E.
Table 6.35. Resistance Uncertainties
Standard Cutoff Standard

Category Deviation CV Bias Sf_Mean Sf_AASHTO Normal Variable
A 1,000.0 0.24 1.43 4,167.40 2,924 1

B 666.7 0.22 1.34 3,077.47 2,289 1
B9 250.0 0.11 1.28 2,336.10 1,827 1
C and C9 454.6 0.21 1.35 2,210.77 1,638 1
D 185.2 0.10 1.36 1,773.69 1,300 1
E 140.9 0.12 1.17 1,207.41 1,032 1
E9 232.6 0.20 1.56 1,140.28 730 1

211
Figure 6.62. Normal probability plot of detail categories C

and C fatigue data.
Determination of Nominal Fatigue Parameter The bias value for each category was determined by taking
and Bias Values the ratio of the mean value to the nominal value of the fatigue
CV and the mean of the fatigue resistance data were deter- parameter, as seen in Equation 6.20; the results are shown in
mined as described in the previous subsection. These values, Table 6.35.
along with the nominal fatigue resistance, were needed to
determine the bias of the data. The nominal value of the cho- Bias = Sf _Mean Sf _AASHTO (6.20)
sen fatigue parameter was calculated using AASHTO LRFD
Equation 6.6.1.2.5-2 and rearranged to achieve the relation- where Sf_Mean is the mean value of the fatigue parameter using
ship in terms of the desired fatigue damage parameter, as seen the fatigue data for each detail category.
in Equation 6.19. The resulting nominal resistance values can
be seen in Table 6.35.
6.6.1.5 Develop Reliability Analysis Procedure
Sf _ AASHTO = (N p Sr )
3 13
=A13
(6.19) In code calibration, it is necessary to develop a process by
which to express the structural reliability or the probability of
where Sf_AASHTO is the nominal value of the fatigue parameter the loads on the member being greater than its resistance;
using AASHTO LRFD specifications for each detail category, in other words, failure of the criteria. The reliability analysis
and A is a constant taken from AASHTO LRFD Table 6.6.1.2.5-1 performed within this project was an iterative process that
for the various detail categories. consisted of Monte Carlo simulations to select load and
Figure 6.63. Normal probability plot of detail categories C

and C truncated fatigue data with best-fit line.

212
resistance factors that would achieve reliability close to the equal to the design fatigue load, which was normalized to a
target reliability index. The Monte Carlo technique samples stress range equal to 1 ksi.
load and resistance parameters from selected statistical distri-
butions, such as a normal distribution. Reliability is mea- R = A 0.753 (6.22)
sured in terms of b, the reliability index or safety index. b is
defined as a function of the probability of failure by using where R is resistance, and A is a constant taken from AASHTO
Equation 6.21. Thus b is the number of standard deviations LRFD Table 6.6.1.2.5-1 for the various detail categories.
that the mean safety margin falls on the safe side. The higher The simulations for both limit states were completed using
the b value, the higher the reliability. a total of 10,000 replicates to achieve a sufficient number of
failures. The resulting reliability indices for each of the eight
β = − Φ −1 (Pf ) (6.21) detail categories are reported in Table 6.36.
where F 1 is the inverse standard normal distribution func-

-
6.6.1.7 Select Target Reliability Index
tion, and Pf is the probability of failure.
Target reliability indices (bT) were based on the inherent
Monte Carlo Simulation reliability of the current specifications (see Table 6.36). The
The Monte Carlo analysis is described more fully in Chapter 3. fatigue limit states were harmonized by selecting a single,
A step-by-step outline of the Monte Carlo simulation using common target reliability index for both steel and concrete
Microsoft Excel is included in Appendix F. members equal to 1.0. This proposed target was selected to
The distribution of loads was assumed to be normally best reflect the inherent reliability of the Fatigue I and II limit
distributed as the loads are a summation of force effects. The states for structural steel members and the Fatigue I limit
fatigue resistance was also assumed to follow normal dis state for reinforcement and concrete.
tributions. These distributions for load and resistance were
developed using determined statistical parameters from the 6.6.1.8 Select Potential Load and
available data. Resistance Factors
When the proposed load factors of 2.0 and 0.8 for the Fatigue I
6.6.1.6 Calculate Reliability Indices for Current and Fatigue II limit states, respectively, and the inherent resis-
Design Code or Current Practice tance factor of 1.0 were applied along with the statistical data,
the reliability indices for each detail category remained essen-
The current reliability indices inherent for the various fatigue
tially unchanged from those reported in Table 6.36. Accepting a
detail categories were determined using the Monte Carlo simu-
range of ±0.2 on the reliability index, three Fatigue I limit state
lation technique with the provisions for the Fatigue I and II
reliability indices appear to be too large: detail Category B′
limit states as specified in AASHTO LRFD. The Fatigue I limit
at b = 1.5, detail Category D at b = 2.0, and detail Category E′
state uses a load factor of 1.5, which is common to all the detail
types. The CV for the Fatigue I limit state was determined to
be 0.12 through the work discussed in Chapter 5. The resistance
parameters for the Monte Carlo simulation were determined Table 6.36. Current Reliability
by equating the nominal load and resistance values and then Indices () Using AASHTO
applying the statistical parameters for each of the detail LRFD Fatigue I and Fatigue II
categories (see Table 6.35). Insufficient fatigue data exist to Limit States
determine the constant amplitude fatigue threshold por-
tions of the fatigue design curves for finite fatigue design life Category Fatigue I Fatigue II
(Fatigue II limit state). In consultation with AASHTO Technical A 1.2 1.0
Committee T-14 and the American Iron and Steel Institute B 1.1 0.9
Bridge Task Force, the research team deemed it acceptable to use
B9 1.5 1.0
the statistical parameters for the sloping portions of these curves
for the constant amplitude fatigue thresholds of the different C 1.2 0.9
bridge detail categories. In AASHTO LRFD, the Fatigue II C9 1.2 0.9
limit state currently has a load factor of 0.75 for all the detail D 2.0 1.3
categories and a CV of 0.07. The nominal resistance values were
E 0.9 0.7
determined using Equation 6.22, which resulted from setting
E9 1.7 1.4
the AASHTO LRFD fatigue resistance Equation 6.6.1.2.5-2

213
Table 6.37. Proposed Fatigue I Limit State Table 6.39. Proposed Revisions to AASHTO
Resistance Factors LRFD Table 6.6.1.2.5-1
Proposed Resistance Current Proposed

Detail Category Factor () Reliability Index (b) Detail Category Constant A 108 Constant A 108
A 1.0 1.2 A 250 250

B 1.0 1.1 B 120 120
B9 1.10 0.9 B9 61 61
C 1.0 1.2 C 44 44
C9 1.0 1.2 C9 44 44
D 1.15 1.1 D 22 21
E 1.0 0.9 E 11 12
E9 1.20 1.0 E9 3.9 3.5
at b = 1.7. Similarly, two Fatigue II limit state reliability indi- categories and in some cases greater than unity. The required
ces appear to be too large (detail Category D at b = 1.3 and revisions to the AASHTO LRFD tables are given in Table 6.39
detail Category E′ at b = 1.4) and one appears to be too small and Table 6.40 with changes shown in boldface.
(detail Category E at b = 0.7).
Proposed resistance factors for the Fatigue I limit state and
6.6.2 Concrete Members
the Fatigue II limit state are given in Table 6.37 and Table 6.38,
respectively. Resistance factors other than the current values This section deals with concrete and reinforcing steel. Pre-
of unity are shown in boldface. Reliability index values using stressing strand is not covered as there are currently no design
the proposed resistance factors are shown in the right-hand checks required for fully prestressed components, as explained
column. in Chapter 2.
6.6.1.9 Calculate Reliability Indices 6.6.2.1 Formulate Limit State Function

With the proposed resistance factors, the reliability indices Two limit states for load-induced fatigue are defined in
were all within ±0.2 of the target reliability index of 1.0. AASHTO LRFD Article 3.4.1; however, only Fatigue I, related
The reliability indices shown in Tables 6.37 and 6.38 can also to infinite load-induced fatigue life, is valid for concrete
be achieved by revising the AASHTO LRFD fatigue resistance members as they are always designed for infinite life.
equations for steel members. This may be a better solution For load-induced fatigue considerations, according to
than including resistance factors for only a few of the detail AASHTO LRFD Article 5.5.3.1, concrete members shall satisfy
Table 6.40. Proposed Revisions to AASHTO LRFD

Table 6.38. Proposed Fatigue II Limit State Table 6.6.1.2.5-3
Resistance Factors
Current Constant Proposed Constant
Proposed Resistance Amplitude Fatigue Amplitude Fatigue
Detail Category Factor () Reliability Index (b) Detail Category Threshold (ksi) Threshold (ksi)
A 1.0 1.0 A 24 24
B 1.0 0.9 B 16 16
B9 1.0 1.0 B9 12 13
C 1.0 0.9 C 10 10
C9 1.0 0.9 C9 12 12
D 0.95 1.0 D 7 8.0
E 1.10 1.0 E 4.5 4.5
E9 0.90 1.0 E9 2.6 3.1

214
Equation 6.23, which is seen as a variation of Equation 6.14 ACI 215R-74 and the supporting literature indicate that steel
applicable only to infinite life: reinforcement exhibits a constant amplitude fatigue threshold.
ACI 215R-74 suggests that the resistances are “a conservative
γ (∆f ) ≤ (∆F)TH (6.23) lower bound of all available test results.” In other words, a hori-
zontal constant amplitude threshold has been drawn beneath
where
all the curves.
g = load factor;
The studies used to define the fatigue resistance of steel
Df = force effect (live load stress range due to the passage
reinforcement (Fisher and Viest 1961; Pfister and Hognestad
of the fatigue load); and
1964; Burton and Hognestad 1967; Hanson et al. 1968;
(DF)TH = constant amplitude fatigue threshold.
Helgason et al. 1976; Lash 1969; MacGregor et al. 1971; Amorn
The general limit state function given by Equation 6.23 will et al. 2007) were reanalyzed to estimate constant amplitude
be used for the calibration of the fatigue limit states for con- fatigue thresholds for every case that could be identified in
crete members. the research to determine their uncertainty in terms of bias,
As discussed in Section 6.6.1.1, the Fatigue I limit state load mean, and CV. The various thresholds were grouped together
factor is currently 1.5, and all resistance factors are inherently to make design practical.
unity for the fatigue limit states.
Concrete in Compression
The compressive stress limit of 0.40 f c′ for fully prestressed
components in other than segmentally constructed bridges
Two fatigue limit states for concrete members can be ratio- of AASHTO LRFD Article 5.5.3.1 applies to a combination of
nally calibrated on the basis of current practice and the the live load specified in the Fatigue I limit state load combi-
available data: steel reinforcement in tension (AASHTO LRFD nation plus one-half the sum of the effective prestress and
Article 5.5.3.2) and concrete in compression (AASHTO LRFD permanent loads after losses (i.e., a load combination derived
Article 5.5.3.1). from a modified Goodman diagram). This suggests that the
compressive stress limit represents an infinite life check, as
the Fatigue I limit state load combination corresponds with
6.6.2.3 Determine Load and Resistance Parameters
infinite fatigue life.
For this study, the research used to define these S-N curves
Steel Reinforcement in Tension (Hilsdorf and Kesler 1966) was reevaluated to estimate the
Steel reinforcement as it is considered here includes straight constant amplitude fatigue threshold, which is the infinite life
reinforcing bars and welded-wire reinforcement. AASHTO fatigue resistance. The uncertainty of the fatigue resistance
LRFD Article 5.5.3.2 specifies the fatigue resistance of these was quantified in terms of bias, mean, and CV.
types of reinforcement.
The fatigue resistance of straight reinforcing bars and 6.6.2.4 Develop Statistical Models
welded-wire reinforcement without a cross weld in the high- for Loads and Resistances
stress region (defined as one-third of the span on each side
of the section of maximum moment) is specified by Equa- Load Uncertainties
tion 6.24: The distribution of fatigue loads was determined on the basis of
studies conducted on the WIM data, as described in Chapter 5.
(∆F)TH = 24 − 20 f min f y (6.24) The fatigue load uncertainties in terms of the mean values and
CVs are tabulated in Table 6.34.
where fmin is the minimum stress.
For welded-wire reinforcement with a cross weld in the Resistance Uncertainties
high-stress region, the fatigue resistance is specified by Equa- As discussed in Section 6.6.1.4.2b, the lower tail of the fatigue
tion 6.25: resistance data plots was used to characterize the uncertainties,
biases, and CVs. Figure 6.64 and Figure 6.65, respectively, show
(∆F)TH = 16 − 0.33 f min (6.25) the normal probability plots of the full fatigue data set and the
truncated data for fatigue resistance of steel reinforcement
Equations 6.24 and 6.25 implicitly assume a ratio of radius in tension. The resulting statistical parameters, along with the
to height (in other words, r/h) of the rolled-in transverse bar cutoff scores, are given in Table 6.41. The probability plots of
deformations of 0.3. the fatigue data and corresponding truncated data for both
These fatigue resistances are defined as constant amplitude steel reinforcement in tension and concrete in compression can
fatigue thresholds in AASHTO LRFD. ACI Committee Report be seen in Appendix E.

215
Figure 6.64. Normal probability plot of fatigue resistance

data for steel reinforcement in tension.
Figure 6.65. Normal probability plot of truncated fatigue

resistance data with best-fit line for steel reinforcement
in tension.
6.6.2.5 Develop Reliability Analysis Procedure inherent b for compression of concrete members is approxi-
mately 1.0. Both fatigue limit states are based on the Fatigue I
As discussed in Section 6.6.1.5, Monte Carlo simulation using
limit state and design for infinite life. The calculated inherent
Microsoft Excel formed the basis of the reliability analysis values of b are given in Table 6.42.
procedure for fatigue of concrete members.
6.6.2.7 Select Target Reliability Index
6.6.2.6 Calculate Reliability Indices for Current
Design Code or Current Practice Theoretically, the target reliability index (bT) should be
identical for all members and all fatigue limit states. Thus, the
Monte Carlo simulation was used to estimate the current work on reinforcement and concrete fatigue was performed
inherent reliability indices by comparing the distribution of concurrently with, and was compared with, the work on
fatigue load with the distribution of fatigue resistance on the structural steel fatigue.
basis of the uncertainties of load and resistance. It is proposed to use a constant bT of 1.0 for steel reinforce-
For steel reinforcement in reinforced concrete members, ment in tension, concrete in compression, and structural steel
the current inherent b is approximately 2.0, and the current members. This proposed target reflects the inherent reliability
Table 6.41. Resistance Uncertainties
Standard Cutoff Standard

Resistance Deviation CV Bias Mean Nominal Normal Variable
Steel reinforcement in tension 769.23 0.24 1.94 3,261.54 1,681.21 2

Concrete in compression 117.65 0.45 1.74 260.35 149.66 2

216
Table 6.42. Current Reliability The reliability indices shown in Table 6.43 can also be
Indices for AASHTO LRFD achieved by revising the AASHTO LRFD constant amplitude
Fatigue I Limit States fatigue thresholds for steel reinforcement in tension. This may
be a better solution than including a resistance factor other than
Resistance b
unity for only one of the concrete member fatigue limit states.
Steel reinforcement in tension 1.9 The required revisions to the AASHTO LRFD equations for the
Concrete in compression 0.9 thresholds are given below.
The revised fatigue resistance of straight reinforcing bars
and welded-wire reinforcement without a cross weld in the
of the current Fatigue I limit state for concrete in compression high-stress region would be specified by Equation 6.26:
and the Fatigue I and II limit states for structural steel members.
This proposed target reduces the reliability of steel reinforce- (∆F)TH = 30 − 25 f min f y (6.26)
ment in tension to levels consistent with the three other cali-
brated fatigue limit states. where fmin is the minimum stress.
For welded-wire reinforcement with a cross weld in the
high-stress region, the fatigue resistance would be specified
6.6.2.8 Select Potential Load
by Equation 6.27:
and Resistance Factors
Proposed resistance factors for the Fatigue I limit state are (∆F)TH = 20 − 0.41 fmin (6.27)
given in Table 6.43. Resistance factors other than the current
values of unity are shown in boldface.
In AASHTO LRFD (2012), the fatigue limit state is addressed
6.6.2.9 Calculate Reliability Indices
in Sections 3, 5, and 6. The articles that require modification
With the proposed resistance factors, the reliability indices to implement the revisions recommended here are indicated
were all within ±0.1 of the target reliability index of 1.0. in Table 6.44.
Table 6.43. Proposed Fatigue I Limit State Resistance Factors
Proposed Resistance
Resistance Factor () Reliability Index (b)
Steel reinforcement in tension 1.25 1.1

Concrete in compression 1.0 0.9
Table 6.44. Summary of Relevant Articles in AASHTO LRFD for Foundation

Fatigue Deformations
Article Title Relates to
3.4.1, Table 3.4.1-1 Load Factors and Load Combinations Fatigue I and II
5.5.3.2 Reinforcing Bars Fatigue threshold
5.5.3.3 Prestressing Tendons Fatigue threshold
6.6.1.2.3 Detail Categories, Table 6.6.1.2.3-1 Constant A
6.6.1.2.5 Fatigue Resistance, Table 6.6.1.2.5-1 Constant A
6.6.1.2.5 Fatigue Resistance, Table 6.6.1.2.5-2 Cycle parameter (n)
6.6.1.2.5 Fatigue Resistance, Table 6.6.1.2.5-3 Constant amplitude fatigue threshold
Note: The proposed article revisions are detailed in Chapter 7.

Chapter 7
Proposed Changes to AASHTO LRFD
In Chapter 6, various articles of AASHTO LRFD were identi- permission of the American Association of State Highway and
fied that would need to be modified to implement the calibrated Transportation Officials. Since the various SLS revisions are
SLS resulting from this research. This chapter contains the independent of each other and could be implemented indi-
suggested modifications formatted in a form suitable for vidually, the suggested provisions are presented in separate
consideration by the affected technical committees that could subsections for each SLS. The article numbering system used
be potential AASHTO Highway Subcommittee on Bridges in AASHTO LRFD has been preserved. The proposed revisions
and Structures agenda items. Excerpted material is used by are underlined and deletions are shown as strikethrough.
217

218
7.1 Foundation Deformations – Service I

7.1.1 Proposed Revisions to Section 3
3.4—LOAD FACTORS AND COMBINATIONS
3.4.1—Load Factors and Load Combinations C3.4.1
•
•
•
•
• Service I—Load combination relating to the Compression in prestressed concrete components

normal operational use of the bridge with a and tension in prestressed bent caps are investigated
55 mph wind and all loads taken at their using this load combination. Service III is used to
nominal values. Also related to deflection investigate tensile stresses in prestressed concrete
control in buried metal structures, tunnel liner components.
plate, and thermoplastic pipe, to control crack
width in reinforced concrete structures, and for
transverse analysis relating to tension in
concrete segmental girders. This load
combination should also be used for the
investigation of slope stability, and settlement
of foundations.
• Service II—Load combination intended to This load combination corresponds to the overload
control yielding of steel structures and slip of provision for steel structures in past editions of the
slip-critical connections due to vehicular live AASHTO Specifications, and it is applicable only to
load. steel structures. From the point of view of load level,
this combination is approximately halfway between that
used for Service I and Strength I Limit States.
• Service III—Load combination for longitudinal The live load specified in these specifications
analysis relating to tension in prestressed reflects, among other things, current exclusion weight
concrete superstructures with the objective of limits mandated by various jurisdictions. Vehicles
crack control and to principal tension in the permitted under these limits have been in service for
webs of segmental concrete girders. many years prior to 1993. For longitudinal loading, there
is no nationwide physical evidence that these vehicles
have caused cracking in existing prestressed concrete
components. The statistical significance of the
0.80 factor on live load is that the event is expected to
occur about once a year for bridges with two traffic
lanes, less often for bridges with more than two traffic
lanes, and about once a day for bridges with a single
traffic lane. Service I should be used for checking
tension related to transverse analysis of concrete
segmental girders.
The principal tensile stress check is introduced in
order to verify the adequacy of webs of segmental
concrete girder bridges for longitudinal shear and
torsion.
• Service IV—Load combination relating only to The 0.70 factor on wind represents an 84 mph wind.
tension in prestressed concrete columns with This should result in zero tension in prestressed concrete
the objective of crack control. columns for ten-year mean reoccurrence winds. The
prestressed concrete columns must still meet strength
requirements as set forth in Load Combination Strength
III in Article 3.4.1.
It is not recommended that thermal gradient be
combined with high wind forces. Superstructure

219
expansion forces are included.

•
•
•
•
The evaluation of overall stability of retained fills, Applying these criteria for the evaluation of the
as well as earth slopes with or without a shallow or deep sliding resistance of walls:
foundation unit should be investigated at the service
limit state based on the Service I Load Combination and • The vertical earth load on the rear of a
an appropriate resistance factor as specified in Article cantilevered retaining wall would be multiplied
11.5.6 and Article 11.6.2.3. by γpmin (1.00) and the weight of the structure
The investigation of foundation settlement shall would be multiplied by γpmin (0.90) because
proceed using the provisions of Article 10.6.2.4 using these forces result in an increase in the contact
the load factor, γSE, specified in Table 3.4.1-4. stress (and shear strength) at the base of the
For structural plate box structures complying with wall and foundation.
the provisions of Article 12.9, the live load factor for the
vehicular live loads LL and IM shall be taken as 2.0.
• The horizontal earth load on a cantilevered
retaining wall would be multiplied by γpmax
(1.50) for an active earth pressure distribution
because the force results in a more critical
sliding force at the base of the wall.
Similarly, the values of γpmax for structure weight (1.25),
vertical earth load (1.35) and horizontal active earth
pressure (1.50) would represent the critical load
combination for an evaluation of foundation bearing
resistance.
Water load and friction are included in all strength
load combinations at their respective nominal values.
For creep and shrinkage, the specified nominal values
should be used. For friction, settlement, and water loads,
both minimum and maximum values need to be
investigated to produce extreme load combinations.
The load factor for temperature gradient, γTG, should The load factor for temperature gradient should be
be considered on a project-specific basis. In lieu of determined on the basis of the:
project-specific information to the contrary, γTG may be
taken as: • Type of structure, and
• 0.0 at the strength and extreme event limit • Limit state being investigated.
states,
Open girder construction and multiple steel box
• 1.0 at the service limit state when live load is girders have traditionally, but perhaps not necessarily
not considered, and correctly, been designed without consideration of
temperature gradient, i.e., γTG = 0.0.
• 0.50 at the service limit state when live load is
considered.

220
The effects of the foundation deformation on the Methods for estimation of settlement based on local
bridge superstructure, retaining walls, or other load geologic conditions and calibration may be used subject
bearing structures shall be evaluated at applicable to approval from the Owner. Calibration of local
strength and service limit states using the provisions of methods should be based on processes as described in
Article 10.5.2.2 and the settlement load factor (γSE) SHRP 2 R19B program report (Kulicki et al., 2013).
specified in Table 3.4.1-4. The value of γSE=1.25 for soil-structure interaction
The load factor for settlement, γSE, should be methods in Table 3.4.1-4 for estimation of lateral
considered on a project-specific basis. In lieu of project- deformations has been established based on judgment at
specific information to the contrary, γSE, may be taken as this time.
1.0. Load combinations which include settlement shall
also be applied without settlement.
For segmentally constructed bridges, the following
combination shall be investigated at the service limit
state:
DC + DW + EH + EV + ES + WA + CR + SH + TG + EL + PS
(3.4.1-2)
Table 3.4.1-1—Load Combinations and Load Factors
DC Use One of These at a Time

DD
DW
EH
EV LL
ES IM
EL CE
Load PS BR
Combination CR PL
Limit State SH LS WA WS WL FR TU TG SE EQ BL IC CT CV
Strength I γp 1.75 1.00 — — 1.00 0.50/1.20 γTG γSE — — — — —
(unless noted)
Strength II γp 1.35 1.00 — — 1.00 0.50/1.20 γTG γSE — — — — —
Strength III γp — 1.00 1.4 — 1.00 0.50/1.20 γTG γSE — — — — —
0
Strength IV γp — 1.00 — — 1.00 0.50/1.20 — — — — — — —
Strength V γp 1.35 1.00 0.4 1.0 1.00 0.50/1.20 γTG γSE — — — — —
0
Extreme γp γEQ 1.00 — — 1.00 — — — 1.00 — — — —
Event I
Extreme γp 0.50 1.00 — — 1.00 — — — — 1.00 1.00 1.00 1.00
Event II
Service I 1.00 1.00 1.00 0.3 1.0 1.00 1.00/1.20 γTG γSE — — — — —
0
Service II 1.00 1.30 1.00 — — 1.00 1.00/1.20 — — — — — — —
Service III 1.00 0.80 1.00 — — 1.00 1.00/1.20 γTG γSE — — — — —
Service IV 1.00 — 1.00 0.7 — 1.00 1.00/1.20 — 1.0 — — — — —
0
Fatigue I— — 1.50 — — — — — — — — — — — —
LL, IM & CE
only
Fatigue II— — 0.75 — — — — — — — — — — — —
LL, IM & CE
only
•
•
•
•

221
Table 3.4.1-4—Load Factors for Permanent Loads Due to Foundation Deformations, γSE
Foundation Deformation and Deformation Estimation Method SE

Immediate Settlement
• Hough method 1.00
• Schmertmann method 1.25
• Local method *
Consolidation settlement 1.00
Lateral Deformation
• Soil-structure interaction method (P-y or Strain Wedge) 1.25
• Local method *
*To be determined by the owner based on local geologic conditions and calibration using a target reliability index of
0.50 for Service I limit state.
•
•
•
•
3.4.2—Load Factors for Construction Loads
3.4.2.2—Evaluation of Deflection at the Service Refer to Article 3.4.1 for evaluation of foundation
Limit State deformations due to construction loads.
In the absence of special provisions to the contrary,

where evaluation of construction deflections are required
by the contract documents, Load Combination Service I
shall apply. Construction dead loads shall be considered
as part of the permanent load and construction transient
loads considered part of the live load. The associated
permitted deflections shall be included in the contract
documents.

222

10.3—NOTATION
Ad = angular distortion (10.5.2)

Adm = modified angular distortion (10.5.2)
C1 = correction factor to incorporate the effect of strain relief due to embedment (10.6.2.4.2b)
C2 = correction factor to incorporate time-dependent (creep) increase in settlement for t (years) after
construction (10.6.2.4.2b)
E = modulus of elasticity of pile material (ksi) (10.7.3.8.2); elastic modulus of layer i based on guidance
provided in Table C10.4.6.3-1
Iz = strain influence factor from Figure 10.6.2.4.2c-1a
Ls = bridge span length over which Ad and Adm are computed (10.5.2)
Sd = differential settlement between two bridge support elements spaced at a distance of Ls (ft) (10.5.2.2)
Sta = total foundation settlement using all applicable loads in the Service I load combination (ft) (10.5.2)
Stp = total foundation settlement using all applicable loads prior to construction of bridge superstructure in the
Service I load combination (ft) (10.5.2.2)
Str = relevant total settlement defined as Sta – Stp (10.5.2.2)
X = width or smallest dimension of pile group (ft) (10.7.3.9); a factor used to determine the value of elastic
modulus (10.6.2.4.2b)
= load factor for settlement (10.5.2.2)
∆p = net uniform applied stress (load intensity) at the foundation depth (Figure 10.6.2.4.2c-1b)
10.5—LIMIT STATES AND RESISTANCE

FACTORS
10.5.1—General
The limit states shall be as specified in

Article 1.3.2; foundation-specific provisions are
contained in this Section.
Foundations shall be proportioned so that the
factored resistance is not less than the effects of the
factored loads specified in Section 3.
10.5.2—Service Limit States
10.5.2.1—General C10.5.2.1
Foundation design at the service limit state shall In bridges where the superstructure and substructure
include: are not integrated, settlement corrections can be made
by jacking and shimming bearings. Article 2.5.2.3
• Settlements, requires jacking provisions for these bridges.
• Horizontal movements, The cost of limiting foundation movements should

be compared with the cost of designing the
• Overall stability, and
superstructure so that it can tolerate larger movements or
• Scour at the design flood. of correcting the consequences of movements through
maintenance to determine minimum lifetime cost. The
Consideration of foundation movements shall be
Owner may establish more stringent criteria.
based upon structure tolerance to total and differential
The foundation movements should be translated to
movements, rideability and economy. Foundation
the deck elevation to evaluate the effect of such
movements shall include all movement from settlement,
movements on the superstructure. In this process,
horizontal movement, and rotation.
deformations of the substructure, i.e., elements between
Bearing resistance estimated using the presumptive
foundation and superstructure, should be added to
allowable bearing pressure for spread footings, if used,
foundation deformations as appropriate.
shall be applied only to address the service limit state.

223
The foundations for retaining walls and other load

bearing structures such as tunnels may also be evaluated
using the provisions of this Article.
The design flood for scour is defined in
Article 2.6.4.4.2, and is specified in Article 3.7.5 as
applicable at the service limit state.
Presumptive bearing pressures were developed for
use with working stress design. These values may be
used for preliminary sizing of foundations, but should
generally not be used for final design. If used for final
design, presumptive values are only applicable at service
limit states.
10.5.2.2—Tolerable Movements and Movement

Criteria
10.5.2.2.1—General C10.5.2.2.1
Foundation movement criteria shall be consistent Experience has shown that bridges can and often do
with the function and type of structure, anticipated accommodate more movement and/or rotation than
service life, and consequences of unacceptable traditionally allowed or anticipated in design. Creep,
movements on structure performance. Foundation relaxation, and redistribution of force effects
movement shall include vertical, horizontal, and accommodate these movements. Some studies have
rotational movements. The tolerable movement criteria been made to synthesize apparent response. These
shall be established by either empirical procedures or studies indicate that angular distortions between
structural analyses, or by consideration of both. adjacent foundations greater than 0.008 radians in
Foundation settlement shall be investigated using simple spans and 0.004 radians in continuous spans
all applicable loads in the Service I Load Combination should not be permitted in settlement criteria (Moulton
specified in Table 3.4.1-1. Transient loads may be et al., 1985; DiMillio, 1982; Barker et al., 1991; Samtani
omitted from settlement analyses for foundations et al. 2010). Other angular distortion limits may be
bearing on or in cohesive soil deposits that are subject to appropriate after consideration of:
time-dependent consolidation settlements.
All applicable service limit state load combinations • cost of mitigation through larger foundations,
in Table 3.4.1-1 shall be used for evaluating horizontal realignment or surcharge,
movement and rotation of foundations. • rideability,
All foundation deformation evaluations shall be
• vertical clearance
based on the geomaterial information obtained in
accordance with Article 10.4. • tolerable limits of deformation of other
The following steps shall be followed to estimate a structures associated with a bridge, e.g.,
practical value of angular distortion of the superstructure approach slabs, wingwalls, pavement
based on foundation settlement; a similar approach can structures, drainage grades, utilities on the
be applied and is recommended for evaluation of bridge, etc.
horizontal movement and rotation of foundations: • roadway drainage
• aesthetics, and
1. Compute total foundation settlement at each support
• safety.
element using an Owner approved method for the
assumed foundation type (e.g., spread footings,
driven piles, drilled shafts, etc.) as follows: The bridge engineer shall add deformations from
a. Determine the total foundation settlement, Sta, the substructure (elements between foundation and
using all applicable loads in the Service I load superstructure) as appropriate in evaluation of angular
combination. distortions at the deck elevation.
b. Determine the total foundation settlement, While the angular distortion is generally applied in
Stp, prior to construction of bridge the longitudinal direction of a bridge, similar analyses
superstructure. This settlement would generally should be performed in transverse direction based on
be as a result of all applicable substructure consideration of bridge width and stiffness. For all
loads computed in accordance with Service I bridges, stiffness should be appropriate to the considered
load combination. limit state. Similarly, the effects of continuity with the
c. Determine relevant total settlement, Str as substructure should be considered. In assessing the
Str = Sta – Stp. structural implications of foundation deformations of
concrete bridges, the determination of the stiffness of

224
the bridge components should consider the effects of

cracking, creep, and other inelastic responses
Example: In Figure C10.5.2.2-1, a hypothetical 4-
span bridge structure with span lengths, Ls1, Ls2, Ls3 and
Ls4. The relevant total settlement, Str, is computed at
each support element and the profile of Str along the
bridge is shown by the solid line. In this example, Str-A1
< Str-P1 > Str-P2 < Str-P3 < Str-A2. The Str profile assumed for
computation of the angular distortion, Ad, for each span
is represented by the dashed lines.
2. At a given support element assume that the actual

relevant settlement could be as large as the value
calculated by the chosen method. At the same time,
assume that the settlement of an adjacent support
element could be zero instead of the relevant
settlement value calculated by the same chosen
method. Thus, differential settlement, Sd, within a
given bridge span is equal to the larger of the
relevant settlement at each of two supports of a Figure C10.5.2.2-1—Example for Computing
bridge span. Compute angular distortion, Ad, as the Angular Distortion, Ad, Based on Relevant Total
ratio of the differential settlement, Sd, to the span Settlement, Str, along a hypothetical 4-span Bridge
length, Ls. Express Ad value in radians. (Modified after Samtani, et al., 2010)
3. Compute modified angular distortion, Adm, by The angular distortion, Ad, within each span is as
multiplying the angular distortion value from Step 2 follows: Ad1 = Str-P1/Ls1; Ad1 = Str-P1/Ls2; Ad3 = Str-P3/Ls3;
with the γSE values for settlement in Table 3.4.1-4 and Ad4 = Str-A2/Ls4. Express Ad value in radians.
based on the method used for computing the total Multiply the Ad values with appropriate γSE as per Step
settlement value. 3.
4. Compare the Adm value with owner specified

angular distortion criteria. If owner specified
criteria is not available then use 0.008 radians for
the case of simple spans and 0.004 radians for the
case of continuous spans as the limiting angular
distortions.
5. Evaluate the structural ramifications of the

computed angular distortions that are within
acceptable limits as per Step 4. Modify foundation
design as appropriate based on structural
ramifications.
The above procedure shall also be used for the cases
where foundations of various support elements are
proportioned for equal total settlement because the
prediction of settlements from any given method is
uncertain by itself.
10.5.2.2.2—Lateral Deformations C10.5.2.2.2
Using a procedure similar to settlement evaluation Rotation movements should be evaluated at the top
specified in Article 10.5.2.2.1, lateral (horizontal) of the substructure unit in plan location and at the deck
movement at foundation level shall also be evaluated. elevation.
Horizontal movement criteria should be established at Tolerance of the superstructure to lateral movement
the top of the foundation based on the tolerance of the will depend on bridge seat or joint widths, bearing
structure to lateral movement, with consideration of the type(s), structure type, and load distribution effects.
column length and stiffness. Table 3.4.1-4 provides

225
values of γSE for lateral deformations.
10.5.2.2.3—Walls
The procedure for computing angular distortions

shall also be applied for evaluating angular distortions
along and transverse to retaining walls as well as the
junction of the approach walls to abutment walls. The
angular distortion values along a retaining wall can be
used to select an appropriate wall type, e.g., MSE walls
can tolerate larger angular distortions compared to cast-
in-place walls
10.5.2.3—Overall Stability
The evaluation of overall stability of earth slopes

with or without a foundation unit shall be investigated at
the service limit state as specified in Article 11.6.2.3.
10.5.2.4—Abutment Transitions C10.5.2.4
Vertical and horizontal movements caused by Settlement of foundation soils induced by

embankment loads behind bridge abutments shall be embankment loads can result in excessive movements of
investigated. substructure elements. Both short and long term
settlement potential should be considered.
Settlement of improperly placed or compacted
backfill behind abutments can cause poor rideability and
a possibly dangerous bump at the end of the bridge.
Guidance for proper detailing and material requirements
for abutment backfill is provided in Cheney and Chassie
Samtani and Nowatzki (20006).
Lateral earth pressure behind and/or lateral squeeze
below abutments can also contribute to lateral
movement of abutments and should be investigated, if
applicable.
10.6.2—Service Limit State Design
10.6.2.1—General C10.6.2.1
•
•
•
•
10.6.2.4—Settlement Analyses
10.6.2.4.1—General C10.6.2.4.1
Foundation settlements should be estimated using Elastic, or immediate, settlement is the

computational methods based on the results of instantaneous deformation of the soil mass that occurs as
laboratory or insitu testing, or both. The soil parameters the soil is loaded. The magnitude of elastic settlement is
used in the computations should be chosen to reflect the estimated as a function of the applied stress beneath a
loading history of the ground, the construction sequence, footing or embankment. Elastic settlement is usually
and the effects of soil layering. small and neglected in design, but where settlement is
Both total and differential settlements, including critical, it is the most important deformation
time dependant effects, shall be considered. consideration in cohesionless soil deposits and for
Total settlement, including elastic, consolidation, footings bearing on rock. For footings located on over-

226
and secondary components may be taken as: consolidated clays, the magnitude of elastic settlement is
not necessarily small and should be checked.
St = S e + S c + S s (10.6.2.4.1-1) In a nearly saturated or saturated cohesive soil, the
pore water pressure initially carries the applied stress.
As pore water is forced from the voids in the soil by the
where:
applied load, the load is transferred to the soil skeleton.
Consolidation settlement is the gradual compression of
Se = elastic settlement (ft)
the soil skeleton as the pore water is forced from the
Sc = primary consolidation settlement (ft)
voids in the soil. Consolidation settlement is the most
important deformation consideration in cohesive soil
Ss = secondary settlement (ft)
deposits that possess sufficient strength to safely support
a spread footing. While consolidation settlement can
occur in saturated cohesionless soils, the consolidation
occurs quickly and is normally not distinguishable from
the elastic settlement.
Secondary settlement, or creep, occurs as a result of
the plastic deformation of the soil skeleton under a
constant effective stress. Secondary settlement is of
principal concern in highly plastic or organic soil
deposits. Such deposits are normally so obviously weak
and soft as to preclude consideration of bearing a spread
footing on such materials.
The principal deformation component for footings
on rock is elastic settlement, unless the rock or included
discontinuities exhibit noticeable time-dependent
behavior.
To avoid overestimation, relevant settlements
should be evaluated using the construction point concept
noted in Samtani et al. (2010). The effect of settlement
on superstructure shall be evaluated based on Article
10.5.2.2.
The effects of the zone of stress influence, or For guidance on vertical stress distribution for
vertical stress distribution, beneath a footing shall be complex footing geometries, see Poulos and Davis
considered in estimating the settlement of the footing. (1974) or Lambe and Whitman (1969).
Spread footings bearing on a layered profile Some methods used for estimating settlement of
consisting of a combination of cohesive soil, footings on sand include an integral method to account
cohesionless soil and/or rock shall be evaluated using an for the effects of vertical stress increase variations. For
appropriate settlement estimation procedure for each guidance regarding application of these procedures, see
layer within the zone of influence of induced stress Gifford et al. (1987).
beneath the footing.
The distribution of vertical stress increase below
circular or square and long rectangular footings, i.e.,
where L > 5B, may be estimated using
Figure 10.6.2.4.1-1.

227
Figure 10.6.2.4.1-1—Boussinesq Vertical Stress Contours

for Continuous and Square Footings Modified after Sowers
(1979)
10.6.2.4.2—Settlement of Footings on Cohesionless

Soils
10.6.2.4.2a—General C10.6.2.4.2a
The settlement of spread footings bearing on Although methods are recommended for the
cohesionless soil deposits shall be estimated as a determination of settlement of cohesionless soils,
function of effective footing width and shall consider the experience has indicated that settlements can vary
effects of footing geometry and soil and rock layering considerably in a construction site, and this variation
with depth. may not be predicted by conventional calculations.
Settlements of cohesionless soils occur rapidly,
essentially as soon as the foundation is loaded.
Therefore, the total settlement under the service loads
may not be as important as the incremental settlement
between intermediate load stages. For example, the total
and differential settlement due to loads applied by
columns and cross beams is generally less important
than the total and differential settlements due to girder
placement and casting of continuous concrete decks.
Settlements of footings on cohesionless soils shall Generally conservative settlement estimates may be
be estimated using elastic theory or empirical obtained using the elastic half-space procedure or the
procedures. empirical method by Hough. Additional information
regarding the accuracy of the methods described herein
is provided in Gifford et al. (1987), and Kimmerling
(2002) and Samtani and Notwazki (2006). This
information, in combination with local experience and
engineering judgment, should be used when determining
the estimated settlement for a structure foundation, as
there may be cases, such as attempting to build a
structure grade high to account for the estimated
settlement, when overestimating the settlement
magnitude could be problematic.
Details of other procedures can be found in

228
textbooks and engineering manuals, including:
• Terzaghi and Peck (1967)

• Sowers (1979)
• U.S. Department of the Navy (1982)
• D’Appolonia (Gifford et al., 1987)—This
method includes consideration for over-
consolidated sands.
• Tomlinson (1986)
• Gifford et al. (1987)
10.6.2.4.2b—Elastic Half-space Method C10.6.2.4.2b
The elastic half-space method assumes the footing For general guidance regarding the estimation of
is flexible and is supported on a homogeneous soil of elastic settlement of footings on sand, see Gifford et al.
infinite depth. The elastic settlement of spread footings, (1987), and Kimmerling (2002), and Samtani and
in feet, by the elastic half-space method shall be Notwazki (2006).
estimated as: The stress distributions used to calculate elastic
settlement assume the footing is flexible and supported
S =
 o
 (
 q 1 −ν 2 A′ 
 ) (10.6.2.4.2b-1)
on a homogeneous soil of infinite depth. The settlement
below a flexible footing varies from a maximum near
e the center to a minimum at the edge equal to about
144 E β
s z 50 percent and 64 percent of the maximum for
rectangular and circular footings, respectively. The
where: settlement profile for rigid footings is assumed to be
uniform across the width of the footing.
qo = applied vertical stress (ksf) Spread footings of the dimensions normally used
for bridges are generally assumed to be rigid, although
A′ = effective area of footing (ft2) the actual performance will be somewhere between
perfectly rigid and perfectly flexible, even for relatively
Es = Young’s modulus of soil taken as specified in thick concrete footings, due to stress redistribution and
Article 10.4.6.3 if direct measurements of Es concrete creep.
are not available from the results of in situ or The accuracy of settlement estimates using elastic
laboratory tests (ksi) theory are strongly affected by the selection of soil
modulus and the inherent assumptions of infinite elastic
βz = shape factor taken as specified in half space. Accurate estimates of soil moduli are
Table 10.6.2.4.2b-1 (dim) difficult to obtain because the analyses are based on
only a single value of soil modulus, and Young’s
ν = Poisson’s Ratio, taken as specified in modulus varies with depth as a function of overburden
Article 10.4.6.3 if direct measurements of ν are stress. Therefore, in selecting an appropriate value for
not available from the results of in situ or soil modulus, consideration should be given to the
laboratory tests (dim) influence of soil layering, bedrock at a shallow depth,
and adjacent footings.
Unless Es varies significantly with depth, Es should For footings with eccentric loads, the area, A′,
be determined at a depth of about 1/2 to 2/3 of B below should be computed based on reduced footing
the footing, where B is the footing width. If the soil dimensions as specified in Article 10.6.1.3.
modulus varies significantly with depth, a weighted
average value of Es should be used.

229
Table 10.6.2.4.2b-1—Elastic Shape and Rigidity Factors,

EPRI (1983)
Flexible, βz βz
L/B (average) Rigid
Circular 1.04 1.13
1 1.06 1.08
2 1.09 1.10
3 1.13 1.15
5 1.22 1.24
10 1.41 1.41
10.6.2.4.2c—Hough Method C10.6.2.4.2c
Estimation of spread footing settlement on The Hough method was developed for normally
cohesionless soils by the empirical Hough method shall consolidated cohesionless soils.
be determined using Eqs. 10.6.2.4.2c-2 and The Hough method has several advantages over
10.6.2.4.2c-3. SPT blow counts shall be corrected as other methods used to estimate settlement in
specified in Article 10.4.6.2.4 for depth, i.e. overburden cohesionless soil deposits, including express
stress, before correlating the SPT blow counts to the consideration of soil layering and the zone of stress
bearing capacity index, C ′. influence beneath a footing of finite size.
The subsurface soil profile should be subdivided
n into layers based on stratigraphy to a depth of about
S=
e ∑ ∆H i (10.6.2.4.2c-1) three times the footing width. The maximum layer
i =1 thickness should be about 10 ft.
While Cheney and Chassie (2000), and Hough
in which: (1959), did not specifically state that the SPT N values
should be corrected for hammer energy in addition to
1  σ′ + ∆σv  overburden pressure, due to the vintage of the original
∆H i =
Hc log  o  (10.6.2.4.2c-2)
C′  σ′o  work, hammers that typically have an efficiency of
approximately 60 percent were in general used to
where: develop the empirical correlations contained in the
method. If using SPT hammers with efficiencies that
n = number of soil layers within zone of stress differ significantly from this 60 percent value, the N
influence of the footing values should also be corrected for hammer energy, in
effect requiring that N160 be used (Samtani and
∆Hi = elastic settlement of layer i (ft) Nowatzki, 2006).
Studies conducted by Gifford et al. (1987) and
HC = initial height of layer i (ft) Samtani and Nowatzki (2006) indicate that Hough’s
procedure is conservative and over-predicts settlement
C′ = bearing capacity index from by a factor of 2 or more. Such conservatism may be
Figure 10.6.2.4.2c-1 (dim) acceptable for the evaluation of the settlement of
embankments. However, in the case of shallow
foundations such conservatism may lead to unnecessary
use of costlier deep foundations in cases where shallow
foundations may be viable.
In Figure 10.6.2.4.2c-1, N1 shall be taken as N160,

Standard Penetration Resistance, N (blows/ft), corrected
for overburden pressure as specified in Article 10.4.6.2.4.
σ′o = initial vertical effective stress at the midpoint of

layer i (ksf)
∆σv = increase in vertical stress at the midpoint of

layer i (ksf)

230
The Hough method is applicable to cohesionless

soil deposits. The “Inorganic Silt” curve should
generally not be applied to soils that exhibit plasticity
because N-values in such soils are unreliable. The
settlement characteristics of cohesive soils that exhibit
plasticity should be investigated using undisturbed
samples and laboratory consolidation tests as prescribed
in Article 10.6.2.4.3.
Figure 10.6.2.4.2c-1—Bearing Capacity Index versus

Corrected SPT (Samtani and Nowatzki, 2006, after Hough,
1959)
10.6.2.4.2d—Schmertmann Method C10.6.2.4.2d
An estimate of the immediate settlement, Si, of To overcome the conservatism of the Hough
spread footings may be made by using Eq. 10.6.2.4.2d-1 method, use of a more rigorous procedure such as
as proposed by Schmertmann, et al. (1978). Schmertmann’s method (1978) may be used for shallow
foundations.
n
Si = C1C2 Dp ∑ DH i (10.6.2.4.2d-1)
• Effect of lateral strain: Schmertmann method is
i=1
based on the results of displacement measurements
within sand masses loaded by model footings, as
in which:
well as finite element analyses of deformations of
materials with nonlinear stress-strain behavior that
 I  expressly incorporated Poisson’s ratio. Therefore,
ΔH i = H c  z  (10.6.2.4.2d-2)
 XE  the effect of the lateral strain on the vertical strain is
included in the strain influence factor diagrams.
 po  • Effect of preloading: The equations used in
C
=1 1 − 0.5   ≥ 0.5 (10.6.2.4.2d-3)
 ∆p  Schmertmann’s method are applicable to normally
loaded sands. If the sand was pre-strained by
previous loading, then the actual settlements will be
 t (years)  (10.6.2.4.2d-4) overpredicted. Schmertmann, et al. (1970) and
C=
2 1 + 0.2 log10  
 0.1  Holtz (1991) recommend a reduction in settlement
after preloading or other means of compaction of
where: half the predicted settlement. Alternatively, in case
of preloaded soil deposits, the settlement can be
Iz = strain influence factor from Figure 10.6.2.4.2d- computed by using the method proposed by
1a. The dimension Bf represents the least D’Appolonia (1968, 1970), which includes explicit
lateral dimension of the footing after correction consideration of preloading.
for eccentricities, i.e. use least lateral effective
footing dimension. The strain influence factor • C2 correction factor and applicability of the
is a function of depth and is obtained from the method: The time duration, t, in Eq. 10.6.2.4.2d-4
strain influence diagram. The strain influence is set to 0.1 years to evaluate the settlement
diagram is easily constructed for the immediately after construction, i.e., C2 = 1. If long-

231
axisymmetric case (Lf/Bf = 1) and the plane term creep deformation of the soil is suspected then
strain case (Lf/Bf ≥ 10) as shown in Figure an appropriate time duration, t, can be used in the
10.6.2.4.2d-1a. The strain influence diagram computation of C2. Creep deformation is not the
for intermediate conditions can be determined same as consolidation settlement. This factor can
by simple linear interpolation. have an important influence on the reported
settlement since it is included in Eq. 10.6.2.4.2d-1
n = number of soil layers within the zone of strain as a multiplier. For example, the C2 factor for time
influence (strain influence diagram). durations of 0.1 yrs, 1 yr, 10 yrs and 50 yrs are 1.0,
1.2, 1.4 and 1.54, respectively. In cohesionless
∆p = net uniform applied stress (load intensity) at the soils and unsaturated fine-grained cohesive soils
foundation depth (see Figure 10.6.2.4.2d-1b). with low plasticity, time durations of 0.1 yr and 1
yr, respectively, are generally appropriate and
E = elastic modulus of layer i based on guidance sufficient for cases of static loads.
provided in Table C10.4.6.3-1.
The C2 parameter shall not be used to estimate time-
X = a factor used to determine the value of elastic dependent consolidation settlements. Where
modulus. If the value of elastic modulus is consolidation settlement can occur within the depth
based on correlations with N160-values or qc of the strain distribution diagram, the magnitude of
from Table C10.4.6.3-1, then use X as follows. the consolidation settlement shall be estimated as
per Article 10.6.2.4.3 and added to the immediate
X = 1.25 for axisymmetric case (Lf/Bf = 1) settlement of other layers within the strain
X = 1.75 for plane strain case (Lf/Bf ≥ 10) distribution diagram where consolidation settlement
may not occur.
Use interpolation for footings with
1 < Lf/Bf ≤ 10
If the value of elastic modulus is estimated

based on the range of elastic moduli in Table
C10.4.6.3-1 or other sources use X = 1.0.
C1 = correction factor to incorporate the effect of

strain relief due to embedment
po = effective in-situ overburden stress at the

foundation depth and ∆p is the net foundation
pressure as shown in Figure 10.6.2.4.2d-1b
C2 = correction factor to incorporate time-dependent

(creep) increase in settlement for t (years) after
construction where:
(a)

232
(b)
Figure 10.6.2.4.2d-1—(a) Simplified vertical strain

influence factor distributions, (b) Explanation of pressure
terms in equation for Izp (after Schmertmann, et al., 1978,
Samtani and Notatzki, 2006).
10.6.2.4.2e—Local Method C10.6.2.4.2e
Methods based on local geologic conditions and Calibration of local methods should be based on
calibration may be used subject to approval from the processes as described in SHRP 2 R19B program report
Owner. (Kulicki et al., 2013).
10.10—REFERENCES
D'Appolonia, D. J., D'Appolonia, E. E., and Brissette, R. F. (1968). "Settlement of Spread Footings on Sand."
American Society of Civil Engineers, Journal of the Soil Mechanics and Foundations Division, 94 (SM3), 735-760.
D'Appolonia, D. J., D'Appolonia, E. E., and Brissette, R. F. (1970). Closure to discussions on "Settlement of Spread
Footings on Sand." American Society of Civil Engineers, Journal of the Soil Mechanics and Foundations Division, 96
(SM2), 754-761.
Holtz, R. D. 1991. “Stress Distribution and Settlement of Shallow Foundations.” In Foundation Engineering
Handbook, 2nd Edition, H. Y. Fang, editor. Van Nostrand Reinhold Co., New York, NY, Chapter 5, pp. 166–185.
Samtani, N. C., and Nowatzki, E. A. 2006. Soils and Foundations, FHWA NHI-06-088 and FHWA NHI 06-089,
Federal Highway Administration, U.S. Department of Transportation, Washington, DC.
Samtani, N. C., Nowatzki, E. A., and Mertz, D.R. 2010. Selection of Spread Footings on Soils to Support Highway
Bridge Structures, FHWA-RC/TD-10-001, Federal Highway Administration, Resource Center, Matteson, IL
Schmertmann, J. H. 1970. "Static Cone to Compute Static Settlement Over Sand." American Society of Civil
Engineers, Journal of the Soil Mechanics and Foundations Division, 96(SM3), 1011-1043.
Schmertmann, J. H., Hartman, J. P., and Brown, P. R. 1978. "Improved Strain Influence Factor Diagrams." American
Society of Civil Engineers, Journal of the Geotechnical Engineering Division, 104 (No. GT8), 1131-1135.

233
7.2 Live Load Response
2.5.2.6—Deformations
2.5.2.6.1—General C2.5.2.6.1
• •
• •
• •
• •
2.5.2.6.2—Criteria for Deflection Live Load C2.5.2.6.2

Response
The criteria in this Section shall be considered These provisions permit, but do not encourage, the
optional, except for the following: use of past practice for deflection control. Designers
were permitted to exceed these limits at their discretion
• The provisions for orthotropic decks shall be in the past. Calculated deflections of structures have
considered mandatory. often been found to be difficult to verify in the field due
to numerous sources of stiffness not accounted for in
• The provisions in Article 12.14.5.9 for precast calculations. Despite this, many Owners and designers
reinforced concrete three-sided structures shall be have found comfort in the past requirements to limit the
considered mandatory. overall stiffness of bridges. The desire for continued
• Metal grid decks and other lightweight metal and availability of some guidance in this area, often stated
concrete bridge decks shall be subject to the during the development of these Specifications, has
serviceability provisions of Article 9.5.2. resulted in the retention of optional criteria, except for
orthotropic decks, for which the criteria are required.
In applying these criteria, the vehicular load shall Deflection criteria are also mandatory for lightweight
include the dynamic load allowance. decks comprised of metal and concrete, such as filled
If an Owner chooses to invoke deflection control, and partially filled grid decks, and unfilled grid decks
the following principles may be applied: composite with reinforced concrete slabs, as provided in
Article 9.5.2.
Additional guidance regarding deflection of steel
bridges can be found in Wright and Walker (1971).
Additional considerations and recommendations for
deflection in timber bridge components are discussed in
more detail in Chapters 7, 8, and 9 in Ritter (1990).
• When investigating the maximum absolute For a straight multibeam bridge, this is equivalent to
deflection for straight girder systems, all design saying that the distribution factor for deflection is equal
lanes should be loaded, and all supporting to the number of lanes divided by the number of beams.
components should be assumed to deflect equally; For curved steel girder systems, the deflection limit is
applied to each individual girder because the curvature
• For curved steel box and I-girder systems, the causes each girder to deflect differently than the adjacent
deflection of each girder should be determined girder so that an average deflection has little meaning. For
individually based on its response as part of a curved steel girder systems, the span used to compute the
system; deflection limit should be taken as the arc girder length
between bearings.
• For composite design, the stiffness of the design

cross-section used for the determination of
deflection and frequency should include the entire
width of the roadway and the structurally
continuous portions of the railings, sidewalks, and
median barriers;
• For straight girder systems, the composite bending
stiffness of an individual girder may be taken as the

234
stiffness determined as specified above, divided by

the number of girders;
• When investigating maximum relative
displacements, the number and position of loaded
lanes should be selected to provide the worst
differential effect;
• The live load portion of Load Combination Service
I of Table 3.4.1-1 should be used, including the
dynamic load allowance, IM;
• The live load shall be taken from Article 3.6.1.3.2;
• The provisions of Article 3.6.1.1.2 should apply;
and
• For skewed bridges, a right cross-section may be
used, and for curved and curved skewed bridges, a
radial cross-section may be used.
In the absence of other criteria, the following Frequency may be determined by refined analysis
deflection limits may be considered for steel, aluminum, methods or by equations available in the literature if
and/or concrete vehicular bridges: should meet the they apply to the girder or structure being analyzed.
criteria shown in Figure 2.5.2.6.1-1 for the anticipated
level of pedestrian usage. Unless otherwise specified
herein, the deflection and frequency may be calculated
on a system or component basis using any recognized
method of analysis.
Figure 2.5.2.6.1-1—Criteria for Live Load Response (Used

with permission of the Canadian Standards Association)
• Vehicular load, general ............................. Span/800,

• Vehicular and pedestrian loads ............... Span/1000,
• Vehicular load on cantilever arms ............................
Span/300, and
• Vehicular and pedestrian loads on cantilever arms...
Span/375.
For steel I-shaped beams and girders, and for steel box

235
and tub girders, the provisions of Articles 6.10.4.2 and

6.11.4, respectively, regarding the control of permanent
deflections through flange stress controls, shall apply.
For pedestrian bridges, i.e., bridges whose primary
function is to carry pedestrians, bicyclists, equestrians,
and light maintenance vehicles, the provisions of
Section 5 of AASHTO’s LRFD Guide Specifications for
the Design of Pedestrian Bridges shall apply.
In the absence of other criteria, the following From a structural viewpoint, large deflections in
deflection limits may be considered for wood wood components cause fasteners to loosen and brittle
construction: materials, such as asphalt pavement, to crack and break.
In addition, members that sag below a level plane
• Vehicular and pedestrian loads ...........Span/425, and present a poor appearance and can give the public a
perception of structural inadequacy. Deflections from
• Vehicular load on wood planks and panels (extreme moving vehicle loads also produce vertical movement
relative deflection between adjacent edges) ...... and vibrations that annoy motorists and alarm
0.10 in. pedestrians (Ritter, 1990).
The following provisions shall apply to orthotropic Excessive deformation can cause premature
plate decks: deterioration of the wearing surface and affect the
performance of fasteners, but limits on the latter have
• Vehicular load on deck plate .................... Span/300, not yet been established.
The intent of the relative deflection criterion is to
• Vehicular load on ribs of orthotropic metal decks
protect the wearing surface from debonding and
Span/1000, and
fracturing due to excessive flexing of the deck.
• Vehicular load on ribs of orthotropic metal decks The 0.10-in. relative deflection limitation is
(extreme relative deflection between adjacent ribs) tentative.
0.10 in.
• •
• •
• •
• •
• Service I—Load combination relating to the normal Compression in prestressed concrete components
operational use of the bridge with a 55 mph wind and tension in prestressed bent caps are investigated
and all loads taken at their nominal values. Also using this load combination. Service III is used to
related to dynamic response of superstructures, investigate tensile stresses in prestressed concrete
deflection control in buried metal structures, tunnel components.
liner plate, and thermoplastic pipe, to control crack
transverse analysis relating to tension in concrete
segmental girders. This load combination should
also be used for the investigation of slope stability.
• Service II—Load combination intended to control This load combination corresponds to the overload
yielding of steel structures and slip of slip-critical provision for steel structures in past editions of the
connections due to vehicular live load. AASHTO Specifications, and it is applicable only to
steel structures. From the point of view of load level,

236
analysis relating to tension in prestressed concrete reflects, among other things, current exclusion weight
superstructures with the objective of crack control limits mandated by various jurisdictions. Vehicles
and to principal tension in the webs of segmental permitted under these limits have been in service for
concrete girders. many years prior to 1993. For longitudinal loading, there
segmental girders.
torsion.
tension in prestressed concrete columns with the This should result in zero tension in prestressed concrete
objective of crack control. columns for ten-year mean reoccurrence winds. The
• Service V—Load combination to be used to
Dead load is included in this load combination
because mass is part of the required calculation of
investigate deflection and vibration response under frequency.
traffic in accordance with Article 2.5.2.6.2.
•
•
•
•

237

DD
DW
EH
EV LL
ES IM
EL CE
Load PS BR
Combination CR PL W
Limit State SH LS WA WS L FR TU TG SE EQ BL IC CT CV
(unless noted)
Strength IV γp — 1.00 — — 1.00 0.50/1.20 — — — — — — —
Extreme γp γEQ 1.00 — — 1.00 — — — 1.00 — — — —
Event I
Extreme γp 0.50 1.00 — — 1.00 — — — — 1.00 1.00 1.00 1.00
Event II
Service I 1.00 1.00 1.00 0.30 1.0 1.00 1.00/1.20 γTG γSE — — — — —
Service II 1.00 1.30 1.00 — — 1.00 1.00/1.20 — — — — — — —
Service IV 1.00 — 1.00 0.70 — 1.00 1.00/1.20 — 1.0 — — — — —
Service V 1.00 1.50 — — — — — — — — — — — —
Fatigue I— — 1.50 — — — — — — — — — — — —
LL, IM & CE
only
Fatigue II— — 0.75 — — — — — — — — — — — —
LL, IM & CE
only

238
7.3 Premature Yielding and Slip of Bolts – Service II

•
•
•
•
• Service I—Load combination relating to the Compression in prestressed concrete components

normal operational use of the bridge with a and tension in prestressed bent caps are investigated
55 mph wind and all loads taken at their using this load combination. Service III is used to
nominal values. Also related to deflection investigate tensile stresses in prestressed concrete
control in buried metal structures, tunnel liner components.
plate, and thermoplastic pipe, to control crack
transverse analysis relating to tension in
concrete segmental girders. Relevant elements
of this load combination should also be used for
the investigation of slope stability.
• Service II—Load combination intended to This load combination corresponds to the overload
control yielding of steel structures and slip of provision for steel structures in past editions of the
slip-critical connections due to vehicular live AASHTO Specifications, and it is applicable only to
load. steel structures. From the point of view of load level,
used for Service I and Strength I Limit States. A recent
evaluation of WIM data from 31 sites around the
country indicated that the load level specified in Table
3.4.1-1 for this limit state could reasonably be expected
to be exceeded less than once every six months on
average. For structures with unique truck loading
conditions, such as access roads to ports or industrial
sites which might lead to a disproportionate number of
permit loads, a site-specific increase in the load factor or
number of loaded lanes should be considered.
analysis relating to tension in prestressed reflects, among other things, current exclusion weight
concrete superstructures with the objective of limits mandated by various jurisdictions. Vehicles
crack control and to principal tension in the permitted under these limits have been in service for
webs of segmental concrete girders. many years prior to 1993. For longitudinal loading, there
segmental girders.
torsion.

239
tension in prestressed concrete columns with This should result in zero tension in prestressed concrete
the objective of crack control. columns for ten-year mean reoccurrence winds. The
•
•
•
•

240
7.4 Cracking of Prestressed Concrete – Currently Service III

The total factored force effect shall …………….. The background for the load factors ………………..
. .
---. .
Service I—Load combination relating to the normal Compression in prestressed concrete components
operational use of the bridge with a 55 mph wind and all and tension in prestressed bent caps are investigated
loads taken at their nominal values. Also related to using this load combination. Service III is used to
deflection control in buried metal structures, tunnel liner investigate tensile stresses in prestressed concrete
plate, and thermoplastic pipe, to control crack width in components.
reinforced concrete structures, and for transverse
analysis relating to tension in concrete segmental
girders. This load combination should also be used for
the investigation of slope stability.
Service II—Load combination intended to control This load combination corresponds to the overload
yielding of steel structures and slip of slip-critical provision for steel structures in past editions of the
connections due to vehicular live load. AASHTO Specifications, and it is applicable only to
steel structures. From the point of view of load level,
Service III—Load combination for longitudinal analysis Prior to 2014, the longitudinal analysis relating to
relating to tension in prestressed concrete tension in prestressed concrete superstructures was
superstructures with the objective of crack control and to investigated using a load factor for live load of 0.8. The
principal tension in the webs of segmental concrete live load specified in these specifications This load
girders. factor reflecteds, among other things, current exclusion
weight limits mandated by various jurisdictions at the
time of the development of the specifications in 1993.
Vehicles permitted under these limits have been in
service for many years prior to 1993. It was concluded at
that time that, for longitudinal loading, there is no
nationwide physical evidence that these vehicles have
caused cracking in existing prestressed concrete
components. The 0.8 load factor was applied regardless
of the method used for determining the loss of
prestressing. The statistical significance of the
traffic lane.
The calibration of the service limit states for
concrete components (Wassef et. al. 2014) concluded
that typical components designed using the Refined
Estimates of Time-Dependent Losses method
incorporated in the specifications in 2005 have a lower
reliability index against flexural cracking in prestressed
components than components designed using the
prestress loss calculation method specified prior to 2005.
For components designed using the currently-specified
methods for instantaneous prestressing losses and the

241
currently-specified Refined Estimates of Time-

Dependent Losses method, an increase in the load factor
for live load from 0.8 to 1.0 was required to maintain the
level of reliability against cracking of prestressed
concrete components inherent in the system.
Components which design satisfies all of the
following conditions:
• A refined time step method is used for
calculating the time-dependent prestressing
losses
• The section properties are determined based on
the concrete gross section, and,
• The force in prestressing steel is determined
without taking advantage of the elastic gain,
were not affected by the changes in the prestressing loss
calculation method introduced in 2005. For these
components, a load factor for live load of 0.8 was
maintained.
Service I should be used for checking tension
related to transverse analysis of concrete segmental
girders.
torsion.
Service IV—Load combination relating only to tension
in prestressed concrete columns with the objective of
crack control.

DD
DW
EH
EV LL
ES IM
EL CE
Load PS BR
Combination CR PL W
Limit State SH LS WA WS L FR TU TG SE EQ BL IC CT CV
(unless noted)
Strength IV γp — 1.00 — — 1.00 0.50/1.20 — — — — — — —
Extreme γp γEQ 1.00 — — 1.00 — — — 1.00 — — — —
Event I
Extreme γp 0.50 1.00 — — 1.00 — — — — 1.00 1.00 1.00 1.00
Event II
Service I 1.00 1.00 1.00 0.30 1.0 1.00 1.00/1.20 γTG γSE — — — — —
Service II 1.00 1.30 1.00 — — 1.00 1.00/1.20 — — — — — — —
γLL
Service IV 1.00 — 1.00 0.70 — 1.00 1.00/1.20 — 1.0 — — — — —
Fatigue I— — 1.50 — — — — — — — — — — — —
LL, IM & CE
only
Fatigue II— — 0.75 — — — — — — — — — — — —
LL, IM & CE
only

242
Table 3.4.1-4—Load Factors for Live Load for Service III Load Combination, γLL
Component γLL
Prestressed concrete components designed using a refined time step method 0.8
to determine the time-dependant prestressing losses in conjunction with the
gross section properties and without taking advantage of the elastic gain
All other prestressed concrete components 1.0

243
7.5 Fatigue
•
•
•
•

DD
DW
EH
EV LL
ES IM
EL CE
Load PS BR
Combination CR PL
Limit State SH LS WA WS WL FR TU TG SE EQ BL IC CT CV
(unless noted)
0
Strength IV γp — 1.00 — — 1.00 0.50/1.20 — — — — — — —
0
Extreme γp γEQ 1.00 — — 1.00 — — — 1.00 — — — —
Event I
Extreme γp 0.50 1.00 — — 1.00 — — — — 1.00 1.00 1.00 1.00
Event II
Service I 1.00 1.00 1.00 0.3 1.0 1.00 1.00/1.20 γTG γSE — — — — —
0
Service II 1.00 1.30 1.00 — — 1.00 1.00/1.20 — — — — — — —
Service IV 1.00 — 1.00 0.7 — 1.00 1.00/1.20 — 1.0 — — — — —
0
Fatigue I— — 1.50 — — — — — — — — — — — —
LL, IM & CE 2.0
only
Fatigue II— — 0.75 — — — — — — — — — — — —
LL, IM & CE 0.80
only
5.5.3 Fatigue Limit State
•
•
•
•

244
5.5.3.2—Reinforcing Bars C5.5.3.2
The constant-amplitude fatigue threshold, (ΔF)TH, Bends in primary reinforcement should be avoided
for straight reinforcement and welded wire in regions of high stress range.
reinforcement without a cross weld in the high-stress Structural welded wire reinforcement has been
region shall be taken as: increasingly used in bridge applications in recent years,
especially as auxiliary reinforcement in bridge I- and
( ∆F )TH =24 − 20 f min / f y (5.5.3.2-1) box beams and as primary reinforcement in slabs.
Design for shear has traditionally not included a fatigue
check of the reinforcement as the member is expected to
( ∆F )TH 30 − 25 f min / f y
= (5.5.3.2-1) be uncracked under service conditions and the stress
range in steel minimal. The stress range for steel bars
The constant-amplitude fatigue threshold, (ΔF)TH, has existed in previous editions. It is based on Hansen et
for straight welded wire reinforcement with a cross weld al. (1976). The simplified form in this edition replaces
in the high-stress region shall be taken as: the (r/h) parameter with the default value 0.3
recommended by Hansen et al. Inclusion of limits for
( ∆F )TH 16 − 0.33 f min
= (5.5.3.2-2) WWR is based on recent studies by Hawkins et al.
(1971, 1987) and Tadros et al. (2004). Coefficients in
Eqs. 5.5.3.2-1 and 5.5.3.2-2 have been updated based on
( ∆F )TH =20 − 0.41 f min (5.5.3.2-2) calibration reported in Kulicki et al. (2013).
Since the fatigue provisions were developed based
where: primarily on ASTM A615 steel reinforcement, their
applicability to other types of reinforcement is largely
fmin = minimum live load stress resulting from the unknown. Consequently, a cautionary note is added to
Fatigue I load combination, combined with the the Commentary.
more severe stress from either the permanent
loads or the permanent loads, shrinkage, and
creep-induced external loads; positive if
tension, negative if compression (ksi)
The definition of the high-stress region for application of

Eqs. 5.5.3.2-1 and 5.5.3.2-2 for flexural reinforcement
shall be taken as one-third of the span on each side of the
section of maximum moment.
6.6.1—Fatigue
6.6.1.1—General C6.6.1.1
• .
•
•
•

245
6.6.1.2.3—Detail Categories C6.6.1.2.3

Components and details shall be designed to satisfy Components and details susceptible to load-induced
the requirements of their respective detail categories fatigue cracking have been grouped into eight
summarized in Table 6.6.1.2.3-1. Where bolt holes are categories, called detail categories, by fatigue resistance.
depicted in Table 6.6.1.2.3-1, their fabrication shall Experience indicates that in the design process the
conform to the provisions of Article 11.4.8.5 of the fatigue considerations for Detail Categories A through
AASHTO LRFD Bridge Construction Specifications. B′ rarely, if ever, govern. Nevertheless, Detail
Where permitted for use, unless specific information is Categories A through B′ have been included in Table
available to the contrary, bolt holes in cross-frame, 6.6.1.2.3-1 for completeness. Investigation of
diaphragm, and lateral bracing members and their components and details with a fatigue resistance based
connection plates shall be assumed for design to be on Detail Categories A through B′ may be appropriate in
punched full size. unusual design cases.
Except as specified herein for fracture critical Table 6.6.1.2.3-1 illustrates many common details
members, where the projected 75-year single lane found in bridge construction and identifies potential
Average Daily Truck Traffic (ADTT)SL is less than or crack initiation points for each detail. In Table 6.6.1.2.3-
equal to that specified in Table 6.6.1.2.3-2 for the 1, “Longitudinal” signifies that the direction of applied
component or detail under consideration, that stress is parallel to the longitudinal axis of the detail.
component or detail should be designed for finite life “Transverse” signifies that the direction of applied stress
using the Fatigue II load combination specified in Table is perpendicular to the longitudinal axis of the detail.
3.4.1-1. Otherwise, the component or detail shall be Category F for allowable shear stress range on the
designed for infinite life using the Fatigue I load throat of a fillet weld has been eliminated from
combination. The single-lane Average Daily Truck Table 6.6.1.2.3-1. When fillet welds are properly sized for
Traffic (ADTT)SL shall be computed as specified in strength considerations, Category F should not govern.
Article 3.6.1.4.2. Fatigue will be governed by cracking in the base metal at
For components and details on fracture-critical the weld toe and not by shear on the throat of the weld.
members, the Fatigue I load combination specified in Research on end-bolted cover plates is discussed in Wattar
Table 3.4.1-1 should be used in combination with the et al. (1985).
nominal fatigue resistance for infinite life specified in Where the design stress range calculated using the
Article 6.6.1.2.5. Fatigue I load combination is less than (ΔF)TH, the detail
Orthotropic deck components and details shall be will theoretically provide infinite life. Except for
designed to satisfy the requirements of their respective Categories E and E′, for higher traffic volumes, the design
detail categories summarized in Table 6.6.1.2.3-1 for the will most often be governed by the infinite life check.
chosen design level shown in the table and as specified Table 6.6.1.2.3-2 shows for each detail category the
in Article 9.8.3.4. values of (ADTT)SL above which the infinite life check
governs, assuming a 75-yr design life and one stress range
cycle per truck.
The values in the second column of Table 6.6.1.2.3-
The constant in the 2 were computed as follows:
denominator of the equation to 75 _ Year ( ADTT ) SL = 3
A
 ( ∆F )TH 
the right has been changed   ( 365 )( 75 )( n )
(C6.6.1.2.3-1)
from 2 to 2.5.  2.5 
using the values for A and (∆F)TH specified in
Tables 6.6.1.2.5-1 and 6.6.1.2.5-3, respectively, a fatigue
design life of 75 yr and a number of stress range cycles per
truck passage, n, equal to one. These values were rounded
up to the nearest five trucks per day. That is, the indicated
values were determined by equating infinite life and finite
life resistances with due regard to the difference in load
factors used with the Fatigue I and Fatigue II load
combinations. For other values of n, the values in Table
6.6.1.2.3-2 should be modified by dividing by the
appropriate value of n taken from Table 6.6.1.2.5-2. For
other values of the fatigue design life, the values in Table
6.6.1.2.3-2 should be modified by multiplying the values
by the ratio of 75 divided by the fatigue life sought in
years. Some of the values of the parameter A and the
threshold (∆F)TH have been revised based on a calibration
reported in Kulicki et al. (2013).

246
The procedures for load-induced fatigue are

followed for orthotropic deck design. Although the local
structural stress range for certain fatigue details can be
caused by distortion of the deck plate, ribs, and
floorbeams, research has demonstrated that load-induced
fatigue analysis produces a reliable assessment of
fatigue performance.
Considering the increased γLL and cycles per truck
passage (n) in orthotropic decks, the 75-yr ADTTSL
equivalent to infinite life (trucks per day) results in 870
for deck plate details and 4350 for all other details,
based on Category C. Thus, finite life design may
produce more economical designs on lower-volume
roadways.

247
Table 6.6.1.2.3-1—Detail Categories for Load-Induced Fatigue

Constant Threshold
A (ΔF)TH Potential Crack
Description Category (ksi3) ksi Initiation Point Illustrative Examples
Section 1—Plain Material away from Any Welding
1.1 Base metal, except A 250 × 108 24 Away from all
noncoated weathering steel, welds or
with rolled or cleaned structural
surfaces. Flame-cut edges connections
with surface roughness value
of 1,000 µ-in. or less, but
without re-entrant corners.
1.2 Noncoated weathering B 120 × 108 16 Away from all
steel base metal with rolled or welds or
cleaned surfaces designed and structural
detailed in accordance with connections
FHWA (1989). Flame-cut
edges with surface roughness
value of 1,000 µ-in. or less,
but without re-entrant corners.
1.3 Member with re-entrant C 44 × 108 10 At any external
corners at copes, cuts, block- edge
outs or other geometrical
discontinuities made to the
requirements of
AASHTO/AWS D1.5, except
weld access holes.
1.4 Rolled cross sections with C 44 x 108 10 In the base
weld access holes made to the metal at the
requirements of re-entrant
AASHTO/AWS D1.5, Article corner of the
3.2.4. weld access
hole
1.5 Open holes in members D 22 × 108 7 In the net

(Brown et al., 2007). 21 × 108 8 section
originating at
the side of the
hole
Section 2—Connected Material in Mechanically Fastened Joints

2.1 Base metal at the gross B 120 × 108 16 Through the
section of high-strength bolted gross section
joints designed as slip-critical near the hole
connections with pretensioned
high-strength bolts installed in
holes drilled full size or
subpunched and reamed to
size—e.g., bolted flange and
web splices and bolted
stiffeners. (Note: see Condition
2.3 for bolt holes punched full
size; see Condition 2.5 for
bolted angle or tee section
member connections to gusset
or connection plates.)

248
Table 6.6.1.2.3-1 (continued)—Detail Categories for Load-Induced Fatigue

Constant Threshold Potential
A (ΔF)TH Crack
Section 2—Connected Material in Mechanically Fastened Joints (continued)
2.2 Base metal at the net section of B 120 × 108 16 In the net
high-strength bolted joints designed section
as bearing-type connections but originating at
fabricated and installed to all the side of the
requirements for slip-critical hole
connections with pretensioned high-
strength bolts installed in holes
drilled full size or subpunched and
reamed to size. (Note: see Condition
2.3 for bolt holes punched full size;
see Condition 2.5 for bolted angle or
tee section member connections to
gusset or connection plates.)
2.3 Base metal at the net section of all D 22 × 108 7 In the net
bolted connections in hot dipped 21 × 108 8 section
galvanized members (Huhn and originating at
Valtinat, 2004); base metal at the the side of the
appropriate section defined in hole or
Condition 2.1 or 2.2, as applicable, of through the
high-strength bolted joints with gross section
pretensioned bolts installed in holes near the hole,
punched full size (Brown et al., 2007); as applicable
and base metal at the net section of
other mechanically fastened joints,
except for eyebars and pin plates, e.g.,
joints using ASTM A307 bolts or
non-pretensioned high-strength bolts.
(Note: see Condition 2.5 for bolted
angle or tee section member
connections to gusset or connection
plates).
2.4 Base metal at the net section of E 11 × 108 4.5 In the net
eyebar heads or pin plates (Note: 12 × 108 section
for base metal in the shank of originating at
eyebars or through the gross section the side of the
of pin plates, see Condition 1.1 or hole
1.2, as applicable.)
2.5 Base metal in angle or tee See See See Through the
section members connected to a applicable applicable applicable gross section L
gusset or connection plate with Category Constant Threshold near the hole, x
above above above
high-strength bolted slip-critical or in the net c.g.
connections. The fatigue stress section

range shall be calculated on the originating at
effective net area of the member, the side of the
L
Ae = UAg, in which U=(1- x /L) and hole, as c.g. x
where Ag is the gross area of the applicable
member. x is the distance from the
centroid of the member to the
surface of the gusset or connection
plate and L is the out-to-out
distance between the bolts in the
connection parallel to the line of
force. The effect of the moment due
to the eccentricities in the
connection shall be ignored in
computing the stress range
(McDonald and Frank, 2009).

249

A (ΔF)TH Crack
2.5 (continued) The fatigue category

shall be taken as that specified for
Condition 2.1. For all other types of
bolted connections, replace Ag with
the net area of the member, An, in
computing the effective net area
according to the preceding equation
and use the appropriate fatigue
category for that connection type
specified for Condition 2.2 or 2.3, as
applicable.
Section 3—Welded Joints Joining Components of Built-Up Members
3.1 Base metal and weld metal in B 120 × 108 16 From surface
members without attachments built or internal
up of plates or shapes connected by discontinuities
continuous longitudinal complete in the weld
joint penetration groove welds away from the
back-gouged and welded from the end of the
second side, or by continuous fillet weld
welds parallel to the direction of
applied stress.
3.2 Base metal and weld metal in B′ 61 × 108 12 From surface
members without attachments built or internal
up of plates or shapes connected by 13 discontinuities
continuous longitudinal complete in the weld,
joint penetration groove welds with including weld
backing bars not removed, or by attaching
continuous partial joint penetration backing bars
groove welds parallel to the
direction of applied stress.
3.3 Base metal and weld metal at D 22 × 108 7 From the weld
the termination of longitudinal 8
termination
welds at weld access holes made to 21 × 10 8 into the web or
the requirements of AASHTO/AWS flange
D1.5, Article 3.2.4 in built-up
members. (Note: does not include
the flange butt splice).
3.4 Base metal and weld metal in B 120 × 108 16 From surface
partial length welded cover plates or internal
connected by continuous fillet discontinuities
welds parallel to the direction of in the weld
applied stress. away from the
end of the
weld

250

A (Δf)TH Crack
Section 3—Welded Joints Joining Components of Built-Up Members (continued)
3.5 Base metal at the termination of In the flange at

partial length welded cover plates the toe of the
having square or tapered ends that end weld or in
are narrower than the flange, with the flange at the
or without welds across the ends, or termination of
cover plates that are wider than the the longitudinal
flange with welds across the ends: weld or in the
edge of the
flange with
wide cover
plates
11 × 108
Flange thickness ≤ 0.8 in. E 4.5
12 × 108
Flange thickness > 0.8 in. E′ 3. 9 × 108 2.6
8
3.5 × 10 3.1
3.6 Base metal at the termination of B 120 × 108 16 In the flange at

partial length welded cover plates the termination
with slip-critical bolted end of the
connections satisfying the longitudinal
requirements of Article 6.10.12.2.3. weld
3.7 Base metal at the termination of E′ 3.9 × 108 2.6 In the edge of
partial length welded cover plates 8
the flange at
that are wider than the flange and 3.5 × 10 3.1 the end of the
without welds across the ends. cover plate
weld
Section 4—Welded Stiffener Connections
4.1 Base metal at the toe of C′ 44 × 108 12 Initiating from

transverse stiffener-to-flange fillet the
welds and transverse stiffener-to- geometrical
web fillet welds. (Note: includes discontinuity
similar welds on bearing stiffeners at the toe of
and connection plates). the fillet weld
extending into
the base metal
4.2 Base metal and weld metal in B 120 × 108 16 From the
longitudinal web or longitudinal surface or
box-flange stiffeners connected by internal
continuous fillet welds parallel to discontinuities
the direction of applied stress. in the weld
away from the
end of the
weld

251

Potential
Constant Threshold Crack
A (Δf)TH Initiation
Description Category (ksi3) ksi Point Illustrative Examples
Section 4—Welded Stiffener Connections (continued)
4.3 Base metal at the termination of
longitudinal stiffener-to-web or
longitudinal stiffener-to-box flange
welds:
With the stiffener attached by fillet In the
welds and with no transition radius primary
provided at the termination: member at
the end of
Stiffener thickness < 1.0 in. the weld at
11 × 108
Stiffener thickness ≥ 1.0 in. E 4.5 the weld toe
12 × 108
E′
3.9 × 108
2.6
3.5 × 108
3.1
With the stiffener attached by welds
and with a transition radius R
provided at the termination with the
weld termination ground smooth:
R ≥ 24 in. B 120 × 108 16

In the
8
24 in. > R ≥ 6 in. C 44 × 10 10 primary
member near
6 in. > R ≥ 2 in. D 22 × 108 7 the point of
21 × 108 8 tangency of
2 in. > R
the radius
11 × 108
E 12 × 108 4.5
Section 5—Welded Joints Transverse to the Direction of Primary Stress
5.1 Base metal and weld metal in or From internal

adjacent to complete joint penetration discontinuities
groove welded butt splices, with weld in the filler
soundness established by NDT and metal or along
with welds ground smooth and flush the fusion
parallel to the direction of stress. boundary or at
Transitions in thickness or width shall the start of the
be made on a slope no greater than transition
1:2.5 (see also Figure 6.13.6.2-1).
Fy < 100 ksi B 120 × 108 16

8
Fy ≥ 100 ksi B′ 61 × 10 12
13
8
5.2 Base metal and weld metal in or B 120 × 10 16 From internal
adjacent to complete joint penetration discontinuities
groove welded butt splices, with weld in the filler
soundness established by NDT and metal or
with welds ground parallel to the discontinuities
direction of stress at transitions in along the
width made on a radius of not less fusion
than 2 ft with the point of tangency at boundary
the end of the groove weld (see also
Figure 6.13.6.2-1).

252
Constant Threshold
A (Δf)TH Potential Crack
8
5.3 Base metal and weld metal in C 44 × 10 10 From the
or adjacent to the toe of complete surface
joint penetration groove welded T discontinuity at
or corner joints, or in complete the toe of the
joint penetration groove welded weld extending
butt splices, with or without into the base
transitions in thickness having metal or along
slopes no greater than 1:2.5 when the fusion
weld reinforcement is not removed. boundary
(Note: cracking in the flange of the
“T” may occur due to out-of-plane
bending stresses induced by the
stem).
5.4 Base metal and weld metal at C as 44 × 108 10 Initiating from

details where loaded discontinuous adjusted the geometrical
plate elements are connected with a in Eq. discontinuity at
pair of fillet welds or partial joint 6.6.1.2.5-4 the toe of the
penetration groove welds on weld extending
opposite sides of the plate normal into the base
metal or
to the direction of primary stress.
initiating at the
weld root subject
to tension
extending up and
then out through
the weld
Section 6—Transversely Loaded Welded Attachments
6.1 Base metal in a longitudinally Near point of

loaded component at a transversely tangency of the
loaded detail (e.g. a lateral radius at the
connection plate) attached by a edge of the
weld parallel to the direction of longitudinally
primary stress and incorporating a loaded
transition radius R with the weld component or at
termination ground smooth. the toe of the
weld at the
weld
termination if
not ground
smooth
R ≥ 24 in. B 120 × 108 16
24 in. > R ≥ 6 in. C 44 × 108 10
6 in. > R ≥ 2 in. D 22 × 108 7

21× 108 8
11 × 108
2 in. > R E 4.5
12 × 108
11 × 108
For any transition radius with the E 4.5
12 × 108
weld termination not ground
smooth (Note: Condition 6.2, 6.3
or 6.4, as applicable, shall also be
checked.)

253

Constant Threshold
A (Δf)TH Potential Crack
Section 6—Transversely Loaded Welded Attachments (continued)
6.2 Base metal in a transversely

loaded detail (e.g. a lateral
connection plate) attached to a
longitudinally loaded component of
equal thickness by a complete joint
penetration groove weld parallel to
the direction of primary stress and
incorporating a transition radius R,
with weld soundness established by
NDT and with the weld termination
ground smooth:
With the weld reinforcement
removed:
R ≥ 24 in. B 120 × 108 16 Near points of

tangency of the
24 in. > R ≥ 6 in. C 44 × 108 10 radius or in the
8 weld or at the
6 in. > R ≥ 2 in. D 22 × 10 7 fusion boundary of
21 × 108 the longitudinally
8
loaded component
or the transversely
11 × 108 loaded attachment
2 in. > R E 4.5
12 × 108
With the weld reinforcement not At the toe of the
removed: weld either along
the edge of the
R ≥ 24 in. longitudinally
C 44 × 108 10
loaded component
24 in. > R ≥ 6 in. C 44 × 10 8
10 or the transversely
loaded attachment
8
6 in. > R ≥ 2 in. D 22 × 10 7
8
21 x 10 8
11 × 108
2 in. > R E 4.5
(Note: Condition 6.1 shall also be 12 × 108
checked.)
6.3 Base metal in a transversely At the toe of the
loaded detail (e.g. a lateral weld along the
connection plate) attached to a edge of the thinner
longitudinally loaded component plate
of unequal thickness by a In the weld
complete joint penetration groove termination of
weld parallel to the direction of small radius weld
primary stress and incorporating a transitions
weld transition radius R, with
weld soundness established by At the toe of the
NDT and with the weld weld along the
termination ground smooth: edge of the thinner
plate
With the weld reinforcement
removed:
D 22 × 108 7
R ≥ 2 in. 8
21 × 10 8
11 × 108
R < 2 in. E 4.5
12 × 108
For any weld transition radius 11 × 108
with the weld reinforcement not E 4.5
removed (Note: Condition 6.1 12 × 108
shall also be checked.)

254

Potential
Section 6—Transversely Loaded Welded Attachments (continued)
6.4 Base metal in a transversely See

loaded detail (e.g. a lateral Condition
connection plate) attached to a 5.4
longitudinally loaded component by
a fillet weld or a partial joint
penetration groove weld, with the
weld parallel to the direction of
primary stress (Note: Condition 6.1
shall also be checked.)
Section 7—Longitudinally Loaded Welded Attachments
7.1 Base metal in a longitudinally In the primary

loaded component at a detail with a member at the
length L in the direction of the end of the
primary stress and a thickness t weld at the
attached by groove or fillet welds weld toe
parallel or transverse to the direction
of primary stress where the detail
incorporates no transition radius:
L < 2 in. C 44 × 108 10

8
2 in. ≤ L ≤ 12t or 4 in D 22 × 10 7
8
21 × 10 8
L > 12t or 4 in.
11 × 108
t < 1.0 in. E 4.5
12 × 108
t ≥ 1.0 in. E′ 3.9 × 108 2.6
8
(Note: see Condition 7.2 for welded 3.5 × 10 3.1
angle or tee section member
connections to gusset or connection
plates.)
7.2 Base metal in angle or tee section E 11x108 4.5 Toe of fillet
members connected to a gusset or 8
welds in
connection plate by longitudinal fillet 12x10 connected L
L
welds along both sides of the element x

connected element of the member c.g.
cross-section. The fatigue stress

range shall be calculated on the
effective net area of the member, Ae L
= UAg, in which U = (1– x /L) and
L
c.g. x
where Ag is the gross area of the
member. x is the distance from the
centroid of the member to the surface
of the gusset or connection plate and
L is the maximum length of the
longitudinal welds. The effect of the
moment due to the eccentricities in
the connection shall be ignored in
computing the stress range
(McDonald and Frank, 2009).

255

A (Δf)TH Crack
Section 8—Miscellaneous
8.1 Rib to Deck Weld—One-sided C 44 × 108 10 See Figure

80% (70% min) penetration weld
with root gap ≤ 0.02 in. prior to
welding
Allowable Design Level
1, 2, or 3
8.2 Rib Splice (Welded)—Single D 22 × 108 7 See Figure

groove butt weld with permanent 8
backing bar left in place. Weld gap 21 × 10 8
> rib wall thickness

1, 2, or 3
8.3 Rib Splice (Bolted)—Base B 120 × 108 16 See Figure

metal at gross section of high
strength slip critical connection

1, 2, or 3
8.4 Deck Plate Splice (in Plane)— D 22 × 108 7 See Figure

Transverse or Longitudinal single 8
groove butt splice with permanent 21 × 10 8
backing bar left in place

1, 2, or 3
8.5 Rib to FB Weld (Rib)—Rib C 44 × 108 10 See Figure

wall at rib to FB weld (fillet or CJP)

1, 2, or 3

256

Potential
8.6 Rib to FB Weld (FB Web)—FB C 44 × 108 10 See Figure

web at rib to FB weld (fillet, PJP, or
CJP) (see
Note 1)

1 or 3
8.7 FB Cutout—Base metal at edge A 250 × 108 24 See Figure

with “smooth” flame cut finish as per
AWS D1.5

1 or 3
8.8 Rib Wall at Cutout—Rib wall at C 44 × 108 10 See Figure

rib to FB weld (fillet, PJP, or CJP)

1 or 3
8.9 Rib to Deck Plate at FB C 44 × 108 10 See Figure

1 or 3
Note 1: Where stresses are dominated by in-plane component at fillet or PJP welds, Eq. 6.6.1.2.5-4 shall be considered. In this case, ∆f should
be calculated at the mid-thickness and the extrapolation procedure as per Article 9.8.3.4.3 need not be applied.
Section 9—Miscellaneous
9.1 Base metal at stud-type shear 44 × 108 10 At the toe of

connectors attached by fillet or the weld in
automatic stud welding the base
metal

257

Constant Threshold
A (ΔF)TH Potential Crack
Section 9—Miscellaneous (continued)
9.2 Nonpretensioned high- At the root of

strength bolts, common bolts, the threads
threaded anchor rods, and extending into
hanger rods with cut, ground, the tensile
or rolled threads. Use the stress stress area
range acting on the tensile
stress area due to live load plus
prying action when applicable.
(Fatigue II) Finite Life E′ 3.9 × 108 N/A
3.5 × 108
(Fatigue I) Infinite Life D N/A 7
8
Table 6.6.1.2.3-2—75-yr (ADTT)SL Equivalent to Infinite

Life
Detail 75-yrs (ADTT)SL Equivalent to Infinite

Category Life (trucks per day)
A 1030
B 1670
B′ 1585
C 2510
C′ 1455
D 2340
E 7515
E′ 6710
6.6.1.2.5—Fatigue Resistance C6.6.1.2.5

•
•
•
•

258
Table 6.6.1.2.5-1—Detail Category Constant, A
Constant, A
Detail Category times 108 (ksi3)
A 250.0
B 120.0
B′ 61.0
C 44.0
C′ 44.0
D 22.0
21.0
E 12.0
E′ 3.9
3.5
M 164 (A325) Bolts in 17.1
Axial Tension
M 253 (A490) Bolts in 31.5
Axial Tension
Table 6.6.1.2.5-2—Cycles per Truck Passage, n For the purpose of determining the stress range
cycles per truck passage for continuous spans, a distance
Longitudinal Span Length equal to one-tenth the span on each side of an interior
Members >40.0 ft ≤40.0 ft support should be considered to be near the support.
Simple Span 1.0 2.0 Values of n for longitudinal members have been revised
Girders based on the calibration reported in Kulicki et al., 2013.
Continuous The number of stress range cycles per passage is
Girders taken as 5.0 for cantilever girders because this type of
bridge is susceptible to large vibrations, which cause
1) near 1.5 2.0 additional cycles after the truck has left the bridge
interior (Moses et al., 1987; Schilling, 1990).
support Orthotropic deck details that are connected to the
2) elsewhere 1.0 2.0 deck plate (e.g., the rib-to-deck weld) are subjected to
Cantilever 5.0 cycling from direct individual wheel loads. Thus, the
Girders passage of one design truck results in five fatigue load
Orthotropic 5.0 cycles as each axle produces one load cycle. The force
Deck Plate effect (∆f) can be conservatively taken as the worst case
Connections from the five wheels or by application of Miner’s Rule
Subjected to determine the effective stress range from the group of
to Wheel wheels.
Load Cycling
Trusses 1.0
Transverse Spacing
Members > 20.0 ft ≤20.0 ft
1.0 2.0

259
Table 6.6.1.2.5-3—Constant-Amplitude Fatigue Thresholds
Detail Category Threshold (ksi)

A 24.0
B 16.0
B′ 12.0
13.0
C 10.0
C′ 12.0
D 7.0
8.0
E 4.5
E′ 2.6
3.1
M 164 (A 325) Bolts in 31.0
Axial Tension
M 253 (A 490) Bolts in 38.0
Axial Tension

Chapter 8
Purpose and Contents of Appendix F
The purpose of Appendix F to the main report is to facilitate calibration of locally preferred methods of predicting set-
future enhancement of service limit state calibration, as well tlement of shallow and deep foundations in lieu of the
as to allow for customization by individual bridge owners. To Hough (1959) and Schmertmann et al. (1978) methods
accomplish this purpose, certain guidelines and databases used in this report.
have been included in Appendix F. • Sections F.5 and F.6—Descriptions of the concrete and
steel bridges, respectively, selected from the NCHRP Proj-
• Section F.2—A step-by-step description of the implementa- ect 12-78 database and used for various studies reported here
tion of the Monte Carlo analysis introduced in Chapter 3. (Mlynarski et al. 2011). Other bridges could be substituted.
• Section F.3—Excerpt from the database of WIM (weigh in • Section F.7—A database of fatigue tests on steel details
motion) data used in the development of live load models (Keating and Fisher 1986; P. B. Keating, personal commu-
presented in Chapter 5. These data are available at http:// nication, 2012).
www.trb.org/Main/Blurbs/170201.aspx. • Section F.8—A database of fatigue tests on plain concrete
• Section F.4—A condensed version of the geotechnical and reinforcement (Ople and Hulsbos 1966; Fisher and
calibration reported in Chapter 6. It is anticipated that Viest 1961; Pfister and Hognestad 1964; Hanson et al. 1968;
the regional nature of geotechnical practice will result in Lash 1969).
260

Chapter 9
Summary and Recommendations
9.1 Summary processes for the various SLSs. These statistical parameters

were tabulated for various average daily truck traffic val-
From a survey of owners and a literature review that included ues from 250 to 10,000, time periods ranging from 1 day
other national and international bridge design specifica- to 100 years, and spans ranging from 30 to 200 ft.
tions, a set of possible service limit states (SLSs) was devel- • Analysis for fatigue that included treating the WIM data
oped. Those SLSs were reviewed to determine which could for each site as a continuous stream of traffic for rainflow
be calibrated using reliability theory. These are identified in counting and estimation of cumulative damage using the
Table 2.21 in Chapter 2. Calibrated, reliability-based load Palmgren–Miner model. This calibration also involved a
factors or resistance factors, or both, were developed for each
reassessment of a comprehensive international database of
of the following SLSs:
fatigue tests.
• Verification of the Service II load factor for permanent defor-
• Foundation deformations;
mations based on a WIM-based multiplier on HL-93 and a
• Cracking of reinforced concrete components;
set of bridges from the NCHRP Project 12-78 database. This
• Live load deflections;
•
verification was further informed by considering the average
Permanent deformations;
•
number of times per year that bending moment for various
Tensile stresses in prestressed concrete components; and
•
spans exceeded various percentages of the HL-93 bending
Fatigue of steel and reinforced concrete components.
moment.
• A statistically calibrated load factor for live load for use in
Although the same general process was followed in calibrat-
ing each of the SLSs, customization was needed in most cases. the Service III load combination for prestressed concrete
The calibration process produced target reliability levels in tension.
that were much different from those used for the strength cali- • A reliability-based comparison of pre- and post-2005 loss
bration. This was expected. Although the strength calibration, calculations for prestressed concrete components.
with some exceptions, was based on a target reliability index • A calibration process for foundation deformations, which
of about 3.5 for a 75-year life, most of the SLS calibration was was demonstrated by calibration of both the Schmertmann
done with a target reliability index of around 1.0 to 1.5 based on et al. (1978) and Hough (1959) methods of predicting set-
an annual probability. Once the target reliabilities were deter- tlement of footings on the basis of measured field data and
mined, changes to load factors, resistance factors, or other the presentation of a methodology to allow for calibration
design parameters were developed. Proposed modifications to of locally preferred settlement models. The same processes
AASHTO LRFD design provisions were drafted for consider- can be applied to horizontal movements, as well as other
ation by the appropriate AASHTO Highway Subcommittee foundation types (e.g., deep foundations). The foundation
on Bridges and Structures Technical Committees. deformation calibration processes are formulated to allow
Several of the more important underlying changes needed adaptation to different levels of reliability as expressed by
to develop the proposed specification changes included the different values of probability of exceedance (or reliability
following: index).
• Use of frequency–deflection–perception–based criteria to
• Use of WIM (weigh in motion) data to develop multipliers calibrated bridge movement. These criteria have been used
for the HL-93 live load model to be used in the calibration by Ontario and Canadian bridge codes for over two decades.
261

262
The objectives of this project were to provide calibrated SLSs 9.2 Recommendations

to provide 100-year life and to develop a framework for further
development of calibrated SLSs. Generally, it has been assumed The following recommendations are made:
that maintenance activities will be sufficient to prevent sig-
nificant loss of the strength and stiffness that would result • Submit the specification proposals in Chapter 7 to the
in unsatisfactory performance. No well-accepted direct link AASHTO Highway Subcommittee on Bridges and Struc-
between the National Bridge Inventory (NBI) condition data tures for possible implementation.
and the types of unsatisfactory performance related to the SLSs • Evaluate the feasibility of establishing a clearinghouse for
calibrated in this study has been found. Several locally devel- observed SLS issues, such as unsatisfactory user perception,
oped predictors of changes in the NBI condition number over permanent distortion, unanticipated cracking of concrete
time have been presented to provide guidance to owners on components, and settlement.
possible changes to the resistance side of the limit states used • Continue to interact with LTBPP to recommend SLS-
here within the context of the direct-link caveat above. oriented data collection, as well as to take advantage of
There are three sources of significant future improvements evolving data and knowledge on behavior and deteriora-
to the reported work: tion to improve SLS calibration and design provisions.
• Assess the impact of any of the specification proposals that
• Improved limit state functions; are adopted by working directly with the states involved or
• Improved knowledge of structural behavior; and by engaging LTBPP to include data gathering in those proj-
• Improved knowledge of the change in strength and stiffness ects. In either event, a set of expectations, associated met-
with time (i.e., response deterioration). rics, and assessment protocols should be developed for
consistent evaluation of impacts.
As discussed in Chapter 2, the limit state function related • Initiate a research project to identify possible ways to more
to permanent deformations is based on six data points from directly link NBI data to changes in resistance, as they may
the AASHO Road Tests (1962) and the anecdotal evidence affect both service and strength behavior.
that there appears to be virtually no record of permanent • Initiate a research project to identify failure modes in decks,
deformation in the superstructures of modern steel bridges potentially differentiating among shear and bending fail-
that has been tied to premature yielding or bolt slip. It is rea- ure modes, failure modes in composite and noncomposite
sonable to ask whether any systematic assessment of this type systems, end diaphragm and flange width and fillet effects,
of response is being made. and possible fatigue effects.
The same question about systematic assessment applies to
live load response, settlement, and cracking. Certainly very Finally, there is much interest nationally and internation-
exceptional cases are probably noted by individual owners, ally on the improved implementation of SLSs that should be
but detailed investigations are few in number and not cen- considered in any continued development of AASHTO LRFD.
trally archived. Fatigue cracking of steel components is prob- Some late developments are described in Walraven and Bicaj
ably the most identified and best recorded of the SLS responses, (2011), Balázs et al. (2013), and Helland (2013), all of which
as well as being the most directly tied to truck traffic density address provisions in the 2010 fib Model Code (Fédération
and time in the design process. Internationale du Béton 2010).
The ongoing Long-Term Bridge Performance Program
(LTBPP) is probably the best source of improved future data
on loads, response, and deterioration (both response deterio- 9.3 Implementation
ration and routine maintenance needs as a function of time
and environment). The initial focus, based on stake-holder If the recommendations in Chapter 7 are accepted by the
input, is on bridge decks, but other aspects of bridge perfor- AASHTO Highway Subcommittee on Bridges and Structures,
mance will also be considered. The project objective to cap- a webinar is probably the most effective way to explain changes
ture research-quality performance data on a systematically in design to state engineers and consultants.
selected national bridge set through visual inspection and A separate webinar or short National Highway Institute
nondestructive evaluation techniques should provide the course would be valuable to owners who want to modify the
information needed to advance calibrated SLSs further. databases in Appendix F and rerun calibration calculations.

References
AASHO Road Test: Report 4: Bridge Research. 1962. Special Report 61D. ACI Committee 318. 1995. Building Code Requirements for Structural
AASHO, Highway Research Board, Washington, D.C. Concrete and Commentary. ACI 318-95/318R-95. American Concrete
AASHTO. 2013. Bridges for Service Life Beyond 100 Years: Innovative Institute, Farmington Hills, Mich.
Systems, Subsystems, and Components. SHRP 2 Project R19A. http:// ACI Committee 318. 1999. Building Code Requirements for Structural
onlinepubs.trb.org/onlinepubs/shrp2/SHRP2prepubR19Areport. Concrete and Commentary. ACI 318-99/318R-99. American Concrete
pdf. Accessed Nov. 25. Institute, Farmington Hills, Mich.
AASHTO Guide Manual for Bridge Element Inspection. 2011. Ameri- ACI Committee 318. 2002. Building Code Requirements for Structural
can Association of State Highway and Transportation Officials, Concrete and Commentary. ACI 318-02/318R-02. American Concrete
Washington, D.C. Institute, Farmington Hills, Mich.
AASHTO LRFD Bridge Design Specifications, 4th ed. 2008. Includes 2008 ACI Committee 318. 2005. Building Code Requirements for Structural
Interim. American Association of State Highway and Transportation Concrete and Commentary. ACI 318-05. American Concrete Institute,
Officials, Washington, D.C. Farmington Hills, Mich.
and 2009 Interims. American Association of State Highway and Concrete and Commentary. ACI 318-08. American Concrete Institute,
Transportation Officials, Washington, D.C. Farmington Hills, Mich.
Interim. American Association of State Highway and Transportation Concrete and Commentary. ACI 318-11. American Concrete Institute,
Officials, Washington, D.C. Farmington Hills, Mich.
AASHTO LRFD Bridge Design Specifications, 6th ed. 2012. Ameri- ACI-ASCE Joint Committee 323. 1958. Tentative Recommendations for
can Association of State Highway and Transportation Officials, Prestressed Concrete. Journal, American Concrete Institute, Vol. 54,
Washington, D.C. No. 1, pp. 545–578.
Abeles, P., F. Barton, and E. Brown, II. 1969. Fatigue Behavior of Pre- Agom International, srl. 2013. Catalog for V-MAX Pot Bearings. http://
stressed Concrete Bridge Beams. In First International Symposium on www.agom.it/index.aspx?m=85&f=3&idf=212. Accessed Nov. 25.
Concrete Bridge Design, SP-23, American Concrete Institute, Detroit, Agrawal, A., and A. Kawaguchi. 2009. Bridge Element Deterioration
Mich., pp. 579–599. Rates. Project Number C-01-51. City College of New York, Trans-
Abeles, P., and E. Brown, II. 1971. Expected Fatigue Life of Prestressed portation Infrastructure Research Consortium, and New York State
Concrete Highway Bridges As Related to the Expected Load Spec- Department of Transportation, New York.
trum. In Second International Symposium on Concrete Bridge Design, Akbas, S., and F. Kulhawy. 2009. Axial Compression of Footings in
SP-26, American Concrete Institute, Detroit, Mich., pp. 962–1010. Cohesionless Soils: I—Load-Settlement Behavior. ASCE Journal of
Abeles, P., E. Brown, II, and C. Hu. 1974. Fatigue Resistance of Under- Geotechnical and Geoenvironmental Engineering, Vol. 135, No. 11,
Reinforced Prestressed Beams Subjected to Different Stress Ranges; pp. 1562–1574.
Miner’s Hypothesis. In Abeles Symposium: Fatigue of Concrete, SP-41, Allen, T., A. Nowak, and R. Bathurst. 2005. Transportation Research Cir-
American Concrete Institute, Detroit, Mich., pp. 279–300. cular E-C079: Calibration to Determine Load and Resistance Factors
ACI Committee 201. 2008. Guide to Durable Concrete. ACI 201.2R-08. for Geotechnical and Structural Design. Transportation Research Board
American Concrete Institute, Farmington Hills, Mich. of the National Academies, Washington, D.C. http://onlinepubs.trb
ACI Committee 215. 1974. Considerations for Design of Concrete Struc- .org/onlinepubs/circulars/ec079.pdf.
tures Subjected to Fatigue Loading. ACI 215R-74. American Concrete Amorn, W., J. Bowers, A. Girgis, and M. Tadros. 2007. Fatigue of
Institute, Detroit, Mich. Deformed Welded-Wire Reinforcement. Journal, Precast/Prestressed
ACI Committee 318. 1963. Building Code Requirements for Reinforced Concrete Institute, Vol. 52, No. 1, pp. 106–120.
Concrete. ACI 318-63. American Concrete Institute, Detroit, Mich. Arizona Department of Transportation (ADOT) Bridge Design Guide-
ACI Committee 318. 1977. Building Code Requirements for Reinforced lines. 2009. Arizona Department of Transportation, Phoenix.
Concrete. ACI 318-77. American Concrete Institute, Detroit, Mich. ASCE Committee on Deflections Limitations of Bridges of the Struc-
ACI Committee 318. 1983. Building Code Requirements for Reinforced tural Division. 1958. Deflection limitations of bridges. Journal of the
Concrete. ACI 318-83. American Concrete Institute, Detroit, Mich. Structural Division, Proc. ASCE, 84(ST 3): 1633-1 to 1633-20.
263

264
ASCE/SEI 7-10: Minimum Design Loads for Buildings and Other Struc- Brockenbrough, R., and F. Merritt (eds.). 2011. Structural Steel Designer’s
tures. 2010. American Society of Civil Engineers, Reston, Va. Handbook, 5th ed. McGraw-Hill, New York.
Azizinamini, A., E. Power, G. Myers, and H. Ozyildirim. 2013. Design Broms, B. 1965. Crack Width and Crack Spacing in Reinforced Concrete
Guide for Bridges for Service Life. Prepublication draft. Transporta- Members. Journal, American Concrete Institute, Vol. 62, No. 10, p. 1237.
tion Research Board of the National Academies, Washington, D.C. Burland, J., and M. Burbridge. 1984. Settlement of Foundations on
Balázs, G., P. Bisch, A. Borosnyói, O. Burdet, C. Burns, F. Ceroni, Sand and Gravel. Proceedings, Part I, Institution of Civil Engineers,
V. Cervenka, M. Chiorino, P. Debernardi, L. Eckfeldt, M. El-Badry, Vol. 78, No. 6, pp. 1325–1381.
E. Fehling, S. Foster, A. Ghali, V. Gribniak, M. Guiglia, G. Kaklauskas, Burton, K., and E. Hognestad. 1967. Fatigue Test of Reinforcing Bars-
R. Lark, P. Lenkei, M. Lorrain, A. Marí, J. Ozbolt, M. Pecce, A. Pérez Tack Welding of Stirrups. Journal, American Concrete Institute, Vol. 64,
Caldentey, M. Taliano, D. Tkalcic, J. Torrenti, L. Torres, F. Toutlemonde, No. 5, pp. 244–252.
T. Ueda, J. Vitek, and L. Vráblík. 2013. Design for SLS According to Canadian Highway Bridge Design Code. 2006. Includes Supplement 1,
fib Model Code 2010. Structural Concrete, Journal of the fib, Vol. 14, Supplement 2, and Supplement 3. Canadian Standards Association
No. 2, pp. 99–123. International, Toronto, Ontario, Canada.
Barker, M., and K. Barth. 2007. Live Load Deflection Serviceability of Chang, L., and Y. Lee. 2001. Evaluation and Policy for Bridge Deck Expan-
HPS Composite Steel Girder Bridges. 2007 World Steel Bridge Sym- sion Joints. FHWA/IN/JTRP-2000/01. Joint Transportation Research
posium Papers, New Orleans, La. Program, Indiana Department of Transportation, Indianapolis, and
Barker, M., L. Gandiaga, and J. Staebler. 2008. Serviceability Limits and Purdue University, West Lafayette, Ind.
Economical Steel Bridge Design. MPC-08-206. Mountain-Plains Con- Choi, Y., and B. Oh. 2009. Crack Width Formula for Transversely Post-
sortium, Fargo, N.Dak. Tensioned Concrete Deck Slabs on Box Girder Bridges. Structural
Barker, R., J. Duncan, K. Rojiani, P. Ooi, C. Tan, and S. Kim. 1991. Journal, American Concrete Institute, Vol. 106, No. 6, pp. 753–761.
NCHRP Report 343: Manuals for the Design of Bridge Foundations: Clark, A. 1956. Cracking in Reinforced Concrete Flexural Members. Jour-
Shallow Foundations, Driven Piles, Retaining Walls and Abutments, nal Proceedings, American Concrete Institute, Vol. 52, No. 4, p. 851.
Drilled Shafts, Estimating Tolerable Movements, Load Factor Design Coastal Engineering Manual. 2002. Engineer Manual 1110-2-1100. U.S.
Specifications, and Commentary. TRB, National Research Council, Army Corps of Engineers, Washington, D.C.
Washington, D.C. Committee on Deflection Limitations of Bridges of the Structural Divi-
Barth, K., and H. Wu. 2007. Development of Improved Natural Frequency sion. 1958. Deflection Limitations. ASCE Journal of the Structural
Equations for Continuous Span Steel I-Girder Bridges. Engineering Division, Vol. 84, No. 3, pp. 1633.1–1633.20.
Structures, Vol. 29, No. 12, pp. 3432–3442. Cooper, P., T. Galambos, and M. Ravindra. 1978. LRFD Criteria for Plate
Bartlett, F., R. Dexter, M. Graeser, J. Jelinek, B. Schmidt, and T. Galambos. Girders. ASCE Journal of the Structural Division, Vol. 104, No. 9,
2003. Updating Standard Shape Material Properties Database for pp. 1389–1407.
Design and Reliability. Engineering Journal, American Institute of D’Appolonia, D., E. D’Appolonia, and R. Brissette. 1968. Settlement of
Steel Construction, Vol. 40, No. 1, pp. 1–14. Spread Footings on Sand. ASCE Journal of Soil Mechanics and Foun-
Beeby, A. 1979. The Prediction of Crack Widths in Hardened Concrete. dations Division, Vol. 94, No. 3, pp. 735–762.
The Structural Engineer, Vol. 57A, No. 1, p. 9. Das, B., and B. Sivakugan. 2007. Settlements of Shallow Foundations on
Biggs, J. 1964. Introduction to Structural Dynamics. McGraw-Hill, Granular Soil: An Overview. International Journal of Geotechnical
New York. Engineering, No. 1, pp. 19–29.
Birrcher, D., R. Tuchscherer, M. Huizinga, O. Bayrak, S. Wood, and Detwiler, R. 2008. Protecting concrete from sulfate attack. 2008. Con-
J. Jirsa. 2009. Strength and Serviceability Design of Reinforced Concrete crete News, January. L&M Construction Chemicals, Inc. http://www
Deep Beams. FHWA/TX-09/0-5253-1. Center for Transportation .lmcc.com/concrete_news/0801/protecting_concrete_from_sulfate_
Research at the University of Texas at Austin and Texas Department attack.asp. Accessed Oct. 8, 2014.
of Transportation, Austin. Dexter, R., and J. Fisher. 2000. Fatigue and Fracture. In Bridge Engineer-
Blackman, D., and R. Frosch. 2005. Epoxy-Coated Reinforcement and ing Handbook (Chen and Duan, eds.), CRC Press, New York,
Crack Control. In Serviceability of Concrete: A Symposium Honoring pp. 53.1–53.23.
Dr. Edward G. Nawy, SP-225, American Concrete Institute, Farming- Dexter, R., M. Graeser, W. Saari, C. Pascoe, C. Gardner, and T. Galambos.
ton Hills, Mich., pp. 163–178. 2000. Structural Shape Material Property Survey. Final report. Uni-
Bolukbasi, M., J. Mohammadi, and D. Arditi. 2004. Estimating the versity of Minnesota, Minneapolis.
Future Condition of Highway Bridge Components Using National Dexter, R., and M. Melendrez. 2000. Through-Thickness Properties of
Bridge Inventory Data. ASCE Practice Periodical on Structural Design Column Flanges in Welded Moment Connections. ASCE Journal of
and Construction, Vol. 9, No. 1, pp. 16–25. Structural Engineering, Vol. 126, No. 1, pp. 24–31.
Bridge Bearings. 2006. Indian Railways Institute of Civil Engineering, Dowling, N. 1972. Fatigue Failure Predictions for Complicated Stress–
Pune, India. Strain Histories. Journal of Mechanics, JMLSA, Vol. 7, No. 1, pp. 71–87.
Bridge Design Manual. 2006. Includes 2006 Interim and Errata. Engi- Downing, S., and D. Socie. 1982. Simple Rainflow Counting Algorithms.
neering Division, Division of Highways, West Virginia Department International Journal of Fatigue, Vol. 4, No. 1, pp. 31–40.
of Transportation, Charleston. Duncan, J. 2000. Factors of Safety and Reliability in Geotechnical
Bridge Design Manual. 2007. Office of Structural Engineering, Ohio Engineering. ASCE Journal of Geotechnical and Geoenvironmental
Department of Transportation, Columbus. Engineering, Vol. 126, No. 4, pp. 307–316.
Bridge Inspector’s Reference Manual. 2002. Federal Highway Adminis- DuraCrete: Final Technical Report: General Guidelines for Durability
tration, U.S. Department of Transportation. Design and Redesign. 2000. Project BE95-1347/R17. European Union,
Bridge Joint Association. 2010. Current Practice Sheet. www.bridge Brite EuRam III, CUR, Gouda, Netherlands.
joints.org.uk/Current%20Practice%20Sheet%20on%20Bridge Elias, V., K. Fishman, B. Christopher, and R. Berg. 2009. Corrosion/
%20Joints.pdf. Accessed April 13. Degradation of Soil Reinforcements for Mechanically Stabilized Earth

265
Walls and Reinforced Soil Slopes. FHWA-NHI-09-087. National High- Gergely, P., and L. Lutz. 1968. Maximum Crack Width in Reinforced
way Institute, Federal Highway Administration, Washington, D.C. Concrete Members. In Causes, Mechanisms and Control of Crack-
Ellingwood, B., J. Galambos, J. MacGregor, and C. Cornell. 1980. Devel- ing in Concrete, SP-20, American Concrete Institute, Detroit, Mich.,
opment of a Probability-Based Load Criterion for American National pp. 87–117.
Standard A58. NBS Special Publication 577. National Bureau of Gifford, D., S. Kraemer, J. Wheeler, and A. McKown. 1987. Spread Foot-
Standards, U.S. Department of Commerce. ings for Highway Bridges. FHWA/RD-86-185. Haley and Aldrich,
EN 1990 (Eurocode 0): Basis of Structural Design. 2002. European Com- Cambridge, Mass.
mittee for Standardization, Brussels, Belgium. Goodpasture, D., and W. Goodwin. 1971. Final Report on the Evaluation
EN 1991-1-6 (Eurocode 1): Actions on Structures: Part 1-6: General Actions: of Bridge Vibration As Related to Bridge Deck Performance. University
Actions During Execution. 2005. European Committee for Standard- of Tennessee, Knoxville; Tennessee Department of Transportation,
ization, Brussels, Belgium. Nashville; and Federal Highway Administration, Washington, D.C.
EN 1991-2 (Eurocode 1): Actions on Structures: Part 2: Traffic Loads on Grant, R., J. Christian, and E. Vanmarcke. 1974. Differential Settlement
Bridges. 2003. European Committee for Standardization, Brussels, of Buildings. ASCE Journal of the Geotechnical Engineering Division,
Belgium. Vol. 100, No. 9, pp. 973–991.
EN 1992-2 (Eurocode 2): Design of Concrete Structures: Part 2: Concrete Gross, S., and S. Burns. 2000. Field Performance of Prestressed High Per-
Bridges: Design and Detailing Rules. 2005. European Committee for formance Concrete Bridges in Texas. FHWA/TX-05/9-580/589-2.
Standardization, Brussels, Belgium. Center for Transportation Research at the University of Texas and
EN 1998-2 (Eurocode 8): Design of Structures for Earthquake Resistance: Texas Department of Transportation, Austin.
Part 2: Bridges. 2005. European Committee for Standardization, Hansell, W., T. Galambos, M. Ravindra, and I. Viest. 1978. Composite
Brussels, Belgium. Beam Criteria in LRFD. ASCE Journal of the Structural Division,
Evaluation of Various Types of Bridge Deck Joints. 2006. FHWA-AZ- Vol. 104, No. 9, pp. 1409–1426.
06-510. Baker Engineering and Energy; Arizona Department of Hanson, J. M., K. T. Burton, and E. Hognestad. 1968. Fatigue Tests of
Transportation, Phoenix; and Federal Highway Administration, Reinforcing Bars: Effect of Deformation Pattern. Journal of the Port-
U.S. Department of Transportation. land Cement Association Research and Development Laboratories,
Vol. 10, No. 3, pp. 2–13.
Fédération Internationale du Béton. 2006. Model Code for Service Life
Hanson, J., C. Hulsbos, and D. VanHorn. 1970. Fatigue Tests of Pre-
Design. fib Bulletin 34, p. 47.
stressed Concrete I-Beams. ASCE Journal of the Structural Division,
Fédération Internationale du Béton. 2010. fib Model Code for Concrete
Vol. 96, No. 11, pp. 2443–2464.
Structures. Ernst and Sohn, Berlin.
Hatami, A., and A. Morcous. 2011. Developing Deterioration Models
Fisher, J. 1977. Bridge Fatigue Guide: Design and Details. American Insti-
for Nebraska Bridges. Project No. SPR-P1(11) M302. University of
tute of Steel Construction, Chicago, Ill.
Nebraska, Lincoln, and Nebraska Department of Roads, Omaha.
Fisher, J., P. Albrecht, B. Yen, D. Klingerman, and B. McNamee. 1974.
Helgason, T., J. Hanson, N. Somes, W. Corley, and E. Hognestad. 1976.
NCHRP Report 147: Fatigue Strength of Steel Beams with Trans-
NCHRP Report 164: Fatigue Strength of High-Yield Reinforcing Bars.
verse Stiffeners and Attachments. TRB, National Research Council,
TRB, National Research Council, Washington, D.C.
Washington, D.C.
Helland, S. 2013. Design for Service Life: Implementation of fib Model
Fisher, J., K. Frank, M. Hirt, and B. McNamee. 1970. NCHRP Report 102: Code 2010, Rules in the Operational Code ISO 16204. Structural
Effect of Weldments on the Fatigue Strength of Steel Beams. HRB, Concrete, Journal of the fib, Vol. 14, No. 1, pp. 10–18.
National Research Council, Washington, D.C. Hilsdorf, H., and C. Kesler. 1966. Fatigue Strength of Concrete Under
Fisher, J., D. Mertz, and A. Zhong. 1983. NCHRP Report 267: Steel Bridge Varying Flexural Stresses. Journal Proceedings, American Concrete
Members Under Variable-Amplitude Long-Life Fatigue Loading. TRB, Institute, Vol. 63, No. 10, pp. 1059–1076.
National Research Council, Washington, D.C. Hough, B. 1959. Compressibility as the Basis for Soil Bearing Value. ASCE
Fisher, J. W., and I. M. Viest. 1961. Special Report 66: Fatigue Tests of Journal of the Soil Mechanics and Foundations Division, Vol. 85,
Bridge Materials of the AASHO Road Test. HRB, National Research No. 4, pp. 11–40.
Council, Washington, D.C. Imbsen, R. A., D. E. Vandershaf, R. A. Schamber, and R. V. Nutt. 1985.
Fishman, K., and J. Withiam. 2011. NCHRP Report 675: LRFD Metal NCHRP Report 276: Thermal Effects in Concrete Bridge Superstructures.
Loss and Service-Life Strength Reduction Factors for Metal-Reinforced TRB, National Research Council, Washington, D.C.
Systems. TRB, National Research Council, Washington, D.C. International Code Council. 2007. 2006 Seattle Building Code. ICC,
Fountain, R., and C. Thunman. 1987. Deflection Criteria for Steel High- Washington, D.C.
way Bridges. Proc., AISC National Engineering Conference, American ISO 2394: General Principles on Reliability for Structures, 3rd ed. 1998.
Institute of Steel Construction, New Orleans, La., pp. 20-1–20-12. International Organization for Standardization, Geneva.
Freyermuth, C. 2009. Service Life and Sustainability of Concrete Bridges. Itoh, Y., and T. Kitagawa. 2001. Bridge Lifecycle Analysis and Dura
ASPIRE, Fall 2009, pp. 12–15. bility Evaluation Using Accelerated Exposure Test. Presented at
Frosch, R. 1999. Another Look at Cracking and Crack Control in U.S.–Japan Seminar on Advanced Stability and Seismicity Concept
Reinforced Concrete. Structural Journal, American Concrete Insti- for Performance-Based Design of Steel and Composite Structures,
tute, Vol. 96, No. 3, pp. 437–442. Kyoto, Japan.
Frosch, R. 2001. Flexural Crack Control in Reinforced Concrete. In Design Jiang, Y., and K. Sinha. 1989. The Development of Optimal Strategies for
and Construction Practices to Mitigate Cracking, SP-204, American Maintenance, Rehabilitation, and Replacement of Highway Bridges:
Concrete Institute, Farmington Hills, Mich., pp. 135–154. Volume 6: Performance Analysis and Optimization. FHWA/IN/JHRP-
Frosch, R. 2002. Modeling and Control of Side Face Beam Cracking. 89/13. Purdue University, West Lafayette, Ind., and Indiana Depart-
Structural Journal, American Concrete Institute, Vol. 99, No. 3, ment of Transportation, Indianapolis.
pp. 376–385. Kaar, P., and A. Mattock. 1963. High-Strength Bars As Concrete
Geotechnical Design Manual. 2012. M 46-03.07. Washington State Depart- Reinforcement, Part 4: Control of Cracking. Portland Cement Asso-
ment of Transportation, Olympia. ciation Development Department Bulletin D59, pp. 15–38.

266
Kaczinski, M. 2008. New Developments in Expansion Joints, Bearings, Miner, M. 1945. Cumulative Damage in Fatigue. Journal of Applied
Cable Corrosion Protection, and Grid Deck Systems. Presented at Mechanics, American Society of Mechanical Engineers, Vol. 12,
2008 IBTTA Fall Maintenance Meeting, New Orleans, La. pp. A159–A164.
Keating, P. B., and J. W. Fisher. 1986. NCHRP Report 286: Evaluation of Mirza, S. 2007. Design of Durable and Sustainable Concrete Bridges.
Fatigue Tests and Design Criteria on Welded Details. TRB, National Proc., CBM-CI International Workshop, NED University of Engineer-
Research Council, Washington, D.C. ing and Technology, Karachi, Pakistan, pp. 333–344.
Klaiber, F., T. Wipf, and F. Russo. 2004. NCHRP Synthesis 327: Cost- Mirza, S., and J. MacGregor. 1979a. Variability of Mechanical Properties
Effective Practices for Off-System and Local Interest Bridges. TRB, of Reinforcing Bars. ASCE Journal of the Structural Division, Vol. 105,
National Research Council, Washington, D.C. No. 5, pp. 921–937.
Krauss, P., and E. Rogalla. 1996. NCHRP Report 380: Transverse Cracking Mirza, S., and J. MacGregor. 1979b. Variations in Dimensions of
in Newly Constructed Bridge Decks. TRB, National Research Council, Reinforced Concrete Members. ASCE Journal of the Structural Divi-
Washington, D.C. sion, Vol. 105, No. 4, pp. 751–766.
Kulicki, J., Z. Prucz, C. Clancy, D. Mertz, and A. Nowak. 2007. Updating Miska. 2013. Compression Seals: Neoprene. www.miska.com.au/ebm3-
the Calibration Report for AASHTO LRFD Code. Report on NCHRP doc/130914/comsealdatasheet.pdf. Accessed May 9.
Project 20-7/186. Transportation Research Board of the National Mlynarski, M., W. Wassef, and A. Nowak. 2011. NCHRP Report 700: A
Academies, Washington, D.C. Comparison of AASHTO Bridge Load Rating Methods. TRB, National
Kulicki, J., W. Wassef, D. Mertz, A. Nowak, N. Samtani, and H. Nassif. 2013. Research Council, Washington, D.C.
Service Load Design for 100-Year Life. SHRP 2 Project R19B. TRB, Moulton, L., H. Ganga Rao, and G. Halvorsen. 1985. Tolerable Movement
National Research Council, Washington, D.C. (Draft.) Criteria for Highway Bridges. FHWA/RD-85-107. West Virginia Uni-
Kupfer, H., and K. Gerstle. 1973. Behavior of Concrete Under Biaxial versity, Morgantown.
Stress. ASCE Journal of the Engineering Mechanics Division, Vol. 99, Murray, T., D. Allen, and E. Ungar. 2003. Design Guide 11: Floor Vibrations
No. 4, pp. 853–866. Due to Human Activity. American Institute of Steel Construction,
Langley, W. 1999. Concrete Mix Proportioning to Meet Durability Chicago, Ill.
Concerns for Confederation Bridge. In High-Performance Concrete: Musso, A., and P. Provenzano. 2003. Discussion of Predicting Settle-
ment of Shallow Foundations Using Neural Networks. ASCE Jour-
Performance and Quality of Concrete Structures, SP-186, American
nal of Geotechnical and Geoenvironmental Engineering, Vol. 129,
Concrete Institute, Detroit, Mich., pp. 129–148.
No. 12, pp. 1172–1175.
Lash, S. 1969. Can High-Strength Reinforcement Be Used in Highway
Nawy, E., and P. Huang. 1977. Crack and Deflection Control of Pre
Bridges? In First International Symposium on Concrete Bridge
tensioned Prestressed Beams. Journal, Precast/Prestressed Concrete
Design, SP-23, American Concrete Institute, Detroit, Mich.,
Institute, Vol. 23, No. 3, pp. 30–43.
pp. 283–300.
NCHRP Research Results Digest 197: Fatigue Behavior of Welded and
Lind, N., and A. Nowak. 1978. Calculation of Load and Performance Factors.
Mechanical Splices in Reinforcing Steel. 1994. TRB, National Research
Ontario Ministry of Transportation and Communications, Toronto,
Council, Washington, D.C.
Canada.
Nevels, J., and D. Hixon. 1973. A Study to Determine the Causes of
LRFD Bridge Design Manual. 2006. Louisiana Department of Transpor-
Bridge Deck Deterioration. Research and Development Division,
tation and Development, Baton Rouge.
Oklahoma Department of Highways, Oklahoma City.
MacGregor, J., I. Jhamb, and N. Nuttall. 1971. Fatigue Strength of Hot- Nowak, A. 1999. NCHRP Report 368: Calibration of LRFD Bridge Design
Rolled Deformed Reinforcing Bars. Journal Proceedings, American Code. TRB, National Research Council, Washington, D.C.
Concrete Institute, Vol. 68, No. 3, pp. 169–179. Nowak, A., and K. Collins. 2013. Reliability of Structures. McGraw-Hill,
Main Roads Specification MRS81: Bridge Bearings. 2012. MRS81. New York.
Queensland Government, Department of Main Roads, Queensland, Nowak, A., E. Szeliga, and M. Szerszen. 2008. Reliability-Based Calibra-
Australia. tion for Structural Concrete, Phase 3. Portland Cement Association
Manual for Bridge Evaluation. 2008. American Association of State Research and Development Serial No. 2849, pp. 1–110.
Highway and Transportation Officials, Washington, D.C. Okeil, A. 2006. Allowable Tensile Stress for Webs of Prestressed Segmen-
Maryland Manual on Uniform Traffic Control Devices for Streets and tal Concrete Bridges. Structural Journal, American Concrete Institute,
Highways. 2006. Includes Revision 1, July 2009. Maryland State Vol. 103, No. 4, pp. 488–495.
Highway Administration, Baltimore. Oluokun, F. 1991. Prediction of Concrete Tensile Strength from Its
Matsuishi, M., and T. Endo. 1968. Fatigue of Metals Subjected to Varying Compressive Strength: Evaluation of Existing Relations for Normal-
Stress. Japan Society of Mechanical Engineers, Fukuoka, Japan. Weight Concrete. Materials Journal, American Concrete Institute,
Maurer Söhne. 2011. Maurer MSM Sliding Bearings. http://www.maurer Vol. 88, No. 3, pp. 302–309.
.co.uk/doc/MSM_Prospekt_1208_engl.pdf. Accessed Nov. 25. Ontario Highway Bridge Design Code. 1979. Ontario Ministry of Trans-
Maurer Söhne. 2013. Head Made of Stainless Steel: Expansion Joints portation, Toronto, Canada.
with Hybrid Profile. http://www.maurer-soehne.com/files/bauwerk Ontario Highway Bridge Design Code. 1991. Ontario Ministry of Trans-
schutzsysteme/pdf/de/news/MAU_PR_Hybrid_Profile_0707_engl portation, Toronto, Canada.
.pdf. Accessed Nov. 25. Ontario Traffic Manual: Book 5: Regulatory Signs. 2000. Ontario Ministry
Merriam-Webster. 2010. Deterioration. http://www.merriam-webster of Transportation, Toronto, Canada.
.com/dictionary/deterioration. Accessed Nov. 25. Ople, F. S., and C. L. Hulsbos. 1966. Probable Fatigue Life of Plain Con-
Merriam-Webster. 2010. Maintenance. http://www.merriam-webster crete with Stress Gradient. Journal Proceedings, American Concrete
.com/dictionary/maintenance. Accessed Nov. 25. Institute, Vol. 63, No. 1, pp. 59–82.
Merriam-Webster. 2010. Serviceable. http://www.merriam-webster Outokumpu. 2013. Sustainable Bridges: 300 Year Design Life for Sec-
.com/dictionary/serviceable. Accessed Nov. 25. ond Gateway Bridge. http://www.outokumpu.com/SiteCollection

267
Documents/SiteCollectionDocumentsSuistanable%20Bridges_ Schopen, D. 2010. Analyzing and Designing for Substructure Movement

44664.pdf. Accessed Nov. 25. in Highway Bridges: An LRFD Approach. Master’s thesis. University
Palmgren, A. 1924. Die Lebensdauer von Kugellagern [Life Length of of Delaware, Newark.
Roller Bearings (in German)]. Zeitschrift des Vereines Deutscher Shahin, M., H. Maier, and M. Jaksa. 2002. Predicting Settlement of
Ingenieure (VDI Zeitschrift), Vol. 68, No. 14, pp. 339–341. Shallow Foundations Using Neural Networks. ASCE Journal of
Peck, R., and A. Bazaraa. 1969. Discussion of Settlement of Spread Foot- Geotechnical and Geoenvironmental Engineering, Vol. 128, No. 9,
ings on Sand. ASCE Journal of the Soil Mechanics and Foundations pp. 785–793.
Division, Vol. 95, No. 3, pp. 900–916. Siriaksorn, A., and A. Naaman. 1980. Reliability of Partially Prestressed
Pelphrey, J., C. Higgins, B. Sivakumar, R. Groff, B. Hartman, J. Charbon- Beams at Serviceability Limit States. University of Illinois at Chicago
neau, J. Rooper, and B. Johnson. 2008. State-Specific LRFR Live Circle, Chicago, Ill.
Load Factors Using Weigh-in-Motion Data. Journal of Bridge Engi- Sivakugan, N., and K. Johnson. 2002. Probabilistic Design Chart for
neering, July/August, pp. 339–350. Settlements of Shallow Foundations in Granular Soils. Australian
Pfister, J. F., and E. Hognestad. 1964. High-Strength Bars As Concrete Civil Engineering Transactions, No. 43, pp. 19–24.
Reinforcement, Part 6: Fatigue Tests. Journal of the Portland Cement Sivakugan, N., and K. Johnson. 2004. Settlement Predictions in Gran-
Association Research and Development Laboratories, Vol. 6, No. 1, ular Soils: A Probabilistic Approach. Geotechnique, Vol. 54, No. 7,
pp. 65–84. pp. 499–502.
Portland Cement Association. 1970. Durability of Concrete Bridge Decks: Sivakumar, B., M. Ghosn, and F. Moses. 2011. NCHRP Report 683:
A Cooperative Study, Final Report. Portland Cement Association Protocols for Collecting and Using Traffic Data in Bridge Design.
(PCA), Skokie, Ill. TRB, National Research Council, Washington, D.C.
Purvis, R. 2003. NCHRP Synthesis 319: Bridge Deck Joint Performance: A Skempton, A., and D. MacDonald. 1956. Allowable Settlement of
Synthesis of Highway Performance. TRB, National Research Council, Buildings. Proceedings, Part III, Institution of Civil Engineers, No. 5,
Washington, D.C. pp. 727–768.
Rakoczy, P. 2011. WIM-Based Load Models for Bridge Serviceability Limit Standard Specifications for Highway Bridges, 6th ed. 1953. American
States. PhD dissertation. University of Nebraska, Lincoln. Association of State Highway Officials, Washington, D.C.
Ramey, G., and R. Wright. 1994. Assessing and Enhancing the Durability/
Standard Specifications for Highway Bridges, 11th ed. 1975. Includes
Longevity Performances of Highway Bridges. Final report. HRC
1974 and 1975 Interims. American Association of State Highway
Research Project 2-13506. Highway Research Center, Auburn Uni-
Officials, Washington, D.C.
versity, Auburn, Ala.
Standard Specifications for Highway Bridges, 15th ed. 1992. American
Ravindra, M., and T. Galambos. 1978. Load and Resistance Factor Design
Association of State Highway and Transportation Officials, Wash-
for Steel. ASCE Journal of the Structural Division, Vol. 104, No. 9,
ington, D.C.
pp. 1337–1353.
Standard Specifications for Highway Bridges, 17th ed. 2002. American
Roeder, C., K. Barth, and A. Bergman. 2002. NCHRP Web Document 46:
Improved Live Load Deflection Criteria for Steel Bridges. TRB, National
ington, D.C.
Research Council, Washington, D.C.
Rolfe, S., and J. Barsom. 1977. Fracture and Fatigue Control in Struc- Steel Bridge Bearing Selection and Design Guide. 1996. In Highway Struc-
tures: Applications of Fracture Mechanics. Prentice-Hall, Englewood tures Design Handbook. American Iron and Steel Institute, National
Cliffs, N.J. Steel Bridge Alliance, Washington, D.C., p. A-3.
Rostam, S. 2005. Design and Construction of Segmental Concrete Bridges Steel Construction Manual, 14th ed. 2011. American Institute of Steel
for Service Life of 100 to 150 Years. American Segmental Bridge Construction, Chicago, Ill.
Institute, Buda, Tex. Steel Design Guide 3: Serviceability Design Considerations for Steel
Samtani, N., A. Nowatzki, and D. Mertz. 2010. Selection of Spread Footings Bridges, 2nd ed. 2003. American Institute of Steel Construction,
on Soils to Support Highway Bridge Structures. FHWA RC/TD-10-001. Chicago, Ill.
Federal Highway Administration Resource Center, Matteson, Ill. Structures Manual. 2013. Florida Department of Transportation,
Samtani, N., and E. Nowatzki. 2006. Soils and Foundations: Volumes I Tallahassee.
and II. FHWA-NHI-06-088 and FHWA-NHI-06-089. Federal High- Stukhart, G., R. James, A. Garcia-Diaz, R. Bligh, J. Sobanjo, and
way Administration, U.S. Department of Transportation. W. McFarland. 1991. Study for a Comprehensive Bridge Management
Sargand, S., and T. Masada. 2006. Further Use of Spread Footing Founda- System for Texas. FHWA/TX-90/1212-1F. Texas Transportation Insti-
tions for Highway Bridges. State Job No. 14747(0), FHWA-OH-2006/8. tute, College Station, and Texas State Department of Highways and
Ohio Research Institute for Transportation and the Environment, Public Transportation, Arlington.
Athens; Ohio Department of Transportation, Columbus; and Office Tabatabai, H., A. Ghorbanpoor, and A. Turnquist-Nass. 2004. Reha-
of Research and Development, Federal Highway Administration, U.S. bilitation Techniques for Concrete Bridges. Project No. 0092-01-06.
Department of Transportation. University of Wisconsin, Milwaukee, and Wisconsin Department
Sargand, S., T. Masada, and R. Engle. 1999. Spread Footing Foundation of Transportation, Madison.
for Highway Bridge Applications. ASCE Journal of Geotechnical and Tachau, H. 1971. Discussion of Fatigue Tests on Prestressed Concrete
Geoenvironmental Engineering, Vol. 125, No. 5, pp. 373–382. I-Beams. ASCE Journal of the Structural Division, Vol. 97, No. 9,
Schilling, C., K. Klippstein, J. Barsom, and G. Blake. 1977. NCHRP pp. 2429–2431.
Report 188: Fatigue of Welded Steel Bridge Members Under Variable- Tadros, M., N. Al-Omaishi, S. Seguirant, and J. Gallt. 2003. NCHRP
Amplitude Loadings. TRB, National Research Council, Washing- Report 496: Prestress Losses in Pretensioned High-Strength Concrete
ton, D.C. Bridge Girders. TRB, National Research Council, Washington, D.C.
Schmertmann, J., P. Brown, and J. Hartman. 1978. Improved Strain Tan, C., and J. Duncan. 1991. Settlement of Footings on Sands: Accuracy
Influence Factor Diagrams. ASCE Journal of the Geotechnical Engi- and Reliability. Proc., Geotechnical Engineering Congress 1991, ASCE
neering Division, Vol. 104, No. 8, pp. 1131–1135. Geotechnical Special Publication No. 27, Vol. 1, pp. 446–455.

268
Technoslide. 2013. Bearings for Bridges and Structures. http://www Warner, R., and C. Hulsbos. 1966. Probable Fatigue Life of Prestressed
.technoslide.com/pdf/elastomeric/elastomeric-plain-sliding-bearings- Concrete Beams. Journal, Precast/Prestressed Concrete Institute,
for-bridges-structures-brochure.pdf. Accessed Nov. 25. Vol. 11, No. 2, pp. 16–39.
Thompson, P., and R. Shepard. 2000. AASHTO Commonly Recognized Wikipedia. 2010. Deterioration. http://en.wikipedia.org/wiki/
Bridge Elements: Successful Applications and Lessons Learned. National Deterioration. Accessed Nov. 25.
Workshop on Commonly Recognized Measures for Maintenance, Wikipedia. 2010. Serviceability (Structure). http://en.wikipedia.org/
American Association of State Highway and Transportation Offi- wiki/Serviceability_(structure). Accessed Nov. 25.
cials, Washington, D.C. Wiktionary. 2010. Maintenance. http://en.wiktionary.org/wiki/
Vincent, G. 1969. Tentative Criteria for Load Factor Design of Steel High- maintenance. Accessed Nov. 25.
way Bridges. Bulletin for the American Iron and Steel Institute, No. 15. Wright, R., and W. Walker. 1972. Vibration and Deflection of Steel Bridges.
Wahls, H. 1983. NCHRP Synthesis of Highway Practice 107: Shallow Engineering Journal, American Institute of Steel Construction, Vol. 9,
Foundations for Highway Structures. TRB, National Research Coun- No. 1, pp. 20–31.
cil, Washington, D.C. Yura, J., T. Galambos, and M. Ravindra. 1978. The Bending Resistance
Walker, W., and R. Wright. 1971. Criteria for the Deflection of Steel Bridges. of Steel Beams. ASCE Journal of the Structural Division, Vol. 104,
Bulletin for the American Iron and Steel Institute, No. 19, pp. 1–75. No. 9, pp. 1355–1370.
Walraven, J., and A. Bicaj. 2011. The 2010 fib Model Code for Concrete Zhang, L., and A. Ng. 2005. Probabilistic Limiting Tolerable Displace-
Structures: A New Approach to Structural Engineering. Structural ments for Serviceability Limit State Design of Foundations. Geotech-
Concrete, Journal of the fib, Vol. 12, No. 2, pp. 139–147. nique, Vol. 55, No. 2, pp. 151–161.

A p p endi x A
SLS Requirements in the Eurocode
A.1 Introduction which comprise the full text of each Eurocode (including any
annexes) and may be followed by a National Annex.
A.1.1 General Information The National Annex only contains information on those
The Structural Eurocode program provides comprehensive parameters which are left open in the Eurocode for national
information for the structural design and verification of build- choice, (known as Nationally Determined Parameters). They
ings and civil engineering works (including geotechnical are to be used for the design of buildings and civil engineer-
aspects). The program comprises the following standards, each ing works to be constructed in the country concerned and are
one consisting of a number of parts. [Often only a limited usually one or more of the following:
number of parts of each standard may be relevant to bridge
structures.] • Values and/or classes where alternatives are given in the
Eurocode;
EN 1990 Eurocode 0: Basis of structural design • Values to be used where a symbol only is given in the
EN 1991 Eurocode 1: Actions on structures Eurocode.
EN 1992 Eurocode 2: Design of concrete structures • Country specific data (geographical, climatic, etc.) e.g.
EN 1993 Eurocode 3: Design of steel structures snow map.
EN 1994 Eurocode 4: Design of composite steel and concrete • The procedure to be used where alternative procedures are
structures given in the Eurocode.
EN 1995 Eurocode 5: Design of timber structures
EN 1996 Eurocode 6: Design of masonry structures The National Annex may also contain the following:
EN 1997 Eurocode 7: Geotechnical design
EN 1998 Eurocode 8: Design of structures for earthquake • Decisions on the application of informative annexes, and
resistance • References to non-contradictory complementary informa-
EN 1999 Eurocode 9: Design of aluminum structures tion to help the user apply the Eurocode.
Following is a description of the serviceability limit state This summary does not include any numeric values pre-
(SLS) requirements in sections relevant to bridges. sented in any National Annex.
The following sections address some of the Structural Euro-
codes in turn and summarize the relevant articles relating to
A.1.2 Structural Eurocodes
the serviceability limit state used in bridge design.
The Structural Eurocode standards provide common structural
design rules for everyday use for the design of whole structures
A.2 EN 1990 Eurocode 0:
and component products of both a traditional and an innova-
Basis of Structural Design
tive nature. Unusual forms of construction or design conditions
are not specifically covered and additional expert consideration Eurocode 0 (Basis of structural design) is the lead document
is required by the designer in such cases. in the Eurocode suite. It describes the principles and require-
The Eurocodes are being implemented by each member ments for safety, serviceability, and durability of structures.
country of the European Union through National Standards It is based on the limit state concept used in conjunction
269

270
with a partial factor method. It provides the basis and gen- NOTE: There may be cases when these assumptions need
eral principles for the structural design and verification of to be supplemented.
buildings and civil engineering works (including geotechni- It should be noted that clauses are listed and enumerated
cal aspects). within each article of the Eurocodes and that distinction is
EN 1990:2002 should be used in conjunction with all the made between clauses that present principles and those that
other Eurocodes (EN 1991 to EN 1999) for design. present Application Rules. This distinction is preserved in the
NOTE: For the design of special construction works (e.g. summaries given in this report.
nuclear installations, dams), other provisions than those in The Principles comprise
EN 1990 to EN 1999 might be necessary.
EN 1990 also gives guidelines for the aspects of structural • General statements and definitions for which there is no
reliability relating to safety, serviceability, and durability: alternative, as well as
• Requirements and analytical models for which no alterna-
• For design cases not covered by EN 1991 to EN 1999 (other tive is permitted unless specifically stated.
actions, structures not treated, other materials);
• To serve as a reference document for other European Com- The Principles are identified by the letter P following the
mittee for Standardization Technical Committees (CEN/TCs) paragraph number. [e.g. (2)P]
concerning structural matters. The Application Rules [identified by a number in brackets,
e.g. (2)] are generally recognized rules which comply with the
EN 1990 is also applicable as a guidance document for the principles and satisfy their requirements.
design of structures where other materials or other actions It is permissible to use alternative design rules different
outside the scope of EN 1991 to EN 1999 are involved. from the Application Rules given in EN 1990 for works, pro-
EN 1990 is applicable for the structural appraisal of exist- vided that it is shown that the alternative rules accord with
ing construction, in developing the design of repairs and the relevant principles and are at least equivalent with regard
to the structural safety, serviceability, and durability which
alterations, or in assessing changes of use.
would be expected when using the Eurocodes.
NOTE: Additional or amended provisions might be neces-
The clauses relating to serviceability limit state design pre-
sary where appropriate.
sented in Eurocode 0 are summarized in Table A.1.
EN 1990 is intended for use by
• Committees drafting standards for structural design and A.3 EN 1991 Eurocode 1:
related product, testing, and execution standards; Actions on Structures
• Clients (e.g., for the formulation of their specific require-
Eurocode 1 (Actions on structures) provides information on all
ments on reliability levels and durability); actions that should normally be considered in the design of
• Designers and constructors; and
buildings and civil engineering works. It is in four main parts.
• Relevant authorities. The first part is divided into seven sub-parts which cover densi-
ties, self-weight, and imposed loads; actions due to fire; snow;
The general assumptions of EN 1990 are as follows: wind; thermal actions; loads during execution; and accidental
actions. The remaining three parts cover traffic loads on bridges,
• The choice of the structural system and the design of the actions by cranes and machinery, and actions for silos and tanks.
structure are made by appropriately qualified and expe- The second part (EN 1991-2:2003) concerns the design of
rienced personnel. bridges. Sections from this standard relating to the service-
• Execution is carried out by personnel having the appropriate ability limit state are summarized in Table A.2.
skill and experience. For the design of bridges, EN 1991-2 defines imposed loads
• Adequate supervision and quality control are provided (models and representative values) associated with road traf-
during execution of the work (i.e. in design offices, facto- fic, pedestrian actions, and rail traffic which include, when
ries, plants, and on site). relevant, dynamic effects and centrifugal, braking, and accel-
• The construction materials and products are used as spec- eration actions and actions for accidental design situations.
ified in EN 1990 or in EN 1991 to EN 1999 or in the rele- For the design of new bridges, EN 1991-2 is intended to be
vant execution standards, or reference material, or product used, for direct application, together with Eurocodes EN 1990
specifications. to EN 1999. The bases for combinations of traffic loads with
• The structure will be adequately maintained; and non-traffic loads are given in EN 1990, A2.
• The structure will be used in accordance with the design A summary of clauses relating to loads and actions in
assumptions. Eurocode EN 1991-2 is presented in Table A.2.

271
Table A.1. Summary of Clauses Relating to Serviceability Limit State Design in Eurocode 0
Eurocode Article Basic Provision Discussion
Eurocode 0 (1)P The limit states that concern

– the functioning of the structure or structural members under normal use;
3.4 Serviceability limit
– the comfort of people;
states
– the appearance of the construction works, shall be classified as serviceability limit states.
NOTE 1: In the context of serviceability, the term “appearance” is concerned with such
criteria as high deflection and extensive cracking, rather than aesthetics.
NOTE 2: Usually the serviceability requirements are specific to each individual project.
(2)P A distinction shall be made between reversible and irreversible serviceability limit states.
NOTE: ‘Reversible’ = where no consequences of actions exceeding the specified service require-
ment will remain when the actions are removed.
‘Irreversible’ = where some consequences of actions will remain when the actions are removed.
(3) The verification of serviceability limit states should be based on criteria concerning the follow-
ing aspects:
a) deformations that affect
– the appearance,
– the comfort of users, or
– the functioning of the structure (including the functioning of machines or services), or
that cause damage to finishes or non-structural members;
b) vibrations
– that cause discomfort to people, or
– that limit the functional effectiveness of the structure;
c) damage that is likely to adversely affect
– the appearance,
– the durability, or
– the functioning of the structure.
NOTE: Additional provisions related to serviceability criteria are given in the relevant EN 1992
to EN 1999.
Eurocode 0 (1)P It shall be verified that
6.5.1 Verifications Ed ≤ Cd (6.13)
where:
Cd is the limiting design value of the relevant serviceability criterion.
Ed is the design value of the effects of actions specified in the serviceability criterion, determined
on the basis of the relevant combination.
Eurocode 0 (1) The deformations to be taken into account in relation to serviceability requirements should be
as detailed in the relevant Annex A according to the type of construction works, or agreed
6.5.2 Serviceability criteria
with the client or the National authority.
NOTE: For other specific serviceability criteria such as crack width, stress or strain limitation,
slip resistance, see EN 1991 to EN 1999.
Eurocode 0 (1) The combinations of actions to be taken into account in the relevant design situations should
be appropriate for the serviceability requirements and performance criteria being verified.
6.5.3 Combination of
(2) The combinations of actions for serviceability limit states are defined symbolically (see also 6.5.4):
actions
NOTE: It is assumed, in these expressions, that all partial factors are equal to 1. See Annex A
and EN 1991 to EN 1999.
a) Characteristic combination: (equation given at 6.14a)
NOTE: The characteristic combination is normally used for irreversible limit states.
b) Frequent combination: (equation given at 6.15a)
NOTE: The frequent combination is normally used for reversible limit states.
c) Quasi-permanent combination: (equation given at 6.16a)
NOTE: The quasi-permanent combination is normally used for long-term effects and the
appearance of the structure.

272
Table A.1. Summary of Clauses Relating to Serviceability Limit State Design in Eurocode 0 (continued)
(3) For the representative value of the prestressing action (i.e. Pk or Pm), reference
should be made to the relevant design Eurocode for the type of prestress under
consideration.
(4)P Effects of actions due to imposed deformations shall be considered where relevant.
NOTE: In some cases expressions (6.14) to (6.16) require modification. Detailed rules are
given in the relevant parts of EN 1991 to EN 1999.
Eurocode 0 (1) For serviceability limit states the partial factors gM for the properties of materials should be
taken as 1.0 except if differently specified in EN 1992 to EN 1999.
6.5.4 Partial factors for
materials
Eurocode 0 (1) This Annex A2 to EN 1990 gives rules and methods for establishing combinations of
Annex A2 actions for serviceability and ultimate limit state verifications (except fatigue verifications)
with the recommended design values of permanent, variable, and accidental actions
A2.1 Field of application
and y factors (applied to actions) to be used in the design of road bridges, footbridges,
and railway bridges. It also applies to actions during execution. Methods and rules
for verifications relating to some material-independent serviceability limit states are
also given.
NOTE 1: Symbols, notations, Load Models, and groups of loads are those used or defined in
the relevant section of EN 1991-2.
NOTE 2: Symbols, notations, and models of construction loads are those defined in
EN 1991-1-6.
NOTE 3: Guidance may be given in the National Annex with regard to the use of Table 2.1
(design working life—for UK bridges this is normally 120 years).
NOTE 4: Most of the combination rules defined in clauses A2.2.2 to A2.2.5 are simplifications
intended to avoid needlessly complicated calculations. They may be changed in the National
Annex or for the individual project as described in A2.2.1 to A2.2.5.
NOTE 5: This Annex A2 to EN 1990 does not include rules for the determination of actions on
structural bearings (forces and moments) and associated movements of bearings or give rules
for the analysis of bridges involving ground-structure interaction that may depend on move-
ments or deformations of structural bearings.
(2) The rules given in this Annex A2 to EN 1990 may not be sufficient for
– bridges that are not covered by EN 1991-2 (e.g. bridges under an airport runway,
mechanically-moveable bridges, roofed bridges, bridges carrying water),
– bridges carrying both road and rail traffic, and
– other civil engineering structures carrying traffic loads (e.g. backfill behind a retaining wall).
Eurocode 0 (1) Effects of actions that cannot occur simultaneously due to physical or functional reasons
Annex A2 need not be considered together in combinations of actions.
(2) Combinations involving actions which are outside the scope of EN 1991 (e.g. due to mining
A2.2 Combination of
subsidence, particular wind effects, water, floating debris, flooding, mud slides, avalanches, fire,
actions
and ice pressure) should be defined in accordance with EN 1990, 1.1(3).
A2.2.1 General
NOTE 1: Combinations involving actions that are outside the scope of EN 1991 may be
defined either in the National Annex or for the individual project.
NOTE 2: For seismic actions, see EN 1998.
NOTE 3: For water actions exerted by currents and debris effects, see also EN 1991-1-6.
(4) The combinations of actions given in expressions 6.14a to 6.16b should be used when
verifying serviceability limit states. Additional rules are given in A2.4 for verifications
regarding deformations and vibrations.

273
Eurocode 0 (1) The infrequent values of variable actions may be used for certain serviceability limit states of
Annex A2 concrete bridges.
A2.2 Combination of NOTE: The National Annex may refer to the infrequent combination of actions.
actions
(6) Wind actions and thermal actions need not be taken into account simultaneously unless
A2.2.2 Combination rules
otherwise specified for local climatic conditions.
for road bridges
NOTE: Depending upon the local climatic conditions, a different simultaneity rule for wind
and thermal actions may be defined either in the National Annex or for the individual project.
Eurocode 0 (2) The serviceability criteria should be defined in relation to the serviceability requirements in
Annex A2 accordance with 3.4 and EN 1992 to EN 1999. Deformations should be calculated in accor-
dance with EN 1991 to EN 1999 by using the appropriate combinations of actions according
A2.4 Serviceability and
to expressions (6.14a) to (6.16b) (see Table A2.6), taking into account the serviceability
other specific limit states
requirements and the distinction between reversible and irreversible limit states.
A2.4.1 General
NOTE: Serviceability requirements and criteria may be defined as appropriate in the National
Annex or for the individual project.
Eurocode 0 (1) Where relevant, requirements and criteria should be defined for road bridges concerning
Annex A2 – uplift of the bridge deck at supports, and
– damage to structural bearings.
A2.4.2 Serviceability criteria
regarding deformation NOTE: Uplift at the end of a deck can jeopardize traffic safety and damage structural and
and vibration for road non-structural elements. Uplift may be avoided by using a higher safety level than usually
bridges accepted for serviceability limit states.
(2) Serviceability limit states during execution should be defined in accordance with EN 1990 to
EN 1999.
(3) Requirements and criteria should be defined for road bridges concerning deformations and
vibrations, where relevant.
NOTE 1: The verification of serviceability limit states concerning deformation and vibration
needs to be considered only in exceptional cases for road bridges. The frequent combination
of actions is recommended for the assessment of deformation.
NOTE 2: Vibrations of road bridges may have various origins, in particular traffic actions and
wind actions. For vibrations due to wind actions, see EN 1991-1-4. For vibrations due to traf-
fic actions, comfort criteria may have to be considered. Fatigue may also have to be taken
into account.
Eurocode 0 (1) The comfort criteria should be defined in terms of maximum acceptable acceleration of any
Annex A2 part of the deck.
A2.4.3.2 Pedestrian comfort NOTE: The criteria may be defined as appropriate in the National Annex or for the individual
criteria (for serviceability) project.
The following accelerations (m/s2) are the recommended maximum values for any part of the
deck:
i) 0.7 for vertical vibrations,
ii) 0.2 for horizontal vibrations due to normal use, and
iii) 0.4 for exceptional crowd conditions.
(2) A verification of the comfort criteria should be performed if the fundamental frequency of the
deck is less than
– 5 Hz for vertical vibrations, or
– 2.5 Hz for horizontal (lateral) and torsional vibrations.
NOTE: The data used in the calculations, and therefore the results, are subject to very high
uncertainties. When the comfort criteria are not satisfied with a significant margin, it may be
necessary to make provision in the design for the possible installation of dampers in the
structure after its completion. In such cases the designer should consider and identify any
requirements for commissioning tests.

274
Eurocode 0 (1) Passenger comfort depends on the vertical acceleration bv inside the coach during travel on
Annex A2 the approach to, passage over, and departure from the bridge.
(2) The levels of comfort and associated limiting values for the vertical acceleration should be
A2.4.4.3 Limiting values for
specified.
the maximum vertical
deflection for passenger NOTE: These levels of comfort and associated limiting values may be defined for the individ-
comfort ual project. Recommended levels of comfort are given in Table A2.9.
A2.4.4.3.1 Comfort criteria
Eurocode 0 (1) Where a vehicle/bridge dynamic interaction analysis is required, the analysis should take
Annex A2 account of the following behaviors:
iv) a series of vehicle speeds up to the maximum speed specified,
A2.4.4.3 Limiting values for
v) characteristic loading of the real trains specified for the individual project in accordance
the maximum vertical
with EN 1991-2, 6.4.6.1.1,
deflection for passenger
vi) dynamic mass interaction between vehicles in the real train and the structure,
comfort
vii) the damping and stiffness characteristics of the vehicle suspension,
A2.4.4.3.3 Requirements for
viii) a sufficient number of vehicles to produce the maximum load effects in the longest span,
a dynamic vehicle/bridge
ix) a sufficient number of spans in a structure with multiple spans to develop any resonance
interaction analysis for
effects in the vehicle suspension.
checking passenger
comfort NOTE: Any requirements for taking track roughness into account in the vehicle/bridge
dynamic interaction analysis may be defined for the individual project.
Table A.2. Summary of Clauses Relating to Loads and Actions in Eurocode EN 1991-2
Eurocode 1 (5) It is permissible to use alternative design rules different from the Application Rules given in
EN 1991-2 for works, provided that it is shown that the alternative rules accord with the rele-
1.3 Distinction between
vant principles and are at least equivalent with regard to the structural safety, serviceability,
Principles and Applica-
and durability which would be expected when using the Eurocodes.
tion Rules
Eurocode 1 (1) For normal conditions of use (i.e. excluding any accidental situation), the traffic and pedes-
trian loads (dynamic amplification included where relevant) should be considered as variable
Section 2 Classification of
actions.
actions
(2) The various representative values are
2.2 Variable actions
– characteristic values, which are either statistical (i.e. corresponding to a limited probability
of being exceeded on a bridge during its design working life) or nominal; see EN 1990,
4.1.2(7);
– frequent values; and
– quasi-permanent values.
(3) For calculation of fatigue lives, separate models, associated values, and where relevant,
specific requirements are given in 4.6 for road bridges, in 6.9 for railway bridges, and in the
relevant annexes.
Eurocode 1 (1) Load models defined in this section should be used for the design of road bridges with
loaded lengths less than 200 m.
Section 4 Road traffic
actions and other actions NOTE 1: 200 m corresponds to the maximum length taken into account for the
specifically for road calibration of Load Model 1 (see 4.3.2). In general, the use of Load Model 1 is
bridges safe-sided for loaded lengths over 200 m.
4.1 Field of application
NOTE 2: Load models for loaded lengths greater than 200 m may be defined in the National
Annex or for the individual project.
(2) The models and associated rules are intended to cover all normally foreseeable traffic situ-
ations (i.e. traffic conditions in either direction on any lane due to the road traffic) to be
taken into account for design [see however (3) and the notes in 4.2.1].
(3) The effects of loads on road construction sites (e.g. due to scrapers, lorries carrying earth) or
of loads specifically for inspection and tests are not intended to be covered by the load mod-
els and should be separately specified, where relevant.

275
Table A.2. Summary of Clauses Relating to Loads and Actions in Eurocode EN 1991-2 (continued)
Eurocode 1 (1) Loads due to the road traffic, consisting of cars, lorries, and special vehicles (e.g. for indus-
trial transport), give rise to vertical and horizontal, static and dynamic forces.
4.2 Representation of
actions NOTE 1: The load models defined in this section do not describe actual loads. They have
4.2.1 Models of road been selected and calibrated so that their effects (with dynamic amplification included where
traffic loads indicated).
NOTE 2: The National Annex may define complementary load models, with associated combi-
nation rules where traffic outside the scope of the load models specified in this section needs
to be considered.
NOTE 3: The dynamic amplification included in the models (except for fatigue), although
established for a medium pavement quality (see Annex B) and pneumatic vehicle suspension,
depends on various parameters and on the action effect under consideration. Therefore, it
cannot be represented by a unique factor. In some unfavorable cases, it may reach 1,7 (local
effects), but still more unfavorable values can be reached for poorer pavement quality, or if
there is a risk of resonance. These cases can be avoided by appropriate quality and design
measures. Therefore, an additional dynamic amplification may have to be taken into account
for particular calculations [see 4.6.1.(6)] or for the individual project.
Eurocode 1 (1) Characteristic loads are intended for the determination of road traffic effects associated with
ultimate limit state verifications and with particular serviceability verifications (see EN 1990 to
4.3 Vertical loads—
EN 1999).
Characteristic values
4.3.1 General and associated NOTE: There are 4 load models described in detail to cover most of the effects of the traffic
design situations of lorries and cars, special vehicles, and pedestrian crowd loading. They are used for general
and local verifications. One of these models is used to represent dynamic effects on short
structural members.
Eurocode 1 (1) Traffic running on bridges produces a stress spectrum which may cause fatigue. The stress
spectrum depends on the geometry of the vehicles, the axle loads, the vehicle spacing, the
4.6 Fatigue load models
composition of the traffic, and its dynamic effects.
4.6.1 General
NOTE: There are 5 load models described in detail. The first two are intended to be used to
check whether the fatigue life may be considered unlimited when a constant stress amplitude
fatigue limit is given. Therefore, they are appropriate for steel constructions and may be inap-
propriate for other materials. The remaining 3 load models are intended to be used for fatigue
life assessment. Each of these last three models is more accurate than its predecessor, cul-
minating in the last model which is based on actual traffic data.
A.4 EN 1992 Eurocode 2: EN 1992-1.2:2004 Design of concrete structures. Fire

Design of Concrete design
Structures EN 1992-2:2005 Design of concrete structures.
Concrete bridges. Design and
Eurocode 2 (Design of concrete structures) is concerned with detailing rules
the requirements for resistance, serviceability, durability, and EN 1992-3:2006 Design of concrete structures.
fire resistance of concrete structures. (Other requirements, Liquid retaining and containing
e.g. concerning thermal or sound insulation, are not consid- structures
ered.) It applies to the design of buildings and civil engineer-
ing works in plain, reinforced, and prestressed concrete. Note also
EN 1992 is presented in three main parts. The first part has PD 6687:2006 Background paper to the UK National
two sub-parts covering buildings and structural fire design. Annexes to BS EN 1992-1
The last two main parts cover concrete bridges and liquid PD 6687-2:2008 Recommendations for the design of
retaining and containing structures, as in the following list. structures to BS EN 1992-2
Those underlined have been reviewed in the compilation of
this report. The second part, EN 1992-2:2005 (Design of concrete struc-
tures. Concrete bridges. Design and detailing rules) is relevant
EN 1992-1.1:2004
Design of concrete structures. General for the design of concrete bridges. Sections from this stan-
rules and rules for buildings dard relating to the serviceability limit state are summarized

276
in Table A.3. It should be noted that EN 1992-2 draws heavily together with specific provisions for bridges. For the design
from the general clauses presented in EN 1992-1.1 (Design of of new bridges, EN 1992-2 is intended to be used, for direct
concrete structures. General rules and rules for buildings); where application, together with other parts of EN 1992 and Euro-
relevant, these clauses are also included in the summaries codes EN 1990, 1991, 1997, and 1998.
given in the table. A summary of clauses relating to the serviceability limit state
EN 1992-2 describes the principles and requirements for design of concrete bridges Eurocode EN 1992-1 is presented in
safety, serviceability, and durability of concrete structures, Table A.3.
Table A.3. Summary of Clauses Relating to the Serviceability Limit State Design
of Concrete Bridges in Eurocode EN 1992-1
Eurocode 2 (3) The basic requirements of EN 1990, Section 2 are deemed to be satisfied for concrete struc-
tures when the following are applied together:
Section 2 Basis of Design
– limit state design in conjunction with the partial factor method in accordance with EN 1990,
2.1 Requirements
– actions in accordance with EN 1991,
2.1.1 Basic requirements
– combination of actions in accordance with EN 1990, and
– resistances, durability and serviceability in accordance with this standard.
NOTE: Requirements for fire resistance (see EN 1990 Section 5 and EN 1992-1.2) may
dictate a greater size of member than that required for structural resistance at normal
temperature.
Eurocode 2 (1) Thermal effects should be taken into account when checking serviceability limit states.
(2) Thermal effects should be considered for ultimate limit states only where they are significant
2.3.1.2 Thermal effects
(e.g. fatigue conditions, in the verification of stability where second order effects are of impor-
tance). In other cases they need not be considered, provided that the ductility and rotation
capacity of the elements are sufficient.
(3) Where thermal effects are taken into account, they should be considered as variable actions
and applied with a partial factor and y factor.
NOTE: The y factor is defined in the relevant annex of EN 1990 and EN 1991-1.5.
Eurocode 2 (2) The effects of differential settlements should generally be taken into account for the verification
of serviceability limit states.
2.3.1.3 Differential
settlements/movements
Eurocode 2 (1) Shrinkage and creep are time-dependent properties of concrete. Their effects should
generally be taken into account for the verification of serviceability limit states.
2.3.2 Material and product
(3) When creep is taken into account, its design effects should be evaluated under the quasi-
properties
permanent combination of actions irrespective of the design situation considered (i.e. persistent,
2.3.2.1 General
transient, or accidental).
2.3.2.2 Shrinkage and creep
NOTE: In most cases the effects of creep may be evaluated under permanent loads and the
mean value of prestress.
Eurocode 2 (2) The values for partial factors for materials for serviceability limit state verification should be
taken as those given in the particular clauses of this Eurocode.
2.4.2 Design values
2.4.2.4 Partial factors for NOTE: The values of gC and gS in the serviceability limit state for use in a country may be
materials found in its National Annex. The recommended value for situations not covered by particular
clauses of this Eurocode is 1.0.
Eurocode 2 (1)P The following clauses give principles and rules for normal and high-strength concrete.
(2) Rules for lightweight aggregate concrete are given in Section 11.
Section 3 Materials
3.1 Concrete
3.1.1 General
Eurocode 2 (1)P This clause applies to wires, bars, and strands used as prestressing tendons in concrete
3.3 Prestressing steel structures.
3.3.1 General (2)P Prestressing tendons shall have an acceptably low level of susceptibility to stress
corrosion.
(3) The level of susceptibility to stress corrosion may be assumed to be acceptably low if the
prestressing tendons comply with the criteria specified in EN 10138 or given in an appropriate
European Technical Approval.

277
of Concrete Bridges in Eurocode EN 1992-1 (continued)
Eurocode 2 (1)P A durable structure shall meet the requirements of serviceability, strength, and stability
throughout its design working life, without significant loss of utility or excessive unforeseen
Section 4 Durability and
maintenance (for general requirements see also EN 1990).
cover to reinforcement
(2)P The required protection of the structure shall be established by considering its intended
4.1 General
use, design working life (see EN 1990), maintenance program, and actions.
(3)P The possible significance of direct and indirect actions, environmental conditions (4.2), and
consequential effects shall be considered.
NOTE: Examples include deformations due to creep and shrinkage (see 2.3.2).
Eurocode 2 (3) Imperfections need not be considered for serviceability limit states.
Section 5 Structural
analysis
5.2 Geometric
imperfections
Eurocode 2 (1) Linear analysis of elements based on the theory of elasticity may be used for both the
serviceability and ultimate limit states.
5.4 Linear elastic analysis
(3) For thermal deformation, settlement, and shrinkage effects at the ultimate limit state (ULS), a
reduced stiffness corresponding to the cracked sections, neglecting tension stiffening but
including the effects of creep, may be assumed. For the serviceability limit state (SLS), a
gradual evolution of cracking should be considered.
Eurocode 2 (2) Verifications in SLS may be carried out using strut-and-tie models (e.g. verification of steel
stresses and crack width control) if approximate compatibility for strut-and-tie models is
5.6 Plastic analysis
ensured (in particular the position and direction of important struts should be oriented
5.6.4 Analysis with
according to linear elasticity theory).
strut-and-tie models
Eurocode 2 (1) Non-linear methods of analysis may be used for both ULS and SLS, provided that equilibrium
and compatibility are satisfied and an adequate non-linear behavior for materials is assumed.
5.7 Non-linear analysis
The analysis may be first or second order.
(105) Non-linear analysis may be used, provided that the model can appropriately cover all failure
modes (e.g. bending, axial force, shear, compression failure affected by reduced effective con-
crete strength) and that the concrete tensile strength is not utilized as a primary load resisting
mechanism. If one analysis is not sufficient to verify all the failure mechanisms, separate addi-
tional analyses should be carried out.
The following design format should be used:
– The resistance should be evaluated for different levels of appropriate actions which should
be increased from their serviceability values by incremental steps, such that the value of
gG.Gk and gQ.Qk are reached in the same step. The incrementing process should be contin-
ued until one region of the structure attains the ultimate strength, evaluated taking
account of aCC, or there is global failure of the structure. The corresponding load is
referred to as qud.
Further steps in the design format that should be used are given.
Eurocode 2 (1)P For serviceability and fatigue calculations, allowance shall be made for possible variations
in prestress. Two characteristic values of the prestressing force at the serviceability limit state
5.10 Prestressed members
are estimated. These are based on the upper characteristic value and the lower characteristic
and structures
value.
5.10.9 Effects of prestress-
ing at serviceability limit
state and limit state of
fatigue
Eurocode 2 (1)P This section covers the common serviceability limit states. These are
– stress limitation (see 7.2),
Section 7 Serviceability
– crack control (see 7.3), and
limit states (SLS)
– deflection control (see 7.4).
7.1 General
Other limit states (such as vibration) may be of importance in particular structures but are not
covered in this standard.

278
(2) In the calculation of stresses and deflections, cross sections should be assumed to be
uncracked, provided that the flexural tensile stress does not exceed fct,eff. The value of fct,eff
may be taken as fctm or fctm,n, provided that the calculation for minimum tension reinforcement
is also based on the same value. For the purposes of calculating crack widths and tension
stiffening, fctm should be used.
Eurocode 2 (1)P The compressive stress in the concrete shall be limited to avoid longitudinal cracks, micro-
cracks, or high levels of creep, where they could result in unacceptable effects on the func-
7.2 Stress limitation
tion of the structure.
(102) Longitudinal cracks may occur if the stress level under the characteristic combination of
loads exceeds a critical value. Such cracking may lead to a reduction of durability. In the
absence of other measures, such as an increase in the cover to reinforcement in the com-
pressive zone or confinement by transverse reinforcement, it may be appropriate to limit the
compressive stress to the value k1fck in areas exposed to environments of exposure classes
XD, XF, and XS (see Table 4.1 of EN 1992-1-1).
NOTE: The value of k1 for use in a country may be found in its National Annex. The recom-
mended value is 0.6. The maximum increase in the stress limit above k1fck in the presence of
confinement may also be found in a country’s National Annex. The recommended maximum
increase is 10%.
NOTE: British National Document PD 6687:2006 (Background paper to the UK National
Annexes to BS EN 1992-1) gives non-contradictory complimentary information for use with EN
1992-1. In particular, when considering stress limitation in serviceability it notes
a) Stress checks in reinforced concrete members have not been required in the UK for the
past 50 years or so, and there has been no known adverse effect. Provided that the
design has been carried out properly for ultimate limit state, there will be no significant
effect at serviceability in respect of longitudinal cracking.
b) There has been no evidence either from research or practice that there is a correlation
between high compressive stress and durability problems.
(3) If the stress in the concrete under the quasi-permanent loads is less than k2fck, linear creep
may be assumed. If the stress in concrete exceeds k2fck, non-linear creep should be consid-
ered (see 3.1.4).
NOTE: The value of k2 for use in a country may be found in its National Annex. The recom-
mended value is 0.45.
(4)P Tensile stresses in the reinforcement shall be limited to avoid inelastic strain, unacceptable
cracking, or deformation.
(5) When structural appearance is considered, unacceptable cracking or deformation may be
assumed to be avoided if, under the characteristic combination of loads, the tensile strength
in the reinforcement does not exceed k3fyk. Where the stress is caused by an imposed defor-
mation, the tensile strength should not exceed k4fyk. The mean value of the stress in pre-
stressing tendons should not exceed k5fyk.
NOTE: The values of k3, k4, and k5 for use in a country may be found in its National Annex.
The recommended values are 0.8, 1, and 0.75, respectively.
Eurocode 2 (1)P Cracking shall be limited to an extent that will not impair the proper functioning or durability
of the structure or cause its appearance to be unacceptable.
7.3 Crack control
(2) Cracking is normal in reinforced concrete structures subject to bending, shear,
7.3.1 General
torsion, or tension resulting from either direct loading or restraint or imposed deformations.
considerations
(3) Cracks may also arise from other causes such as plastic shrinkage or expansive chemical
reactions within the hardened concrete. Such cracks may be unacceptably large, but their
avoidance and control lie outside the scope of this section.
(4) Cracks may be permitted to form without any attempt to control their width, provided they do
not impair the functioning of the structure.

279
(105) A limiting calculated crack width wmax, taking account of the proposed function and nature
of the structure and the costs of limiting cracking, should be established. Due to the random
nature of the cracking phenomenon, actual crack widths cannot be predicted. However, if the
crack widths calculated in accordance with the models given in this standard are limited to
the values given in Table 7.101N, the performance of the structure is unlikely to be impaired.
NOTE: The value of wmax and the definition of decompression and its application for use in a
country may be found in its National Annex. The recommended value for wmax and the appli-
cation of the decompression limit are given in Table 7.101N. The recommended definition of
decompression is noted in the text under the table.
NOTE: British National Document PD 6687-2:2008 (Recommendations for the design of
structures to BS EN 1992-2:2005) gives non-contradictory complimentary information for
use with EN 1992-2. In particular, it contains a Section 8, Serviceability limit states. Under
8.2.1 it makes recommendations for the values of wmax and notes a lack of clarity. Under
8.2.2 it offers a simplification in crack calculation methods. Under 8.2.3 it gives guidance
on calculating crack widths due to early age restraint of imposed deformations, which can
arise due to early thermal contraction and shrinkage. Such effects should be taken into
account in design.
(6) For members with only unbonded tendons, the requirements for reinforced concrete ele-
ments apply. For members with a combination of bonded and unbonded tendons, require-
ments for prestressed concrete members with bonded tendons apply.
(7) Special measures may be necessary for members subjected to exposure class XD3. The
choice of appropriate measures will depend upon the nature of the aggressive agent
involved.
(8) When using strut-and-tie models with the struts oriented according to the compressive stress
trajectories in the uncracked state, it is possible to use the forces in the ties to obtain the
corresponding steel stresses to estimate the crack width [see 5.6.4 (2)].
(9) Crack widths may be calculated according to 7.3.4. A simplified alternative is to limit the bar
size or spacing according to 7.3.3.
(110) In some cases it may be necessary to check and control shear cracking in webs.
NOTE: Further information may be found in Annex QQ.
Eurocode 2 (1)P If crack control is required, a minimum amount of bonded reinforcement is required
to control cracking in areas where tension is expected. The amount may be estimated
7.3 Crack control
from equilibrium between the tensile force in concrete just before cracking and the
7.3.2 Minimum reinforce-
tensile force in reinforcement at yielding or at a lower stress if necessary to limit the
ment areas
crack width.
(102) Unless a more rigorous calculation shows lesser areas to be adequate, the required
minimum areas of reinforcement may be calculated; a procedure is given.
(3) Bonded tendons in the tension zone may be assumed to contribute to crack control within a
distance 5 150 mm from the centre of the tendon.
(4) In prestressed members, no minimum reinforcement is required in sections where, under
the characteristic combination of loads and the characteristic value of prestress, the con-
crete is compressed or the absolute value of the tensile stress in the concrete is below a
given value.
Eurocode 2 (101) The control of cracking without direct calculation may be performed by means of simplified
methods. A recommended method is given with several sub-clauses indicating where crack
7.3 Crack control
control is deemed to be adequate, provided relevant detailing rules have been followed.
7.3.3 Control of cracking
without direct calculation
Eurocode 2 (101) The evaluation of crack width may be performed using recognized methods.
7.3 Crack control NOTE: Details of recognized methods for crack width control may be found in a country’s
7.3.4 Calculation of National Annex. The recommended method is that in EN 1992-1-1, 7.3.4.
crack widths

280
Eurocode 2 (1)P The deformation of a member or structure shall not be such that it adversely affects its
proper functioning or appearance.
7.4 Deflection control
(2) Appropriate limiting values of deflection taking into account the nature of the structure, of
7.4.1 General
the finishes, partitions and fixings and upon the function of the structure should be
considerations
established.
Eurocode 2 (1)P Where a calculation is deemed necessary, the deformations shall be calculated under load
conditions which are appropriate to the purpose of the check.
7.4 Deflection control
(2) P The calculation method adopted shall represent the true behavior of the structure under rel-
7.4.3 Checking deflections
evant actions to an accuracy appropriate to the objectives of the calculation.
by calculation
(3) Members which are not expected to be loaded above the level which would cause the tensile
strength of the concrete to be exceeded anywhere within the member should be considered to
be uncracked. Members which are expected to crack, but may not be fully cracked, will behave
in a manner intermediate between the uncracked and fully cracked conditions. And for mem-
bers subjected mainly to flexure, an adequate prediction of behavior is given by Expression
(7.18) presented in EN 1992-1.1.
(4) Deformations due to loading may be assessed using the tensile strength and modulus of
elasticity of the concrete [see (5)].
(5) For loads with a duration causing creep, the total deformation including creep may be calcu-
lated by using an effective modulus of elasticity for concrete according to Expression (7.20)
presented in EN 1992-1.1.
(6) Shrinkage curvatures may be assessed using Expression (7.21) presented in EN 1992-1.1.
(7) The most rigorous method of assessing deflections using the method given in (3) above is to
compute the curvatures at frequent sections along the member and then calculate the deflec-
tion by numerical integration. In most cases it will be acceptable to compute the deflection
twice, assuming the whole member to be in the uncracked and fully cracked condition in
turn, and then interpolate using Expression (7.1 8).
Eurocode 2 No rules peculiar to the serviceability limit state are given.
Section 8 Detailing of
reinforcement and
prestressing tendons—
General
Eurocode 2 (103) Minimum areas of reinforcement are given to prevent a brittle failure and wide cracks and
also to resist forces arising from restrained actions.
Section 9 Detailing of
members and NOTE: Additional rules concerning the minimum thickness of structural elements and the
particular rules minimum reinforcement for all surfaces of members in bridges, with minimum bar diameter
9.1 General and maximum bar spacing for use in a country may be found in its National Annex. No addi-
tional rules are recommended in this standard.
Eurocode 2 (1) For precast products in continuous production, subjected to an appropriate quality
control system according to the product standards, with the concrete tensile strength
Section 10 Additional rules
tested, a statistical analysis of test results may be used as a basis for the evaluation
for precast concrete
of the tensile strength that is used for serviceability limit states verifications, as an
elements and structures
alternative to Table 3.1.
10.3 Materials
(2) Intermediate strength classes within Table 3.1 may be used.
10.3.1 Concrete
10.3.1.1 Strength
Eurocode 2 (1)P The basic ratios of span/effective depth for reinforced concrete members without axial
compression, given in 7.4.2, should be reduced by a factor when applied to LWAC [light-
Section 11 Lightweight
weight aggregate concrete].
aggregate concrete
structures
states

281
Eurocode 2 (4) Members using plain concrete do not preclude the provision of steel reinforcement needed to
satisfy serviceability and/or durability requirements, nor reinforcement in certain parts of the
Section 12 Plain and lightly
members. This reinforcement may be taken into account for the verification of local ultimate
reinforced concrete
limit states as well as for the checks of the serviceability limit states.
structures
12.1 General
Eurocode 2 (1) Since plain concrete members have limited ductility, linear analysis with redistri-
bution or a plastic approach to analysis (e.g. methods without an explicit check
12.5 Structural analysis:
of the deformation capacity) should not be used unless their application can be
ultimate limit states
justified.
(2) Structural analysis may be based on the non-linear or the linear elastic theory. In the
case of a non-linear analysis (e.g. fracture mechanics) a check of the deformation
capacity should be carried out.
Eurocode 2 (1) Stresses should be checked where structural restraint is expected to occur.
(2) The following measures to ensure adequate serviceability should be considered:
a) with regard to crack formation,
states
– limitation of concrete tensile stresses to acceptable values,
– provision of subsidiary structural reinforcement (surface reinforcement, tying system
where necessary),
– provision of joints,
– choice of concrete technology (e.g. appropriate concrete composition, curing), and
– choice of appropriate method of construction.
b) with regard to limitation of deformations,
– a minimum section size, and
– limitation of slenderness in the case of compression members.
(3) Any reinforcement provided in plain concrete members, although not taken into account for
load bearing purposes, should comply with 4.4.1.
Eurocode 2 (101) The verifications for the execution stage should be the same as those for the completed
structure, with the following exceptions.
Section 113 Design for the
(102) Serviceability criteria for the completed structure need not be applied to intermediate exe-
execution stages
cution stages, provided that durability and final appearance of the completed structure are
113.3 Verification criteria
not affected (e.g. deformations).
113.3.2 Serviceability limit
(103) Even for bridges or elements of bridges in which the limit state of decompression is
states
checked under the quasi-permanent or frequent combination of actions on the completed
structure, tensile stresses less than k.fctm(t) under the quasi-permanent combination of
actions during execution are permitted.
NOTE: The value of k to be used in a country may be found in its National Annex. The
recommended value of k is 1.0.
(104) For bridges or elements of bridges in which the limit-state of cracking is checked
under frequent combination on the completed structure, the limit state of cracking
should be verified under the quasi-permanent combination of actions during
execution.
Eurocode 2 (101) This Annex may be used for calculating creep and shrinkage, including development with
time. However, typical experimental values can exhibit a scatter of ±30% around the values
Annex B (informative)
of creep and shrinkage predicted in accordance with this Annex. Where greater accuracy is
Creep and shrinkage strain
required due to the structural sensitivity to creep and/or shrinkage, an experimental assess-
B.100 General
ment of these effects and of the development of delayed strains with time should be under-
taken. Section B.104 includes guidelines for the experimental determination of creep and
shrinkage coefficients.

282
Eurocode 2 (1) The choice of adequately durable concrete for corrosion protection of reinforcement and pro-
tection of concrete attack requires consideration of the composition of concrete. This may
Annex E (informative)
result in a higher compressive strength of the concrete than is required for structural design.
Indicative strength classes
The relationship between concrete strength classes and exposure classes (see Table 4.1)
for durability
may be described by indicative strength classes.
E.1 General
(2) When the chosen strength is higher than that required for structural design, the value
of fctm should be associated with the higher strength in the calculation of minimum
reinforcement according to 7.3.2 and 9.2.1.1 and crack width control according to
7.3.3 and 7.3.4.
Eurocode 2 To avoid unacceptable cracks for the serviceability limit state, and to ensure the required deforma-
tion capacity for the ultimate limit state, the reinforcement derived from Expressions (F.8) and
Annex F (informative)
(F.9) for each direction should not be more than twice and not less than half the reinforcement
Tension reinforcement
determined by Expressions (F.2) and (F.3) or (F.5) and (F.6).
expressions for in-plane
stress conditions
F.1 General
Eurocode 2 (1) The interaction between the ground, the foundation, and the superstructure should be con-
sidered. The contact pressure distribution on the foundations and the column forces are both
Annex G (informative)
dependent on the relative settlements. More guidance is given in this annex.
Soil structure interaction
G.1 Shallow foundations
G.1.1 General
Eurocode 2 (1) For design purposes, various levels of analysis are permitted, depending on
conditions at both the serviceability and the ultimate limit states. More guidance
G.1.2 Levels of analysis
is given.
Eurocode 2 This Annex describes different methods of evaluating the time-dependent effects of concrete
behavior.
Annex KK (informative)
Structural effects of
time-dependent behavior
of concrete
KK.1 Introduction
Eurocode 2 (101) Structural effects of time-dependent behavior of concrete, such as variation of
deformation and/or of internal actions, shall be considered, in general, in serviceability
KK.2 General
conditions.
considerations
NOTE: In particular cases (e.g. structures or structural elements sensitive to second order
effects or structures in which action effects cannot be redistributed), time-dependent effects
may also have an influence at ULS.
(102) It is noted that for higher compressive stresses, non-linear creep effects should be
considered.
(104) Different types of analysis and their typical applications are shown in a table.
Brief outline details of some of the analysis methods are given in the sections that follow.
Eurocode 2 At present, the prediction of shear cracking in webs is accompanied by large model uncer-
tainty. Where it is considered necessary to check shear cracking, particularly for prestressed
Annex QQ (informative)
members, the reinforcement required for crack control can be determined. Some detailed
Control of shear cracks
guidance is given.
within webs

283
A.5 EN 1993 Eurocode 3: EN 1993-4.1:2007 Design of steel structures. Silos, tanks,

Design of Steel Structures and pipelines. Silos
EN 1993-4.2:2007 Design of steel structures. Silos, tanks,
The scope of EN 1993 is wider than most of the other design and pipelines. Tanks
Eurocodes due to the diversity of steel structures. This Euro- EN 1993-4.3:2007 Design of steel structures. Silos, tanks,
code covers both bolted and welded joints, and the possible and pipelines. Pipelines
slenderness of construction. The differences when compared EN 1993-5:2007 Design of steel structures. Piling
with existing British standards are that the Eurocode brings new EN 1993-6:2007 Design of steel structures. Crane sup-
methods into the scope. For example, the design of semi-rigid porting structures
joints in buildings is explained, and more advanced methods of
design for cold-formed steelwork are included. The rules for Note also
shell structures and for the design of piles, sheet piling, and silos EN 1090-2:2008 Execution of steel structures and
are new, and rules for stainless steel appear for the first time. aluminum structures. Technical
EN 1993 has 20 parts covering common rules for fire requirements for the execution of steel
design, bridges, buildings, tanks, silos, pipelined piling, crane structures
supported structures, chimneys, towers and masts, and so on, PD 6695-1.9:2008 Recommendations for the design of
as in the following list. Those underlined have been reviewed structures to BS EN 1993-1.9
in the compilation of this report. PD 6695-1.10:2009 Recommendations for the design of
structures to BS EN 1993-1.10
EN 1993-1.1:2005 Design of steel structures. General PD 6695-2:2008 Recommendations for the design of
rules and rules for buildings bridges to BS EN 1993
EN 1993-1.2:2005 Design of steel structures. General
rules. Structural fire design EN 1993-2:2006 (Design of steel structures. Steel bridges) pro-
EN 1993-1.3:2006 Design of steel structures. General vides a general basis for the structural design of steel bridges and
rules. Supplementary rules for cold- steel parts of composite bridges. EN 1993-2 gives design rules
formed members and sheeting which are supplementary to the generic rules in EN 1993-1-1.
EN 1993-1.4:2006 Design of steel structures. General EN 1993-2 is intended to be used with Eurocodes EN 1990 (Basis
rules. Supplementary rules for stainless of design), EN 1991 (Actions on structures), and the Parts 2 of
steels EN 1992 to EN 1998 when steel structures or steel components
EN 1993-1.5:2006 Design of steel structures. Plated struc- for bridges are referred to. Matters that are already covered in
tural elements those documents are not repeated within EN 1993-2.
EN 1993-1.6:2007 Design of steel structures. General. Sections from this standard (and, where noted, the other
Strength and stability of shell structures highlighted parts of EN 1993) relating to the serviceability
EN 1993-1.7:2007 Design of steel structures. General. limit state are summarized in Table A.4.
Plated structures subject to out of
plane loading A.6 EN 1994 Eurocode 4:
EN 1993-1.8:2005 Design of steel structures. Design of Design of Composite Steel
joints and Concrete Structures
EN 1993-1.9:2005 Design of steel structures. Fatigue
strength Eurocode 4 applies to the design of composite structures and
EN 1993-1.10:2005 Design of steel structures. Material members for buildings and civil engineering works. It com-
toughness and through-thickness plies with the principles and requirements for the safety and
properties serviceability of structures, the basis of their design, and veri-
EN 1993-1.11:2006 Design of steel structures. Design of fication that are given in EN 1990:2002 (Basis of structural
structures with tension components design). Eurocode 4 is concerned only with requirements for
EN 1993-1.12:2007 Design of steel structures. Additional resistance, serviceability, durability, and fire resistance of
rules for the extension of EN 1993 up composite structures. Other requirements (e.g. concerning
to steel grades S 700 thermal or sound insulation) are not considered.
EN 1993-2:2006 Design of steel structures. Steel bridges Eurocode 4 is intended to be used in conjunction with
EN 1993-3.1:2007 Design of steel structures. Towers, masts,
and chimneys. Towers and masts EN 1990 Basis of structural design
EN 1993-3.2:2008 Design of steel structures. Towers, EN 1991 Actions on structures
masts, and chimneys. Chimneys (text continues on page 289)

284
of Steel Bridges in Eurocode EN 1993-1 and 1993-2
Eurocode 3 EN 1993-1.5 (Plated structural elements) stipulates

(1)P The effects of shear lag and plate buckling shall be taken into account at the ultimate, service
Section 2 Basis of design
ability, or fatigue limit states.
and modeling
2.1 General EN 1993-1.11 (Tension components) stipulates
2.2(1)P Due to the difficulties in modeling the excitation characteristics of tension elements, service-
ability limit state checks should be carried out in addition to fatigue checks.
2.2(3) Any attachments to prefabricated tension components, such as saddles or clamps,
should be designed for ultimate limit states and serviceability limit states using the
breaking strength or proof strength of cables as actions; see Section 6. For fatigue see EN
1993-1-9.
NOTE: Fatigue action on the ropes is governed by the radius in the saddle or anchorage area
(see Figure 6.1 for minimum radius).
(2) For the calculation of stresses at the serviceability and fatigue limit state the effective area
2.3 Plate buckling effects
may be used if the condition in 3.1 is fulfilled. For ultimate limit states the effective area
on uniform members
according to 3.3 should be used with b replaced by bult.
(1) Shear lag in flanges may be neglected if b0 < Le/50 where b0 is taken as the flange outstand
Section 3 Shear lag in
or half the width of an internal element and Le is the length between points of zero bending
member design
moment; see 3.2.1(2).
3.1 General
(2) Where the above limit for b0 is exceeded, the effects due to shear lag in flanges should be
considered at serviceability and fatigue limit state verifications by the use of an effective
width according to 3.2.1 and a stress distribution according to 3.2.2. For the ultimate limit
state verification an effective area according to 3.3 may be used.
(1) The effective width beff for shear lag under elastic conditions should be determined from
3.2 Effective width for
elastic shear lag beff = b b0 (3.1)
3.2.1 Effective width
where the effective factor b is given in Table 3.1.
This effective width may be relevant for serviceability and fatigue limit states.
EN 1993-1.5 Annex E gives alternative methods for determining effective cross sections. It gives
a calculation for the serviceability limit slenderness to give effective area for stiffness. It also
gives that
(3) The effective second moment of area Ieff may be taken as variable along the span according
to the most severe locations. Alternatively a uniform value may be used based on the maxi-
mum absolute sagging moment under serviceability loading.
Eurocode 3 (6) Components that cannot be designed with sufficient reliability to achieve the total design
working life of the bridge should be replaceable. These may include
Section 4 Durability
– stays, cables, hangers;
– bearings;
– expansion joints;
– drainage devices;
– guardrails, parapets;
– asphalt layer and other surface protection;
– wind shields; and
– noise barriers.
EN 1993-1.11 (Tension components) contains a section (4) devoted to durability of wires,
ropes, and strands. This gives details of corrosion protection for the three classes of tension
components defined at the beginning of this standard.
EN 1993-1.11 (Tension components) covers transport, storage, and handling of tension components
in its Annex B (informative).

285
of Steel Bridges in Eurocode EN 1993-1 and 1993-2 (continued)
Eurocode 3 (1) A steel structure should be designed and constructed such that all relevant serviceability cri-
teria are satisfied.
(2) The basic requirements for serviceability limit states are given in 3.4 of EN 1990.
limit states
(3) Any serviceability limit state and the associated loading and analysis model should be speci-
7.1 General
fied for a project.
(4) The following serviceability criteria should be met:
a) Restriction to elastic behavior to limit
– excessive yielding, see 7.3(1);
– deviations from the intended geometry by residual deflections, see 7.3(1); and
– excessive deformations, see 7.3(4).
b) Limitation of deflections and curvature to prevent
– unwanted dynamic impacts due to traffic (combination of deflection and natural fre-
quency limitations), see 7.7 and 7.8;
– infringement of required clearances, see 7.5 or 7.6;
– cracking of surfacing layers, see 7.8; and
– damage of drainage, see 7.12.
c) Limitation of natural frequencies (see 7.8 and 7.9) to
– exclude vibrations due to traffic or wind which are unacceptable to pedestrians or pas-
sengers in cars using the bridge;
– limit fatigue damages caused by resonance; and
– limit excessive noise emission.
d) Restriction of plate slenderness (see 7.4) to limit
– excessive rippling of plates;
– breathing of plates; and
– reduction of stiffness due to plate buckling, resulting in an increase of deflection, see
EN 1993-1-5.
e) Improved durability by appropriate detailing to reduce corrosion and excessive wear; see
7.11.
f) Ease of maintenance and repair (see 7.11) to ensure
– accessibility of structural parts for maintenance and inspection, renewal of corrosion
protection and asphaltic pavements; and
– replacement of bearings, anchors, cables, expansion joints with minimum disruption to
the use of the structure.
(5) In most situations serviceability aspects should be dealt with in the conceptual design of the
bridge, or by suitable detailing. However, in appropriate cases, serviceability limit states may
be verified by numerical assessment (e.g. for calculating deflections or Eigen frequencies).
NOTE: The National Annex may give guidance on serviceability requirements for specific
types of bridges.
EN 1993-1.11 (Tension components) stipulates
7.1(1) The following serviceability criteria should be considered:
1. Deformations or vibrations, and
2. Elastic service conditions.
NOTE 1: Limits for deformations or vibrations may result in a stiffness requirement governed
by the structural system, the dimensions and the preloading of high-strength tension compo-
nents, and by the slipping resistance of attachments.
NOTE 2: Limits to retain elastic behavior and durability are related to maximum and minimum
values of stresses for serviceability load combinations.
7.1(2) Bending stresses in the anchorage zone may be reduced by suitable measures (e.g.
neoprene pads for transverse loading).
Eurocode 3 (1) Stresses at serviceability limit states should be determined from a linear elastic analysis,
using the appropriate section properties; see EN 1993-1-5.
7.2 Calculation models
(2) In modeling the structure, the non-uniform distribution of loads and stiffness resulting from
the changes in plate thickness, stiffening, etc. should be taken into account.
(3) Deflections should be determined by linear elastic analysis using the appropriate section
properties; see EN 1993-1-5.
NOTE: Simplified calculation models may be used for stress calculations, provided that the
effects of the simplification are conservative.

286
Eurocode 3 (1) The nominal stresses sEd.ser and tEd.ser resulting from the characteristic load combinations, cal-
culated making due allowance for the effects of shear lag in flanges and the secondary
7.3 Limitations for stress
effects caused by deflections (e.g. secondary moments in trusses), should be limited. The
standard gives equations for maximum allowable stresses.
NOTE 1: Where relevant, the above checks should include stresses sz from transverse loads;
see EN 1993-1-5.
NOTE 2: The National Annex may give the value for gMser. gMser = 1,00 is recommended.
NOTE 3: Plate buckling effects may be ignored as specified in EN 1993-1-5, 2.2(5).
(2) The nominal stress range Dsfre, due to the frequent load combination, should be limited to
1.5 fy/gMser; see EN 1993-1-9.
(3) For non-preloaded bolted connections subject to shear, the bolt forces due to the character-
istic load combination should be limited to
Fb.Rd.ser ≤ 0.7 Fb.Rd (7.4)
where Fb.Rd is the bearing resistance for ultimate limit states verifications.
(4) For slip-resistant preloaded bolted connections category B (slip-resistant at serviceability,
see EN 1993-1-8), the assessment for serviceability should be carried out using the charac-
teristic load combination.
EN 1993-1.11 (Tension components) stipulates
7.2(1) Limiting stress may be specified for the characteristic load combination for the following
purposes:
– to keep stresses in the elastic range for the relevant design situations during construc-
tion and in the service phase;
– to limit strains such that corrosion control measures are not affected (i.e., cracking of
sheaths, hard fillers, opening of joints) and also to cater for uncertainty in the fatigue
design;
– ULS verifications for linear and sub-linear structural response to actions.
Numeric values of limiting stress in the serviceability limit state are given.
Eurocode 3 (1) The slenderness of web plates should be limited to avoid excessive breathing that might
result in fatigue at or adjacent to the web-to-flange connections.
7.4 Limitation of web
breathing NOTE: The National Annex may define cases where web breathing checks are not necessary.
(2) Web breathing may be neglected for web panels without longitudinal stiffeners or for subpan-
els of stiffened webs, where certain criteria are met. (Criteria for road bridges and for rail
bridges are given in the standard. If the criteria are not met, a method for checking web
breathing is given.)
NOTE also EN 1993-1.7 (Plated structures subject to out of plane loading) gives a note
(8.2) on the out of plane deflection limit as the condition in which the effective use of a plate
segment is ended.
Eurocode 3 (1) To achieve a satisfactory appearance of the bridge, consideration should be given to
precambering.
7.6 Limits for visual
(2) In calculating camber, the effects of shear deformation and slip in riveted or bolted connec-
impression
tions should be considered.
(3) For connections with rivets or fitted bolts, a fastener slip of 0.2 mm should be assumed. For
preloaded bolts, slip does not need to be considered.
Eurocode 3 (1) Excessive deformation should be avoided where it could
– endanger traffic by excessive transverse slope when the surface is iced;
7.8 Performance criteria for
– affect the dynamic load on the bridge by impact from wheels;
road bridges
– affect the dynamic behavior causing discomfort to users;
7.8.1 General
– lead to cracks in asphaltic surfacing;
– adversely affect the drainage of water from the bridge deck.
NOTE: For durability requirements, see Annex C.

287
(2) Deformations should be calculated using the frequent load combination.

(3) The natural frequency of vibrations and deflections of the bridge should be limited to avoid
discomfort to users.
NOTE also EN 1993-1.7 (Plated structures subject to out of plane loading) gives guid-
ance on excessive vibration limits:
8.3(1) Excessive vibrations should be defined as the limit condition in which either the failure
of a plated structure occurs by fatigue caused by excessive vibrations of the plate or
serviceability limits apply.
EN 1993-1.11 (Tension components) contains a section (8) devoted to vibration of cables. A

general section is followed by a section on measures to limit vibration of cables and a section
giving estimation of risks.
Eurocode 3 (1) The deck structure should be designed to ensure that its deflection along the length is uni-
form and that there is no abrupt change in cross section giving rise to impact. Sudden
7.8.2 Deflection limits to
changes in the slope of the deck and changes of level at the expansion joints should be elimi-
avoid excessive impact
nated. Any transverse girders at the end of the bridge should be designed to ensure that the
from traffic
deflection does not exceed
– the limit specified for the proper functioning of the expansion joint;
– 5 mm under frequent loads unless other limits are specified for the particular type of
expansion joint.
NOTE: Guidance on the deflection limit of expansion joints is given in Annex B.
(2) Where the deck structure is irregularly supported (e.g. by additional bracings at intermediate
bridge piers), the deck area adjacent to these additional deck supports should be designed
for the enhanced impact factors given in EN 1991-2 for the area close to the expansion joints.
Eurocode 3 (1) Mechanical resonance should be taken into account when relevant. Where light bracing
members, cable stays, or similar components have natural frequencies that are close to the
7.8.3 Resonance effects
frequency of any mechanical excitation due to regular passage of vehicles over deck joints,
consideration should be given to either increasing the stiffness or providing artificial dampers
(i.e. oscillation dampers).
NOTE: Guidance on members supporting expansion joints is given in Annex B.
Eurocode 3 (1) All decks should be waterproofed, and the surfaces of carriageways and footpaths should be
sealed to prevent the ingress of water.
7.12 Drainage
(2) The layout of the drainage should take into account the slope of the bridge deck as well as
the location, diameter, and slope of the pipes.
(3) Free fall drains should carry water to a point clear of the underside of the structure to prevent
water entering into the structure.
(4) Drainage pipes should be designed so that they can be cleaned easily. The distance between
centers of the cleaning openings should be shown on drawings.
(5) Where drainage pipes are used inside box girder bridges, provisions should be made to pre-
vent accumulation of water during leaks or breakage of pipes.
(6) For road bridges, drains should be provided at expansion joints on both sides where [it] is
appropriate.
(7) Provision should be made for the drainage of all closed cross sections, unless these are fully
sealed by welding.
Eurocode 3 EN 1993-1.8 (Design of joints) stipulates
3.4 Categories of bolted connections
Section 8 Fasteners, welds,
3.4.1 Shear connections
connections, and joints
(1) Bolted connections loaded in shear should be designed as one of the following:
8.1 Connections made of
a) Category A: Bearing type
bolts, rivets, and pins
In this category, bolts from Class 4.6 up to and including Class 10.9 should be used. No
8.1.1 Categories of bolted
preloading and special provisions for contact surfaces are required. The design ultimate
connections
shear load should not exceed the design shear resistance, obtained from 3.6, nor the
design bearing resistance, obtained from 3.6 and 3.7.

288
b) Category B: Slip-resistant at serviceability limit state

In this category, preloaded bolts in accordance with 3.1.2(1) should be used. Slip should
not occur at the serviceability limit state. The design serviceability shear load should not
exceed the design slip resistance, obtained from 3.9. The design ultimate shear load
should not exceed the design shear resistance, obtained from 3.6, nor the design bearing
resistance, obtained from 3.6 and 3.7.
c) Category C: Slip-resistant at ultimate limit state
In this category, preloaded bolts in accordance with 3.1.2(1) should be used. Slip should
not occur at the ultimate limit state. The design ultimate shear load should not exceed the
design slip resistance, obtained from 3.9, nor the design bearing resistance, obtained
from 3.6 and 3.7. In addition, for a connection in tension, the design plastic resistance of
the net cross section at bolt holes Nnet.Rd , (see 6.2 of EN 1993-1-1), should be checked, at
the ultimate limit state.
3.4.2 Tension connections
(1) Bolted connection loaded in tension should be designed as one of the following:
a) Category D: Non-preloaded
In this category, bolts from Class 4.6 up to and including Class 10.9 should be used. No
preloading is required. This category should not be used where the connections are fre-
quently subjected to variations of tensile loading. However, they may be used in connec-
tions designed to resist normal wind loads.
b) Category E: Preloaded
In this category, preloaded 8.8 and 10.9 bolts with controlled tightening in conformity with
1.2.7 Reference Standards: Group 7 should be used.
The design checks for these connections are summarized in Table 3.2.
Eurocode 3 (1) Fatigue assessments should be carried out for all critical areas in accordance with EN
1993-1.9.
Section 9 Fatigue
(2) Fatigue assessment is not applicable to
assessment
– pedestrian bridges, bridges carrying canals, or other bridges that are predominantly stati-
9.1 General
cally loaded, unless such bridges or parts of them are likely to be excited by wind loads or
9.1.1 Requirements for
pedestrians;
fatigue assessment
– parts of railway or road bridges that are neither stressed by traffic loads nor likely to be
excited by wind loads.
NOTE that EN 1993-1.9 (Fatigue) specifies that the actions applied for a fatigue limit state
verification are different from those used for ultimate limit state or for serviceability state.
However, the stresses should be calculated at the serviceability state [Clause 5(1)].
NOTE that EN 1993-1.11 (Tension components) comments [2.2(1)P] that, due to the diffi-
culties in modeling the excitation characteristics of tension elements, serviceability limit state
checks should be carried out in addition to fatigue checks.
Eurocode 3 (1) Fatigue assessments should be carried out for all bridge components unless the structural
detailing complies with standard requirements for durable structures established through
9.1.2 Design of road
testing.
bridges for fatigue
NOTE: The National Annex may give guidance on the conditions where no fatigue assess-
ment is necessary.
(2) Fatigue assessment should be carried out using the procedure given in this section and EN
1993-1-9.
Eurocode 3 (1) Design can be assisted by testing. If so, it should be in accordance with EN 1990, supple-
mented by the additional provisions given in 10.2 and 10.3.
Section 10 Design assisted
by testing
10.1 General

289
(continued from page 283) EN 1994-2:2005

Design of composite steel and concrete
European Standards (ENs), Harmonized European Standards structures. General rules and rules for
(hENs), Guidelines for European Technical Approval (ETAGs), bridges
and European Technical Approvals (ETAs) for construction
Note also
products relevant for composite structures
PD 6696-2:2007
Recommendations for the design of
EN 1090 Execution of steel structures and aluminum structures to BS EN 1994-2:2005
structures
EN 13670 Execution of concrete structures EN 1994-2 describes the principles and requirements for
EN 1992 Design of concrete structures safety, serviceability, and durability of composite steel and con-
EN 1993 Design of steel structures crete structures, together with specific provisions for bridges. It
EN 1997 Geotechnical design is based on the limit state concept used in conjunction with a
EN 1998 Design of structures for earthquake resistance partial factor method. It gives design rules for steel-concrete
composite bridges or members of bridges, additional to the
EN 1994 has three parts covering common rules and rules general rules in EN 1994-1-1. Cable stayed bridges are not fully
for buildings, structural fire design, and bridges, as in the fol- covered by this part.
lowing list. Those underlined have been reviewed in the com- EN 1994-2 contains the general rules from EN 1994-1-1 as
pilation of this report. well as the specific rules for the design of composite steel and
concrete bridges or composite members of bridges.
EN 1994-1.1:2004 Design of composite steel and concrete EN 1994-2 is intended to be used with EN 1990, the relevant
structures. General rules and rules for parts of EN 1991, EN 1992 for the design of concrete structures,
buildings and EN 1993 for the design of steel structures.
EN 1994-1.2:2005 Design of composite steel and concrete Sections from this standard (and, where noted, the other
structures. General rules. Structural highlighted recommendations) relating to the serviceability
fire design limit state are summarized in Table A.5.
Table A.5. Summary of Clauses Relating to Serviceability Limit State Design of Composite Steel
and Concrete Bridges in Eurocode EN 1994-2
Eurocode 4 (3) The basic requirements of EN 1990:2002, Section 2 are deemed to be satisfied for
composite structures when the following are applied together:
Section 2 Basis of design
– limit state design in conjunction with the partial factor method in accordance with
2.1 Requirements
EN 1990:2002;
– actions in accordance with EN 1991;
– combination of actions in accordance with EN 1990:2002; and
– resistances, durability, and serviceability in accordance with this standard.
Eurocode 4 (1)P The effects of shrinkage and creep of concrete and non-uniform changes of
temperature result in internal forces in cross sections, and curvatures and longitudi-
2.3.3 Classification of
nal strains in members; the effects that occur in statically determinate structures,
actions
and in statically indeterminate structures when compatibility of the deformations is
not considered, shall be classified as primary effects.
(2)P In statically indeterminate structures the primary effects of shrinkage, creep, and
temperature are associated with additional action effects, such that the total effects
are compatible; these shall be classified as secondary effects and shall be consid-
ered as indirect actions.
Eurocode 4 (1) The corrosion protection of the steel flange should extend into the steel-concrete
interface at least 50 mm. For additional rules for bridges with precast deck slabs,
4.2 Corrosion protection
see Section 8.
at the steel-concrete
interface in bridges
Eurocode 4 (1)P Calculation of stresses for beams at the serviceability limit state shall take into
account the following effects, where relevant:
– shear lag;
limit states
– creep and shrinkage of concrete;
7.2 Stresses
– cracking of concrete and tension stiffening of concrete;
7.2.1 General
– sequence of construction;

290
and Concrete Bridges in Eurocode EN 1994-2 (continued)
– increased flexibility resulting from significant incomplete interaction due to slip of

shear connection;
– inelastic behavior of steel and reinforcement, if any;
– torsional and distorsional warping, if any.
(2) Shear lag may be taken into account according to 5.4.1.2.
(3) Unless a more accurate method is used, effects of creep and shrinkage may be
taken into account by use of modular ratios according to 5.4.2.2.
(4) In cracked sections, the primary effects of shrinkage may be neglected when verify-
ing stresses.
(5)P In section analysis, the tensile strength of concrete shall be neglected.
(6) The influence of tension stiffening of concrete between cracks on stresses in rein-
forcement and prestressing steel should be taken into account. Unless more accu-
rate methods are used, the stresses in reinforcement should be determined
according to 7.4.3.
(7) The influences of tension stiffening on stresses in structural steel may be neglected.
(8) Stresses in the concrete slab and its reinforcement caused by simultaneous global
and local actions should be added.
Eurocode 4 (1)P Excessive creep and microcracking shall be avoided by limiting the compressive
stress in concrete.
7.2 Stresses
(2) Stress limitation for concrete to the value k1 fck should be in accordance with EN
7.2.2 Stress limitation for
1991-1-1:2002, 7.2, as modified by EN 1992-2.
bridges
(3)P The stress in reinforcing steel and in prestressing tendons shall be such that
inelastic strains in the steel are avoided.
(4) Under the characteristic combination of actions, the stresses should be limited to
k1 fsk in reinforcing steel and to k5 fpk in tendons, where the values k1 and k5 are given
in EN 1992-1-1:2004, 7.2(5).
(5) The stresses in structural steel should be in accordance with EN 1993-2, 7.3.
(6) For serviceability limit states, the longitudinal shear force per connector should be
limited according to 6.8.1 (3).
Eurocode 4 (1) The slenderness of unstiffened or stiffened web plates of composite girders should
be limited according to 7.4 of EN 1993-2.
7.2.3 Web breathing
Eurocode 4 (1) For the limit state of deformation, EN 1990:2002; A2.4 of Annex A2; and EN 1993-2,
7.5 to 7.8 and 7.12 apply where relevant.
7.3 Deformations in bridges
(2) Deflections should be calculated using elastic analysis in accordance with Section 5.
7.3.1 Deflections
(3) Deformations during construction should be controlled such that the concrete is not
impaired during its placing and setting by uncontrolled displacements and the
required long-term geometry is achieved.
Eurocode 4 (1) For the limit state of vibration, EN 1990:2002; A2.4 of Annex A2; EN 1991-2:2003,
5.7 and 6.4; and EN 1993-2, 7.7 to 7.10 apply where relevant.
7.3.2 Vibrations
Eurocode 4 (1) For the limitation of crack width in bridges, the general considerations of EN 1992-1-
1:2004, 7.3.1 as modified in EN 1992-2 apply to composite structures. The limitation
7.4 Cracking of concrete
of crack width depends on the exposure classes according to EN 1992-2, 4.
7.4.1 General
(2) An estimation of crack width can be obtained from EN 1992-1-1:2004, 7.3.4, where the
stress ss should be calculated by taking into account the effects of tension stiffening.
Unless a more precise method is used, ss may be determined according to 7.4.3(3).
(3) As a simplified and conservative alternative, crack width limitation to acceptable
width can be achieved by ensuring a minimum reinforcement defined in 7.4.2, and
bar spacing or diameters not exceeding the limits defined in 7.4.3.
(4) Application rules for the limitation of crack widths to wk are given in 7.4.2 and 7.4.3.
NOTE: The values of wk and the combination of actions may be found in the National
Annex. The recommended values for relevant exposure classes are as given (as wmax)
in the note to EN 1992-2, 7.3.1(105).

291
and Concrete Bridges in Eurocode EN 1994-2 (continued)
(5) Where composite action becomes effective as concrete hardens, effects of heat of

hydration of cement and corresponding thermal shrinkage should be taken into
account only during the construction stage for the serviceability limit state to define
areas where tension is expected.
(6) Unless specific measures are taken to limit the effects of heat of hydration of
cement, for simplification a constant temperature difference between the concrete
section and the steel section (concrete cooler) should be assumed for the determi-
nation of the cracked regions according to 7.4.2 (5) and for limitation of crack width
according to 7.4.2 and 7.4.3. For the determination of stresses in concrete, the
short-term modulus should be used.
NOTE: The National Annex may give specific measures and a temperature differ-
ence. The recommended value for the temperature difference is 20K.
Eurocode 4 (1) Unless a more accurate method is used in accordance with EN 1992-1-1:2004,
7.3.2(1), in all sections without prestressing by tendons and subjected to significant
7.4.2 Minimum
tension due to restraint of imposed deformations (e.g. primary and secondary effects
reinforcement
of shrinkage), in combination or not with effects of direct loading, the required mini-
mum reinforcement area As for the slabs of composite beams is given in this section
together with rules on its placement.
Eurocode 4 (1) Where at least the minimum reinforcement given by 7.4.2 is provided, the limitation
of crack widths to acceptable values may generally be achieved by limiting bar
7.4.3 Control of cracking
spacing or bar diameters.
due to direct loading
Maximum bar diameter and maximum bar spacing depend on the stress ss in the
reinforcement and the design crack width are given in tables.
(2) The internal forces should be determined by elastic analysis in accordance with Sec-
tion 5, taking into account the effects of cracking of concrete. The stresses in the
reinforcement should be determined taking into account effects of tension stiffening
of concrete between cracks. Unless a more precise method is used, the stresses
may be calculated according to the method given in (3).
Eurocode 4 (1) The action effects for the serviceability limit states should be determined according
to paragraphs given earlier in the standard (5.4.2.9).
7.5 Filler beam decks
Rules for cracking of concrete, minimum reinforcement, and control of cracking
(a deck consisting of a rein-
due to direct loading are also given for filler beam decks in this section.
forced concrete slab and
Guidance for transverse filler beams is given in PD 6696-2:2007.
partially concrete-encased
It indicates that for the determination of stresses in the concrete slab and steel
rolled or welded steel
beams, the concrete slab and steel beams should be considered to be
beams, having their bot-
• non-composite at ultimate limit state (ULS)
tom flange on the level of
• composite and to have equal deflections at serviceability limit state (SLS).
the slab bottom)
7.5.1 General
Eurocode 4 (1)P Resistance to fatigue and requirements for serviceability limit states shall be
verified for the combined local and simultaneous global effect.
Section 9 Composite plates
in bridges
9.1 General
9.4 Design of shear
connectors

Appendix B
SHRP 2 R19B Survey of Bridge Owners
292

293
Questionnaire #1 Summary of Experience with Service Issues
Your Name and Title:
Your Phone Number and Email Address:
Please address the following questions as they relate to summarizing your experience by material type, structure type and subsystem, component
and element type. The following are possible examples of these various features:
• Material Type: steel, plain concrete, reinforced concrete, prestressed concrete, etc.
• Structure Type: I girder bridge, box girder bridge, segmental bridges, truss, cable stayed, etc.
• Subsystems: superstructure, substructure, foundations, drainage systems, etc.
• Component: bearing, expansion joint
Please make as many copies of the appropriate questions as needed for the structure types, materials, subsystems or components for which
you are responding.
General Questions
1. What are the five or ten most costly maintenance/durability items in your structural maintenance budget?
2. Does your agency utilize deterioration models other than those in Pontis? If so, What and why?
3. The current LRFD service limit states include limits on:
• live load deflection of bridges,
• cracking of reinforced-concrete components,
• tensile stresses of prestressed-concrete components,
• compressive stresses of prestressed concrete components,
• permanent deformations of compact steel components,
• slip of slip-critical friction bolted connections, and
• settlement of shallow and deep foundations.
In your experience, are these service limit states adequate for your needs or are further safeguards required? If more are required, what should
they guard against for what types of members, systems or details?
Specific Questions for Structure Type, Material Subsystem or Component
Structure Type:
Material:
Subsystem or Component if Appropriate:
4. What have you seen as the important service and durability issues (not strength) and bridge age affects that impact the serviceability of bridge
components?
5. Have you seen issues resulting in reduced serviceability (or a trend in that direction) that could have been avoided if the design specifications
had additional service (not strength) design requirements? If so, what?
6. What type of foundation/wall settlement or other movements have resulted in maintenance issues or reduced serviceability?
7. Do you make Quantitative/Qualitative condition assessments beyond what is in Pontis? If so, has this data provided insight into serviceability
requirements?
8. Have you seen a direct correlation between deterioration and reduced serviceability (not nuisance maintenance)? If so, in what types of
structures/components? Have you been able to quantify the rate of reduced serviceability or service life?
9. Are there other questions we should have asked to gain more insight into your experience with service limit states? If so, what are they and
what would your responses have been?

294
Questionnaire #2 Geotechnical Service Issues
Form for Bridge Movements and Observed Distress
1. Preparer Information (fill one form per structure)
Prepared by/Title/Agency (Dept.)

Phone/E-mail
2. Bridge Information
State/County/Town Route No./Structure No.

Year built Crossing (over/under)
No. of spans Type of spans (simple,
continuous, cantilever, etc.)
Type of superstructure (steel, Type of abutments (integral,
concrete, girder, slab, box spill-through, full height,
beam, etc.) perched, stub, etc.)
Pier foundation type (spread Approach fill or wall type (Fill
footings, driven piles, drilled with side slope, MSE wall, etc.)
shafts, etc.)
Approach height As-built drawings available?
Geotechnical report available? Boring logs available?
Were repairs performed? Maintenance records available?
Any instrumentation data Any photos of bridge damage
available? and/or repairs available?
3. Construction Sequence: Fill in 1, 2 and 3 based on sequence of construction of following elements
Substructure _________________ Superstructure _________________ Approach fill/wall _________________
4. Geologic Information: Describe generalized geologic strata including soil types, water table location, Standard Penetration Test (SPT)
N-values, consolidation parameters, etc. If geotechnical report including boring logs is provided, refer to the report and no further
information is necessary.

295
5. Bridge Movements (horizontal movements are movements in longitudinal direction of bridge)
Vertical Movements Abutments Piers Horizontal Movements Abutments Piers

Estimated Estimated
Observed Observed
Note: If varied movement was observed at different support elements, provide additional information on a separate page as appropriate.
6. Effect of Movements on Bridge Structure: Indicate if distress types were tolerable or not based on the following definition:
“Movement is NOT tolerable if damage requires costly maintenance and/or repairs AND a more expensive construction to avoid this would have
been preferable.”
(Use additional pages if necessary to provide detailed information on any distress type)
# Distress Type Tolerable? (Yes/No), Description

1 Damage to abutments: cracking and spalling of abutments, abutment
footings, abutment pile caps, or abutment slope protection; also
included in this category are the opening, closing or damage to
abutment joints, the separation of the wingwall from abutment, and the
rupturing or exposure of abutment foundations.
2 Damage to piers: cracking and spalling of piers, pier footing, pier pile
caps, or struts of diaphragms between pier columns.
3 Vertical displacement: raising or lowering of the superstructure above
or below planned grade or a sag or heave in the deck; structures
requiring shimming or jacking as well as truss structures with increase
camber are also included.
4 Horizontal displacement: structures with a misalignment of bearings
and superstructure, or beams jammed against abutments; bridges where
superstructure extended beyond the abutment, where beams required
cutting, or where there was horizontal movement of the floor system.
5 Distress in superstructure: cracks or other evidence of excessive stress in
beams, girders, struts, and diaphragms as well as cracking and spalling
or the deck; shearing of anchor bolts; opening, closing or damage of
deck joints and cases where the cutting or relief joints were required.
6 Damage to railings, curbs, sidewalks or parapets: cracking, deformation,
or misalignment of railing, curb, sidewalks, or parapets; jammed curbs
and crushed concrete and open, closed or damaged portions of these
elements.
7 Damage to bearings: tilting or jamming or rockers as well as cases
where rockers have pulled off bearing, or where movement results
in an improper fit between bearing shoes and rockers requiring
re-positioning; deformed neoprene/elastomeric bearing pads, sheared
anchor bolts in the bearing shoes, damage to expansion devices, and
cracking of concrete at the bearings.
8 Poor riding quality: reported noticeable driver discomfort.
9 Not given or corrected during construction: those cases where any
mention of structural effects was omitted or where foundation
movement was corrected prior to construction of the superstructure.
10 None: no damage or repairs.
11 Other: other observed distress types not included in above categories.

296
Agency’s Tolerable Movement Criteria for New Bridge Structures
1. Article C10.5.5.2 of AASHTO (2007) allows angular distortion of 0.004 for multiple (continuous) spans and 0.008 for single-span bridges.
Does your agency follow these criteria? (Yes/No) _____.
[Note: Angular distortion is defined as DS/L where DS is the differential settlement between two support elements and L is the distance
between support elements (i.e., span length). Example: limiting angular distortion of 0.004 permits a differential settlement of 4.8-inches
over a 100-ft span length.]
If answer to above question is “No,” please provide following information. If criteria vary by span length, number of spans and/or structure
type (steel vs. concrete, girder vs. box beam, etc.) provide additional information as appropriate:
Permissible total vertical movement (settlement), S, at any given support element

Permissible differential vertical movement (settlement), DS, within a given span
Permissible angular distortion, DS/L (where L is span length)
Permissible horizontal movement, H (in longitudinal direction of the bridge)
Additional Information:
2. Does the agency have criteria for permissible horizontal movement (in longitudinal direction of the bridge)? (Yes/No) _______.
If answer to above question is “Yes,” please provide following information. If criteria vary by span length, number of spans and/or structure
type (steel vs. concrete, girder vs. box beam, etc.) provide additional information as appropriate:
Permissible horizontal movement, H (in longitudinal direction of the bridge)

Additional Information:

297
Response to Questionnaire 1 on Superstructure Issues Most respondents felt that the current SLS are adequate as cur-
rently specified. Two responses felt that at least one limit state
Question 1: What are the five or ten most costly maintenance/ was over conservative, with one believing the L/800 limit for live
durability items in your structural maintenance budget? load deflections is over conservative while the other related to
whether AASHTO LRFD 5.7.3.4-Control of Cracking by Distri-
The most costly maintenance/durability items within the
bution of Reinforcement is over conservative. The responses
structural maintenance budget as reported in the survey
suggested adding serviceability limit states with regards to:
responses are:
• foundation settlement of approach pavement;
• Expansion joints and steel coating systems (13 each);
• relative movement between adjacent girders and determina-
• Concrete decks (cracking, repair, and sealing for cracking)
tion the factored out-of-plane resistance for this condition;
(8);
• consider steel corrosion/section loss based on type of steel
• Deck overlays and bearings (7 each);
coating and a corrosion model and then rechecking stresses
• Concrete and steel repair or replacement (5 each);
based on reduced section;
• Abutment maintenance (3);
• consider local deflections or incompatible deformations at
• Timber components, movable bridges, approach slabs, rail-
component interfaces;
ings and curbs, reinforcement, fatigue, built-up steel corro-
• requirements for stress and cracking of reinforced concrete
sion, and scour (2 each); and
flexural members (current method of designing for strength
• Weld cracking, slope maintenance, anchor cables, deck
and then checking crack control reinforcement is not ade-
drains, concrete coatings, header joints, and concrete spalling
quate and members crack resulting in reduced service life,
(1 each).
this agency no longer uses reinforced concrete bridge beams);
• serviceability of connections; and
Question 2: Does your agency utilize deterioration models
• SLS for expansion joints and bearings.
other than those in Pontis? If so, what and why?
Fifteen survey responses were received and indicate that the Question 4: What have you seen as important service and
following are used to estimate deterioration: durability issues (not strength) and bridge age affects that
impact the serviceability of bridge components?
• Utilize Pontis only (8);
The fourteen responses to the above question included:
• Use a DOT created program (5); and
• Use experience or use no deterioration models (1 each).
• Deck cracking (6);
• Corrosion of steel, steel coating systems, and leaking expan-
The responses indicate that most agencies use models to esti-
sion joints (5 each);
mate deterioration. Those that do not use either Pontis or a • Fatigue (4);
DOT specific program use engineering judgment or do not • Bearing failure, chloride penetration, and deterioration of
attempt to estimate deterioration. Some programs consist of
beam ends (3 each);
computer software created using data combined with experi- • Preparation for steel painting, ADT combined with salt
ence in estimating remaining service life. Other programs
(2 each); and
combine condition assessments completed for NBI inspec- • Slope failure, end bent movement, deck drainage, deck mem-
tions and curves developed based on the structure type. Several
brane durability, bond and splice lengths, deterioration of
of the DOT programs use different elements in the deteriora-
non-composite bridge decks, exodermic bridge decks, bent
tion models than are used in Pontis.
cracking, concrete mix design, foundation movement, con-
crete deterioration, deflections and vibrations, adequate
Question 3: In your experience, are the current SLS adequate
detailing, and the requirements for appropriate cover ver-
for your needs or are further safeguard required? If more are
sus the requirements for crack control steel (1 each).
required, what should they guard against for what types of
members, systems or details?
The results are combined for all bridge types and components.
The responses to this question were as follows: The most common issue mentioned was deck cracking closely
followed by corrosion of steel in reinforced concrete and steel
• Adequate but a need for additional requirements (9); superstructures and painting of steel. Often times both were
• Adequate (6); and mentioned as “proper painting of the steel girders will slow
• Some of the current SLS are over conservative (2). down the corrosion process.” In addition, leaking expansion

298
joints was a common response, which may be related to many • Poor soil, movement of piles, and movement of abutments
of the other issues mentioned. As would be expected, the chlo- (2 each); and
ride penetration and “ADT and salt” categories were focused • Movement of end bents and rotation of spread footings
in the northern part of the country where winter weather con- (1 each).
ditions require the use of salt for traffic safety.
Foundation problems varied greatly, with some responses indi-
Question 5: Have you seen issues resulting in reduced ser- cating no foundation problems while others had many prob-
viceability (or a trend in that direction) that could have lems. One response said that they had no foundation problems
been avoided if the design specification had additional ser- but indicated that they had previously had some but these were
vice (not strength) design requirements? If so, what? eliminated using rules of thumb or by setting guidelines that
require all foundations to be at the same level. They also said
The thirteen responses to the above question included:
that the service issues with foundations were caused by con-
• Reduced serviceability could not have been avoided with struction issues. Those that noted movement of MSE or retain-
additional service requirements (7) ing walls as causing reduced serviceability noted that in one
• Additional requirements that would have helped avoid case the movement was caused by thermal loading and the
reduced serviceability were as follows: other was a result of the bridge being supported by piling while
44 Crack control requirements for concrete decks and pro- the retaining walls were not.
visions to limit corrosion of deck reinforcement;
44 Requirements to check for expansion caused by thermal Question 7: Do you make Quantitative/Qualitative condi-
loading (this is already included in the design specification assessments beyond what is in Pontis? If so, has this
tion for ULS and SLS); data provided insight into serviceability requirements?
44 Provisions for use of proprietary deck systems; Thirteen responses were received indicating whether addi-
44 Specifications for proper fill materials to prevent slope tional assessments are completed and if they provide insight
failures; and into serviceability requirements:
44 Cracking of cantilevered portions of bents—for sec-
tions with shear span to depth ratios of approximately • No additional quantitative or qualitative assessments (7);
1.5, limit service load stresses in longitudinal reinforce- • No additional quantitative but do complete additional
ment are limited to 30 ksi at column face (up to 36 ksi qualitative assessments (3); and
at column center) for moderate exposures and up to • Complete both additional quantitative and qualitative
24 ksi and 30 ksi at the column face and center, respec- assessments (3).
tively, for severe exposures.
The first agency that completes both additional assessments
Most respondents believe that no, new additional serviceabil- used various methods, such as, bridge deck condition sur-
ity requirements are necessary. Those that believed reduced veys, measurement of chloride penetration depth, and ultra-
serviceability could have been avoided suggested new service- sonic testing to measure corrosion combined with engineering
ability requirements for the bridge deck and foundation and judgment to determine priority for replacement or rehabili-
substructure. Others were not sure whether issues resulting in tation. The second agency completing additional assessments
reduced serviceability could have been avoided had there uses a condition scale for each bridge component and provides
been additional service requirements. relevant notes and figures that portray the overall condition
and what deficiencies exist. This agency also noted that the use
Question 6: What type of foundation/wall settlement or
of additional inspections show the same conditions over and
other movements have resulted in maintenance issues or
over in their bridge inventory but did not state what conditions
reduced serviceability?
these were, suggesting that there is a need for enhanced service-
Twelve respondents indicated different types of foundation ability requirements. The third agency completing additional
or wall settlement and movement that have resulted in quantitative/qualitative assessments while providing insight
reduced serviceability or maintenance include: into the deterioration of the structure have not been used with
regard to additional serviceability requirements.
• No foundation problems (3);
• Scour at spread footings and retaining structures, settle- Question 8: a. Have you seen a direct correlation between
ment or movement of approach slabs, MSE and retaining deterioration and reduced serviceability (not nuisance main-
walls, and wingwalls (3 each); tenance)? If so, in what types of structures or components?

299
b. Have you been able to quantify the rate of reduced service- • Provide SLS for permit trucks similar to the Strength II
ability or service life? limit state.
• Load test all bridge type/material combinations except
Thirteen responses were received in regards to Question 8a.
for steel.
The responses included:
Additional questions regarding the following items could
• Expansion joints and associated deterioration of beam
have been asked:
ends and substructure units (11);
• Bridge decks (5); • What alternative design loads (alternatives to HL-93) are
• No correlation (2); and being used to check limit states by other agencies? (This was
• Joints and timber piles (1 each). covered by NCHRP 12-83)
• What types of corrosion protection systems are required?
The results suggest that many of the responding agencies have • What types of exposures and associated environmental
determined that there is a correlation between deterioration distresses have you observed?
in different bridge components and reduced serviceability. • How have locked in connection forces due to permanent
The most common response was corrosion or section loss of deformations affected serviceability?
a steel girder/beam resulting in higher stresses under service • What effects on bearings and joints have you seen due to
loads. Additionally, correlations between deterioration and creep, shrinkage, and uniform and gradient temperature
reduced serviceability were noted for all components of a changes? and
bridge, from the foundation to the superstructure and deck. • How do live load deformations affect connections?
The results noted that there was a reduction in service life for
bridge decks and load carrying capacity for girders. Responses to Questionnaire #2 on Geotechnical Issues
The fourteen responses were received for Question 8b
regarding the quantification of reduced serviceability are: Only three respondents provided significant information in
response to this part of the survey. All respondents framed their
• No (12); and response in terms of experience with specific bridges. The rel-
• Yes (2). evant information is summarized in the following table. The
first four bridges are from the same state. They indicated that
The second part of question 8 was whether the different agen- they had observed distress in almost all cases where an integral
cies had tried to quantify the rate of reduced serviceability. The abutment and an unisolated MSE mass were used and foun-
overwhelming response was No, but two indicated that they dation movement occurred.
had, or were trying to, quantify the reduction in serviceability. Four responses were obtained to Question 6 regarding
One agency used engineering judgment to quantify the reduc- effects of movement on the structure. Three of the four are
tion in serviceability and service life of bridge decks while the from the agencies that provided responses to Questions 1
other was trying to determine a method to quantify the rate of through 5 of this questionnaire.
A majority of responses believe that the different types of
deterioration for deck systems which incorporated a sealer and
movement listed are not tolerable while some believe that
also to quantify the difference in deterioration between steel
movement is tolerable until it starts affecting other structural
girders with and without a protective coating.
components. The effects of movement resulted in varying
degrees of damage or distress to the structure.
Question 9: Are there other questions we should have asked
The first response indicated that movement and damage
to gain more insight into your experience with SLS? If so,
caused by movement is not tolerable according to the defini-
what are they and what would your responses have been?
tion provided. While movement and associated damage is not
Seven agencies answered Question 9 regarding other questions tolerated, they believe that poor riding quality is tolerable.
that should have been asked. Their responses are as follows: The second response provided repair methods for the differ-
ent types of movement listed. These include repairing vertical
• No (4) and settlement due to beam deterioration by jacking the beam and
• Yes (3). repairing the section loss. Poor riding quality was improved by
mudjacking the approach slabs. Damage to bearings is repaired
Approximately half of the agencies that provided a response by either replacing bearings with elastomeric bearing pads or
to the question (i.e., 3 agencies) had additional questions that resetting rocker bearings.
could have been asked or comments/suggestions. The follow- The third response indicated that movement occurred caus-
ing are their suggestions: ing a hinge to form over the pier resulting in the loss of bearing

300
Table B.1. Bridge Characteristics, Geological Information, and Movements
Question 1: Preparer Assistant Geotechnical Engineer (State A) DOT (State B) DOT (State C)
Question 2: Bridge Characteristics
Year built 1999 1998 2007 2002 1963 1999
No. of spans 5 5 5 3 3 6
Type of superstructure Steel Concrete Steel Concrete CIP Concrete Prestressed
(steel, concrete, girder, Slab Span concrete
slab, box beam, etc.) I-Beam
Pier foundation type Driven Piles Drilled Shafts Drilled Shafts Driven Piles Spread footing Pier footing on
(spread footings, Piles
driven piles, drilled
shafts, etc.)
Approach height 23 27 23 23 20
Geotechnical report Yes Yes No Yes No Yes
available?
Were repairs performed? Yes Yes Yes Yes Yes No
Any instrumentation No No No No No Yes
data available?
Route No./Structure No. I-135 US 75 I-35 US 50 I-40 15725 6015
Crossing (over/under) Broadway and Over 46th St. Under 87th St. Over Mary St. SH-9A Over Johnson
1st St. Creek and
Lower Arnot
Rd.
Type of spans (simple, Continuous Continuous Continuous Continuous Continuous Continuous
continuous,
cantilever, etc.)
Type of abutments Integral Integral Integral Integral Stub Stub
(integral, spill-
through, full height,
perched, stub, etc.)
Approach fill or wall MSE Wall MSE Wall MSE Wall MSE Wall Fill w/ side slope Fill w/ side
type (Fill with side slope
slope, MSE wall, etc.)
As-built drawings No No No No No Yes
available?
Boring logs available? Yes Yes No Yes Yes Yes
Maintenance records No No No No Yes Yes
available?
Any photos of bridge Probably ??? No No Yes Yes
damage and/or
repairs available?

301
Table B.1. Bridge Characteristics, Geological Information, and Movements (continued)
Question 3: Construction Sequence

Substructure, Substructure, Substructure, Substructure, Approach Fill/
Approach Fill/ Approach Fill/ Approach Fill/ Approach Fill/ Wall, Sub-
Wall, Wall, Wall, Wall, structure,
Superstructure Superstructure Superstructure Superstructure Superstructure
Question 4: Geological Info
Alluvial and Glacial drift and Residual clay Loessial soil and Stiff to hard Silt with gravel
residual clay alluvial soils soil over fill above allu- sandy clay and cobbles
overlying sand over glacial till Pennsylvanian vial sand (Liquid Limit overlying
or shale followed by bedrock (LL) ~50, Plas- non-plastic
bedrock Pennsylvanian ticity Index silt and silty
bedrock (PI)~30, #200 sand and
75%) gravel
Question 5: Bridge Movements
Vertical Movement NA NA NA NA 12″ at abutment 1-4″ prior to
construction
Horizontal Movement NA NA NA NA NA NA
Table B.2. Effects of Movement on Structure
Question 6: Effects of Movement on Structure

Indicate if distress types were tolerable or not based on the following definition:
“Movement is NOT tolerable if damage requires costly maintenance and/or repairs AND a more expensive construction to avoid this
would have been preferable.”
State A Damage to abutments: NO
Damage to piers: NO
Vertical Displacement: NO
Horizontal Displacement: NO
Distress in superstructure: NO
Damage to railings, curbs, sidewalks or parapets: NO
Damage to bearings: NO
Poor riding quality: YES
Not given or corrected: NA
None: YES
Other: YES
State A Damage to abutments: Tolerable until it effects bearings or severe deterioration of rebar
Damage to piers: same as abutments
Vertical Displacement: not much of a problem, typically caused by deterioration of beam end, fixed by jacking beam and repairing
section loss
Horizontal Displacement: not much of a problem, let it go until it gets excessive
Distress in superstructure: not tolerated if cracks in steel beam or shear cracks in concrete beam. Joint normally tolerated unless
pushing bents or abutments and causing damage
Damage to railings, curbs, sidewalks or parapets: tolerated
Damage to bearings: not tolerated-repair by replacing bearings with elastomeric pads or resetting rocker bearings
Poor riding quality: generally tolerated, mud-jack approach slabs if excessive
None: NA
Other: NA

302
Table B.2. Effects of Movement on Structure (continued)
State B Damage to abutments: NO, settlement resulted in end span cantilevered from pier, loss of bearing
Damage to piers:
Vertical Displacement: NO, shims added, caused crack to form over pier
Horizontal Displacement:
Distress in superstructure: NO, crack/hinge over pier
Damage to railings, curbs, sidewalks or parapets: NO, cracks in parapet
Damage to bearings: NO
Poor riding quality: NO
None: NA
Other: After hinge/crack in slab span, end span acted as simply supported. End span was not designed or reinforced to act as a
simple span
State C Damage to abutments: Not tolerable, except for slope protection and minor separation of wing wall from the abutment
Damage to Piers: Not tolerable, tolerable if crack widths are less than 0.025″ based on Oregon DOT cracking guidelines
Vertical Displacement: tolerable
Horizontal Displacement: assumed tolerable
Distress in superstructure: not tolerable, tolerable if crack width is less than 0.025″ based on Oregon DOT crack guidelines
Damage to railings, curbs, sidewalks, or parapets: tolerable
Damage to bearings: tolerable
Poor Riding quality: tolerable
Not given or corrected during construction: potential settlement was addressed in construction
None: NA
Other: NA
for the end span. This resulted in the end span becoming a sim- Those agencies that do not follow the criteria of Article
ple span which was not considered in the design. The move- C10.5.5.2 of the AASHTO LRFD Specification were asked
ment was a result of drilled shafts for the abutments being to provide the deflection limits that they use. These limits
founded in fill instead of bedrock. are provided in Table B.3.
The fourth response indicated that damage to railings,
curbs, sidewalks, parapets, and bearings are tolerable. Cracking Question 2: Does the agency have criteria for permissible
of piers and superstructure components are tolerable up to a horizontal movement (in longitudinal direction of the
certain limit based upon guidelines used by the Oregon DOT. bridge)?
Ten agencies responded to two questions about criteria for
The responses to the above question are:
movements permitted on new bridges. Their responses are
below.
• Yes (3)
• No (7)
Question 1: Does your agency follow the criteria of Article
• No response (6)
C10.5.5.2 of AASHTO (2007)?
Responses are: The allowable horizontal movement limits are shown in the
table on the following page Agency D limits horizontal move-
• Yes, we use the criteria of Article C10.5.5.2 of the AASHTO ment to that which can be accommodated by the bearings
LRFD Specification (2) and joints within the structure while others provide a finite
• No, we do not use the criteria of Article C10.5.5.2 of the value.
AASHTO LRFD Specification (8) Limits provided by various agencies which do not follow
• No response (6) Article C10.5.5.2 of AASHTO are shown in Table B.3.

303
Table B.3. Allowable Vertical and Horizontal Movements
Movement Agency A Agency B Agency C Agency D Agency E Agency F Agency G
Total Vertical Axial Elastic Only 1″ Assessed on a 1″ 1″ Follows local/ 1.5″

Movement case by case state
basis, and if likely requirements
Differential 0″ 1″ Case by case 1″ Structure
to continue steps
Vertical Specific
are taken to stop
Movement
settlement.
Angular 0″ 0.1% of Span 0.0025 for Structure
Distortion Length simple span Specific
0.0015 for
continuous
span
Horizontal <2″ due to thermal 0″ Case by case, Not specified 1″
Movement loading for inte- thermal, in design, is
gral abutments shrinkage, specified for
founded on a and live load construction
single row of movement
piles must be less
than that
allowed by
bearings and
expansion
joints

A ppen d i x C
Comparison of Crack Width Prediction Equations

for Prestressed Concrete Members
This appendix presents a review and comparison of various Based on a regression analysis of the test data, the authors
prediction equations for the maximum crack width in pre- proposed Equation C.3:
stressed concrete members. Test data from various sources
A 
14
were used in the comparisons. The equations are presented w max = 1.13 × 10−6  t  ac ∆f s31 (C.3)
in chronological order.  As 
where
C.1 CEB-FIP (1970) Equation Dfs1 = [fs - fd - 3.75] (ksi);
The 1970 Euro-International Committee for Concrete ac = stabilized crack spacing (in.);
and International Federation for Prestressing (CEB-FIP) At = area of concrete in tension (in.2);
recommended adopting the following equation (C.1) to As = total area of reinforcement (in.2);
predict the maximum crack width in partially prestressed E = 27.5 × 103 ksi was used;
beams: fs = stress in prestressing steel after cracking (ksi); and
fd = stress in the prestressing steel when the modulus of
w max = ( Dfs − 4000) × 10−6 (C.1) rupture of concrete at the extreme tensile fibers is
reached (ksi).
For static loads, the equation is this (C.2): After further simplification of Equation C.3, Nawy and
Potyondy (1971) recommended the following expression
w max = ∆ fs × 10−6 (C.2) (Equation C.4):
w max = 1.44 (∆ fs − 8.3) (C.4)
where Dfs is the stress change in steel after decompression of
concrete at centroid of steel. Please note that the Dfs in the where Dfs is the net stress in prestressing steel, or the magni-
CEB-FIP equation is in N/cm2. tude of tensile stress in normal steel at any crack width level.
Note the units for Dfs in Equation C.4 are ksi, and the units
for crack width are inches.
C.2 Nawy and Potyondy
(1971) Equation
C.3 Bennett and
Nawy and Potyondy (1971) conducted a research program Veerasubramanian
to study the flexural cracking behavior of pretensioned (1972) Equation
I- and T-beams. Table C.1 shows the geometric and
Bennett and Veerasubramanian (1972) investigated the behav-
mechanical properties of the prestressed beam specimens.
ior of nonrectangular beams with limited prestress after flexural
As represents the area of tension reinforcement comprising
cracking. They tested 34 prestressed concrete beams with the
both prestressing and normal steel reinforcement, As′ repre-
following cross sections:
sents the area of compression reinforcement, fc′ is the con-
crete cylinder compressive strength, and ft′ is the concrete • Rectangular: 12-in. deep × 6-in. wide;
tensile splitting strength. • I-Beam: 12-in. deep with 6-in. wide top and bottom flanges;
304

305
Table C.1. Geometrical Properties of Prestressed Beams
As As′
Width b Depth d r5 (%) r′5 (%)
Beam Section (in.) (in.)a As (in. )
2 bd A9s (in. )
2 b
bd fc9 (psi) ft9 (psi) Slump (in.)
B1 T 8 8.75 0.271 0.389 — — 4865 400 3

B2 I 6 8.90 0.271 0.518 — — 4865 400 3
B3 T 8 8.75 0.271 .0389 — — 4330 430 4
B4 I 6 8.90 0.271 0.518 — — 4290 430 4
B5 I 6 8.90 0.271 0.518 — — 4340 430 4
B6 T 8 8.75 0.271 0.389 — — 4375 430 4
B7 T 8 8.75 0.271 0.389 — — 4290 390 6
B8 I 6 8.90 0.271 0.518 — — 4260 390 6
B9 I 6 8.90 0.271 0.518 — — 4190 390 6
B10 T 8 8.75 0.271 0.389 — — 4280 390 6
B11 T 8 8.75 0.271 0.389 — — 4150 370 8
B12 I 6 8.90 0.271 0.518 — — 3920 370 8
B13 I 6 8.90 0.281 0.518 — — 3890 370 8
B14 T 8 8.75 0.271 0.389 — — 4110 370 8
B15 T 8 8.75 0.271 0.389 0.93 1.332 3490 340 5½
B16 I 6 8.90 0.271 0.518 0.33 0.631 3400 340 5½
B17 I 6 8.90 0.271 0.518 0.93 1.776 3390 340 5½
B18 T 8 8.75 0.271 0.389 0.33 0.473 3510 340 5½
B19c I 6 8.90 0.235 0.448 — — 3610 385 6
B20c I 6 8.90 0.235 0.448 — — 3495 385 6
B21 c I 6 8.90 0.235 0.448 — — 3430 355 6½
B22 c
I 6 8.90 0.235 0.448 — — 3280 355 6½
B23 I 6 8.90 0.271 0.518 — — 4060 380 5
B24 I 6 8.90 0.271 0.518 — — 4095 380 5
B25 I 6 8.90 0.271 0.518 — — 3950 380 5
B26 I 6 8.90 0.271 0.518 — — 4000 380 5
Note: — = no compression steel in the specimen.
a Total depth h of all beams = 12 in.
b A includes two 3⁄ -in.-diameter high-strength steel wire (f = 96,000 psi) cage bars in addition to prestressing strands.
s 16 y
c Beams B19–B22 were continuous beams and were not included in the cracking analysis.
Source: Nawy and Potyondy (1971); used by permission of the American Concrete Institute.
• I-Beam: 12-in. deep with 12-in. wide top flange and 6-in. where
wide bottom flange; and b1 = a constant representing the residual crack width mea-
• I-Beam: 8-in. deep. A slab 24-in. wide was cast later to rep- sured after the first cycle of loading. The value sug-
resent the deck. gested for deformed bars is 0.02 mm.
b2 = a constant depending on bond characteristics of the
All beams were simple spans with a span length of 10 ft.
nonprestressed steel. The value recommended for
Two concentrated loads spaced 6 ft. apart and centered on
deformed bars is 6.5.
the span were used for loading.
es = increase in strain in nonprestressed steel from stage
Bennett and Veerasubramanian recommended a predic-
of decompression of concrete at tensile face of beam
tion equation for the maximum crack width as follows in
Equation C.5: (µe).
dc = clear cover over the nearest reinforcing bar to the ten-
w max = β1 + β2 εs dc (C.5) sile face (mm).

306
Note that this equation uses the International System of Dfps = increase in stress in the prestressing steel beyond
Units (SI). decompression state (ksi); and
Σ0 = sum of reinforcing element circumferences (in.).
C.4 Nawy and Huang Table C.2 presents a comparison of the crack widths mea-
(1977) Equation sured from the beam tests performed by Nawy and Huang
Nawy and Huang (1977) studied crack and deflection con- (1977) and the ones predicted using the equation devel-
trol in pretensioned prestressed beams. They performed oped by Nawy and Huang (1977). On average, Equa-
tests on 20 single-span and four continuous beams. Based tion C.6 provides prediction results that are within 20%
on a detailed statistical analysis of the test data, they proof the measured maximum crack width of prestressed con-
posed the following equation (Equation C.6): crete beams.
At
w max = 5.85 × 10−5
βΣ0
(∆ fps ) (C.6)
C.5 Rao and Dilger
where (1992) Equation
At = area of concrete in tension (in.2);
b = ratio of distance from neutral axis of beam to con- Rao and Dilger (1992) developed a detailed crack control
crete outside tension face to distance from neutral procedure for prestressed concrete members. The authors
axis to steel reinforcement centroid; studied the prediction equation of maximum crack width
Table C.2. Observed Versus Theoretical Maximum Crack Width at Tensile Face of Beam
Net Steel Stress
30 ksi 40 ksi 60 ksi 80 ksi
wobs. wtheory Error (%) wobs. wtheory Error (%) wobs. wtheory Error (%) wobs. wtheory Error (%)
0.0111 0.0131 -15.3 0.0151 0.0175 -13.7 0.0261 0.0262 -0.4 0.04 0.0349 14.6
0.0127 0.0118 7.6 0.0204 0.0157 29.9 0.0275 0.0236 16.5 0.0409 0.0313 30.7
0.0131 0.0128 2.3 0.0166 0.0172 -3.5 0.0304 0.0256 18.8 0.0382 0.0344 11.0
0.0097 0.013 -25.4 0.0158 0.0174 -9.2 0.0226 0.0259 -12.7 0.0304 0.0347 -12.4
0.0091 0.0147 -38.1 0.0117 0.0197 -40.6 0.0205 0.0294 -30.3 0.032 0.0393 -18.6
0.0124 0.0148 -16.2 0.0181 0.0199 -9.0 0.0213 0.0297 -28.3 0.0364 0.0397 -8.3
0.0052 0.0051 2.0 0.0068 0.0069 -1.4 0.0117 0.0103 13.6 0.0188 0.0137 37.2
0.0049 0.0051 -3.9 0.0061 0.0069 -11.6 0.0111 0.0103 7.8 0.0146 0.0137 6.6
0.0051 0.0045 13.3 0.0064 0.0061 4.9 0.0107 0.009 18.9 0.0165 0.0121 36.4
0.0058 0.0045 28.9 0.0082 0.0061 34.4 0.0134 0.009 48.9 0.0185 0.0121 52.9
0.0054 0.0059 -8.5 0.0069 0.0079 -12.7 0.0112 0.0119 -5.9 0.0172 0.0158 8.9
0.0048 0.0059 -18.6 0.0076 0.0079 -3.8 0.0134 0.0119 12.6 0.0192 0.0158 21.5
0.0043 0.0046 -6.5 0.0058 0.0062 -6.5 0.0105 0.0092 14.1 0.0138 0.0123 12.2
0.0052 0.0046 13.0 0.0059 0.0062 -4.8 0.0103 0.0092 12.0 0.0145 0.0123 17.9
0.0039 0.0057 -31.6 0.0061 0.0076 -19.7 0.0115 0.0114 0.9 0.0181 0.0153 18.3
0.0038 0.0057 -33.3 0.0057 0.0076 -25.0 0.0093 0.0114 -18.4 0.016 0.0153 4.6
0.0039 0.0056 -30.4 0.006 0.0074 -18.9 0.0098 0.0112 -12.5 0.0159 0.0148 7.4
0.003 0.0056 -46.4 0.0045 0.0074 -39.2 0.0086 0.0112 -23.2 0.0147 0.0148 -0.7
0.0057 0.0061 -6.6 0.0085 0.0081 4.9 0.0129 0.0121 6.6 0.0202 0.0163 23.9
0.0034 0.0045 -24.4 0.0045 0.0059 -23.7 0.0089 0.0089 0.0 0.0139 0.0119 16.8
Average 18.6 Average 15.9 Average 15.1 Average 18.0
Source: Adapted from Nawy and Huang (1977).

307
developed by various previous researchers and proposed a ss = stress in the tension reinforcement assuming a
new equation (C.7) expressed as follows: cracked section. For pretensioned members, ss may
be replaced by Dsp, the stress variation in prestress-
w max = k1 fs dc ( At As )0.5 (C.7) ing tendons from the state of zero strain of the con-
crete at the same level.
where x1 = adjusted ratio of bond strength, taking into account
k1 = the bond coefficient defined for each combination of the different diameters of prestressing and reinforc-
prestressed and nonprestressed reinforcement; φ
fs = stress in steel after decompression (MPa); ing steel, calculated as ξ i s , where x is the ratio
φp
dc = concrete cover measured from surface to the center of
of bond strength of prestressing and reinforcing steel,
nearest reinforcement bar (mm);
fs is the largest bar diameter of reinforcing steel, and
At = area of concrete in tension (mm2);
fp is equivalent diameter of tendon. Equation C.10
As = total area of reinforcement (mm2).
follows.
C.6 Eurocode 2 (2004) sr,max = k3c + k1k2k4 φ ρ p ,eff (C.10)

Provisions
where
Eurocode 2 (2004) provides the following provisions to f = bar diameter;
calculate the crack widths (Equation C.8): c = cover to the longitudinal reinforcement;
k1 = coefficient that takes account of the bond properties of
w k = sr,max (εsm − εcm) (C.8) the bonded reinforcement;
k2 = coefficient that takes account of the distribution of
where strain;
sr,max = maximum crack spacing. k3 = coefficient that can be found in the National Annex
wk = crack width. according to different countries (recommended value
esm = mean strain in the reinforcement under the rele- is 3.4); and
vant combination of loads, including the effect of k4 = coefficient that can be found in the National Annex
imposed deformations and taking into account the according to different countries (recommended value
effects of tension stiffening. Only the additional is 0.425).
tensile strain beyond the state of zero concrete
strain at the same level is considered. C.7 Comparison Between
ecm = mean strain in the concrete between cracks. Measured and Predicted
In Equation C.8, the quantity (esm - ecm) can be calculated Maximum Crack Width
from the following expression (C.9): Using Various Equations
Figure C.1 through Figure C.4 present a comparison of the
fct,eff
σ s − kt
ρp ,eff
(1 + αeρ p,eff ) equation developed by Nawy and Huang (1977) and four
σs other prediction equations. Any points that fall on the 45o line
(εsm − εcm) = ≥ 0.6 (C.9)
Es Es plotted on the figures indicate agreement between sources. The
equations used in Eurocode were not compared with the test-
where ing data since there is not sufficient information to apply this
A′p = area of pre- or posttensioned tendons within Ac,eff. equation. Figure C.1 indicates that the equation developed by
Ac,eff = effective area of concrete in tension surrounding the Nawy and Potyondy (1971) did not provide good prediction
reinforcement or prestressing tendons of depth, hc,ef , results compared with the measured data since it relates the
where hc,ef is the lesser of 2.5(h - d), (h - x) / 3 or maximum crack width with the Dfps only. The equation devel-
h / 2, where h is the height of the beam, d is the effec- oped by Nawy and Huang (1977) exhibited excellent correla-
tive depth of a cross section, and x is the neutral axis tion at low values of crack width. The predicted values are
depth. slightly different from the measured data when the loading
kt = factor dependent on the duration of the load. increases, but the results are still close to the measured data.
ae = Es / Ecm, where Ecm is the secant modulus of elasticity Figure C.2 indicates the equation developed by Bennett
of concrete and Es is the design value of modulus of and Veerasubramanian (1972) does not exhibit good correla-
elasticity of reinforcing steel. tion with measured results when the maximum crack width
rp,eff = (As + x12A′p) / Ac,eff. increases.

308
0.05 0.05
Nawy and Huang Nawy and Huang
CEB-FIP
Predicted Maximum Crack Width (in.)

Nawy and Potyondy

Perfect Correlation Perfect Correlation
0.04 0.04
0.03 0.03
0.02 0.02
0.01 0.01
0 0
0 0.01 0.02 0.03 0.04 0.05 0 0.01 0.02 0.03 0.04 0.05
Measured Maximum Crack Width (in.) Measured Maximum Crack Width (in.)
Figure C.1. Comparison of measured and Figure C.3. Comparison of measured and
predicted maximum crack widths using predicted maximum crack widths using
equations developed by Nawy and Huang equations developed by Nawy and Huang
(1977) and Nawy and Potyondy (1971). (1977) and CEB-FIP (1970).
0.05 0.05
Nawy and Huang Nawy and Huang
Bennett and Rao and Dilger
Perfect Correlation
Veerasubramanian
0.04 Perfect Correlation 0.04
0.03 0.03
0.02 0.02
0.01 0.01
0 0
0 0.01 0.02 0.03 0.04 0.05 0 0.01 0.02 0.03 0.04 0.05
Measured Maximum Crack Width (in.) Measured Maximum Crack Width (in.)
Figure C.2. Comparison of measured and Figure C.4. Comparison of measured

predicted maximum crack widths using and predicted maximum crack widths
equations developed by Nawy and Huang using equations developed by Nawy
(1977) and Bennett and Veerasubramanian and Huang (1977) and Rao and Dilger
(1972). (1992).

309
Figure C.3 indicates that the equation recommended by References

CEB-FIP (1970) overestimates the crack width prediction at Bennett, E., and N. Veerasubramanian. 1972. Behavior of Nonrectangular
small load. A number of beam specimens had fully prestressed Beams with Limited Prestress after Flexural Cracking. Journal Proceed-
tendons, and the measured data does not compare well with ings, American Concrete Institute, Vol. 69, No. 9, pp. 533–542.
CEB-FIP Joint Committee. 1970. International Recommendations for
the predicted value.
the Design and Construction of Concrete Structures. Cement and
Figure C.4 indicates that the equation recommended by Concrete Association, London, England.
Rao and Dilger (1992) underestimates the crack width pre- EN 1992-2 (Eurocode 2): Design of Concrete Structures–Part 2: Concrete
diction, especially under heavy load. Bridges–Design and Detailing Rules. 2004. European Committee for
In summary, based on the comparisons, the equation Standardization, Brussels, Belgium.
Nawy, E., and P. Huang. 1977. Crack and Deflection Control of Preten-
developed by Nawy and Huang (1977) provides the best cor- sioned Prestressed Beams. Journal, Precast/Prestressed Concrete
relation with measured data. Furthermore, this equation takes Institute, Vol. 23, No. 3, pp. 30–43.
the effect of bar size and steel stress into account and can be Nawy, E., and J. Potyondy. 1971. Flexural Cracking Behavior of Preten-
easily incorporated into the calibration procedure. The equa- sioned Prestressed Concrete I- and T-Beams. Journal Proceedings,
American Concrete Institute, Vol. 68, No. 5, pp. 355–360.
tion by Nawy and Hwang (1977) was used in the calibration Rao, S., and W. Dilger. 1992. Control of Flexural Crack Width in Cracked
of the tension in prestressed concrete when the crack width Prestressed Concrete Members. Structural Journal, American Concrete
was considered. Institute, Vol. 89, No. 2, pp. 127–138.

Appendix D
Derivation of the Resistance Prediction Equation

of Prestressed Concrete Bridge Girders
The derivation of the resistance prediction equation for a By rearranging Equation D.2, the stress in the concrete at
prestressed concrete girder subjected to flexural loading is the top fiber can be calculated as follows (Equation D.3):
shown in this appendix. Figure D.1 displays the stress dis-
 f ps − Eps (ε se + ε ce ) Ec c
tribution diagram for a typical prestressed concrete bridge fct =  (D.3)
girder at various stages of loading. In this study, decom- Eps (dp − c )
pression is considered as the stress state producing a zero
stress at the extreme bottom fibers of the prestressed girder. From strain compatibility, we have Equation D.4,
Using axial force equilibrium, we have Equation D.1:
c − hf
f s = Es εs =
Es
Ec ( )
d −c
fct s
c
(D.4)
fct
Aps f ps + As f s = b0c − 2 (b0 − b)(c − h f ) fct where
2 c
c = depth of neutral axis from the from extreme com-
c − hf − hf 1 pression fiber;
− 2 (b − bw )(c − h f − h f 1) fct dp = distance from extreme compression fiber to centroid
c of prestressing steel;
ds = distance from extreme compression fiber to centroid
f ct  b0c − (c − h f ) (b0 − b) 
 2 
2
of nonprestressing steel;
= (D.1)
2c   Ec = modulus of elasticity of concrete;
− (c − h f − h f 1) (b − bw )
2
Eps = modulus of elasticity of prestressing steel;

where Es = modulus of elasticity of nonprestressing steel;
As = area of nonprestressing steel; fct = stress in the concrete at the top of the beam after
Aps = area of prestressing steel in tension zone; losses at service load;
b = prestressed beam top flange width; ece = strain in concrete at the level of prestressing steel after
b0 = effective deck width transformed to the beam material; losses at dead load state;
bw = web thickness; eps = strain in prestressing steel after losses at service load;
c = depth of neutral axis from the from extreme comes = strain in nonprestressing steel after losses at service
pression fiber; load; and
fct = calculated stress in concrete at the top fiber; ese = strain in prestressing steel after losses at dead load state.
fps = calculated stress in prestressing steel;
fs = calculated stress in nonprestressing steel; Substitute Equation D.3 and Equation D.4 into Equa-
hf = deck thickness; and tion D.1 to obtain Equation D.5:
hf1 = top flange thickness.
 f ps − Eps (ε se + ε ce ) Ec  2
Aps f ps =  i b0c − (c − h f ) (b0 − b)
2
2 Eps (d p − c)
The stress in the prestressing steel can be calculated as
follows (Equation D.2): 
Eps  dp − c 2 As Es (ds − c)
− (c − h f − h f 1) (b − bw ) −
2
f ps = Eps ε ps = Eps (ε se + ε ce ) + fct  (D.2) (D.5)
Ec  c  Ec 

310

311
b0
b Tension Compression Compression Compression Compression
hf
hf1
c
dp bw
h ds
hf2
b1 Compression Compression Tension

(c) Decompression (d) Decompression
(a) Initial State (b) Dead Load State @ Bottom of Girder @ Level of Prestressing (e) Service Load State
Figure D.1. Stress distribution diagrams for a typical prestressed concrete bridge girder at various stages
of loading.
Equation D.5 can be simplified and rewritten as a quadratic Then Equation D.9 follows,
equation with unknown c, neutral axis depth, as follows in Equa-
tion D.6: Aps f se  1 e02  M D e0
ε ce = + − (D.9)
Ec  Ac I  Ec I
 Aps f ps Eps 
  f − E (ε + ε ) E + b0h f + bh f 1  where
2  ps ce  c
c2 +  ic Ac = area of concrete at the cross section considered;
ps se
bw  As Es  e0 = eccentricity of the prestressing force with respect to

 − bw h f − bw h f 1 +  the centroid of the section;
 Ec 
Ec = modulus of elasticity of concrete;
 Aps f ps Eps 1 2 Eps = modulus of elasticity of prestressing steel;
  f − E (ε + ε ) E − 2 (b − b0 ) h f  fse = effective stress in prestressing steel after losses;
2  ps ce  c
=0
− 
ps se
(D.6) I = moment of inertia; and
bw  AEd 
 + (b − bw ) (h f + h f 1) + s s s 
1 2
MD = dead load moment.
 2 Ec 
Considering an uncracked section under service loads and
plane section remains plane, the linear strain distribution
The moment resistance can be expressed as follows (Equa-
diagram is as follows in Figure D.2:
tion D.7):
From Figure D.2, the relationship between top and bottom
strain is as follows (Equations D.10 and D.11):
(c − hf ) (b0 − b) f  c + 2hf 
2
1
M n = Aps fps dp + As fs ds − f ctb0c 2 +
6 2c  3 
ct  dp − c
i εc = ε ps (D.10)
h−c
(c − h − h f 1) (b − bw )
2
f  c + 2h f + 2h f 1 
+ f ct   (D.7)
2c  3 dp − c Eps
i f cb = ∆f pt (D.11)
h−c Ec
where
Mn = nominal moment resistance. Compression
fs is calculated using Equation D.3, and fct is calculated using

Equation D.4.
The depth of neutral axis from the compression face, c, can
c
be computed from Equation D.6. dp h
Also, assuming a linear elastic relationship in the behavior
of the prestressing steel, we have Equation D.8,
Figure D.2. Strain
f distribution at
ε se = se (D.8)
Eps Tension service loads.

312
where The decompression moment at the level of prestressing

ec = strain in concrete at bottom fiber; strands, MDecp, can be calculated using Equation D.17:
Ec = elastic modulus of concrete;
Dfpt =
change in prestressing tendons stress between
 
decompression and the stress in concrete at the bot-  
MD e0 Eps
tom of the girder reaching fct assuming uncracked  f se − 
section; and   Aps Eps  1 e0   
2
 I + E  A + I   IEc 
fcb = concrete allowable tensile stress at the bottom of the M Decp =  c c   (D.17)
girder. e0 e0 Eps
−
According to the current AASHTO LRFD Bridge Design  1 e02   Aps Eps  1 e02  
I  +  Aps 1 + + IE
Specifications (2012), fcb = 0.19 fc′ or fcb = 0.0948 fc′, depend-  Ac I   Ec  Ac I   c
ing on the exposure conditions.
Then fps for the uncracked section should be as follows The decompression moment at the bottom fiber of
(Equation D.12): the concrete girder, M Decb, can be calculated using
f ps = ∆f pt + f ps (M Dec) (D.12) Equation D.18:
For the cracked section, the fps can be calculated by  

Equation D.13:  MD e0 Eps 
 f se − 
f ps = ∆f ps + f ps(M Dec) (D.13)   Aps Eps  1 e02   
 I + E  A + I   IEc 
M Decb =  c c  
where yb e0 Eps
fps(MDec) = the stress in prestressing steel at decompression; −
1 e y   Aps Eps  1 e02  
Dfps = the increase in the prestressing steel stress beyond I  + 0 b  Aps 1 + + IE
 Ac I   Ec  Ac I   c
the decompression state for cracked members; and
MDec = the decompression moment. (D.18)
Dfps in Equation D.13 can be calculated based on the equa-
tion of maximum crack width at the bottom of prestressed where
concrete girder. In this study, Equation D.14, developed by Ac = area of concrete at the cross section considered;
Nawy and Huang (1977), was used: Aps = area of prestressing steel in tension zone;
e0 = eccentricity of the prestressing force with respect
At
w max = 5.85 × 10−5
βΣ0
(∆fps ) (D.14) to the centroid of the section at supports;
yb = distance from centroidal axis to extreme bottom
where fiber;
S0 = sum of reinforcing element circumferences; Ec = modulus of elasticity of concrete;
At = area of concrete in tension; and Eps = modulus of elasticity of prestressing steel;
b = ratio of distance from neutral axis of beam to concrete fse = effective stress in prestressing steel after losses;
outside tension face to distance from neutral axis to I = moment of inertia;
steel reinforcement centroid. MD = dead load moment;
By rearranging Equation D.14, Dfps can be calculated using MDecb = decompression moment at the bottom of the
Equation D.15: girder; and
MDecp = decompression moment at the level of the pre-
w max i β i Σ0 stressing strands.
∆f ps = (D.15)
5.85 × 10−5 i At
Dfpt varies according to the maximum allowable tensile References
stress at the bottom of the concrete girder.
Moreover, fps(MDec) can be calculated using Equation D.16: AASHTO LRFD Bridge Design Specifications, 6th ed. 2012. American
Association of State Highway and Transportation Officials, Washing-
Eps [M Dec − M D ] ton, D.C.
f ps(M Dec) = f se + (D.16) Nawy, E., and P. Huang. 1977. Crack and Deflection Control of Pre-
 Aps Eps  1 e02  
IEc 1 + + tensioned Prestressed Beams. Journal, Precast/Prestressed Concrete
 Ec  Ac I   Institute, Vol. 23, No. 3, pp. 30–43.

Appendix E
Normal Probability Plots of Fatigue Data for the

Various Detail Categories
The data in the various steel-category plots (A through E′) 1967; Hanson et al. 1968; Helgason et al. 1976; Lash 1969;
are from Keating and Fisher 1986. The data on the plots for MacGregor et al. 1971; and Amorn et al. 2007. The data for
steel reinforcement in tension are from Fisher and Viest the plots of concrete in compression are from Hilsdorf and
1961; Pfister and Hognestad 1964; Burton and Hognestad Kesler 1966.
Figure E.1. Normal probability plot of Detail Category A

fatigue data.
313

314
Figure E.2. Normal probability plot of Detail Category A

truncated fatigue data with best fit line.
Figure E.3. Normal probability plot of Detail Category B fatigue data.

315
Figure E.4. Normal probability plot of Detail Category B truncated

fatigue data with best fit line.
Figure E.5. Normal probability plot of Detail Category B fatigue data.

316
Figure E.6. Normal probability plot of Detail Category B truncated

Figure E.7. Normal probability plot of Detail Categories C and C

fatigue data.

317
Figure E.8. Normal probability plot of Detail Categories C and C truncated

Figure E.9. Normal probability plot of Detail Category D fatigue data.

318
Figure E.10. Normal probability plot of Detail Category D truncated

Figure E.11. Normal probability plot of Detail Category E fatigue data.

319
Figure E.12. Normal probability plot of Detail Category E truncated

Figure E.13. Normal probability plot of Detail Category E

fatigue data.

320
Figure E.14. Normal probability plot of Detail Category E truncated

Figure E.15. Normal probability plot for concrete in compression.
Figure E.16. Normal probability plot for concrete in compression


321
Figure E.17. Normal probability plot for steel reinforcement in tension.
Figure E.18. Normal probability plot for steel reinforcement in tension

References Hilsdorf, H., and C. Kesler. 1966. Fatigue Strength of Concrete Under
Varying Flexural Stresses. Journal Proceedings, American Concrete
Amorn, W., J. Bowers, A. Girgis, and M. Tadros. 2007. Fatigue of Deformed Institute, Vol. 63, No. 10, pp. 1059–1076.
Welded-Wire Reinforcement. Journal, Precast/Prestressed Concrete Keating, P., and J. Fisher. 1986. NCHRP Report 286: Evaluation of Fatigue
Institute, Vol. 52, No. 1, pp. 106–120. Tests and Design Criteria on Welded Details. TRB, National Research
Burton, K., and E. Hognestad. 1967. Fatigue Test of Reinforcing Bars-Tack Council, Washington, D.C.
Welding of Stirrups. Journal, American Concrete Institute, Vol. 64, Lash, S. 1969. Can High-Strength Reinforcement Be Used in Highway
No. 5, pp. 244–252. Bridges? In First International Symposium on Concrete Bridge Design,
Fisher, J., and I. Viest. 1961. Special Report 66: Fatigue Tests of Bridge SP-23, American Concrete Institute, Detroit, Mich., pp. 283–300.
Materials of the AASHO Road Test. HRB, National Research Council, MacGregor, J., I. Jhamb, and N. Nuttall. 1971. Fatigue Strength of Hot-
Washington, D.C. Rolled Deformed Reinforcing Bars. Journal Proceedings, American
Hanson, J., K. Burton, and E. Hognestad. 1968. Fatigue Tests of Re Concrete Institute, Vol. 68, No. 3, pp. 169–179.
inforcing Bars: Effect of Deformation Pattern. Journal of the Portland Pfister, J., and E. Hognestad. 1964. High-Strength Bars As Concrete
Cement Association Research and Development Laboratories, Vol. 10, Reinforcement, Part 6: Fatigue Tests. Journal of the Portland Cement
No. 3, pp. 2–13. Association Research and Development Laboratories, Vol. 6, No. 1,
Helgason, T., J. Hanson, N. Somes, W. Corley, and E. Hognestad. 1976. pp. 65–84.
NCHRP Report 164: Fatigue Strength of High-Yield Reinforcing Bars.
TRB, National Research Council, Washington, D.C.

A ppe n d i x F
Data Used for Calibration
F.1 Purpose and Contents F.2 Monte Carlo Analysis Using

of this Appendix Microsoft Excel
The purpose of this appendix is to facilitate future enhance- Determination of the failure rate and the associated reliability
ment of service limit state (SLS) calibration as well as to allow index, b, through Monte Carlo simulation using Microsoft
for customization by individual bridge owners. To accom- (MS) Excel involves the following 15-step computational
plish this purpose, certain guidelines and databases have been procedure adapted from Nowak and Collins (2013):
included. (Section F.1 references are included in the Refer-
1. Determine the nominal dead load, Dn, the nominal live
ences, page 263.)
load plus impact, Ln, and the nominal resistance, Rn, for
the subject bridge according to the AASHTO LRFD
• Section F.2: a step-by-step description of the imple
Bridge Design Specifications (2012).
mentation of the Monte Carlo analysis introduced in 2. Assume i = 1.
Chapter 3. 3. Generate a uniformly distributed random number 0 ≤
• Section F.3: an excerpt from the database of WIM (weigh- uDi ≤ 1 using the command RAND.
in-motion) data used in the development of live load 4. Calculate the corresponding value of Di (a normal ran-
models presented in Chapter 5. These data are available dom variable). Equation F.2.1 follows:
at http://www.trb.org/Main/Blurbs/170201.aspx.
• Section F.4: a condensed version of the geotechnical cali- Di = µ D + σ D φ−1 (u Di ) (F.2.1)
bration reported in Chapter 6 with additional informa-
tion provided on the geotechnical properties of the test where f-1 is the inverse standard normal distribution
sites discussed in Chapter 6. It is anticipated that the function calculated using the command NORMSINV.
regional nature of geotechnical practice will result in Equations F.2.2 and F.2.3 follow as shown:
calibration of locally preferred methods of predicting µ D = λ D Dn (F.2.2)
settlement of shallow and deep foundations in lieu of the
Hough (1959) and Schmertmann et al. (1978) methods σD = VD µ D (F.2.3)
used herein.
• Sections F.5 and F.6: descriptions of the concrete and steel 5. Generate a uniformly distributed random number 0 ≤ uLi
bridges selected from the NCHRP Project 12-78 database ≤ 1 using the command RAND.
and used for various studies reported here (Mlynarski et al. 6. Calculate the corresponding value of Li (a normal ran-
2011). Other bridges could be substituted. dom variable). Equation F.2.4 follows:
• Section F.7: a database of fatigue tests on steel details
Li = µ L + σ L φ−1 (u Li ) (F.2.4)
(Keating and Fisher 1986; P. B. Keating, personal com-
munication, 2012). where f-1 is the inverse standard normal distribution
• Section F.8: a database of fatigue tests on plain concrete function calculated using the command NORMSINV.
and reinforcement (Ople and Hulsbos 1966; Fisher and Equations F.2.5 and F.2.6 follow as shown:
Viest 1961; Pfister and Hognestad 1964; Hanson et al. 1968;
Lash 1969). µ L = λ L Ln (F.2.5)
322

323
σL = VL µ L (F.2.6) 9. Calculate the limit state function, Yi = Ri - (Di + Li), and

save the value.
7. Generate a uniformly distributed random number 0 ≤ uRi 10. Assume i = i + 1; go back to Step 3 and iterate until the
≤ 1 using the command RAND. desired number of simulations, N, is obtained.
8. Calculate the corresponding value of Ri (a lognormal 11. Rearrange the values of Yi in ascending order using the
random variable). Equation F.2.7 follows. command RANK and reassign the values of i in ascending
order also.
12. Calculate the probabilities. Equation F.2.8 follows.
Ri = exp (µ lnR + σ lnR φ−1 (u Ri )) (F.2.7)
pi = i (1 + N ) (F.2.8)
where f-1 is the inverse standard normal distribution
function calculated using the command NORMSINV. 13. Calculate the corresponding values of the inverse stan-
dard normal distribution function, f-1(pi) using the
µ1nR = ln (µ R ) − 1 2 σ1nR2 command NORMSINV.
14. Plot f-1(pi) versus Yi; the resulting curve is the cumulative
σ1nR = (ln (VR2 + 1)) . distribution function of Y.
12
15. The reliability index, b, is equal to the negative value of

the plotted cumulative distribution function for Y = 0.
Table F.3.1. Filtered WIM Data Format
Entry Entry
Number Description Number Description
1 Direction 23 11th spacing between axles [ft]

2 Lane 24 12th spacing between axles [ft]
3 Year 25 13th spacing between axles [ft]
4 Day 26 1st axle weight [kip]
5 Hour 27 2nd axle weight [kip]
6 Min 28 3rd axle weight [kip]
7 Sec 29 4th axle weight [kip]
8 Total weight 30 5th axle weight [kip]
9 Total length 31 6th axle weight [kip]
10 Speed 32 7th axle weight [kip]
11 Vehicle class 33 8th axle weight [kip]
12 Number of axles 34 9th axle weight [kip]
13 1st spacing between axles [ft] 35 10th axle weight [kip]
14 2nd spacing between axles [ft] 36 11th axle weight [kip]
15 3rd spacing between axles [ft] 37 12th axle weight [kip]
16 4th spacing between axles [ft] 38 13th axle weight [kip]
17 5th spacing between axles [ft] 39 14th axle weight [kip]
18 6th spacing between axles [ft] 40 Ratio of truck moment to HL93
moment for 30-ft span

324
Reference for F2 9,1,2008,2,7,30,6,11.3,21,68,5,2,11.8,0,0,0,0,0,0,0,0,0,0,0,0,

4.9,6.4,0,0,0,0,0,0,0,0,0,0,0,0,0.1435,0.12857,0.11296,
AASHTO LRFD Bridge Design Specifications, 6th ed. 2012. American
Association of State Highway and Transportation Officials, Wash- 0.10204,0.082125
ington, D.C. 9,1,2008,2,10,25,54,10.8,22,68,5,2,13.4,0,0,0,0,0,0,0,0,0,0,0,
0,5.2,5.5,0,0,0,0,0,0,0,0,0,0,0,0,0.12131,0.11636,0.1039,
0.09454,0.076756
F.3 Excerpt from WIM Data 9,1,2008,2,10,29,31,11.4,32,72,5,2,17.5,0,0,0,0,0,0,0,0,0,0,0,
Following is a sample of the filtered WIM data used in the 0,5,6.4,0,0,0,0,0,0,0,0,0,0,0,0,0.11595,0.11838,0.10749,
development of live load models presented in Chapter 5. In 0.098578,0.080754
addition to entries copied from the raw WIM records (Entry 9,1,2008,2,10,37,31,10.1,22,72,5,2,11.8,0,0,0,0,0,0,0,0,0,0,0,
Numbers 1 through 39 in Table F.3.1), the last five entries 0,4.2,5.9,0,0,0,0,0,0,0,0,0,0,0,0,0.13057,0.11576,0.10144,
(Entry Numbers 40 through 44) give the ratio between the 0.091485,0.073538
moment produced by the truck divided by the HL-93 moment 9,1,2008,2,11,11,58,10.4,25,70,5,2,14,0,0,0,0,0,0,0,0,0,0,0,0,
for spans 30, 60, 90, 120, and 200 ft long. 4.7,5.6,0,0,0,0,0,0,0,0,0,0,0,0,0.11947,0.11282,0.10043,
0.091266,0.074002
Sample of Filtered WIM Data 9,1,2008,2,11,13,37,11.9,30,58,5,2,15.7,0,0,0,0,0,0,0,0,0,0,0,
0,5.6,6.3,0,0,0,0,0,0,0,0,0,0,0,0,0.12381,0.12489,0.113,
Each line of raw data goes from Entry 1 through Entry 44. 0.10344,0.084559
9,1,2008,2,11,42,16,11.4,27,68,5,2,12.6,0,0,0,0,0,0,0,0,0,0,0,
9,1,2008,1,1,34,52,11.7,28,52,5,2,14.7,0,0,0,0,0,0,0,0,0,0,0,0, 0,4.6,6.8,0,0,0,0,0,0,0,0,0,0,0,0,0.14485,0.12974,0.11407,
4,7.7,0,0,0,0,0,0,0,0,0,0,0,0,0.14847,0.13308,0.11692, 0.10301,0.082877
0.10561,0.085011 9,1,2008,2,11,46,34,10.3,23,61,5,2,13,0,0,0,0,0,0,0,0,0,0,0,
9,1,2008,1,6,8,4,11.6,26,63,5,2,14.9,0,0,0,0,0,0,0,0,0,0,0,0, 0,4,6.3,0,0,0,0,0,0,0,0,0,0,0,0,0.1338,0.11829,0.10358,
4.9,6.7,0,0,0,0,0,0,0,0,0,0,0,0,0.13449,0.12703,0.11309, 0.093389,0.075018
0.10278,0.083337 9,1,2008,2,13,33,29,10.4,25,65,5,2,14.6,0,0,0,0,0,0,0,0,0,0,0,
9,1,2008,1,12,58,28,10.6,25,65,5,2,14.6,0,0,0,0,0,0,0,0,0,0,0, 0,5.3,5.1,0,0,0,0,0,0,0,0,0,0,0,0,0.11326,0.10983,0.099237,
0,5.5,5,0,0,0,0,0,0,0,0,0,0,0,0,0.11635,0.11172,0.10068, 0.090756,0.074085
0.091959,0.074956 9,1,2008,2,14,46,51,10.5,26,65,5,2,13.8,0,0,0,0,0,0,0,0,0,0,0,
9,1,2008,1,14,24,20,11.2,26,61,5,2,12.8,0,0,0,0,0,0,0,0,0,0,0,
0,5,5.5,0,0,0,0,0,0,0,0,0,0,0,0,0.11809,0.11355,0.10165,
0,4.5,6.7,0,0,0,0,0,0,0,0,0,0,0,0,0.14191,0.12732,0.11188,
0.092585,0.075241
0.10104,0.081333
9,1,2008,2,14,58,50,11.2,25,58,5,2,14.6,0,0,0,0,0,0,0,0,0,0,0,
9,1,2008,1,18,43,59,10.6,24,61,5,2,13.5,0,0,0,0,0,0,0,0,0,0,0,
0,4.8,6.5,0,0,0,0,0,0,0,0,0,0,0,0,0.13022,0.12372,0.11025,
0,5.3,5.3,0,0,0,0,0,0,0,0,0,0,0,0,0.11943,0.11426,0.10236,
0.10021,0.081243
0.093277,0.075859
9,1,2008,2,15,41,47,11.8,27,71,5,2,13.4,0,0,0,0,0,0,0,0,0,0,0,
9,1,2008,1,22,19,18,11.7,25,57,5,2,11.3,0,0,0,0,0,0,0,0,0,0,0,
0,6.1,5.8,0,0,0,0,0,0,0,0,0,0,0,0,0.13647,0.12883,0.11533,
0,4.4,7.3,0,0,0,0,0,0,0,0,0,0,0,0,0.16047,0.13746,0.11951,
0.10734,0.085825 0.10503,0.085331
9,1,2008,2,0,24,39,10.4,25,63,5,2,11.5,0,0,0,0,0,0,0,0,0,0,0, 9,1,2008,2,23,31,12,10.3,23,70,5,2,11.3,0,0,0,0,0,0,0,0,0,0,0,
0,4.8,5.6,0,0,0,0,0,0,0,0,0,0,0,0,0.1308,0.11787,0.10376, 0,4.4,5.9,0,0,0,0,0,0,0,0,0,0,0,0,0.1352,0.1188,0.10392,
0.093758,0.075501 0.093631,0.075142
9,1,2008,2,3,59,48,10.3,25,66,5,2,13.3,0,0,0,0,0,0,0,0,0,0,0,0, 9,1,2008,3,0,20,40,11.3,26,64,5,2,11.5,0,0,0,0,0,0,0,0,0,0,0,
4.5,5.8,0,0,0,0,0,0,0,0,0,0,0,0,0.12427,0.11481,0.10162, 0,4.9,6.4,0,0,0,0,0,0,0,0,0,0,0,0,0.14567,0.12936,0.11349,
0.092083,0.074401 0.10238,0.082278
9,1,2008,2,4,50,41,11.5,25,65,5,2,11.1,0,0,0,0,0,0,0,0,0,0,0,0, 9,1,2008,3,5,56,24,11.8,26,70,5,2,11.4,0,0,0,0,0,0,0,0,0,0,0,
5.1,6.4,0,0,0,0,0,0,0,0,0,0,0,0,0.15061,0.13252,0.11593, 0,5.7,6,0,0,0,0,0,0,0,0,0,0,0,0,0.14463,0.13168,0.11618,
0.10446,0.083851 0.1051,0.084756
9,1,2008,2,6,9,26,25.1,52,65,9,5,16.8,4.2,24,9,0,0,0,0,0,0,0, 9,1,2008,3,7,10,36,10.9,26,67,5,2,14.9,0,0,0,0,0,0,0,0,0,0,0,
0,0,9,2.5,3.7,5.3,4.7,0,0,0,0,0,0,0,0,0,0.16931,0.15654, 0,4.8,6.1,0,0,0,0,0,0,0,0,0,0,0,0,0.12369,0.11821,0.1056,
0.16411,0.16992,0.15632,2 0.096125,0.078094
9,1,2008,2,6,44,32,10.3,25,68,5,2,13.3,0,0,0,0,0,0,0,0,0,0,0, 9,1,2008,3,9,0,56,11.5,29,71,5,2,14.9,0,0,0,0,0,0,0,0,0,0,0,0,
0,4.3,5.9,0,0,0,0,0,0,0,0,0,0,0,0,0.1252,0.11448,0.10109, 3.8,7.7,0,0,0,0,0,0,0,0,0,0,0,0,0.14971,0.13219,0.11572,
0.091493,0.073824 0.10431,0.083779

325
9,1,2008,3,9,12,48,10.2,24,67,5,2,14,0,0,0,0,0,0,0,0,0,0,0,0, 9,1,2008,4,15,39,32,11.7,25,78,5,2,13.7,0,0,0,0,0,0,0,0,0,0,0,
5.5,4.7,0,0,0,0,0,0,0,0,0,0,0,0,0.11773,0.11079,0.099058, 0,4.3,7.4,0,0,0,0,0,0,0,0,0,0,0,0,0.14835,0.13304,0.11697,
0.09016,0.073203 0.10566,0.085035
9,1,2008,3,11,10,23,11.5,27,68,5,2,15.6,0,0,0,0,0,0,0,0,0,0,0, 9,1,2008,4,15,50,43,10.4,25,76,5,2,13.1,0,0,0,0,0,0,0,0,0,0,0,
0,5.2,6.3,0,0,0,0,0,0,0,0,0,0,0,0,0.12322,0.1222,0.11006, 0,4.7,5.7,0,0,0,0,0,0,0,0,0,0,0,0,0.12495,0.11573,0.10247,
0.10054,0.081989 0.092873,0.075068
9,1,2008,3,14,29,59,11.2,27,71,5,2,13.3,0,0,0,0,0,0,0,0,0,0,0, 9,1,2008,4,16,18,35,11.4,28,66,5,2,13.5,0,0,0,0,0,0,0,0,0,0,0,
0,3.8,7.4,0,0,0,0,0,0,0,0,0,0,0,0,0.15011,0.13031,0.11366, 0,6.2,5.3,0,0,0,0,0,0,0,0,0,0,0,0,0.13515,0.12608,0.11233,
0.10226,0.081922 0.10208,0.082733
9,1,2008,3,14,30,45,11.6,32,72,5,2,16.4,0,0,0,0,0,0,0,0,0,0,0, 9,1,2008,4,17,4,44,10.8,24,71,5,2,12.9,0,0,0,0,0,0,0,0,0,0,0,
0,4.2,7.4,0,0,0,0,0,0,0,0,0,0,0,0,0.1371,0.12826,0.11387, 0,3.9,6.9,0,0,0,0,0,0,0,0,0,0,0,0,0.14371,0.12528,0.10933,
0.10333,0.083614
0.098404,0.07889
9,1,2008,3,16,44,3,10.9,27,65,5,2,13.4,0,0,0,0,0,0,0,0,0,0,0,
9,1,2008,4,17,41,43,10.9,26,68,5,2,13.9,0,0,0,0,0,0,0,0,0,0,0,
0,5.8,5,0,0,0,0,0,0,0,0,0,0,0,0,0.12711,0.11849,0.10554,
0,4.4,6.4,0,0,0,0,0,0,0,0,0,0,0,0,0.13191,0.12097,0.10685,
0.095896,0.077711
0.096735,0.078092
9,1,2008,3,16,50,40,11.1,28,65,5,2,13.5,0,0,0,0,0,0,0,0,0,0,0,
9,1,2008,4,17,45,52,10.7,30,57,5,2,13.4,0,0,0,0,0,0,0,0,0,0,0,
0,5.6,5.5,0,0,0,0,0,0,0,0,0,0,0,0,0.12568,0.1194,0.10715,
0.097687,0.079461 0,4.8,5.9,0,0,0,0,0,0,0,0,0,0,0,0,0.12654,0.11834,0.10504,
9,1,2008,3,17,1,0,10.9,25,61,5,2,11.2,0,0,0,0,0,0,0,0,0,0,0,0, 0.095312,0.077126
4.5,6.4,0,0,0,0,0,0,0,0,0,0,0,0,0.14579,0.12671,0.11053, 9,1,2008,4,18,47,2,12.4,9,70,5,2,10.4,0,0,0,0,0,0,0,0,0,0,0,0,
0.099454,0.079693 7.5,4.9,0,0,0,0,0,0,0,0,0,0,0,0,0.17374,0.14753,0.12755,
9,1,2008,4,0,50,45,11.3,26,71,5,2,11.6,0,0,0,0,0,0,0,0,0,0,0, 0.11431,0.091195,2
0,4.6,6.7,0,0,0,0,0,0,0,0,0,0,0,0,0.14809,0.13025,0.11402, 9,1,2008,4,20,37,10,10.5,25,80,5,2,11.6,0,0,0,0,0,0,0,0,0,0,0,
0.10274,0.082455 0,4.2,6.4,0,0,0,0,0,0,0,0,0,0,0,0,0.14019,0.12264,0.10723,
9,1,2008,4,5,15,17,11.1,23,75,5,2,11.9,0,0,0,0,0,0,0,0,0,0,0, 0.096559,0.077435
0,4.8,6.3,0,0,0,0,0,0,0,0,0,0,0,0,0.14182,0.12661,0.11113, 9,1,2008,4,20,41,14,11.2,27,71,5,2,13.3,0,0,0,0,0,0,0,0,0,0,0,
0.10031,0.080683 0,4.9,6.3,0,0,0,0,0,0,0,0,0,0,0,0,0.13503,0.12481,0.11048,
9,1,2008,4,6,4,0,11,26,65,5,2,11.2,0,0,0,0,0,0,0,0,0,0,0,0, 0.10012,0.080895
4.1,6.8,0,0,0,0,0,0,0,0,0,0,0,0,0.15046,0.12841,0.11152,
0.10012,0.08001
9,1,2008,4,6,58,19,11.5,24,65,5,2,11.5,0,0,0,0,0,0,0,0,0,0,0, F.4 Procedure for Calibration
0,3.8,7.7,0,0,0,0,0,0,0,0,0,0,0,0,0.16159,0.13652,0.11832, of Service I Limit State for
0.10609,0.084648 Geotechnical Features
9,1,2008,4,8,41,26,11.1,26,63,5,2,14.8,0,0,0,0,0,0,0,0,0,0,0,
0,4.4,6.6,0,0,0,0,0,0,0,0,0,0,0,0,0.13031,0.12165,0.10793, Information is presented in this section as an aid for an owner
0.097923,0.079249 to perform a calibration of the Service I limit state for geo-
9,1,2008,4,9,25,42,10.8,24,68,5,2,13.4,0,0,0,0,0,0,0,0,0,0,0, technical features by using an analytical method to predict
0,4.7,6.1,0,0,0,0,0,0,0,0,0,0,0,0,0.12968,0.12016,0.10644, (estimate) deformation based on local geologic conditions. A
0.096482,0.077981 step-by-step format is provided with the intention that end
9,1,2008,4,9,41,50,11.2,24,67,5,2,13.4,0,0,0,0,0,0,0,0,0,0,0, users can simply substitute the appropriate data for the
0,4.8,6.4,0,0,0,0,0,0,0,0,0,0,0,0,0.13549,0.12497,0.1106, method and the mode of foundation deformation that they
0.1002,0.080938 are trying to calibrate. The theoretical background is inten-
9,1,2008,4,14,8,35,11.7,35,66,5,2,18.1,0,0,0,0,0,0,0,0,0,0,0, tionally omitted in favor of providing specific computational
0,4.3,7.4,0,0,0,0,0,0,0,0,0,0,0,0,0.13209,0.12671,0.11327, guidance.
0.10313,0.083806 In general, the vertical and lateral deformations for all
9,1,2008,4,14,8,39,11.5,28,70,5,2,14.2,0,0,0,0,0,0,0,0,0,0,0, structural foundation types such as footings, drilled shafts,
0,4.3,7.3,0,0,0,0,0,0,0,0,0,0,0,0,0.14684,0.13181,0.11588, and driven piles can be calibrated by using the process
0.10467,0.084249 described in this section. For the purpose of demonstration
9,1,2008,4,14,20,33,10.5,28,67,5,2,15.9,0,0,0,0,0,0,0,0,0,0,0, of the calibration process, the case of immediate vertical set-
0,4.4,6.1,0,0,0,0,0,0,0,0,0,0,0,0,0.11818,0.11345,0.10148, tlement of spread footings is used.
0.092434,0.07515 For convenience, reference is made to the widely used com-
9,1,2008,4,14,32,53,10.2,26,64,5,2,13.3,0,0,0,0,0,0,0,0,0,0,0, mercial software Microsoft Excel (references to Microsoft
0,3.8,6.3,0,0,0,0,0,0,0,0,0,0,0,0,0.13025,0.11564,0.10142, Excel are applicable to its 2007 and 2010 versions). This has
0.091488,0.073528 been done to help simplify the calibration process without

326
complicating the process with esoteric probabilistic princi- the case in which the data points do not align closely with
ples which in the end lead to the same result. the 1:1 diagonal line. In the geotechnical literature (e.g., Tan
and Duncan 1991), accuracy is defined as the mean value of
the ratio of the predicted (calculated) to the measured set-
F.4.1 Step 1: Express the Service Limit State
tlements. Table F.4.3 shows the values of accuracy, denoted
in Terms of Load and Resistance
by X, where X = dP/dM, for each footing based on the data in
In the context of deformations, tolerable deformations, dT, Table F.4.2.
are considered as resistances while the predicted deforma- As noted in Step 1 of the calibration process, the value of dM
tions, dP, are considered as loads. For service limit state cali- can be considered to be the resistance and equivalent to the
bration for foundation deformation, the limit state, g, is tolerable settlement, dT. The accuracy, X = dP/dM (or dP/dT), is
expressed as follows (Equation F.4.1): a random variable that can be modeled by an appropriate
probability distribution function (PDF). The data for X in
g = δ P δT (F.4.1) Table F.4.3 were used to develop a histogram of accuracy (X)
values as shown in Figure F.4.2.
For the example of immediate settlements, the load and The arithmetic mean (M) and standard deviation (SD) of
the resistance parameters are as follows: the data in Table F.4.3 are 1.381 and 1.006, respectively, and are
noted in Column 2 of Table F.4.3 and Table F.4.4. The histo-
• Load: Predicted (estimated or calculated) immediate verti- gram of the data in Column 2 of Table F.4.3 is shown in Fig-
cal settlement, dP ure F.4.2. The histogram does not resemble the classical bell
• Resistance: Tolerable (limiting or measured) immediate shape that is characteristic of normally distributed data. To
vertical settlement, dT evaluate the deviation of the data from a classical normal PDF,
the data for the value of accuracy (X) in Table F.4.3 were
F.4.2 Step 2: Develop Statistical Model plotted against the standard normal variable, z, to generate a
for Load (dP) cumulative distribution function (CDF) as shown in Fig-
ure F.4.3 [see Chapter 5 in Allen et al. (2005) for the definition
The usefulness of this appendix will be demonstrated by of standard normal variable, z, and procedures to develop Fig-
using the analytical method proposed by Schmertmann et al. ure F.4.3]. As noted in Allen et al. (2005), the benefit of plot-
(1978). Tables F.4.1 and F.4.2 show a data set for spread foot- ting the data in this manner on a CDF plot is that normally
ings based on vertical settlements of footings measured at distributed data plot as a straight line with a slope equal to
20 footings for 10 instrumented bridges in the northeastern 1/SD, and the intercept on the horizontal axis corresponding
United States (Gifford et al. 1987). More detailed subsurface to z = 0 is equal to the mean value M. As can be seen in Figure
data are given in that report. Each of the footing designations F.4.3, the data points based on Table F.4.3 do not plot on the
in the tables represents a footing supporting a single sub- straight line, which confirms the observation made based on
structure unit (abutment or pier). Four of the instrumented the histogram in Figure F.4.2.
bridges were single-span structures. Two 2-span and three By using procedures described in Allen et al. (2005), a
4-span bridges were also monitored in addition to a single lognormal distribution is used to evaluate the nonnormal
5-span structure. Nine of the structures were designed to data. As seen in Figure F.4.3, the lognormal distribution fits
carry highway traffic, while the one instrumented bridge con- the data better than the normal distribution. The lognor-
sisted of a 4-span railroad bridge across an Interstate high- mal distribution, which is valid between values of 0 and +∞,
way. The bridges included 5 simple-span and 5 continuous- is used in Figure F.4.3 because (a) immediate settlement
beam structures. Additional information on the instrumen- cannot have negative values, and (b) lognormal PDFs have
tation and data collection at the 10 bridges can be found in been used in the past for nonnormal distributions during
Gifford et al. (1987). calibration of the strength limit state for geotechnical as
Figure F.4.1 shows a plot of the data in Table F.4.2 and the well as structural features in the AASHTO LRFD frame-
spread of the data about the 1:1 diagonal line, which defines work (2012). For service limit state, a PDF with an upper
the case in which the predicted and measured values are bound and lower bound (e.g., beta distribution) instead of
equal. Such a plot provides a visual frame of reference to open tail(s) (e.g., normal or lognormal distribution) may
judge the accuracy of the prediction method, in this case be more appropriate since the conditions represented by an
Schmertmann’s method. For example, if the data points open tail PDF are not physically possible when one consid-
align closely with the 1:1 diagonal line, then the predictions ers foundation deformations. As noted, the lognormal PDF
based on the analytical method being evaluated are close to is used here simply to be consistent with the PDFs that have
the measured values and are more accurate compared with been used in the LRFD calibration processes to date.

327
Table F.4.1. Study Bridge Data
Element q Nf Nc Water s–
vo
Designation (ksf) (blows/ft) (blows/ft) g (kcf) Level (ft) B (ft) L (ft) D (ft) qc (kg/cm2) H (ft) max
S1 3.20 23 to 36 44 0.120 12 17.0 63.7 F NA 117 2(F)

S2 2.67 60 58 0.120 12 17.0 63.7 F NA 117 2(F)
S3 2.32 32 to 44 43 0.120 11.5 15.25 52.5 F NA 35 2(F)
S4 2.44 18 to 24 19 0.120 9 16.75 52.5 4 NA 42 3
S5 1.88 12 to 13 12 0.120 4 12.50 41.0 5 NA 40 3
S6 1.70 18 to 20 34 0.120 31 11.0 74.6 F 28,61,90,125 155 1(F)
S7 2.34 22 22 0.115 12 18.5 79.0 5 NA 130 1
S8 2.10 18 to 19 18 0.120 6 21.0 21.0 5 NA 150 1
S9 1.50 18 to 19 18 0.120 6 21.0 30.4 5 NA 150 1
S10 2.34 16 to 17 20 0.115 12 16.0 26.8 5 90,70,88 153 1
S11 2.48 18 to 22 22 0.115 12 16.0 18.5 5 74 155 1
S12 1.48 13 to 14 15 0.115 12 21.0 33.0 5 NA 120 1
S13 1.60 23 to 28 25 0.115 12 21.0 30.0 5 NA 120 1
S14 3.30 21 21 0.120 28 8.1 42.9 F 165 197 2(F)
S15 3.43 8 8 0.120 26.5 8.1 42.9 F 53 >150 2(F)
S16 2.40 34 to 37 42 0.120 10 16.75 76.9 6 NA 52 2
S17 2.34 21 to 27 24 0.125 8 15.25 76.1 6.5 NA 51 2
S18 1.88 37 to 53 55 0.120 2 15.25 61.7 9 NA 10 2
S19 1.79 25 to 34 39 0.120 2 15.25 67.3 9 NA 10 2
S20 2.14 19 to 22 24 0.113, 0.115 44 28.0 28.0 0 62,131 >2B 2
S21 3.01 25 to 26 23 0.115 0 20.0 100.8 22 NA >2B 3
S22 3.25 26 to 31 38 0.115 1 20.0 100.8 5 NA >2B 3
S23 3.51 33 to 40 39 0.115, 0.120 17 21.75 44.4 F NA 41 2(F)
S24 3.37 37 to 38 49 0.115 13 16.0 44.7 0 114,183 48 2
Notes:
q = footing bearing pressure (average)
Nf = field standard penetration test N-value (range of N-values is due to different depths of influence for different settlement calculation methods)
Nc = corrected N-value (corrected for overburden per Peck and Bazaraa 1969)
g = soil total unit weight (assumed) (kips per cubic foot)
Water Level = depth of water table (below footing bearing elevation)
B = footing width
L = footing length
D = depth of footing embedment below ground surface; (F) indicates footing is on new fill
qc = static cone penetration test cone resistance (multiple values indicate that profile was subdivided into layers with corresponding qc values)
H = depth below footing to (relatively) incompressible stratum (H > 2B indicates an incompressible stratum is located below the depth of influence)
_
σvo max = soil stress history: 1 = preloaded; 2 = normally loaded; 3 = partially preloaded
NA = not available
Source: Gifford et al. (1987).

328
Table F.4.2. Data for Measured and Table F.4.3. Values and Statistics
Calculated Settlements of Accuracy (X 5 dP/dM) and Natural
Log of X Values [ln(X)] Based on
Settlement (in.) Data Shown in Table F.4.2
Footing
Designation Measured (dM) Calculated (dP)
Footing Designation X ln(X)
S1 0.35 0.79
S1 2.257 0.814
S2 0.67 1.85
S2 2.761 1.016
S3 0.94 0.86
S3 0.915 -0.089
S4 0.76 0.46
S4 0.605 -0.502
S5 0.61 0.30
S5 0.492 -0.710
S6 0.42 0.52
S6 1.238 0.214
S7 0.61 0.18
S7 0.295 -1.221
S8 0.28 0.30
S8 1.071 0.069
S9 0.26 0.18
S9 0.692 -0.368
S10 0.29 0.29
S10 1.000 0.000
S11 0.25 0.36
S11 1.440 0.365
S14 0.46 0.41
S14 0.891 -0.115
S15 0.34 1.57
S15 4.618 1.530
S16 0.23 0.26
S16 1.130 0.123
S17 0.44 0.40
S17 0.909 -0.095
S20 0.64 1.21
S20 1.891 0.637
S21 0.46 0.29
S21 0.630 -0.461
S22 0.66 0.54
S22 0.818 -0.201
S23 0.61 1.02
S23 1.672 0.514
S24 0.28 0.64
S24 2.286 0.827
Note: Gifford et al. (1987), the source for the table, notes that
data for Footings S12, S13, and S18 were not included Mean 1.381 0.1173
because construction problems at those sites resulted in Standard deviation 1.006 0.6479
disturbance of the subgrade soils and short term settlement
was increased. Data for Footing S19 appear to be anomalous Coefficient of variation 0.728
and have been excluded from this table and from Figure F.4.1.
9
8
7
6
5
4
3
2
1
0
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Figure F.4.1. Comparison of measured
and calculated settlements based on
service load data in Table F.4.2 for Figure F.4.2. Histogram for accuracy (X) of
Schmertmann’s method. Schmertmann et al. (1978) method.

329
Table F.4.4. Statistics for Accuracy Based on arithmetic values of the mean and standard deviation for the
Normal and Lognormal Distributions lognormal distribution are shown in Table F.4.4. The corre-
lated and the arithmetic values of mean (MLNC and MLNA) and
Normal Lognormal Lognormal
standard deviation (SDLNC and SDLNA) for lognormal distri-
Statistic Arithmetic Correlated Arithmetic butions are similar but not equal. This is because the corre-
lated values (MLNC and SDLNC) are based on derivations for an
Mean M = 1.381 MLNC = 0.1100 MLNA = 0.1173
idealized lognormal distribution and not a sample distribu-
Standard SD = 1.006 SDLNC = 0.6525 SDLNA = 0.6479
tion from actual data which may not necessarily fit an ideal-
deviation
ized lognormal distribution. In contrast, the arithmetic values
Coefficient of CV = 0.728
(MLNA and SDLNA) are obtained by taking the arithmetic mean
variation
and standard deviation directly from the ln(X) value of each
Notes:
1. The correlated mean (MLNC) and standard deviation (SDLNC) values for log-
data point noted in Column 3 of Table F.4.3.
normal distribution were calculated from the normal (arithmetic) mean (M) It is important to recognize the use of appropriate values
and standard deviation (SD) values of 1.381 and 1.006, respectively, by using of mean and standard deviation based on the syntax for a
the following equations based on idealized normal and lognormal PDFs:
lognormal distribution function used by a particular compu-
MLNC = LN(M) - 0.50(SDLNC)2 tational program. For example, if one is using the @RISK
SDLNC = [LN{(SD/M)2 + 1}]0.5 program by Palisade Corporation, then the RISKLOGNORM
2. The arithmetic mean (MLNA) and standard deviation (SDLNA) values of 0.1173 function in that program is based on arithmetic values
and 0.6479, respectively, were calculated for the lognormal distribution (M and SD) of the normal distribution. In contrast, the
directly from the ln(X) values shown in Column 3 of Table F.4.3.
Microsoft Excel LOGNORMDIST (or LOGNORM.DIST)
function uses the arithmetic mean (MLNA) and standard
deviation (SDLNA) values of ln(X). Use of improper values
Guidance for the selection of an appropriate PDF and of mean and standard deviation can lead to drastically dif-
development of the distribution parameters shown in Table ferent results. This issue is of critical importance since the
F.4.4 is provided in Nowak and Collins (2000) and other presented process is based on use of Microsoft Excel as
similar books that deal with the subject of probabilistic mentioned earlier.
methods. Figure F.4.4 and Figure F.4.5, respectively, show the PDF
Table F.4.4 presents the values of the mean and standard and CDF based on use of the LOGNORM.DIST function in
deviation based on the normal (straight line) and lognormal the 2010 version of Microsoft Excel with MLNA = 0.1173 and
(curved line) PDFs shown in Figure F.4.3. Correlated and SDLNA = 0.6479.
Accuracy Data for Schmertmann et al. (1978) - Full Data Fit

4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Data Points Predicted LN Fitted from Normal statistics Predicted Normal Distribution
Figure F.4.3. Standard normal variable, z, as a function of the

accuracy (X) of Schmertmann et al. (1978) method.

330
Service I limit state the load factor gSE, which represents the
uncertainty in estimated settlement, is only one of the many
load factors. Furthermore, the primary moments due to the
dead and live loads are much larger than the additional (sec-
ondary) moments due to settlement. Based on a consideration
of reversible and irreversible service limit states for bridge
superstructures, a target reliability index, bT, in the range of
0.50 to 1.00 for calibration of load factor gSE for foundation
deformation in the Service I limit state is recommended. For
demonstration purposes, a value of bT = 0.50 is used in the
included example. Using the procedure demonstrated here, the
Figure F.4.4. Probability distribution function end user can develop the appropriate values of gSE for other
for accuracy (X) of Schmertmann et al. (1978) values of bT. Additional discussion on the meaning and use of
method. gSE is presented in Chapter 6.
F.4.4 Step 4: Compute the Probability of

F.4.3 Step 3: Select the Target Exceedance, Pe, of Predicted Value
Reliability Index, bT for bT
For strength limit states, reliability index values in the range The reliability index, b, can be expressed in terms of probabil-
of 3.09 to 3.54 are used. Ultimate or strength limit states per- ity of exceedance, Pe, of a predicted value by using the follow-
tain to structural safety and the loss of load-carrying capacity. ing formula in Microsoft Excel (Equation F.4.2).
In contrast, service limit states are user-defined limiting β = NORMSINV (1 − Pe ) (F.4.2)
conditions that affect the function of the structure under
expected service conditions. Violation of service limit states Table F.4.5 and Figure F.4.6 were generated by using Equa-
occurs at loads much smaller than those for strength limit tion F.4.2. In Table F.4.5, linear interpolation may be used as
states. Since there is no danger of collapse if a service limit an approximation for values intermediate to those shown.
state is violated, a smaller value of target reliability index
may be used for service limit states. In the case of settle-
ment, the structural load effect is manifested in terms of
increased moments and potential cracking. The load effect Table F.4.5. Reliability Index, b, and Corresponding
due to the settlement relative to the load effect due to dead Values of Probability of Exceedance, Pe, Based
and live loads would generally be small because in the on Normally Distributed Data
Pe Pe Pe Pe
b (%) b (%) b (%) b (%)
2.00 2.28 1.50 6.68 1.00 15.87 0.50 30.85

1.95 2.56 1.45 7.35 0.95 17.11 0.45 32.64
1.90 2.87 1.40 8.08 0.90 18.41 0.40 34.46
1.85 3.22 1.35 8.85 0.85 19.77 0.35 36.32
1.80 3.59 1.30 9.68 0.80 21.19 0.30 38.21
1.75 4.01 1.25 10.56 0.75 22.66 0.25 40.13
1.70 4.46 1.20 11.51 0.70 24.20 0.20 42.07
1.65 4.95 1.15 12.51 0.65 25.78 0.15 44.04
1.60 5.48 1.10 13.57 0.60 27.43 0.10 46.02
1.55 6.06 1.05 14.69 0.55 29.12 0.05 48.01
Figure F.4.5. Cumulative distribution function
(CDF) for accuracy (X) of Schmertmann et al. 0.00 50.00
(1978) method. Note: Probability of Success, Ps = 1 - Pe.

331
Reliability Index, β As per the AASHTO LRFD framework, the load factor is
1%
0.0 0.5 1.0 1.5 2.0
1 in 100 rounded up to the nearest 0.05. Thus, use gSE = 1.25.
F.4.6 Summary
10% 1 in 10
The presented information describes the process for develop-
ing the load factor gSE which incorporates the uncertainty in
the model that is used for predicting foundation deforma-
tions and its effect on the bridge superstructure. The key to
successful calibration is the development and maintenance of
100% 1 in 1
a quality database of deformation measurements on founda-
Figure F.4.6. Relationship between b and Pe tions for transportation bridge structures. Using the process
for case of single load and single resistance. described, an owner can develop the load factor of gSE for any
foundation type (e.g., spread footing, driven piles, drilled
shafts), for any mode of deformation (e.g., vertical, lateral, rota-
tion) and for any appropriate analytical method based on local
Based on either Equation F.4.2 or Table F.4.5, a value of Pe ≈
geologic conditions—as long as the owner has established a
0.3085 or 30.85% is obtained for b = 0.50.
quality database of measurements of foundation deformations
in the local geologic conditions. Finally, care must be taken to
F.4.5 Step 5: Compute the Load Factor, gSE, use appropriate load types in the calibration process. For exam-
due to Foundation Deformations ple, as per Article 10.6.2.3 of AASHTO (2012), for immediate
settlements both permanent and transient loads are used, while
The load factor, gSE, is a function of the probability of exceed-
for long-term settlements only the permanent loads are used
ance, Pe, of the foundation deformation under consideration,
since transient loads are not there long enough to affect the
which in this example is the immediate settlement of spread
long-term settlements. The consideration of load types will be
footings calculated by using the analytical method of
based on the type of the foundation and the foundation defor-
Schmertmann et al. (1978). Equation F.4.3 is the formula in
mation being evaluated in the calibration process.
Microsoft Excel for determining a value of accuracy, X, in
terms of Pe, the arithmetic mean (MLNA), and the standard
deviation (SDLNA) of the lognormal distribution function as References for F.4
computed in Step 2. The value of X represents the probability
of the accuracy value (dP/dT) being less than a specified value. AASHTO LRFD Bridge Design Specifications, 6th ed. 2012. American
Pe = LOGNORMDIST (X , M LNA ,SDLNA ) (F.4.3) ington, D.C.
Allen, T., A. Nowak, and R. Bathurst. 2005. Transportation Research Circu-
From Table F.4.4, for the Schmertmann method, MLNA = lar E-C079: Calibration to Determine Load and Resistance Factors for
Geotechnical and Structural Design. Transportation Research Board of
0.1173 and SDLNA = 0.6479. The goal is to determine the value the National Academies, Washington, D.C. http://onlinepubs.trb.org/
of X that gives Pe = 0.3085. Thus, for this example, the expres- onlinepubs/circulars/ec079.pdf.
sion for Pe can be written as follows: Gifford, D., S. Kraemer, J. Wheeler, and A. McKown. 1987. Spread Foot-
ings for Highway Bridges. FHWA/RD-86-185. Haley and Aldrich,
Pe = LOGNORMDIST (X ,0.1173,0.6479) = 0.3085 or 30.85% FHWA, Cambridge, Mass.
Nowak, A. S., and K. C. Collins. 2000. Reliability of Structures. University
Using Goal Seek in Microsoft Excel, X(= dP/dT) ≈ 0.813. [Note of Michigan, Ann Arbor, and McGraw-Hill, New York.
that in the 2010 version of Microsoft Excel, another function Nowak, A., and K. Collins. 2013. Reliability of Structures. McGraw-Hill,
LOGNORM.DIST is also available that can be used. In this case, New York.
Peck, R., and A. Bazaraa. 1969. Discussion of Settlement of Spread Foot-
the same result (X ≈ 0.813) is obtained by using the following ings on Sand. ASCE Journal of the Soil Mechanics and Foundations
syntax and the Goal Seek function to determine X (“TRUE” Division, Vol. 95, No. 3, pp. 900–916.
indicates the use of cumulative distribution function): Pe = Schmertmann, J., P. Brown, and J. Hartman. 1978. Improved Strain
LOGNORM.DIST(X,0.1173,0.6479,TRUE) = 0.3085.] Influence Factor Diagrams. ASCE Journal of the Geotechnical Engi-
In the context of the AASHTO LRFD framework, the load neering Division, Vol. 104, No. 8, pp. 1131–1135.
Tan, C., and J. Duncan. 1991. Settlement of Footings on Sands—
factor, gSE, is the reciprocal of X. Thus, for immediate settle- Accuracy and Reliability. Proceedings of the Geotechnical Engineering
ment of spread footings based on the method of Schmert- Congress 1991, ASCE Geotechnical Special Publication No. 27, Vol. 1,
mann et al. (1978), gSE = 1/0.813 ≈ 1.23. pp. 446–455.

332
F.5 Concrete Girder Database

This section includes descriptions of the concrete bridges
selected from the NCHRP Project 12-78 database and used
for various studies reported here (Mlynarski et al. 2011, in
References, p. 266). Other bridges could be substituted.
Strand Legend
Tendon Types
LR = Low Relaxation
SR = Stress Relieved

Prestressed Concrete I-Beams
Bridge 18067
Dist. to Extreme Strands

Span Girder Overhang Materials (in.) Harp
Bridge Virtis BID Length tslab Spacing Width # of Skew f c' fc' I fc' deck Point Beam
# # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons (ksi) (ksi) (ksi) Bottom Top (ft) Section
50-0.6" Gr. 270 AL BT-54
18067 562 131'-0 ¼" 6 5'-3 ½" 3'-0 ½" 17 112.5 LR 7.5 6.5 3.0 2.5 3.5 44.51 Mod
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 50
Number of Harped Strands: 20
Center of gravity (CG) from bottom at Midspan: 8.9"
CG from Bottom at Support: 20.0"
Debonded Strands: 2
Debonded Length: 60 inches
CG from Bottom at 60": 18.78"
Strand Layout at Midspan Strand Layout at 60"" from Support Strand Layout at Support Cross-Section
333
334
Bridge 8891
Overhang Dist. to Extreme Strands Harp
Virtis BID Span Length tslab Girder Spacing Width # of Skew Materials (in.) Point Beam
Bridge # # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons fc' (ksi) fc'I (ksi) fc'deck (ksi) Bottom Top (ft) Section
1@ 7'-10, 14-0.5" Gr. 270
8891 571 47'-2" 8.5 12@10'-8" 3'-4" 14 90 LR 6.0 5.0 4.0 2.0 14.0 19.33 Beam Type 6
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 14
CG from bottom at Midspan: 2.57"
Strand Layout at Midspan Strand Layout at Support Cross-Section
Bridge 8832
20-0.5" Gr. 270 36"
8832 570 43'-3⅛" 8.5 10 2'-4" 16 87.5 LR 8.0 6.0 4.0 2.0 3.0 na I-Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 20
Strand Layout at Midspan Cross-Section Cross-Section

Bridge 12603
Overhang # of Dist. to Extreme Strands Harp
Bridge Virtis Span Length tslab Girder Spacing Width Girder Skew Materials (in.) Point
# BID # (ft) (in.) (ft) (ft) s (deg) P/S Tendons fc' (ksi) fc'I (ksi) fc'deck (ksi) Bottom Top (ft) Beam Section
7.87 14-0.6" Gr. 270 AASHTO Type
12603 572 37'-8¾" 5 11'-5¾" 3'-6½" 4 90 LR 7.25 5.5 4.0 2.0 4.0 na II
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 14
Bridge 10740
Span Girder Overhang Dist. to Extreme Strands Harp
Bridge Virtis BID Length tslab Spacing Width # of Skew Materials (in.) Point
# # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons fc' (ksi) fc'I (ksi) fc'deck (ksi) Bottom Top (ft) Beam Section
32-0.5" Gr. 270 AASHTO Type III
10740 575 78'-6½" 6.25 7 3'-10½" 6 90 LR 6.0 5.4 3.0 3.0 na
Strand Spacing:
Horizontal: 2"
Vertical: 2" (4" to top pair of strands)
# of Strands: 32
CG from bottom at 48": 6.60"
335
@48" from left support @ 120" from left support @ 168" from left support Strand Layout at Midspan Cross-Section
336
Bridge 10269

Virtis BID Span Length tslab Girder Spacing Width # of Skew Materials (in.) Point
Bridge # # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons fc' (ksi) fc'I (ksi) fc'deck (ksi) Bottom Top (ft) Beam Section
28-0.5" Gr. 270 AASHTO Type
10269 0576 78'-0" 6.25 6'-8" 3'-4 ½" 7 90 LR 6.0 5.0 3.3 3.0 6.0 28.75 III
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 28

Bridge 5624

Virtis BID Length tslab Spacing Width # of Skew Materials (in.) Point Beam
20-0.5" Gr. 270 Beam Type
5624 0588 59'-4⅜" 9 7'-3" 2'-10" 21 109.9 LR 6.0 4.5 4.0 2.0 3.0 24.15 4
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 20
337
338
Bridge 5794

Bridge Virtis BID Length tslab Spacing Width # of Skew fc'deck Point Beam
# # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons fc' (ksi) fc'I (ksi) (ksi) Bottom Top (ft) Section
5794 0589 72'-0" 8.5 5'-10" 2'-9" 6 100.0 LR 9.0 7.0 4.0 2.0 3.0 29.13 3
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 20

Bridge 9378
Girder Overhang Dist. to Extreme Strands Harp

Bridge Virtis BID Span Length tslab Spacing Width # of Skew Materials (in.) Point
# # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons fc' (ksi) fc'I (ksi) fc'deck (ksi) Bottom Top (ft) Beam Section
40-0.5" Gr. 270
9378 0598 101'-10" 9 10'-5" 2'-7½" 5 90 LR 6.0 5.1 4.0 2.0 2.0 41.17 Wisconsin 70"
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 40
CG from bottom at Left Support": 8.13"
@ left support @ 78" from left support @ 318" from left support Strand Layout at Midspan Cross-Section
339
340
Bridge 5884

Virtis BID Span Length tslab Spacing Width # of Skew Materials (in.) Point Beam
5884 0602 90'-0" 8.5 8'-2" 3'-1" 4 110.0 LR 7.0 5.0 4.0 2.0 4.0 27.28 6
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 38
CG from Bottom at Midspan: 5.26"

Bridge 8885
36-0.5" Gr. 270
8885 0603 90'-0" 8.5 10'-7" 3'-5½" 4 105.8 LR 7.0 5.5 4.0 2.5 2.5 36.00 BT-63
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 36
341
342
Bridge 8957

8957 0604 98'-0" 8.5 8'-8" 3'-0" 5 90.0 LR 8.0 6.4 4.0 2.0 4.0 39.67 6
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 26

Bridge 12596

4'-0 50-0.6" Gr. 270 AASHTO Type
12596 0610 96'-9⅜" 7.875 11'-1⅞" 7/16" 4 90 LR 10.2 7.1 4.0 2.0 2.0 na IV
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 50
CG from bottom at Left Support to 102.36": 12.55"
CG from bottom at 133.86": 12.09"
@ 102.36" from left @ 133.86" from left @ 157.48" from left Strand Layout
@ left support
support support support at Midspan Cross-Section
343
344
Bridge 10803

46-0.5" Gr. 270
10803 0611 138'-3" 6.25 6'-0" 3'-4½" 7 90.0 LR 7.0 6.0 4.0 2.5 3.5 54.75 BT-72
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 46

Bridge 8890
Virtis BID Span Length tslab Spacing Width # of Skew Materials (in.) Point
8890 0613 143'-6" 8.5 8'-0" 2'-9" 14 92.8 LR 8.0 6.0 4.0 2.5 3 57.90 VI
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 48
345
346
Bridge 10755

10755 0411 52'-6" 6.25 7'-0" 3'-10½" 6 90.0 SR 6.0 5.0 3.0 3.0 9.0 na II
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 24
CG from Bottom at 60": 8.82"
Strand Layout at Midspan @ 60" from left support Cross-Section

Bridge 3107
2'-9 12-0.5" Gr. 270 36"
3107 0416 49'-6½" 7.6875 5'-9¼" 7/16" 6 90.0 LR 6.1 5.1 3.5 2 na I-Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 12
Bridge 4827
4827 0418 50'-7" 8.5 7'-2" 3'-0" 5 125.0 LR 6.0 4.5 4.0 2 3 20.54 2
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 18
CG from bottom at Left Support: 13.56"
Strand Layout at Midspan Strand Layout at Left Support Cross-Section
347
348
Bridge 10599

10599 0476 62'-10" 8.0 6'-9" 3'-3" 4 90 LR 7.0 6.5 4.0 3.0 13.0 na II
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 28
@ 66.75" from left support @ 138.75" from left support Strand Layout at Midspan Cross-Section
Bridge 12589
Virtis BID Length tslab Spacing Width # of Skew Materials (in.) Point
12589 0478 73'-2 ½" 9.0 8'-9" 3'-0" 5 90 LR 6.0 4.8 4.0 2.0 4.0 29.58 IV
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 30
349
350
Bridge 5840

5840 0489 85'-0" 8.5 9'-0" 3'-4" 5 70.0 LR 7.5 5.25 4.0 2.0 4.0 34.47 6
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 30

Bridge 8330
8330 0491 76'-4 ½" 8.5 8'-8" 3'-0" 5 75.5 LR 6.0 4.5 4.0 2 4 31.14 6
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 28
351
352
Bridge 82

34-0.5" Gr. 270 MN Type
82 0498 82'-9" 8.75 10'-4" 3'-6" 4 105.0 LR 5.0 4.8 4.5 2.0 2.0 33.38 63
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 34

Bridge 4794
3@7'-4", 33-0.5" Gr. 270 Beam Type
4794 0497 66'-8" 8.5 7@9'-4" 3'-7" 11 90 LR 10.0 7.0 6.0 2 3 27.92 4
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 33
353
354
Bridge 12610

Virtis BID Length tslab Spacing Width # of Skew Materials (in.) Point
3'-2 ½"
(L)
8@7'-1 ½", 3'-0½" 52-0.5" Gr. 270 AASHTO Type
12610 0539 108'-6⅜" 8.0 4@6'-11" (R) 13 90.0 LR 7.0 5.4 4.0 2.0 4.0 44.85 IV
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 52

Bridge 11938
46-0.5" Gr. 270
11938 0545 116'-6¼" 7 7' – 3 ¾" 3'-9" 8 117.9 LR 7.0 6.0 3.0 2.5 2.5 47.64 BT-63
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 46
355
356
Bridge 11030
50-0.5" Gr. 270
11030 0551 136'-0" 6.25 6'-4 ½" 3'-10½" 9 90.0 LR 7.0 6.0 3.0 2.5 3.5 48.75 BT-72
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 50

Bridge 8889
3'-7" (L) 36-0.5" Gr. 270
8889 0549 90'-10¼" 8.5 10'-7" 4'-1" (R) 4 70.6 LR 7.0 5.5 4.0 2.5 2.5 36.74 BT-63
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 36
357
358
Bridge 8783

8783 0553 141'-1¾" 8.5 7'-9" 3'-1" 8 96.9 LR 8.0 6.0 4.0 2.0 2.0 57.78 VI
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 46

Bridge 15620
48-0.6" Gr. 270
15620 0561 119'-9 ¾" 9.4375 5'-4⅛" 3'-3⅜" 11 90 LR 10.0 8.0 3.0 2.7559 na Bulb Tee
Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685"
# of Strands: 48
@ 51.57" from Left Support @75.20" from Left Support Strand Layout at Midspan Cross-Section
359
360
Adjacent Precast Box Girders
Bridge 12807
3@4'-0½", 34-0.5" Gr. 270

1@3'-6 ½", SR 2.0 na BIII-48
2@3'-0 ½",
1@3'-6 ½", 28-0.5" Gr. 270
12807 0780 84'-0¼" 6.0 3@4'-0 ½" 2'-4 ½" 11 48.0 SR 6.0 4.8 3.0 2.0 na BIII-36
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 34
Strand Layout at Midspan Cross-Section – BIII-48 Cross-Section – BIII-48

Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 28

Bridge 13788
32-0.5" Gr. 270
13788 0781 83'-0" 6.0 4'-0½" 2'-5" 9 56.8 SR 6.0 4.8 3.0 2.0 na BIV-48
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 32
361
362
Bridge 15238
34-0.5" Gr. 270
1@3'-6⅜", SR 1.9685 5.118 na BII-48
9@4'-0⅜", 1'-11 26-0.5" Gr. 270
15238 0782 73'-9⅞" 5.875 1@3'-6⅜" 15/16" 12 90.0 SR 6.5 5.2 3.0 1.9685 5.118 na BII-36
Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685"
# of Strands: 34
Strand Layout at Midspan @94.49" from Left Support Cross-Section– BII-48

Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685"
# of Strands: 26
Strand Layout at Midspan Cross-Section– BII-36 Cross-Section– BII-36

Bridge 9314
22-0.6" Gr. 270 27x36 Box
9314 0783 83'-8" 6.0 3'-1½" 1'-6" 16 90.0 LR 5.2 5.0 4.0 2 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 22
CG from bottom at Left Support to 96" : 3.14"
CG from bottom at 156" : 2.89"
CG from bottom at 240" : 3.00"
@ left support to 96" from left

support @ 156" from left support @ 240" from left support Strand Layout at Midspan Cross-Section (Interior)
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 22
CG from bottom at Left Support to 96": 3.14"
@ left support to 96" from left @ 156" from left support @ 240" from left support Strand Layout at Midspan Cross-Section (Exterior)
support
363
364
Bridge 17075
44-0.5" Gr. 270
17075 0785 107'-0" 6.0 4'-0" 2'-7½" 10 90 SR 6.0 4.8 3.3 2.0 na BIV-48
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 44
CG from bottom at left support to 138": 4.74"
Strand Layout at Midspan Strand Layout at Left Support to 138" Cross-Section

Bridge 17175

Virtis BID Span Length tslab Girder Spacing # of Skew Materials Beam
Bridge # Width (in.) Point
# (ft) (in.) (ft) Girders (deg) Section
(ft) P/S Tendons fc' (ksi) fc'I (ksi) fc'deck (ksi) Bottom Top (ft)
1@4'-0",
44-0.5" Gr. 270
1@3'-6", 2.0 7.0 na BII-48
SR
17175 0786 88'-9" 6.0 2@3'-0", 2'-6" 7 90.0 6.0 4.8 3.0
1@3'-6", 38-0.5" Gr. 270
1@4'-0" 2 7.0 na BII-36
SR
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 44
Strand Layout at Midspan – 4' Box Cross-Section – BII 48 Cross-Section – BII 48
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 38
Strand Layout at Midspan -3' Interior Box Cross-Section – BII-36 Cross-Section – BII-36
365
366
Bridge 13805
2'-0" (L),
1'-11 3/16" 24-0.5" Gr. 270
13805 0681 52'-6" 3.5 4'-0 11/16" (R) 7 75 SR 5.0 4.0 3.0 2.0 3.0 na BI-48
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 24
Bridge 14246
24-0.5" Gr. 270
14246 0684 52'-0" 6.0 4'-0½" 2'-4¼" 8 90.0 SR 5.0 4.0 3.0 2.0 3.0 na BI-48
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 24

Bridge 9180
13-0.6" Gr. 270 MDOT
9180 0690 44'-8⅜" 5.875 3'-1 ½" 1'-6" 27 78.6 LR 5.0 4.0 4.0 1.9685 na 535x915
Strand Spacing:
Horizontal: 2"
# of Strands: 13
CG from bottom at left support: 1.97"
Strand Layout at Midspan Strand layout at 59.055" Strand layout at 98.425" Cross-Section
367
368
Bridge 17042
24-0.5" Gr. 270
1@4'-0", SR 1.9685 na 4' Box
4@3'-6", 20-0.5" Gr. 270
17042 0693 50'-0⅜" 5.875 1@4'-0" 2'-5 7/16" 7 57.0 SR 6.0 4.8 3.0 1.9685 5.118 na 3' Box
Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685"
# of Strands: 24
Strand Layout at Midspan – 4' Box Cross-Section – 4' Box Cross-Section – 4' Box
Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685"
# of Strands: 20
Strand Layout at Midspan -3' Interior Box Cross-Section – 3' Box Cross-Section – 3' Box
Bridge 9191
20-0.5" Gr. 270 27"x36"
9191 0736 72'-6" 6.0 3'-1 11/16" 1'-7 ½" 15 60.0 LR 5.5 4.0 4.0 2.0 na Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 20
@ 66.00" from left support @ 114" from left support Strand Layout at Midspan Cross-Section
369
370
Bridge 16799
38-0.5" Gr. 270
16799 0737 84'-0" 0.0 4'-0" 2'-1 ½" 5 90.0 LR 5.0 4.0 3.3 2.0 na
Strand Spacing:
Horizontal: 2.0952"
Vertical: 2"
# of Strands: 38
Interior:
Exterior:
Strand Layout at Midspan Strand layout at Support to 13.25" Cross-Section

Bridge 14987
3@4'-0¾", 2'-3" (L) 34-0.5" Gr. 270
14987 0738 73'-0" 6.0 1@4'-7¼" 2'-6" (R) 5 99.7 SR 5.0 4.0 3.3 2.0 8 na BIV-48
Strand Spacing:
Horizontal: 2"
Vertical: 2", 4"
# of Strands: 34
371
372
Bridge 12809
Virtis BID Span Length tslab Girder Spacing # of Skew Materials Beam
Bridge # Width (in.) Point
# (ft) (in.) (ft) Girders (deg) Section
(ft) P/S Tendons fc' (ksi) fc'I (ksi) fc'deck (ksi) Bottom Top (ft)
3@4'-0 28-0.5" Gr. 270
2.0 na BII-36
½",1@ 3'-6 SR
½", 2@3'-0
12809 0739 82'-0" 6.0 2'-4 ¾" 11 90.0 5.0 4.0 3.0
½", 1@3'-6 36-0.5" Gr. 270
2.0 na BII-48
½", 3@4'-0 SR
½"
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 28
Strand Layout at 72" from Left Support Strand Layout at 120" from Left Support Strand Layout at Midspan- BII-36 Cross-Section – BII-36
Strand Spacing:
Horizontal: 2"
Vertical: 2", 8"
# of Strands: 36
Strand Layout at 72" from Left Support Strand Layout at 144" from Left Support Strand Layout at Midspan - BII-48 Cross-Section – BII-48
Bridge 12952

42-0.5" Gr. 270
1@4'-0½", 2.0 2.0 na BIV-48
LR
12952 0741 79'-11 ½" 8.25 10@3'-6½", 2'-4" 13 59.8 6.0 4.8 3.0
30-0.5" Gr. 270
1@4'-0½" 2.0 2.0 na BIV-36
LR
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 42
Strand Layout at Midspan – BIV-48 Cross-Section – BIV-48 Cross-Section – BIV-48
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 30
Strand Layout at Midspan – BIV-36 Cross-Section – BIV-36 Cross-Section – BIV-36
373
374
Bridge 17143
30-0.5" Gr. 270
1@4'-0", 2.0 7.0 na BIII-48
SR
17143 0742 70'-0" 6.0 2@3'-6", 2'-6" 5 90 5.0 4.0 3.0
26-0.5" Gr. 270

1@4'-0"
SR 2.0 7.0 na BIII-36
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 30
Strand Layout at Midspan – 4' Box Cross-Section – BIII-48 Cross-Section – BIII-48
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 26

Bridge 9181
21-0.5" Gr. 270
9181 0743 60'-4⅜" 5.875 3'-1⅝" 1'-7⅛" 15 90.0 LR 6.5 5.0 4.0 1.9685 na MDOT 535x915
Strand Spacing:
Horizontal: 2"
Vertical: 2.0079"
# of Strands: 21
375
376
Bridge 9071
16-0.6" Gr. 270
9071 0745 83'-7⅝" 5.875 3'-1⅜" 1'-6⅛" 15 120.0 LR 5.0 3.5 4.0 2.0 na MDOT 840x915
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 16
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 16
Bridge 9167
Overhang Materials (in.) Harp
Virtis BID Span Length tslab Girder Spacing Width # of Skew fc'I Point
Bridge # # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons fc' (ksi) (ksi) fc'deck (ksi) Bottom Top (ft) Beam Section
22-0.6" Gr. 270 MDOT
9167 0746 75'-6¾" 5.875 4'-1½" 2'-1⅝" 12 120.0 LR 5.0 4.0 4.0 1.9685 na 685x1220
Strand Spacing:
Horizontal: 1.9685"
Vertical: 2.0079"
# of Strands: 22
@ 63.0" from left support @ 118.11" from left support @ 185.04" from left support Strand Layout at Midspan Cross-Section - Interior
Strand Spacing:
Horizontal: 1.9685"
Vertical: 2.0079"
# of Strands: 22
@ 63.0" from left support @ 118.11" from left support @ 185.04" from left support Strand Layout at Midspan Cross-Section - Exterior
377
Bridge 17008
378
38-0.5" Gr. 270 AASHTO BII-
17008 0747 82'-6⅛" 6.25 4'-0" 2'-6" 8 115.0 SR 7.0 5.6 3.0 1.9685 na 1220
Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685"
# of Strands: 38
Bridge 5125
1@3'-6",
7@4'- 24-0.5" Gr. 270 36"x33" Box
0",1@3'-6", LR 2.0 3.0 Girder
2@3'-0",
1@3'-6",
7@4'-0", 26-0.5" Gr. 270 48"x33" Box
5125 0748 66'-0" 5.5 1@3'-6" 1'-6" 22 108.0 LR 6.0 4.5 4.0 2.0 3.0 na Girder
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 24
Strand Layout at Midspan Cross-Section Cross-Section – 36"x33" Box Girder

Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 26
Strand Layout at Midspan Cross-Section Cross-Section – 48"x33" Box Girder

Bridge 13118
Virtis BID Span Length tslabGirder Spacing Width # of Skew Materials (in.) Point
1@3'-7 3/16",
2@3'-1 3/16", 18-0.6" Gr. 270 AASHTO BI-
1.9685 4.9213 na
1@3'-7 ¼", LR 915
2'-3
13118 0749 69'-2¾" 5.875 2@4'1 3/16", 11 60.0 10.2 8.0 3.0
15/16"
1@3'-7 ¼", 19-0.6" Gr. 270 AASHTO BI-
2@3'-1 3/16", 1.9685 4.9213 na
LR 1220
1@3'-7 3/16"
Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685"
# of Strands: 18
Strand Layout at Midspan Cross-Section Cross-Section – AASHTO BI-915
Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685"
# of Strands: 19
Strand Layout at Midspan Cross-Section Cross-Section – AASHTO BI-1220
Bridge 5911
14-0.5" Gr. 270
5911 0750 59'-5" 5.5 3'-0" 1'-6" 50 125 LR 5.0 4.0 4.0 1.75 na 27x36 Box
Strand Spacing:
Horizontal: 1.375"
Vertical: 1.5"
# of Strands: 14
379
Strand Layout at Midspan Cross-Section Cross-Section – 27x36 Box
380
Bridge 3805
15-0.5" Gr. 270 27"x36" IDOT
3805 0751 59'-0½" 0.0 3'-0" 1'-6" 14 90.0 LR 5.0 4.0 3.5 1.75 na Beam
Strand Spacing:
Horizontal: 1.375", 3"
Vertical: 1.5", 1.25", 3"
# of Strands: 15
Strand Layout at Midspan Cross-Section Cross-Section – 27"x36" IDOT Beam
Bridge 3819
18-0.5" Gr. 270
3819 0752 74'-10½" 0.0 3'-0" 1'-6" 11 128.8 LR 5.0 4.0 3.5 1.75 na 33"x36" Box Beam
Strand Spacing:
Vertical: 1.5", 2.75", 3"
# of Strands: 18

Bridge 9240
28-0.5" Gr. 270
9240 0763 97'-11" 6.0 3'-1 ½" 1'-6" 25 90.0 LR 5.5 3.6 4.0 2.0 na 33"x36" MDOT Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 28
CG from bottom at Left Support to 132": 3.75"
@ left support to 132" from left

support @ 252" from left support @ 312" from left support Strand Layout at Midspan Cross-Section
381
382
Bridge 9103
25-0.6" Gr. 270 MDOT
9103 0764 111'-2⅝" 5.875 4'-1⅝" 2'-1 ½" 16 100.8 LR 5.0 3.5 4.0 2.0 na 1220x1220
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 25
Strand Layout at Midspan Strand layout at Support to 157.48" Cross-Section - Interior

Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 25
Strand Layout at Midspan Strand layout at Support to 157.5" Cross-Section - Exterior

Bridge 9228
24-0.6" Gr. 270
9228 0765 110'-5¼" 5.875 4'-2 13/16" 2'-1 ½" 11 72.2 LR 6.1 4.6 4.0 1.9685 na MDOT 1220x1220
Strand Spacing:
Horizontal: 2.0079"
Vertical: 1.9685"
# of Strands: 24
Bridge 14070
42-0.5" Gr. 270
14070 0766 115'-0" 0.0 3'-1 3/16" 1'-6" 11 60.0 LR 7.5 5.5 2.0 4.5 na 36"x45" Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 42
@ 36" from left support @ 72" from left support Strand Layout at Midspan Cross-Section
383
384
Bridge 16538
38-0.5" Gr. 270 AASHTO BIV-48
16538 0767 101'-8½" 5.875 4'-0¾" 2'-4" 9 90.0 LR 7.25 5.1 3.0 1.9685 na mod
Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685"
# of Strands: 38
Spread Precast Box Girders

Bridge 9310
20-0.6" Gr. 270
9310 0774 51'-11⅝" 9.0 5'-11⅝" 2'-2¼" 12 90.0 LR 6.0 5.5 4.0 2.0 na 21x36 Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 20
Bridge 9384
14-0.6" Gr. 270
9384 0622 44'-1½" 9.0 6'-7¼" 3'-3⅜" 10 89.4 LR 6.1 4.6 4.0 2 na 21x36 Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 14
@ 59.0" from left support @86.6" from left support Strand Layout at Midspan Cross-Section
Bridge 9380
12-0.5" Gr. 270
9380 0629 32'-1" 9.0 9'-1 3/16" 2'-6' 10 85.6 LR 5.0 3.0 4.0 2.0 na 21x36 Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 12
385
386
Bridge 17338
18-0.5" Gr. 270

17338 0675 49'-0" 8.5 8'-0" 4'-0" 4 135 SR 5.0 4.0 3.0 2.0 na AASHTO BII-48
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 18
Bridge 9282
3'-0" (L), 16-0.5" Gr. 270

9282 0679 36'-3⅝" 9.0 7'-10" 4'-0" (R) 12 67.3 LR 5.0 4.0 4.0 2 na 17"x36" Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 16
CG from bottom at left support to 120": 2.33"

Bridge 9192
3@6'-6 11/16" 18-0.5" Gr. 270

9192 0686 38'-8" 9.0 5@9'-10" 3'-9 ½" 9 80 LR 5.5 4.4 4.0 2.0 na 21"x36" Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 18
CG from bottom at Support to 24": 3.00"
Strand Layout at Midspan Strand Layout at Support to 24" Cross-Section – 21"x36" Box Beam
Bridge 9286
20-0.6" Gr. 270
9286 0695 50'-8" 9.0 8'-0 ½" 3'-2⅛" 6 90 LR 6.0 5.0 4.0 2.0 2.5 na 27"x36" Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 20
CG from bottom at 0.5" from LS: 3.43"
CG from bottom at 48" from LS: 6.06"
387
@ 0.5" from left support @ 48" from left support @ 140.0" from left support Strand Layout at Midspan Cross-Section
388
Bridge 9368
30-0.5" Gr. 270
9368 0707 71'-3" 9.0 7'-4 ½" 2'-11½" 5 97.4 LR 6.3 4.9 4.0 2.0 na 33"x36" Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 30
@ 63" from left support @ 102" from left support @ 147.0" from left support Strand Layout at Midspan Cross-Section
Bridge 9328
26-0.5" Gr. 270
9328 0740 57'-3¼" 9.0 6'-11" 3'-3⅜" 6 90.0 LR 6.5 5.3 4.0 2 na 27x36 Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 26
Bridge 9376
3'-5 15-0.6" Gr. 270
9376 0744 53'-4¼" 9.0 7'-4⅝" 5/16" 12 90.0 LR 7.0 6.2 4.0 2 na 27x36 Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 15
Bridge 3577
27-0.4375" Gr. 248 27"x36" IDOT
3577 0222003 38'-7" 7.5 6'-8⅜" 4'-0 7/16" 4 101.3 SR 5.0 4.0 3.5 1.75 na Beam
Strand Spacing:
Vertical: 1.5", 1.25", 3"
# of Strands: 27
389
390
Bridge 3754
1@6'-6",
30 (or 28) -0.4375" Gr. 248 33x36 IDOT
3754 0980015 53'-7¼" 7.0 7@6' 2 9/16", 1'-6" 11 151.7 5.0 4.0 3.5 1.75 na
SR Beam
2@6'-0"
Strand Spacing:
Vertical: 1.5", 2.75", 3"
All girders except far left and two far right:

# of Strands: 30
Far left and two far right girders:

# of Strands: 28
Strand Layout at Midspan Cross-Section
Bridge 8875
23-0.5" Gr. 270 27x48 Box
8875 A8029 38'-0" 8.5 11'-3" 3'-1½" 12 96.4 LR 8.0 5.0 4.0 1.75 na Beam
Strand Spacing:
Horizontal: 2.75"
Vertical: 1.5", 2.75", 3"
# of Strands: 23

Bridge 12870
30-0.5" Gr. 270
12870 2219470 77'-6" 8.5 6'-6" 2'-0" 5 105.0 SR 6.0 4.8 3.0 2.0 na 36x48 Box Beam
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 30
Bridge 14969
Overhang Materials Dist. to Extreme Strands (in.) Harp
Virtis BID Span Length tslab Girder Spacing Width # of Skew Point
4'-5 5/16" (L), 48-0.5" Gr. 270 AASHTO BIV-

14969 1023430 78'-8⅞" 9.4375 7'-10⅝" 4'-11 3/16" (R) 7 90.0 LR 6.5 4.5 3.0 1.9685 na 48
Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685"
# of Strands: 48
391
@ 98.4" from left support @ 141.75" from left support @ 216.5" from left support Strand Layout at Midspan Cross-Section
392
Bridge 16293
26-0.6" Gr. 270
16293 BID_2751 57'-3¼" 9.4375 8'-10 5/16" 4'-1⅝" 7 96.0 LR 10.2 7.1 3.0 2.2441 4.252 na 1220x1220 Box Beam
Strand Spacing:
Horizontal: 1.9685"
Vertical: 2"
# of Strands: 26
Bridge 16366
Based on
38-0.5" Gr. 270 AASHTO BI-
16366 2223270 60'-4⅝" 9.4375 6'-7" 4'-0 1/16" 6 102.4 LR 7.5 6.0 3.0 1.9685 na 48
Strand Spacing:
Horizontal: 1.9685"
Vertical: 1.9685", 3.937", 1.9685", 5.9055", 1.9685"
# of Strands: 38

Bridge 17240
24 (or 26)-0.5" Gr. 270 AASHTO BII-
17240 3300870 51'-6" 8.5 8'-0" 4 '-0" 4 90.0 SR 6.0 4.8 3.0 2.0 na 48
Strand Spacing:
Horizontal: 2.0"
Vertical: 2.0"
Interior:
# of Strands: 24
Exterior:
# of Strands: 26
Strand Layout at Midspan – Interior Girder Strand Layout at Midspan – Exterior Girder Cross-Section
393
394
Bridge 9090
Span Length tslab Girder Spacing Width # of Skew Materials (in.) Point
Bridge # Virtis BID # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons fc' (ksi) fc'I (ksi) fc'deck (ksi) Bottom Top (ft) Beam Section
MDOT 33" Box Beam
5.3 5.3 (Int)
20-0.6" Gr. 270 MDOT 33" Box Beam
9090 13113081000S053 66'-0½" 9.0 7'-1" 3'-5" 9 68.6 LR 5.6 5.6 4.0 2.0 na (Ext)
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 20
Interior:
Exterior:
Bridge 9091

MDOT 33" Box Beam
5.3 5.3 (Int.)
20-0.6" Gr. 270 MDOT 33" Box Beam
9091 13113081000S054 66'-2½" 9.0 7'-1" 3'-5" 9 68.3 LR 5.5 5.5 4.0 2.0 na (Ext.)
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 20
Interior:
Exterior:
395
396
Bridge 9128
4-0.6" Gr. 270
9128 29129011000S140 33'-8⅜" 9.0625 7'-5⅞" 3 '-1¼" 8 90.0 LR 6.0 4.5 4.0 2.28 na 1525 Box Beam
Strand Spacing:
Horizontal: 2.402"
Vertical: 2.402"
# of Strands: 4

Bridge 9217
5.8 5.3 17" Box Beam (A)
17" Box Beam (B and
5.6 5.2 C)
5.2 5.1 17" Box Beam (D)
17" Box Beam (E thru
3@6'-2 ½", 16-0.6" Gr. 270 5.0 4.9 K)
9217 45145071000B010 42'-3½" 9.25 7@5'-8⅛" 3'-3" 11 90.0 LR 5.9 5.3 4.0 2.0 2.5 na 17" Box Beam (L)
Strand Spacing:
Horizontal: 2"
Vertical: 2"
Beams: (A) is far left girder, (L) is far right girder
# of Strands: 16
Girders: A and L
Girders: B, C, D
Girders: E thru K
Strand Layout at Midspan (All girders) Strand Layout at Support to 96" (E thru K), 54" (B thru D) Cross-Section
397
398
Bridge 9219
16116032000B03 16-0.6" Gr. 270 MDOT 21" Box
9219 0 53'-2" 9.0 5'-3" 2'-7½" 9 90.0 LR 5.7 5.6 4.0 2.0 na Beam
Strand Spacing:
Horizontal: 1"
Vertical: 2"
# of Strands: 16
Strand Layout at Midspan Strand Layout at Support to 72" Cross-Section
Bridge 9243
21-0.6" Gr. 270 33"x36" Box
9243 82182041000S020 73'-4" 9.0 6'-2" 2'-3¾" 13 90.0 LR 7.0 6.0 4.0 2.0 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 21
Bridge 9248
18-0.6" Gr. 270
9248 02102041000B020 37'-9" 9.0 8'-0" 3'-2½" 6 90.0 LR 6.0 5.0 4.0 2.0 3.0 na 21" Box Beam
Strand Spacing:
Horizontal: 2.0",
Vertical: 2.0", 10.0"
# of Strands: 18
Bridge 9284
17"x36" Box
9284 33133082000B030 31'-6¾" 9.0625 6'-7⅞" 2'-5½" 10 90.0 9-0.6" Gr. 270 LR 5.0 3.0 4.0 2.0 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 9
CG from bottom at Support to 78.74": 2.57"
Strand Layout at Midspan Strand Layout at Support to 78.74" Cross-Section
399
400
Bridge 9289
3'-1 22-0.6" Gr. 270
9289 45145012000B010 63'-7⅝" 9.0625 6'-4" 3/16" 7 45.0 LR 5.0 4.0 4.0 2.0 3.0 na 27"x36" Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 22
@ 62.99" from left support @ 86.61" from left support @ 141.73" from left support @ 173.91" from left support Strand Layout at Midspan Cross-Section
Bridge 9324
20-0.5" Gr. 270 27"x36" Box
9324 14114041000B010 42'-4" 9.0 10'-6" 2'-7½" 5 90.0 LR 5.5 4.4 4.0 2.0 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 20

Bridge 9349
22-0.5" Gr. 270 21"x36" Box
9349 72172041000B010 48'-8" 9.0 7'-0" 3'-6" 6 90.0 LR 6.0 4.9 4.0 2.0 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 22
Bridge 9355
26-0.6" Gr. 270 39"x36" Box
9355 38138111000R021 75'-2⅜" 9.0625 7'-11" 3'-6" 6 109.6 LR 7.0 6.2 4.0 2.0 2.75 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 26
401
402
Bridge 9356

Length tslab Spacing Width # of Skew fc' fc'deck Point
Bridge # Virtis BID # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons (ksi) fc'I (ksi) (ksi) Bottom Top (ft) Beam Section
9.062 26-0.6" Gr. 270 39"x36" Box
9356 38138111000R022 75'-2⅜" 5 7'-11" 3'-6" 6 109.6 LR 7.0 6.2 4.0 2.0 2.75 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 26

Bridge 9361
30-0.5" Gr. 270
9361 47147065000R033 65'-9½" 9.0 6'-0" 2'-7½" 11 137.2 LR 6.5 5.5 4.0 2.0 na 27"x36" Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 30

Bridge 9369
14-0.6" Gr. 270 27"x36" Box
9369 11111015000R033 51'-3⅞" 9.0 6'-6" 3'-4½" 10 115.7 LR 5.4 4.5 4.0 2.0 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 14
403
404
Bridge 9370
Bridge Length tslab Spacing Width # of Skew fc' fc'I fc'deck Point
# Virtis BID # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons (ksi) (ksi) (ksi) Bottom Top (ft) Beam Section
14-0.6" Gr. 270 27"x36" Box
9370 11111015000R034 51'-3⅞" 9.0 6'-5" 3'-4" 12 115.7 LR 5.2 4.4 4.0 2.0 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 14
Bridge 9383
16-0.6" Gr. 270 27"x36" Box
9383 71171073000B010 46'-9⅞" 9.0 10'-7" 2'-6" 5 45.0 LR 6.0 5.0 4.0 2.0 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 16
Bridge 9394
Bridge Length tslab Spacing Width # of Skew fc' fc'deck Point
# Virtis BID # (ft) (in.) (ft) (ft) Girders (deg) P/S Tendons (ksi) fc'I (ksi) (ksi) Bottom Top (ft) Beam Section
34-0.5" Gr. 270 27"x48" Box
9394 25125031000S010 66'-10⅞" 9.0 7'-6" 4'-1½" 11 59.8 LR 6.0 5.0 4.0 2 na Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 34
Bridge 1150
39-0.5" Gr. 270
1150 550A0490001 70'-7" 8.25 8'-6" 3'-8" 3 118.0 LR 6.5 6.0 3.0 2.0 na 33"x36" Box Beam
Strand Spacing:
Horizontal: 2"
Vertical: 2"
# of Strands: 39
405
F.6 Steel Girder Database
406
This section includes descriptions of the steel bridges selected from the NCHRP Project 12-78 database and used for various studies reported here
(Mlynarski et al. 2011). Other bridges could be substituted.
Materials
Span Girder Overhang Cross-Frame
Bridge Length tslab Spacing # of Width Skew f′c deck Spacing
# (ft) (in.) (ft) Girders (ft) (deg) fyw (ksi) fyf (ksi) (ksi) (ft) Girder Description
0032 31.00 6.75 7.21 6 1.06 125.0 33 33 3.0 Fascia Rolled Beam
1 @ 19.00 W27X91
1 @ 12.00
Interior
1 @ 13.95
1 @ 5.05
1 @ 12.00
0053 38.00 5.50 4.75 8 0.63 90.0 30 30 2.5 None Rolled Beam
Fascia
B28, 28X10X85
Interior
CB213, 21X13X112
0058 25.80 11.00 1 @ 8.25 9 0.56 at G1 60.0 33 33 3.3 None Rolled Beam
4 @ 8.00 0.58 at G9 W21X83
1 @ 8.25
1 @ 2.68
1 @ 7.71
0075 32.00 10.00 3.91 11 0.03 66.0 30 30 2.5 None Rolled Beam
S20X75F
0076 40.09 7.00 6.00 4 1.67 115.0 30 30 2.5 G1 Rolled Beam
1 @ 16.69 G1
1 @ 16.00 24WF74
1 @ 7.40 G2
G2 24WF80
1 @ 13.90
1 @ 16.00
1 @ 10.19
0078 51.92 8.00 4.13 7 0.83 at G1 80.0 33 33 2.5 None Rolled Beam
1.08 at G7 CB30X180
0146 34.67 7.25 8.75 7 2.08 102.0 33 33 3.3 2 @ 17.33 Rolled Beam
33WF(B33), 33X11.5X130

(continued)
Materials
0151 39.79 7.00 6 4 3.33 130.0 36 36 4.0 G1 Rolled Beam

1 @ 22.41 G1
1 @ 17.38 W24X68
G2 G2
1 @ 17.38 W24X76
1 @ 5.03
1 @ 17.38
0179 47.50 9.50 9.75 5 1.27 75.0 36 36 3.0 G1 Rolled Beam
1 @ 23.50 0 ft. to 6.89 ft. W30X99
1 @ 24.00 W30X99
G2 6.89 ft. to 8 1/2X9/16 Bott
1 @ 23.50 23.75 ft. Cover PI
1 @ 2.61 (Symm.)
1 @ 21.39
0199 43.50 5.88 1 @ 1.60 13 0.37 90.0 36 36 3.0 2 @ 21.75 Rolled Beam
10 @ W24X94
2.00
1 @ 1.60
0208 48.75 8.50 5.00 7 2.00 at G1 90.0 30 30 3.3 2 @ 24.38 Rolled Beam
1.25 at G7 W28X145
0224 38.60 1.31 1 @ 1.58 13 0.34 90.0 36 36 2.4 2 @ 19.30 Rolled Beam
10 @ 2.00 W21X63
1 @ 1.58
0256 79.45 6.50 8.00 4 4.00 45.0 33 33 3.0 G1 Built-up I-Shape
1 @ 9.87 0 ft. to 10.22 ft. L4X7X3/4 Top and
2 @ 19.75 Bott Angles
1 @ 13.17 47X3/8 Web
1 @ 16.91
10.22 ft. to L4X7X3/4 Top and
G2
20.22 ft. Bott Angles
1 @ 16.46
16X1/2 Top and
2 @ 19.75
Bott Cover Pls
1 @ 13.17
47X3/8 Web
1 @ 10.32
G3 20.22 ft. to L4X7X3/4 Top and
1 @ 9.87 39.72 ft. Bott Angles
1 @ 13.17 (Symm.) 16X15/16 Top and
2 @ 19.75 Bott Cover Pls
1 @ 16.91 47X3/8 Web
G4
1 @ 16.46
1 @ 13.17
2 @ 19.75
407
1 @ 10.32

(continued)
408
Materials
0260 85.00 8.50 9.50 4 3.75 at G1 40.0 50 50 3.3 G1 Welded I-Shape

3.92 at G4 1 @ 2.35 0 ft. to 25.00 ft. 13X5/8 Top Flg
2 @ 23.17 17X3/4 Bott Flg
1 @ 19.33 52X1/2 Web
1 @ 16.98
25 ft. to 42.50 ft. 13X5/8 Top Flg
G2
(Symm.) 17X1 Bott Flg
1 @ 13.68
52X1/2 Web
2 @ 23.17
1 @ 19.33
1 @ 5.65
G3
1 @ 5.65
1 @ 19.34
2 @ 23.17
1 @ 13.67
G4
1 @ 16.98
1 @ 19.33
2 @ 23.17
1 @ 2.35
0267 63.00 6.00 3.50 11 1.38 52.9 36 36 3.0 G1 Welded I-Shape
4 @ 15.75 G1/G2/G10/G11
G2 15x1 1/8 Top Flg
2 @ 15.75 15X1 1/8 Bott Flg
1 @ 13.00 19 3/4X5/8 Web
1 @ 2.75
1 @ 13.00
1 @ 2.75
1 @ 13.00
G10
1 @ 15.75
1 @ 2.75
1 @ 13.00
1 @ 2.75
1 @ 13.00
1 @ 2.75
1 @ 13.00
G11
1 @ 18.50
2 @ 15.75
1 @ 13.00

(continued)
Materials
0268 82.00 7.25 8.52 7 1.60 65.9 36 36 3.3 G1 Welded I-Shape

1 @ 2.54 G1/G7
3 @ 18.00 0 ft. to 20.00 ft. 12X3/4 Top Flg
1 @ 25.46 14X3/4 Bott Flg
G2 54X3/8 Web
1 @ 6.36
20.00 ft. to 12X3/4 Top Flg

3 @ 18.00
41.00 ft. 14X1 1/4 Bott Flg
1 @ 21.64
(Symm.) 54X3/8 Web
G3
1 @ 10.18 G2/G3/G4/G5/G6
3 @ 18.00 0 ft. to 16.00 ft. 12X3/4 Top Flg
1 @ 17.82 14X3/4 Bott Flg
G4 54X3/8 Web
1 @ 14.00
3 @ 18.00 16.00 ft. to 12X3/4 Top Flg
1 @ 14.00 41.00 ft. 14X1 1/2 Bott Flg
G5 (Symm.) 54X3/8 Web
1 @ 17.81
3 @ 18.00
1 @ 10.19
G6
1 @ 21.63
3 @ 18.00
1 @ 6.37
G7
1 @ 25.44
3 @ 18.00
1 @ 2.56
0272 75.00 8.50 9.00 5 3.00 125.0 50 50 3.0 G1 Welded I-Shape
1 @ 15.00 0 ft. to 16.00 ft. 12X1/2 Top Flg
1 @ 12.60 16X3/4 Bott Flg
2 @ 22.50 38X7/16 Web
1 @ 2.40
G2
26.50 ft. 16X1 1/4 Bott Flg
1 @ 8.75
38X7/16 Web
1 @ 12.55
2 @ 22.50 26.50 ft. to 12X5/8 Top Flg
1 @ 8.70 37.50 ft. 16X1 1/4 Bott Flg
(Symm.) 38X7/16 Web
409
(continued)
410
Materials
0273 68.00 8.75 10.33 4 3.83 60.0 36 36 (Top) 4.5 Fascia Welded I-Shape
50 (Bott) 1 @ 19.04 12X3/4 Top Flg
1 @ 23.96 15X1 1/4 Bott Flg
1 @ 25.00 42X3/8 Web
Interior
1 @ 19.04
1 @ 5.97
1 @ 18.00
1 @ 5.97
1 @ 19.03
0277 80.00 8.50 10.50 4 2.08 at 90.0 50 50 3.0 G1 Welded I-Shape
G12.11 at 11 @ 6.67 0 ft. to 14.50 ft. 10X1/2 Top Flg
G4 1 @ 6.63 14X3/4 Bott Flg
G2 50X3/8 Web
2 @ 6.67
1 @ 6.66
21.50 ft. 14X1 1/2 Bott Flg
1 @ 0.01
50X3/8 Web
2 @ 6.67
1 @ 6.66 21.50 ft. to 10X7/8 Top Flg
1 @ 0.02 40.00 ft. 14X1 1/2 Bott Flg
2 @ 6.67 (Symm.) 50X3/8 Web
1 @ 6.64
1 @ 0.03
2 @ 6.67
1 @ 6.62
G3/G4
4 @ 20.00
0283 58.48 6.00 6.00 4 2.17 65.0 33 33 3.0 G1 Rolled Beam
1 @ 13.26 G1/G2
2 @ 14.58 30WF108
1 @ 16.06
G2
1 @ 13.26
1 @ 2.80
1 @ 11.78
1 @ 2.80
1 @ 11.78
1 @ 2.80
1 @ 13.26
0304 55.67 7.00 7.92 5 2.33 58.0 33 33 3.5 1 @ 13.56 Rolled Beam
1 @ 4.95 36WF(CB362), 36X16.5X230
1 @ 13.68
1 @ 4.95
1 @ 18.53

(continued)
Materials
0314 84.00 8.50 7.75 9 3.08 at G1 90.0 32 (G1– 32 4.0 1 @ 16.76 Rolled Beam
2.42 at G9 G5)36 (G1– 3 @ 16.83 G1/G5
(G6–G9) G5)36(G6– 1 @ 16.75 0 ft. to 13.00 ft. 36WF(B36a),
G9) 36X16.5X300
13.00 ft. to 36WF(B36a),

41.00 ft. 36X16.5X300
(Symm.) 15X1 1/2 Bott
Cover Pl
G2/G3/G4
0 ft. to 13.00 ft. 36WF(CB362),
36X16.5X230
13.00 ft. to 36WF(CB362),

41.00 ft. 36X16.5X230
Cover Pl
G6/G7/G8/G9
36WF(CB362), 36X16.5X230
14 1/2X1 1/8 Bott Cover Pl
0317 84.31 7.50 4.60 6 2.34 at G1 133.5 33 33 3.0 G1 Rolled Beam
2.16 at G6 3 @ 21.08 G1/G3/G6
1 @ 21.07 0 ft. to 15.53 ft. W36X182
G3
15.53 ft. to W36X182
1 @ 16.72
42.18 ft. 11X1 Bott Cover Pl
1 @ 4.36
(Symm.)
1 @ 16.72

1 @ 4.36
1 @ 16.72
1 @ 4.36
1 @ 21.07
G6
1 @ 16.72
2 @ 21.08
1 @ 25.43
0329 75.75 6.50 7.08 6 2.55 90.0 36 36 3.0 4 @ 18.94 Rolled Beam
0 ft. to 13.25 ft. W36X160

13.25 ft. to W36X160

37.88 ft. 11X3/4 Bott
(Symm.) Cover Pl
411
(continued)
412
Materials
0332 71.58 6.00 1 @ 4.52 5 0.96 at G1 90.0 36 36 3.0 1 @ 1.17 Rolled Beam
1 @ 4.50 1.08 at G2 1 @ 13.66 W36X160
1 @ 4.52 3 @ 13.75
1 @ 4.54 1 @ 14.00
1 @ 1.50
0337 78.45 9.44 5.49 11 2.54 130.7 50 50 3.0 G1 Rolled Beam
G1/G2/G3
1 @ 16.35
1 @ 6.56 W690X323
2 @ 20.01
1 @ 15.52
G2
1 @ 11.62
1 @ 6.56
2 @ 20.01
1 @ 20.25
G3
1 @ 13.46
3 @ 20.01
1 @ 4.96
0345 100.00 8.00 9.50 5 3.00 90.0 50 46/50 (Top)46 3.0 G1/G2 Welded I-Shape
(Bott) 4 @ 25.00 G1/G2
0 ft. to 22.75 ft. 12X3/4 Top Flg
(50 ksi)
22X1 Bott Flg
44X3/8 Web

25.75 ft. (50 ksi)
22X1 1/2 Bott Flg
44X3/8 Web
25.75 ft. to 12X1 1/2 Top Flg

74.75 ft. (46 ksi)
22X1 1/2 Bott Flg
44X3/8 Web

77.75 ft. (50 ksi)
22X1 1/2 Bott Flg
44X3/8 Web

100.00 ft. (50 ksi)
22X1 Bott Flg
44X3/8 Web

(continued)
Materials
0347 109.58 6.50 7.00 6 2.79 90.0 36 36 3.3 G1/G2 Welded I-Shape
5 @ 21.92 G1/G2
0 ft. to 21.92 ft. 16X1 1/4 Top Flg
18X1 1/8 Bott Flg
44X1/2 Web
21.9 ft. to 16X1 1/4 Top Flg

54.79 ft. 18X1 7/8 Bott Flg
(Symm.) 44X1/2 Web
0348 109.00 7.50 7.67 8 3.50 at G1 56.0 36 36 3.5 G11/G15 Welded I-Shape
3.38 at G8 1 @ 18.16 G11/G15
1 @ 5.18 0 ft. to 28.00 ft. 14X3/4 Top Flg
1 @ 17.32 16X1 1/2 Bott Flg
1 @ 5.18 54X1/2 Web
1 @ 17.32
28.00 ft. to 14X1 1/4 Top Flg
1 @ 5.18
54.50 ft. 16X2 Bott Flg
1 @ 17.32
(Symm.) 54X1/2 Web
1 @ 5.18
1 @ 18.16
0349 115.00 8.75 10.33 4 3.83 90.0 50 36 4.5 5 @ 23.00 Welded I-Shape
0 ft. to 45.00 ft. 14X1 Top Flg
18X1 1/2 Bott Flg
48X3/8 to 70X3/8
Web (Varies)
45.00 ft. to 14X1 Top Flg

57.50 ft. 18X1 1/2 Bott Flg
(Symm.) 70X3/8 Web
0356 99.48 9.00 5.36 12 1.21 at G1 50.5 33 32 4.0 G1 Built-up I-Shape
1.19 at 1 @ 14.26 G1
G12 3 @ 23.78 0 ft. to 24.74 ft. L3.10X6.33X1 Top
1 @ 12.72 Angles
1 @ 1.16 L4X6X5/8 Bott
G2 Angles
1 @ 18.66 14X3/8 Bott
3 @ 23.78 Cover Pl
1 @ 8.30 53 1/4X3/8 Web
1 @ 1.16 24.74 ft. to L3.10X6.33X1 Top
G3 49.74 ft. Angles
1 @ 23.08 (Symm.) L4X6X5/8 Bott
3 @ 23.78 Angles
1 @ 3.90 14X3/4 Bott
1 @ 1.16 Cover Pl
53 1/4X3/8 Web
413
(continued)
414
Materials
0356 G2
(cont.) 0 ft. to 23.24 ft. L3.10X6.33X1 Top
Angles
L4X6X5/8 Bott
Angles
14X1/2 Bott
Cover Pl
53 1/4X3/8 Web
23.24 ft. to L3.10X6.33X1 Top

49.74 ft. Angles
(Symm.) L4X6X5/8 Bott
Angles
14X1 Bott Cover Pl
53 1/4X3/8 Web
G3
0 ft. to 22.74 ft. L3.10X6.33X1 Top
Angles
L4X6X5/8 Bott
Angles
14X1/2 Bott
Cover Pl
53 1/4X3/8 Web
22.74 ft. to L3.10X6.33X1 Top

49.74 ft. Angles
Angles
14X1 Bott Cover Pl
53 1/4X3/8 Web

(continued)
Materials
0357 110.00 10.00 6.00 5 1.50 90.0 36 36 3.0 G1/G212 @ Built-up I-Shape
18.33 G1/G2
0 ft. to 15.50 ft. L4.60X6.27X1.14
Top Angles
L6X6X3/4 Bott
Angles
58 1/2X1/2 Web
15.50 ft. to L4.60X6.27X1.14

22.50 ft. Top Angles
L6X6X3/4 Bott
Angles
14X7/16 Bott
Cover Pl
58 1/2X1/2 Web
22.50 ft. to L4.60X6.27X1.14

L6X6X3/4 Bott
Angles
14X7/8 Bott
Cover Pl
58 1/2X1/2 Web
29.00 ft. to L4.60X6.27X1.14

Angles
14X1 5/16 Bott
Cover Pl
58 1/2X1/2 Web
415
(continued)
416
Materials
0358 90.00 9.00 7.04 10 2.13 at G1 128.5 33 33 3.0 G1/G2 Rolled Beam
3.00 at 4 @ 22.50 G1
G10 G10 0 ft. to 0.40 ft. 36WF(CB362),
1 @ 16.60 36X16.5X230
2 @ 22.50
0.40 ft. to 36WF(CB362),
1 @ 28.40
90.00 ft. 36X16.5X230
18X1 Bott Cover Pl
G2
0 ft. to 10.00 ft. 36WF(CB362),
36X16.5X230
10.00 ft. to 36WF(CB362),
45.00 ft. 36X16.5X230
Cover Pl
G10
0 ft. to 11.00 ft. 36WF(CB362),
36X16.5X230
11.00 ft. to 36WF(CB362),
45.00 ft. 36X16.5X230
Cover Pl
0359 115.00 7.50 8.50 6 2.60 144.9 33 33 3.0 G1 Rolled Beam
1 @ 0.43 G1/G6
1 @ 15.05 0 ft. to 1.00 ft. 21X1 1/8 Top Flg
1 @ 24.21 21X1 13/16 Bott Flg
1 @ 24.07 56X0.319 Web
2 @ 24.08
1 @ 2.65 1.00 ft. to 21X1 1/8 Top Flg
1 @ 0.43 25.17 ft. 21X1 13/16 Bott Flg
G2 56X3/8 Web
1 @ 0.43 25.17 ft. to 21X1 9/16 Top Flg
1 @ 2.93 57.50 ft. 21X2 3/8 Bott Flg
1 @ 12.12 (Symm.) 56X3/8 Web
1 @ 12.09
1 @ 0.08 G2/G3/G4 21x15/16 Top Flg
1 @ 24.00 0 ft. to 25.17 ft. 21X1 3/4 Bott Flg
1 @ 0.21 56X3/8 Web
1 @ 23.87
1 @ 0.21 25.17 ft. to 21X1 3/8 Bott Flg
1 @ 23.87 57.50 ft. 21X2 5/16 Top Flg
1 @ 0.21 (Symm.) 56X3/8 Web
1 @ 11.97
1 @ 2.58
1 @ 0.43

(continued)
Materials
0359 G3 G5
(cont.) 1 @ 0.43 0 ft. to 25.17 ft. 21x15/16 Top Flg
1 @ 0.04 21X1 3/4 Bott Flg
1 @ 2.89 56X3/8 Web
1 @ 12.12
1 @ 0.05 25.17 ft. to 21X15/16 Bott Flg
1 @ 24.16 57.50 ft. 21X2 5/16 Top Flg
1 @ 0.05 (Symm.) 56X3/8 Web

1 @ 24.03
1 @ 0.05
1 @ 24.03
1 @ 0.05
1 @ 11.98
1 @ 12.05
1 @ 2.64
1 @ 0.43
G4
1 @ 0.47
1 @ 2.89
1 @ 12.12
1 @ 12.09
1 @ 0.08
1 @ 24.00
1 @ 0.08
1 @ 24.00
1 @ 0.08
1 @ 24.00
1 @ 0.08
1 @ 12.16
1 @ 2.52
1 @ 0.43
G5
1 @ 0.47
1 @ 2.89
1 @ 12.12
1 @ 0.05
1 @ 24.03
1 @ 0.05
1 @ 24.03
1 @ 0.05
1 @ 24.03
1 @ 0.05
1 @ 12.16
1 @ 11.87
1 @ 2.78
1 @ 0.42
417
(continued)
418
Materials
0359 G6
(cont.) 1 @ 0.47
1 @ 2.89
4 @ 24.08
1 @ 14.90
1 @ 0.43
0366 100.00 9.50 6.83 12 2.42 90.0 50 50 4.0 G1/G4/G5/G6 Rolled Beam
5 @ 20.00 0 ft. to 15.00 ft. W36X230
G2/G3/G7/
15.00 ft. W36X230
G8/
to 50.00 ft. 15X1 1/2 Bott
G9/G10
(Symm.) Cover Pl
10 @ 10.00
G11/G12
20 @ 5.00
0367 138.58 8.00 8.45 7 4.25 121.7 50 50 4.0 Fascia Welded I-Shape
1 @ 19.75 0 ft. to 34.42 ft. 13X1 Top Flg
1 @ 20.00 13X1 Bott Flg
4 @ 22.50 71X1/2 Web
1 @ 8.83
Interior
93.92 ft. 13X1 1/2 Bott Flg
1 @ 14.55
71X1/2 Web
1 @ 20.00
4 @ 22.50 93.92 ft. to 13X1 Top Flg
1 @ 14.04 138.58 ft. 13X1 Bott Flg
71X1/2 Web
0368 178.20 9.00 8.33 7 2.33 at G1 84.2 50 50 4.0 1 @ 3.00 Welded I-Shape
1.37 at G7 1 @ 19.26 0 ft. to 28.10 ft. 18x1 1/4 Top Flg
6 @ 22.28 18X1 1/4 Bott Flg
1 @ 19.26 72X5/8 Web
1 @ 3.00
28.10 ft. to 18x1 3/4 Top Flg
89.10 ft. 24X1 7/8 Bott Flg
(Symm.) 72X1/2 Web
0369 136.85 9.00 7.25 7 2.00 91.1 50 50 4.0 1 @ 3.59 Welded I-Shape
1 @ 19.23 0 ft. to 28.00 ft. 14X3/4 Top Flg
4 @ 22.81 16X1 3/8 Bott Flg
1 @ 20.79 54X1/2 Web
1 @ 2.00
28.00 ft. to 14X1 1/8 Top Flg
68.43 ft. 16X1 3/4 Bott Flg
(Symm.) 54X1/2 Web

(continued)
Materials
0370 153.61 6.50 5 @ 7.252 8 2.50 at G1 22.5 36 36 3.0 G1 Welded I-Shape

@ 6.38 2.46 at G8 1 @ 3.00 G1
5 @ 25.00 0 ft. 19.25 ft. 20X1 Top Flg
1 @ 25.61 22X1 Bott Flg
G2 60X3/8 Web
1 @ 3.95 19.25 ft. to 20X2 1/4 Top Flg
1 @ 16.72 76.81 ft. 22X2 1/4 Bott Flg

1 @ 8.28 (Symm.) 60X3/8 Web
1 @ 16.72
1 @ 8.28 G2
1 @ 16.72 0 ft. 20.50 ft. 20X1 Top Flg
1 @ 8.28 22X1 Bott Flg
1 @ 16.72 60X3/8 Web
1 @ 8.28 20.50 ft. to 20X2 1/4 Top Flg
1 @ 16.72 76.81 ft. 22X2 1/4 Bott Flg
1 @ 8.28 (Symm.) 60X3/8 Web
1 @ 16.72
1 @ 7.94 G5
G5 0 ft. 20.50 ft. 20X1 Top Flg
1 @ 0.83 22X1 Bott Flg
1 @ 17.36 60X3/8 Web
1 @ 7.64 20.50 ft. to 20X2 Top Flg
1 @ 17.36 76.81 ft. 22X2 Bott Flg
1 @ 7.64 (Symm.) 60X3/8 Web
1 @ 17.36
1 @ 7.64 G6
1 @ 17.36 0 ft. 21.50 ft. 20X1 Top Flg
1 @ 7.64 22X1 Bott Flg
1 @ 17.36 60X3/8 Web
1 @ 7.64
1 @ 17.36
1 @ 10.42 (Symm.) 60X3/8 Web
G6
1 @ 0.98 Welded I-Shape
1 @ 17.36 G7
1 @ 7.64 0 ft. 20.00 ft. 20X1 Top Flg
1 @ 17.36 22X1 Bott Flg
1 @ 7.64 60X3/8 Web
1 @ 17.36 20.00 ft. to 20X1 3/4 Top Flg
1 @ 7.64 76.81 ft. 22X2 1/8 Bott Flg
1 @ 17.36 (Symm.) 60X3/8 Web
1 @ 7.64
1 @ 17.36 G8
1 @ 7.64 0 ft. 20.00 ft. 20X1 Top Flg
1 @ 17.64 22X1 Bott Flg
60X3/8 Web
419
1 @ 10.27

(continued)
420
Materials
0370 G7 20.00 ft. to 20X1 3/4 Top Flg

(cont.) 1 @ 1.11 76.81 ft. 22X2 1/8 Bott Flg
1 @ 15.25 (Symm.) 60X3/8 Web
1 @ 9.75
1 @ 15.25
1 @ 9.75
1 @ 15.25
1 @ 9.75
1 @ 15.25
1 @ 9.75
1 @ 15.25
1 @ 9.75
1 @ 15.25
1 @ 12.25
G8
1 @ 16.50
5 @ 25.00
1 @ 12.11
0371 120.00 9.00 7.40 10 2.98 at G1 90.0 50 50 4.0 G1/G6/G10 Welded I-Shape
3.00 at 1 @ 2.50 G1/G6/G10
G10 5 @ 23.50 0 ft. to 21.00 ft. 12X1 1/8 Top Flg
16X1 1/4 Bott Flg
42X1/2 Web
21.00 ft. to 12X1 1/8 Top Flg

60.00 ft. 16X2 1/2 Bott Flg
(Symm.) 42X1/2 Web
0372 163.38 7.00 8.50 4 3.08 45.0 33 33 3.0 G2 Welded I-Shape
1 @ 10.92 G2
1 @ 8.50 0 ft. to 21.67 ft. 12X13/16 Top Flg
1 @16.50 20X1 3/8 Bott Flg
1 @ 8.50 84X1/2 Web
1 @ 16.50 21.67 ft. to 20X1 3/4 Top Flg
1 @ 8.50 46.67 ft. 20X2 3/4 Bott Flg
1 @ 16.50 84X1/2 Web
1 @ 8.50
1 @ 16.50 46.67 ft. to 20X2 Top Flg
1 @ 8.50 84.00 ft. 20X3 Bott Flg
1 @ 16.50 84X1/2 Web
1 @ 8.50
84.00 ft. to 20X1 3/4 Top Flg
1 @ 18.96
141.71 ft. 20X2 3/4 Bott Flg
84X1/2 Web
141.71 ft. to 20X13/16 Top Flg
163.38 ft. 20X1 3/8 Bott Flg
84X1/2 Web

(continued)
Materials
0373 146.92 9.00 6.50 8 2.25 120.0 50 50 4.0 G1 Welded I-Shape

1 @ 10.95 15X1 Top Flg
7 @ 18.91 21X1 5/8 Bott Flg
1 @ 3.60 59X5/8 Web
G2
1 @ 7.20
7 @ 18.91
1 @ 7.35
0374 150.75 8.00 7.50 5 1.44 56.5 36 36 3.0 G1 Welded I-Shape
1 @ 1.75 G1
1 @ 3.82 0 ft. to 25.00 ft. 22X15/16 Top Flg
1 @ 21.50 22X27/32 Bott Flg
4 @ 25.00 78X0.48 Web
1 @ 23.68
25.00 ft. to 22X15/16 Top Flg
41.00 ft. 22X1.74 Bott Flg
78X0.48 Web
41.00 ft. to 22X1 3/16 Top Flg

109.75 ft. 22X1.74 Bott Flg
78X0.48 Web
109.75 ft. to 22X15/16 Top Flg

150.75 ft. 22X27/32 Bott Flg
78X0.48 Web
0375 172.50 7.50 9.08 7 1.75 81.7 50 46 (Top)46/42 3.0 G2 Welded I-Shape
(Bott) 1 @ 23.00 G2
1 @ 1.32 0 ft. to 30.25 ft. 22X1 Top Flg
1 @ 23.68 24X1 Bott Flg
1 @ 1.32 (46 ksi)
1 @ 23.68 76X9/16 Web
1 @ 1.32
1 @ 23.68
48.25 ft. 24X1 7/8 Bott Flg
1 @ 1.32
(42 ksi)
1 @ 23.68
76X9/16 Web
1 @ 1.32
1 @ 23.68 48.25 ft. to 22X1 3/8 Top Flg
1 @ 1.32 86.25 ft. 24X1 7/8 Bott Flg
1 @ 23.18 (Symm.) (42 ksi)
76X9/16 Web
421
(continued)
422
Materials
0376 132.89 6.50 2 @ 6.386 9 3.13 at G1 25.0 36 36 3.0 G1 Welded I-Shape

@ 8.04 1.83 at G9 1 @ 10.62 G1/G2
4 @ 24.88 0 ft. to 22.00 ft. 16X1 Top Flg
1 @ 22.75 20X1 Bott Flg
G2 60X3/8 Web
1 @ 10.62
22.00 ft. to 16X1 5/8 Top Flg
1 @ 13.67
1 @ 11.20
(Symm.) 60X3/8 Web
1 @ 13.67
1 @ 11.20 G3/G4/G9
1 @ 13.67 0 ft. to 22.00 ft. 16X1 Top Flg
1 @ 11.20 20X1 Bott Flg
1 @ 13.67 60X3/8 Web
1 @ 11.20
1 @ 13.67 22.00 ft. to 16X2 Top Flg
1 @ 9.08 66.46 ft. 20X1 7/8 Bott Flg
G3 (Symm.) 60X3/8 Web
1 @ 7.28
1 @ 17.01
1 @ 7.81
1 @ 17.07
1 @ 7.76
1 @ 17.11
1 @ 7.72
1 @ 17.16
1 @ 7.67
1 @ 17.21
1 @ 9.09
G4
1 @ 7.28
1 @ 17.25
1 @ 7.58
1 @ 17.25
1 @ 7.58
1 @ 17.25
1 @ 7.58
1 @ 17.25
1 @ 7.58
1 @ 17.25
1 @ 9.04
G9
1 @ 24.53
4 @ 24.83
1 @ 9.04

(continued)
Materials
0377 169.83 7.00 9.50 5 2.00 31.5 50 50 3.0 G2 Welded I-Shape

1 @ 4.38 G2/G3
1 @ 16.13 0 ft. to 30.00 ft. 24X1 Top Flg
4 @ 25.00 24X1 Bott Flg
1 @ 16.13 78X9/16 to
1 @ 0.31 73X9/16 Web
1 @ 15.83 (Varies)
1 @ 0.61
1 @ 16.44
G3
73X9/16 Web
1 @ 3.76
2 @ 16.13 49.00 ft. to 24X1 7/16 Top Flg
4 @ 25.00 84.92 ft. 24X2 Bott Flg
2 @ 16.13 (Symm.) 73X9/16 Web
1 @ 1.55
0378 140.30 8.50 1 @ 8.004 11 2.42 at G1 76.5 50 45/50 3.3 G8 Welded I-Shape
@ 6.855 2.04 at (Top)42/46 1 @ 24.41 G8
@ 8.25 G11 (Bott) 4 @ 26.00 0 ft. to 28.65 ft. 18X3/4 Top Flg
1 @ 11.89 (50 ksi)
G10 17X7/8 Bott Flg
1 @ 28.37 (45 ksi)
4 @ 26.00 86X5/8 Web
1 @ 7.93 28.65 ft. to 18X3/4 Top Flg
G11 45.65 ft. (50 ksi)
1 @ 30.35 17X1 3/4 Bott Flg
4 @ 26.00 (42 ksi)
1 @ 5.95 86X5/8 Web
70.15 ft. (45 ksi)
(Symm.) 17X1 3/4 Bott Flg
(42 ksi)
86X5/8 Web
G10
0 ft. to 29.15 ft. 18X3/4 Top Flg
(50 ksi)
16X3/4 Bott Flg
(50 ksi)
86X5/8 Web
70.15 ft. (50 ksi)
(45 ksi)
86X5/8 Web
423
(continued)
424
Materials
0378 G11
(cont.) 0 ft. to 35.65 ft. 18X3/4 Top Flg
(50 ksi)
18X3/4 Bott Flg
(50 ksi)
86X5/8 Web

70.15 ft. (50 ksi)
(45 ksi)
86X5/8 Web
0379 212.00 9.50 10.25 12 2.21 66.5 70 70 5.0 1 @ 22.00 Welded I-Shape
8 @ 21.00 0 ft. to 1.25 ft. 21X1 1/2 Top Flg
1 @ 22.00 16X1 3/4 Bott Flg
84X3/4 Web
1.25 ft. to 21X1 1/2 Top Flg

48.00 ft. 27X1 3/4 Bott Flg
84X3/4 Web
48.00 ft. to 21X2 1/8 Top Flg

106.00 ft. 27X2 3/8 Bott Flg
(Symm.) 84X3/4 Web
0380 126.00 8.50 9.75 6 2.96 90.0 50 50 4.0 G1/G2 Welded I-Shape
6 @ 21.00 G1/G2
0 ft. to 37.80 ft. 16X1 Top Flg
20X1 5/8 Bott Flg
60X9/16 Web
37.80 ft. to 18X1 5/8 Top Flg

(Symm.) 60X9/16 Web

(continued)
Materials
0381 204.00 8.50 9.67 6 3.42 95.0 50 50 4.0 G1 Welded I-Shape

9 @ 22.67 G1/G2
G2 0 ft. to 62.00 ft. 18X1 3/4 Top Flg
1 @ 21.82 21X2 3/4 Bott Flg
1 @ 0.85 78X5/8 Web
1 @ 21.82
62.00 ft. to 25X1 3/4 Top Flg

1 @ 0.85 102.00 ft. 25X2 3/4 Bott Flg
1 @ 21.82 (Symm.) 78X5/8 Web
1 @ 0.85
1 @ 21.82
1 @ 0.85
1 @ 21.82
1 @ 0.85
1 @ 21.82
1 @ 0.85
1 @ 21.82
1 @ 0.84
1 @ 21.82
1 @ 0.84
1 @ 22.66
425
426
F.7 Steel Fatigue Database
Table F.7.1. Fatigue Test Data
Units Load Type Detail ID Number Stress Range Cycles Result Min. Stress Max. Stress Yield Stress
Imp. Const. 40 0NC1CRA131 16 393000 crack -6 10 36

Imp. Const. 40 0NC1CWA132 16 393000 crack -6 10 36
Imp. Const. 40 0NC1CRA221 12 798000 crack 2 14 36
Imp. Const. 40 0NC1CWA222 12 655000 crack 2 14 36
Imp. Const. 40 0NC1CRB131 16 418000 crack -6 10 50
Imp. Const. 40 0NC1CWB132 16 356000 crack -6 10 50

427
Table F.7.1. Fatigue Test Data (continued)

Imp. Const. 40 0NC1CBW143 20 171000 crack -6 14 50
Imp. Const. 40 0NC1CRB221 12 842000 crack 2 14 50
Imp. Const. 40 0NC1CWB222 12 667000 crack 2 14 50
Imp. Const. 40 0NC1CWC313 8 2278000 crack 10 18 50
Imp. Const. 40 0NC1CRC131 16 395000 crack -6 10 100
Imp. Const. 40 0NC1CWC132 16 483000 crack -6 10 100

428

Imp. Const. 40 0NC1CRC221 12 844000 crack 2 14 100

429

Imp. Const. 30 0NC*CRB141 20 187000 crack -6 14 50
Imp. Const. 30 0NC*CWB222 12 667000 crack 2 14 50

430

Imp. Const. 30 0NC*CWC222 12 848000 crack 2 14 100

431

Imp. Const. 40 0NC1CMA131 16 427000 crack -6 10 36
Imp. Const. 40 0NC1CMA221 12 904000 crack 2 14 36

432

Imp. Const. 40 0NC1CTA131 16 320000 crack -6 10 36
Imp. Const. 40 0NC1CTA221 12 949000 crack 2 14 36
Imp. Const. 60 0NC1CBA131 16 353000 crack -6 10 36
Imp. Const. 60 0NC*CBA132 16 276000 crack -6 10 36

433

Imp. Const. 60 0NC1CBA221 12 1769000 crack 2 14 36

434

Imp. Const. 27 0NC1PWA131 30 677000 crack -10 20 36
Imp. Const. 20 0NC1PWA221 24 1577000 crack 2 26 36
Imp. Const. 20 0NC1PWA311 18 10200000 runout 14 32 36
Imp. Const. 20 0NC1PWB131 30 855000 crack -10 20 50
Imp. Const. 26 0NC1PWB221 24 1292000 crack 2 26 50

435

Imp. Const. 27 0NC1PWC131 30 783000 crack -10 20 100
Imp. Const. 20 0NC1PWC221 24 2228000 crack 2 26 100
Imp. Const. 15 0NC1PRA131 30 1505000 crack -10 20 36
Imp. Const. 10 0NC1PRA132 30 4910000 runout -10 20 36

436
Imp. Const. 10 0NC1PRA231 30 12200000 runout 2 32 36

Imp. Const. 10 0NC1PRA232 30 10500000 runout 2 32 36
Imp. Const. 10 0NC1PRA241 36 855000 crack 2 38 36
Imp. Const. 10 0NC1PRA242 36 998000 crack 2 38 36
Imp. Const. 15 0NC1PRB141 36 1207000 crack -10 26 50
Imp. Const. 10 0NC1PRB231 30 2677000 crack 2 32 50
Imp. Const. 20 0NC1FSA221 15.8 5808000 crack 1.7 -1 36
Imp. Const. 20 0NC1FSA212 17.8 7754000 crack 2 -1 36
Imp. Const. 20 0NC1FSA213 18 6413000 crack 2 -1 36
Imp. Const. 20 0NC1FSA311 18 10500000 runout 14 -1 36
Imp. Const. 20 0NC1FSA121 22.7 2908000 crack -9.5 -1 36

437

Imp. Const. 21 0NC1FSA141 34 883000 crack 9.4 -1 36
Imp. Const. 20 0NC1FSA143 31 914000 crack -8.6 -1 36
Imp. Const. 20 0NC1FSB312 13 10550000 crack 10.2 -1 50
Imp. Const. 20 0NC1FSB221 22.6 1859000 crack 1.9 -1 50
Imp. Const. 20 0NC1FSB222 24 1917000 crack 2 -1 50
Imp. Const. 20 0NC1FSB223 24.1 1035000 crack 2 -1 50
Imp. Const. 20 0NC1FSB322 24 10500000 runout 14 -1 50
Imp. Const. 20 0NC1FSB133 27.9 1523000 crack -9.3 -1 50

438

Imp. Const. 20 0NC1FSC311 12.5 3704000 crack 9.7 -1 100
Imp. Const. 20 0NC1FSC222 24.7 2479000 crack 2 -1 100
Imp. Const. 20 0NC1FSC223 15 5706000 crack 1.2 -1 100
Imp. Const. 20 0NC1FSC323 24 1102000 crack 14 -1 100
Imp. Const. 20 0NC1FSC133 29.2 887000 crack -9.7 -1 100

439

Imp. Const. 80 0NC1FSA142 36 734000 crack -10 -1 36
Imp. Const. 80 0NC1FSB131 28 1288000 crack -9.3 -1 50
Imp. Const. 80 0NC1FSB141 36 839000 crack -10 -1 50
Imp. Const. 80 0NC1FSB142 36 733000 crack -10 -1 50
Imp. Const. 70 0NC1FSC312 18.4 9840000 runout 14 -1 100
Imp. Const. 70 0NC1FSC313 18 12200000 runout 14 -1 100
Imp. Const. 80 0NC1FSC313 18 12200000 runout 14 -1 100
Imp. Const. 80 0NC1FSC132 26 1592000 crack -8.7 -1 100

440
Imp. Const. 80 0NC1FSC133 30 1239000 crack -10 -1 100

Imp. Const. 80 0NC1FSC142 36 483000 crack -10 -1 100
Imp. Const. 70 0NC1FSC143 35.8 463000 crack -10 -1 100
Imp. Const. 17 0NC1PRB141 36 4799000 crack -10 -1 50
Imp. Const. 10 0NC1PRB231 30 10800000 runout 2 -1 50
Imp. Const. 10 0NC1PRB241 36 1857000 crack 2 -1 50
Imp. Const. 90 0NC2SCB211 13.7 2893000 crack 1.5 -1 50
Imp. Const. 90 0NC2SCB212 13.7 9740000 runout 1.5 -1 50
Imp. Const. 90 0NC2SCB131 22.8 1119000 crack -7.6 -1 50

441

Imp. Const. 90 0NC2SAB111 17.6 4770000 crack 4.4 -1 50
Imp. Const. 90 0NC2SGB211 15.4 4433000 crack 1.7 -1 50
Imp. Const. 90 0NC2SGC212 15.4 3016000 crack 1.7 -1 100
Imp. Const. 90 0NC2SGB311 15.4 4869000 crack 12 -1 50
Imp. Const. 90 0NC2SGC333 25.7 401000 crack 12 -1 100
Imp. Const. 90 0NC2SBB221 20.5 1264000 crack 1.7 -1 50

442

Imp. Const. 90 0NC2SGB211 13.7 4430000 runout 1.5 -1 50

443

Imp. Const. 100 0NC2SCB131 23.9 584000 crack -8 -1 50
Imp. Const. 100 0NC2SGC212 13.8 3850000 runout 1.5 -1 100

444

Imp. Const. 130 0NC2AQB221 12 10800000 runout 2 -1 50
Imp. Const. 130 0NC2AQB321 12 15600000 runout 2 -1 50
Imp. Const. 130 0NC2AQB131 16 3095000 crack -6 -1 50
Imp. Const. 130 0NC2AQB231 16 3113000 crack 2 -1 50
Imp. Const. 130 0NC2A2B221 12 3812000 crack 2 -1 50
Imp. Const. 130 0NC2A8B221 8.9 2960000 crack 1.5 -1 50

445

Imp. Const. 130 0NC2A8B131 16 595000 crack -6 -1 50
Imp. Const. 135 0NC2A4B311 5.9 13600000 runout 7.4 -1 50

446

Imp. Const. 11 0NC2PRC041 36 3113000 crack -22 -1 100
Imp. Const. 10 0NC2PRC241 34.3 10000000 runout 1.9 -1 100
Imp. Const. 10 0NC2PRC261 45.7 702000 crack 1.9 -1 100
Imp. Const. 16 0NC2PRC262 45 338000 crack 2 -1 100

447
Imp. Const. 10 0NC2PRC243 36 10000000 runout 2 -1 100

Imp. Const. 10 0NC2PRC344 36 10000000 runout 14 -1 100
Imp. Const. 16 0NC2PRC442 20.5 2846000 crack 26 -1 100
Imp. Const. 20 0NC2SGB211 18 4433000 crack 2 -1 50
Imp. Const. 20 0NC2SGC212 18 2194000 crack 2 -1 100
Imp. Const. 20 0NC2SBB221 24 1197000 crack 2 -1 50
Imp. Const. 40 1NC3010002 8 950000 crack -1 -1 50

448

Imp. Const. 40 1NC3010025 6 21500000 runout -1 -1 50
Imp. Const. 230 1NC4A10101 21 160000 crack -1 -1 36

449
Imp. Const. 230 1NC4A10108 9 9480000 runout -1 -1 36

Imp. Const. 256 1NC4A50203 12.2 2380000 runout -1 -1 36

450

Imp. Const. 120 1NC4A60301 5.7 4330000 crack -1 -1 36

451

Imp. Var. 120 1NC600A1WW 3.6 22900000 crack -2 -1 50
Imp. Var. 120 1NC600A1MW 3.6 17800000 crack -2 -1 50
Imp. Var. 120 1NC600A1EW 3.6 17800000 crack -2 -1 50
Imp. Var. 120 1NC600A1EE 3.6 15500000 crack -2 -1 50
Imp. Var. 120 1NC600A2WE 4.2 19000000 crack -2 -1 50
Imp. Var. 120 1NC600A2EE 4.2 16500000 crack -2 -1 50
Imp. Var. 120 1NC600B1WW 2.2 71000000 crack -2 -1 50
Imp. Var. 120 1NC600B1EE 2.2 70000000 crack -2 -1 50
Imp. Var. 120 1NC600B2WW 2.6 40500000 crack -2 -1 50
Imp. Var. 120 1NC600B2WE 2.6 31200000 crack -2 -1 50
Imp. Var. 120 1NC600B2MW 2.6 31200000 crack -2 -1 50
Imp. Var. 120 1NC600B2ME 2.6 35700000 crack -2 -1 50
Imp. Var. 120 1NC600B2EW 2.6 28100000 crack -2 -1 50
Imp. Var. 120 1NC600C1WW 2.6 13600000 crack -2 -1 50
Imp. Var. 120 1NC600C1WW 3.5 13600000 crack -1 -1 50
Imp. Var. 120 1NC600C1WE 3.5 11900000 crack -1 -1 50
Imp. Var. 120 1NC600C1MW 3.5 9100000 crack -1 -1 50
Imp. Var. 120 1NC600C1EW 3.5 13000000 crack -1 -1 50
Imp. Var. 120 1NC600C1EE 3.5 16800000 crack -1 -1 50
Imp. Var. 120 1NC600C2WW 4 16100000 crack -1 -1 50
Imp. Var. 120 1NC600C2WE 4 15700000 crack -1 -1 50
Imp. Var. 120 1NC600C2MW 4 13000000 crack -1 -1 50
Imp. Var. 120 1NC600C2EW 4 13600000 crack -1 -1 50
Imp. Var. 120 1NC600C2EE 4 10700000 crack -1 -1 50
Imp. Var. 120 1NC600D1WW 2.9 103600000 crack -1 -1 50
Imp. Var. 120 1NC600D1WE 2.9 116200000 crack -1 -1 50
Imp. Var. 120 1NC600D1MW 2.9 112200000 crack -1 -1 50
Imp. Var. 120 1NC600D1ME 2.9 112200000 crack -1 -1 50
Imp. Var. 120 1NC600D1EE 2.9 103600000 crack -1 -1 50
Imp. Var. 120 1NC600D1EW 2.9 127000000 crack -1 -1 50
Imp. Var. 120 1NC600D2WW 2.9 110000000 crack -1 -1 50
Imp. Var. 120 1NC600D2WE 2.9 110000000 crack -1 -1 50
Imp. Var. 120 1NC600D2MW 2.9 110000000 crack -1 -1 50
Imp. Var. 120 1NC600D2ME 2.9 110000000 crack -1 -1 50
Imp. Var. 120 1NC600D2EW 2.9 110000000 crack -1 -1 50
Imp. Var. 120 1NC600D2EE 2.9 110000000 crack -1 -1 50
Imp. Var. 40 1NC600A101 3.4 25900000 crack -2 -1 50

452
Imp. Var. 40 1NC600A201 3.6 17500000 crack -2 -1 50

Imp. Var. 40 1NC600B101 2.1 200000000 crack -2 -1 50
Imp. Var. 40 1NC600B201 2.2 120000000 crack -2 -1 50
Imp. Var. 40 1NC600C101 3.7 7250000 crack -2 -1 50
Imp. Var. 40 1NC600C201 3.9 10700000 crack -2 -1 50
Imp. Var. 40 1NC600D101 2.8 150000000 crack -2 -1 50
Imp. Var. 40 1NC600D201 2.8 130000000 crack -2 -1 50
Imp. Var. 30 1NC600A102 3.4 30600000 crack -2 -1 50
Imp. Var. 30 1NC600A202 3.6 19500000 crack -2 -1 50
Imp. Var. 30 1NC600B102 2.1 200000000 crack -2 -1 50
Imp. Var. 30 1NC600B202 2.2 150000000 crack -2 -1 50
Imp. Var. 30 1NC600C102 3.7 7250000 crack -2 -1 50
Imp. Var. 30 1NC600C202 3.9 44000000 crack -2 -1 50
Imp. Var. 30 1NC600D102 2.8 140000000 crack -2 -1 50
Imp. Var. 30 1NC600D202 2.8 130000000 crack -2 -1 50
S.I. Const. 150 1JPN01N022 408 211000 crack 20 -1 600
S.I. Const. 150 1JPN01N007 300 2000000 runout 20 -1 600
S.I. Const. 150 1JPN01S014 303 878000 crack 20 -1 600
S.I. Const. 150 1JPN01SA33 300 819000 crack 20 -1 600
S.I. Const. 150 1JPN01M005 302 290000 crack 20 -1 600
S.I. Const. 150 1JPN01M0A1 249 2290000 crack 20 -1 600
S.I. Const. 150 1JPN01M0B5 250 1490000 crack 20 -1 600

453

S.I. Const. 150 1JPN01MC18 284 332000 crack 20 -1 600
S.I. Const. 150 1JPN01M0C5 236 696000 crack 20 -1 600
S.I. Const. 150 1JPN01LB18 300 348000 crack 20 -1 600
S.I. Const. 150 1JPN01L0C7 250 456000 crack 20 -1 600
S.I. Const. 150 1JPN01LC21 251 530000 crack 20 -1 600
S.I. Const. 150 1JPN01L0B6 181 1180000 crack 20 -1 600
S.I. Const. 150 1JPN01SBA1 140 2200000 crack 20 -1 600
S.I. Const. 150 1JPN01SBA2 120 4800000 runout 20 -1 600
S.I. Const. 150 1JPN01SBB1 300 266000 crack 20 -1 600
S.I. Const. 150 1JPN01SBC1 299 417000 crack 20 -1 600
S.I. Const. 160 1JPN1100S1 196 370000 crack -98 -1 580

454

S.I. Const. 160 1JPN1100S4 196 321000 crack 0 0 580
S.I. Const. 160 1JPN1100S9 147 4700000 runout -123 -5 580
S.I. Const. 150 1JPN0200T1 247 1762000 crack 10 -1 800
S.I. Const. 170 1JPN05AB01 135 1200000 crack -1 -1 800
S.I. Const. 170 1JPN5DEF01 150 1100000 crack -1 -1 580
S.I. Const. 170 1JPN05G001 115 2150000 crack -1 -1 800

455
S.I. Const. 150 1JPN09SB01 395 180000 crack 10 1 800

S.I. Const. 150 1JPN09SB07 250 2100000 runout 10 1 800
S.I. Const. 160 1JPN05JS01 200 470000 crack 300 0.6 800
S.I. Const. 160 1JPN05JS02 132 4000000 runout 198 0.6 800
S.I. Const. 160 1JPN05JS03 310 183000 crack 0 0 800
S.I. Const. 160 1JPN05JS09 135 3200000 runout 0 0 800
S.I. Const. 160 1JPN05JS11 390 112000 crack -195 -1 800
S.I. Const. 170 1JPN05BA01 156 890000 crack 0 0 800

456

S.I. Const. 170 1JPN05BB01 174 1200000 crack 0 0 800
S.I. Const. 170 1JPN05BB03 127 3000000 runout 0 0 800
S.I. Const. 170 1JPN05BE01 219 1010000 crack 0 0 800
S.I. Const. 171 1JPN05BD01 228 1040000 crack 0 0 800
S.I. Const. 172 1JPN05BF01 202 807000 crack 0 0 800
S.I. Const. 170 1JPN05BG01 175 1190000 crack -1 -1 500
S.I. Const. 170 1JPN05BG02 182 3000000 runout -1 -1 500
S.I. Const. 150 1JPN100201 392 405000 crack 10 -1 800
S.I. Const. 150 1JPN100204 284 2000000 runout 10 -1 800

457

S.I. Const. 150 1JPN12T102 449 128000 crack 10 0 800

458

S.I. Const. 160 1JPN130P01 245 172000 crack 27 -1 370
S.I. Const. 160 1JPN130P08 127 10000000 runout 27 -1 370
S.I. Const. 160 1JPN130W01 245 242000 crack 27 -1 370
S.I. Const. 160 1JPN130W09 147 5030000 runout 27 -1 370

459


460

S.I. Const. 200 1JPN13OW11 127 5340000 runout 27 -1 370

461

S.I. Const. 26 1ORE11156 188 2940000 crack 9 -1 331
S.I. Const. 20 1ORE11157 154 774000 runout 7 -1 331

462
S.I. Const. 26 1ORE11155S 166 666000 crack 6 -1 331

S.I. Const. 145 1ORE11157S 181 1150000 runout 8 -1 331
S.I. Const. 170 1ORE212321 166 818000 crack -1 -1 331
S.I. Var. 20 1ORE312572 155 5200000 crack -1 -1 331
S.I. Var. 25 1ORE312573 178 4118000 crack -1 -1 331
S.I. Var. 20 1ORE312574 93 10500000 runout -1 -1 331
S.I. Var. 140 1ORE312574 178 10450000 crack -1 -1 331
S.I. Const. 132 1ORE50A201 156 1633000 crack -1 -1 331
S.I. Const. 132 1ORE50B101 134 618000 crack -1 -1 331
S.I. Const. 190 1ORE6OR101 103 3121000 crack 10 -1 331
S.I. Const. 190 1ORE60R102 118 1702000 crack 10 -1 331
S.I. Const. 190 1ORE6OR104 118 1818000 crack 10 -1 331
S.I. Var. 100 1ORE80A601 55.9 24300000 crack -1 -1 331
S.I. Var. 100 1ORE80A602 55.9 27620000 crack -1 -1 331
S.I. Var. 140 1ORE712811 74.7 38500000 runout -1 -1 331
S.I. Var. 140 1ORE712812 74.7 19100000 runout -1 -1 331
S.I. Var. 140 1ORE712813 74.7 18200000 runout -1 -1 331
S.I. Var. 20 1ORE712811 65.6 38500000 runout -1 -1 331
S.I. Var. 21 1ORE712811 65.6 20750000 crack -1 -1 331
S.I. Var. 20 1ORE712812 65.6 19100000 runout -1 -1 331

463
S.I. Var. 20 1ORE712813 65.6 18200000 runout -1 -1 331

S.I. Var. 132 1ORE712811 36.5 12770000 crack -1 -1 331
S.I. Var. 132 1ORE712812 43.3 2335000 crack -1 -1 331
S.I. Var. 132 1ORE712813 45.6 2345000 crack -1 -1 331
S.I. Var. 132 1ORE712812 69 6367000 crack -1 -1 331
S.I. Const. 27 10RE11156 160 2833000 crack -5 -1 331
S.I. Const. 20 1ORE11156S 160 503000 runout 5 -1 331
S.I. Const. 170 1ORE212324 175 921000 runout -1 -1 331
S.I. Const. 160 1ENG014301 200 145000 crack 0 -1 262
S.I. Const. 160 1ENG014303 100 10000000 runout 0 -1 262
S.I. Const. 160 1ENG015013 260 100000 crack -130 -1 727
S.I. Const. 160 1ENGO15021 116 543000 crack 116 -1 727
S.I. Const. 200 1ENG01F301 220 120000 crack 0 -1 392

464

S.I. Const. 200 1ENG01F312 70 7800000 runout 0 -1 392
S.I. Const. 256 1GDR010301 220 226000 crack -1 -1 -1
S.I. Const. 256 1GDR010312 135 4100000 runout -1 -1 -1
S.I. Const. 256 1GDR010318 145 336000 crack -1 0.5 -1
S.I. Const. 256 1GDR010326 105 3280000 runout -1 0.5 -1

465


466


467


468

Imp. Const. 41 0NC1CWA133 16 336700 crack -10 -1 36
Imp. Const. 41 0NC1CRA141 20 192200 crack -14 -1 36
Imp. Const. 41 0NC1CRB131 16 418100 crack -10 -1 50
Imp. Const. 41 0NC1CWB132 16 356300 crack -10 -1 50
Imp. Const. 41 0NC1CRC131 16 394700 crack -10 -1 100
Imp. Const. 41 0NC1CWC132 16 482800 crack -10 -1 100

469

Imp. Const. 41 0NC1CTA141 20 160300 crack -14 -1 36
Imp. Const. 41 0NC1CTA243 20 226400 crack -22 -1 36
Imp. Const. 61 0NC1CBA132 16 275700 crack -10 -1 36
S.I. Const. 220 1MAR01F111 104 3517000 crack 3 107 345
S.I. Const. 220 1MAR01F113 104 10000000 runout 3 107 345

470


471

Imp. Const. 251 1USA01BA00 9 3370000 crack 10.8 0.55 36
Imp. Const. 251 1USA01B4A0 9 3277000 crack 10.8 0.55 36
Imp. Const. 251 1USA01B6A0 9 2823000 crack 10 0.53 50
Imp. Const. 251 1USA01B200 8.7 2024000 crack 17.3 0.67 100
Imp. Const. 251 1USA01B2A0 8.7 2866000 crack 17.3 0.67 100
Imp. Const. 130 1USA01B600 8 2955000 crack 16.4 0.67 50
Imp. Const. 40 1USA01B100 8 1806000 crack 18 0.7 100
Imp. Const. 40 1USA01B1A0 8 1134000 crack 18 0.7 100
Imp. Const. 260 1USA01B8A0 18 4180000 runout 8.7 0.33 100
Imp. Const. 100 1USA01B11A 15.7 1825000 crack 5 0.24 50
Imp. Const. 100 1USA01B700 13.6 8360000 runout 14.4 0.51 100
Imp. Const. 40 1USA02CP01 4 98000000 runout 10 -1 50
Imp. Const. 40 1USA02CP05 4.6 98000000 runout 10 -1 50

472
Imp. Const. 40 1USA02CM09 4.7 34930000 crack 9.3 -1 36

Imp. Const. 30 1USA01CM10 4.7 37710000 crack 9.3 -1 36
Imp. Const. 40 1USA02CP11 5 16610000 crack 10 -1 50
Imp. Const. 40 1USA02CW15 5 89200000 runout 24 -1 50
Imp. Const. 40 1USA02CT17 6 11420000 crack 10 -1 36
Imp. Const. 30 1USA01CT18 6 12160000 crack 10 -1 36
Imp. Const. 40 1USA02CW19 6 4327000 crack 10 -1 50
Imp. Const. 30 1USA01CW20 6 12160000 crack 10 -1 50
S.I. Const. 240 CDN010001 119 1190000 crack -1 -1 -1
Imp. Const. 120 1CDN021W01 5 15900000 runout 2.78 -1 36
Imp. Const. 120 1CDN022W01 10 1442000 crack 2.78 -1 36

473

Imp. Const. 232 1CDN021RG1 11 1329000 crack 2.78 -1 36
Imp. Const. 232 1CDN021RG2 11 2240000 runout 2.78 -1 36
Imp. Const. 232 1CDN024RG2 10 8090000 runout 2.78 -1 36
Imp. Const. 233 1CDN021T01 11 875000 crack 2.78 -1 36
Imp. Const. 233 1CDN02P101 13 714000 crack 2.78 -1 36
Imp. Const. 233 1CDN02P102 13 744000 crack 2.78 -1 36
Imp. Const. 234 1CDN021TG1 11 1163000 crack 2.78 -1 36
Imp. Const. 234 1CDN021TG2 11 1754000 crack 2.78 -1 36
Imp. Const. 10 2NC1620001 33.6 734000 crack 1 -1 33
Imp. Const. 10 2NC1620002 30 1760000 crack 1 -1 33
Imp. Const. 10 2NC1520001 53.8 336000 crack 1 -1.01 0
Imp. Const. 10 2NC1520002 45.8 3000000 runout 1 -1.01 0
Imp. Const. 10 2NC1420004 56 483000 crack 1 -1.01 0

474

Imp. Const. 10 2NC1420013 36 7340000 runout 1 -1.01 0
Imp. Const. 100 1USA040103 16.3 1460000 crack 1 -1 36
Imp. Const. 251 1USA03010H 11.5 760000 crack 1 -1 36
Imp. Const. 251 1USA03010I 11.5 950000 crack 1 -1 36
Imp. Const. 251 1USA03020E 11 1270000 crack 1 -1 36
Imp. Const. 251 1USA03020G 10.8 1860000 crack 1 -1 36
Imp. Const. 251 1USA03030E 10.8 1450000 crack 1 -1 36
Imp. Const. 251 1USA0303EE 8.7 2070000 crack 1 -1 36
Imp. Const. 251 1USA03040J 8.5 1320000 crack 1 -1 36
Imp. Const. 251 1USA03050O 11 1270000 crack 1 -1 36
Imp. Const. 251 1USA03050R 11.1 1300000 crack 1 -1 36
Imp. Const. 251 1USA0305BB 9.6 2140000 crack 1 -1 36
Imp. Const. 130 1USA03010S 11.8 1610000 crack 1 -1 36
Imp. Const. 130 1USA03010Y 11.9 1630000 crack 1 -1 36
Imp. Const. 130 1USA03030C 13.2 1020000 crack 1 -1 36
Imp. Const. 130 1USA03030L 13.9 1600000 crack 1 -1 36
Imp. Const. 130 1USA03040Q 10 1320000 crack 1 -1 36
Imp. Const. 130 1USA03050K 13.3 2140000 crack 1 -1 36
Imp. Const. 230 1USA03020T 10.7 1380000 crack 1 -1 36
Imp. Const. 230 1USA0303II 10.6 1600000 crack 1 -1 36
Imp. Const. 230 1USA03040S 9.2 1640000 crack 1 -1 36
Imp. Const. 230 1USA03040T 10.6 1640000 crack 1 -1 36
Imp. Const. 230 1USA03050U 11.2 2140000 crack 1 -1 36
Imp. Const. 230 1USA03050W 10.4 2140000 crack 1 -1 36
Imp. Const. 230 1USA04010C 14.6 460000 crack 1 -1 36
Imp. Const. 230 1USA0401CC 15.1 770000 crack 1 -1 36
Imp. Const. 230 1USA04010F 9.5 1077000 crack 1 -1 36
Imp. Const. 230 1USA0401FF 9.1 1460000 crack 1 -1 36

475
Imp. Const. 230 1USA0401GG 9 1671000 crack 1 -1 36

Imp. Const. 230 1USA0401HH 9.5 1460000 crack 1 -1 36
Imp. Const. 230 1USA0401GG 8.7 1761000 crack 1 -1 36
Imp. Const. 253 1USA03040B 15.3 1640000 crack 1 -1 36
Imp. Const. 253 1USA03040Z 16.4 1320000 crack 1 -1 36
Imp. Const. 253 1USA0305DD 15.4 1560000 crack 1 -1 36
Imp. Const. 253 1USA0305EE 17 930000 crack 1 -1 36
Imp. Const. 253 1USA0305FF 16.5 2140000 crack 1 -1 36
Imp. Const. 233 1USA0402HH 16.2 1021000 crack 1 -1 36
S.I. Const. 250 1ICOM10501 150 100000 crack -1 -1 -1
S.I. Const. 255 1ICOM10606 120 1700000 runout -1 -1 -1
S.I. Const. 120 1IC0M10703 105 360000 crack -1 -1 -1

476

S.I. Const. 200 1IC0M10321 125 5000000 runout -1 -1 -1

477

S.I. Const. 220 1ICOM20201 180 179000 crack -1 0.1 470
S.I. Const. 200 1IC0M20301 180 164000 crack -1 0.1 470
S.I. Const. 130 1ICOM3PSR1 120 1234000 crack 20 140 510
S.I. Const. 130 1ICOM3PRG1 80 2764000 crack 20 100 510

478

S.I. Const. 251 1ICOM3PSG1 120 297000 crack 20 140 510
S.I. Const. 256 1ICOM3PSG2 120 1720000 runout 20 140 510
S.I. Const. 100 1ICOM3PSR1 120 3050000 runout 20 140 510
S.I. Const. 20 1ICOM3PLT1 220 469000 crack 20 240 510
Imp. Const. 220 1NC5CPS91A 10 25600000 runout 0 0 100
Imp. Const. 220 1NC5CPS25A 10 9244000 crack 0 0 100
Imp. Var. 220 1NC5CPS84A 12 3179000 crack 0 0.5 100
Imp. Var. 220 1NC5CPS116 12 5010000 crack 0 0.5 100
Imp. Var. 220 1NC5CPS102 15 2445000 crack 0 1 100
Imp. Var. 220 1NC5CPS98A 15 2443000 crack 0 1 100
Imp. Const. 220 1NC5CPS115 30 269000 crack 0 0 100

479


480

Imp. Var. 40 1NC5WBC141 4.5 21660000 crack 15 1 100
Imp. Var. 40 1NC5WBC141 3 103700000 crack 10 1 100
Imp. Var. 40 1NC5WBC141 3 104000000 runout 10 1 100
Imp. Const. 40 1NC5WBC148 10 1128000 crack 10 0 100
Imp. Var. 40 1NC5WBC149 12 1431000 crack 10 0.5 100

481

Imp. Const. 20 1NC5WBP141 20 3377000 crack 0 0 100
Imp. Var. 20 1NC5WBP141 24 1542000 crack 0 0.5 100
Imp. Var. 20 1NC5WBP140 30 8060000 crack 0 1 100

482

Imp. Var. 40 1NC5WBC362 10.95 886000 crack 0 0.25 36

483

Imp. Const. 27 1NC5WBP360 20 4700000 crack -10 0 36
Imp. Var. 20 1NC5WBP361 21.9 2944000 crack 0 0.25 36
Imp. Var. 27 1NC5WBP361 24 1469000 crack -10 0.5 36
Imp. Var. 25 1NC5WBP362 32.85 556000 crack -10 0.25 36
Imp. Var. 120 1NC7000001 2.4 43600000 crack 2 4.1 36

484

S.I. Const. 150 1JPN01SA33 300 819000 crack 20 -1 600
S.I. Const. 150 1JPN01MC18 284 332000 crack 20 -1 600
S.I. Const. 150 1JPN01M0C5 236 696000 crack 20 -1 600

485

S.I. Const. 150 1JPN01L0B6 181 1180000 crack 20 -1 600
S.I. Const. 150 1JPN01SBA2 120 4800000 runout 20 -1 600
S.I. Const. 160 1JPN1100S9 147 4700000 runout -123 -5 580

486

S.I. Const. 160 1JPN05JS01 200 470000 crack 300 0.6 800
S.I. Const. 160 1JPN05JS02 132 4000000 runout 198 0.6 800
S.I. Const. 160 1JPN05JS09 135 3200000 runout 0 0 800

487

S.I. Const. 170 1JPN05BB03 127 3000000 runout 0 0 800
S.I. Const. 170 1JPN05BG02 182 3000000 runout -1 -1 500
S.I. Const. 150 1JPN09SB07 250 2100000 runout 10 1 800

488


489


490


491


492

S.I. Const. 200 1JPN13OW11 127 5340000 runout 27 -1 370

493

Imp. Var. 30 1NC8004W01 4.09 34700000 crack 4.17 7.88 36
Imp. Var. 30 1NC8004W01 4.09 35000000 runout 4.17 7.88 36
Imp. Var. 30 1NC8004E01 4.09 34700000 crack 4.17 7.88 36
Imp. Var. 30 1NC8004E01 4.09 35000000 runout 4.17 7.88 36
Imp. Var. 30 1NC8004E17 4.09 32000000 crack 4.17 7.88 36
Imp. Var. 30 1NC8002N01 3.25 86900000 crack 4.17 12.5 36
Imp. Var. 30 1NC8002N01 3.25 120000000 runout 4.17 12.5 36
Imp. Var. 30 1NC8002S01 3.25 120000000 runout 4.17 12.5 36
Imp. Var. 30 1NC8002S01 3.25 87200000 crack 4.17 12.5 36

494

Imp. Var. 40 1NC8003W17 1.99 104000000 crack 2.65 7.09 36
Imp. Const. 90 1NC8003E08 6.29 109000000 crack 7.5 20 36
Imp. Const. 90 1NC8002N08 7.02 89470000 crack 9 27 36
Imp. Const. 90 1NC8102N08 16 110000 crack 9 27 36
Imp. Const. 90 1NC8003E10 6.29 112000000 runout 7.5 20 36
Imp. Const. 90 1NC8001N08 7.12 107000000 runout 8.3 22.3 36
Imp. Const. 90 1NC8001S08 7.12 107000000 runout 8.3 22.3 36
Imp. Const. 90 1NC8002S08 7.02 125000000 runout 9 27 36
Imp. Const. 90 1NC8002S10 7.02 125000000 runout 9 27 36
Imp. Const. 90 1NC8002N10 7.02 120000000 crack 9 27 36
Imp. Const. 90 1NC8003W08 6.7 104000000 runout 9 24 36
Imp. Const. 90 1NC8004E08 8.83 34700000 runout 9 25 36

495

S.I. Var. 90 0ORE80A601 55.9 24290000 crack 9 1 311
S.I. Var. 90 0ORE80A602 55.9 27620000 crack 2 1 311
Imp. Var. 120 1NC8001S03 1.74 107200000 crack 2.07 5.54 36
Imp. Var. 120 1NC8001S06 2.15 86000000 crack 2.56 6.85 36
Imp. Var. 120 1NC8001S11 2.42 43600000 crack 2.88 7.71 36
Imp. Var. 120 1NC8002S02 1.36 120000000 crack 1.74 5.23 36
Imp. Var. 120 1NC8002S03 1.74 120000000 crack 2.23 6.69 36
Imp. Var. 120 1NC8002S04 1.76 120000000 crack 2.26 6.77 36
Imp. Var. 120 1NC8002S05 2.08 120000000 crack 2.67 8 36
Imp. Var. 120 1NC8002S07 2.42 69500000 crack 3.1 9.31 36
Imp. Var. 120 1NC8002S11 2.42 120000000 crack 3.1 9.31 36
Imp. Var. 120 1NC8002S13 2.08 99600000 crack 2.67 8 36

496

Imp. Var. 120 1NC8002N02 1.36 94300000 crack 1.74 5.23 36
Imp. Var. 120 1NC8002N05 2.08 120000000 crack 2.67 8 36
Imp. Var. 120 1NC8002N06 2.15 120000000 crack 2.76 8.27 36
Imp. Var. 120 1NC8002N07 2.42 120000000 crack 3.1 9.31 36
Imp. Var. 120 1NC8002N11 2.42 71900000 crack 3.1 9.31 36
Imp. Var. 120 1NC8002N12 2.15 120000000 crack 2.76 8.27 36
Imp. Var. 120 1NC8002N13 2.08 120000000 crack 2.67 8 36
Imp. Var. 120 1NC8003W07 2.31 37800000 crack 3.11 8.29 36
Imp. Var. 120 1NC8003W12 2.05 67200000 crack 2.74 7.33 36
Imp. Var. 120 1NC8003E07 1.93 104000000 crack 2.59 6.91 36
Imp. Var. 120 1NC8003E11 2.17 109000000 crack 2.59 6.91 36
Imp. Var. 120 1NC8003E12 1.92 109000000 crack 2.3 6.13 36

497

Imp. Var. 120 1NC8004W11 3.05 22600000 crack 3.12 8.65 36
Imp. Var. 120 1NC8004W12 2.7 33000000 crack 2.76 5.21 36
Imp. Var. 120 1NC8004W13 2.61 18600000 crack 2.67 5.04 36
Imp. Var. 120 1NC8004W14 2.22 33000000 crack 2.26 4.27 36
Imp. Var. 120 1NC8004E05 2.61 34700000 crack 2.67 5.04 36
Imp. Var. 120 1NC8004E06 2.7 32000000 crack 2.76 5.21 36
Imp. Var. 120 1NC8004E13 2.61 34700000 crack 2.67 5.04 36
Imp. Var. 120 1NC8004E16 1.71 17400000 crack 1.74 3.29 36
S.I. Const. 100 3CHIS104.0 146 3900000 crack 390
S.I. Const. 50 3ICOM001.0 180 170300 crack 355

498
S.I. Const. 100 3BELG011.0 95 1650000 crack 355

S.I. Const. 100 3BELG031.0 109 2217000 runout 355
S.I. Const. 100 3BELG012.1 103 2060000 runout 355
S.I. Var. 90 3TFHD009.1 45.2 9262000 crack
S.I. Var. 90 3TFHD009.2 40.3 13765700 crack
S.I. Var. 90 3TFHD009.3 45.2 10400700 crack
S.I. Var. 90 3TFHD009.4 40.3 12185400 crack
S.I. Var. 90 3TFHD009.5 45.2 10217900 crack
S.I. Var. 90 3TFHD009.6 40.3 17105200 crack
S.I. Var. 90 3TFHD007.1 56.7 11737900 crack
S.I. Var. 90 3TFHD007.2 63.6 4637600 crack
S.I. Var. 90 3TFHD007.3 56.7 4486950 crack
S.I. Var. 90 3TFHD007.4 63.6 4407900 crack
S.I. Var. 90 3TFHD007.5 56.7 7733800 crack
S.I. Var. 90 3TFHD007.6 63.6 6120000 crack
S.I. Var. 90 3TFHC001.0 63.2 21400000 runout
S.I. Var. 90 3TFHC002.0 80 4500000 runout
S.I. Var. 90 3TFHC012.1 70.7 7900000 crack
S.I. Var. 90 3TFHC012.2 70.7 7850000 crack
S.I. Var. 90 3TFHC012.3 70.7 6019000 crack
S.I. Var. 90 3TFHC012.4 70.7 6300000 crack
S.I. Var. 90 3TFHC012.5 70.7 6200000 crack
S.I. Var. 90 3TFHC012.6 70.7 5650000 crack
S.I. Var. 90 3TFHC011.1 70.7 7654000 crack
S.I. Var. 90 3TFHC011.2 70.7 8600000 crack
S.I. Var. 90 3TFHC011.3 70.7 7300000 crack
S.I. Var. 90 3TFHC011.4 70.7 10400000 crack
S.I. Var. 90 3TFHC011.5 70.7 10850000 crack

499

S.I. Var. 90 3TFHD008.1 37.1 18737000 crack
S.I. Var. 90 3TFHD008.2 37.1 24962000 crack
S.I. Var. 90 3TFHD008.3 37.1 25406000 crack
S.I. Var. 90 3TFHD008.4 37.1 42000000 runout
S.I. Var. 90 3TFHC006.1 54.1 20100000 crack
Imp. Const. 140 1DTRG001 20 12634300 crack 550
Imp. Const. 140 1DTRG013 20 26195300 runout 550
Imp. Const. 130 1DTRAC06 33.2 152950 crack 550
Imp. Const. 130 1DTRAC14 26 392300 crack 550
Imp. Const. 130 1DTRAC11 13.9 25467460 runout 550

500
Imp. Const. 130 2DTRAC17 14 2030830 crack 550

Sources: Keating and Fisher 1986; P. B. Keating, personal communication requesting to locate and use the fatigue test database, 2012.
Reference for F.7

Keating, P. B., and J. W. Fisher. 1986. NCHRP Report 286: Evaluation of
Fatigue Tests and Design Criteria on Welded Details. TRB, National
Research Council, Washington, D.C.
F.8 Concrete Fatigue Database
Table F.8.1. Fatigue Data for Plain Concrete in Compression
S_max/fc S_min (ksi) Sr (ksi) N Reference No. Notes
0.75 0.6 3.9 17000 1 Group 2A; e=0in

0.75 0.6 3.9 24000 1 Group 2A; e=0in
0.75 0.6 3.9 36000 1 Group 2A; e=0in
0.75 0.6 3.9 39000 1 Group 2A; e=0in
0.75 0.6 3.9 40000 1 Group 2A; e=0in
0.75 0.6 3.9 47000 1 Group 2A; e=0in
0.75 0.6 3.9 53000 1 Group 2A; e=0in
0.75 0.6 3.9 59000 1 Group 2A; e=0in
0.75 0.6 3.9 65000 1 Group 2A; e=0in
0.75 0.6 3.9 70000 1 Group 2A; e=0in
0.725 0.6 3.75 39000 1 Group 2A; e=0in
0.725 0.6 3.75 60000 1 Group 2A; e=0in
0.725 0.6 3.75 107000 1 Group 2A; e=0in
0.725 0.6 3.75 110000 1 Group 2A; e=0in
0.725 0.6 3.75 130000 1 Group 2A; e=0in
0.725 0.6 3.75 136000 1 Group 2A; e=0in
0.725 0.6 3.75 192000 1 Group 2A; e=0in
0.725 0.6 3.75 275000 1 Group 2A; e=0in
0.7 0.6 3.6 55000 1 Group 2A; e=0in
0.7 0.6 3.6 106000 1 Group 2A; e=0in
0.7 0.6 3.6 135000 1 Group 2A; e=0in
0.7 0.6 3.6 152000 1 Group 2A; e=0in
0.7 0.6 3.6 155000 1 Group 2A; e=0in
0.7 0.6 3.6 206000 1 Group 2A; e=0in

501

(continued)
0.7 0.6 3.6 269000 1 Group 2A; e=0in

0.7 0.6 3.6 313000 1 Group 2A; e=0in
0.7 0.6 3.6 320000 1 Group 2A; e=0in
0.7 0.6 3.6 356000 1 Group 2A; e=0in
0.7 0.6 3.6 429000 1 Group 2A; e=0in
0.7 0.6 3.6 492000 1 Group 2A; e=0in
0.675 0.6 3.45 159000 1 Group 2A; e=0in
0.675 0.6 3.45 256000 1 Group 2A; e=0in
0.675 0.6 3.45 270000 1 Group 2A; e=0in
0.675 0.6 3.45 655000 1 Group 2A; e=0in
0.675 0.6 3.45 779000 1 Group 2A; e=0in
0.675 0.6 3.45 970000 1 Group 2A; e=0in
0.675 0.6 3.45 1048000 1 Group 2A; e=0in
0.675 0.6 3.45 1051000 1 Group 2A; e=0in
0.675 0.6 3.45 1318000 1 Group 2A; e=0in
0.675 0.6 3.45 1661000 1 Group 2A; e=0in
0.675 0.6 3.45 2000000 1 Group 2A; e=0in
0.675 0.6 3.45 2000000 1 Group 2A; e=0in
0.9 0.6 4.8 28000 1 Group 2B; e=1in
0.9 0.6 4.8 31000 1 Group 2B; e=1in
0.9 0.6 4.8 35000 1 Group 2B; e=1in
0.9 0.6 4.8 45000 1 Group 2B; e=1in
0.9 0.6 4.8 46000 1 Group 2B; e=1in
0.9 0.6 4.8 58000 1 Group 2B; e=1in
0.9 0.6 4.8 61000 1 Group 2B; e=1in
0.9 0.6 4.8 129000 1 Group 2B; e=1in
0.875 0.6 4.65 81000 1 Group 2B; e=1in
0.875 0.6 4.65 120000 1 Group 2B; e=1in
0.875 0.6 4.65 131000 1 Group 2B; e=1in
0.875 0.6 4.65 141000 1 Group 2B; e=1in
0.875 0.6 4.65 156000 1 Group 2B; e=1in
0.875 0.6 4.65 180000 1 Group 2B; e=1in
0.875 0.6 4.65 190000 1 Group 2B; e=1in
0.875 0.6 4.65 226000 1 Group 2B; e=1in
0.875 0.6 4.65 242000 1 Group 2B; e=1in
0.875 0.6 4.65 317000 1 Group 2B; e=1in
0.875 0.6 4.65 351000 1 Group 2B; e=1in
0.875 0.6 4.65 527000 1 Group 2B; e=1in

502

(continued)
0.85 0.6 4.5 305000 1 Group 2B; e=1in

0.85 0.6 4.5 684000 1 Group 2B; e=1in
0.85 0.6 4.5 730000 1 Group 2B; e=1in
0.85 0.6 4.5 859000 1 Group 2B; e=1in
0.85 0.6 4.5 860000 1 Group 2B; e=1in
0.85 0.6 4.5 1045000 1 Group 2B; e=1in
0.85 0.6 4.5 2105000 1 Group 2B; e=1in
0.85 0.6 4.5 2751000 1 Group 2B; e=1in
0.85 0.6 4.5 2000000 1 Group 2B; e=1in
0.85 0.6 4.5 16000 1 Group 2C; e=1/3in
0.85 0.6 4.5 26000 1 Group 2C; e=1/3in
0.85 0.6 4.5 35000 1 Group 2C; e=1/3in
0.85 0.6 4.5 37000 1 Group 2C; e=1/3in
0.85 0.6 4.5 46000 1 Group 2C; e=1/3in
0.85 0.6 4.5 65000 1 Group 2C; e=1/3in
0.8 0.6 4.2 108000 1 Group 2C; e=1/3in
0.8 0.6 4.2 206000 1 Group 2C; e=1/3in
0.8 0.6 4.2 224000 1 Group 2C; e=1/3in
0.8 0.6 4.2 249000 1 Group 2C; e=1/3in
0.8 0.6 4.2 270000 1 Group 2C; e=1/3in
0.8 0.6 4.2 364000 1 Group 2C; e=1/3in
0.8 0.6 4.2 542000 1 Group 2C; e=1/3in
0.8 0.6 4.2 2000000 1 Group 2C; e=1/3in
0.775 0.6 4.05 464000 1 Group 2C; e=1/3in
0.775 0.6 4.05 888000 1 Group 2C; e=1/3in
0.775 0.6 4.05 941000 1 Group 2C; e=1/3in
0.775 0.6 4.05 1198000 1 Group 2C; e=1/3in
0.775 0.6 4.05 2000000 1 Group 2C; e=1/3in

503
Table F.8.2. Concrete Fatigue Data for Reinforced

Concrete
f_min Sr N Reference No. Notes/Specimen No.
5.0 39.0 216,400 2

5.0 39.0 288,100 2
5.0 39.0 315,600 2
5.0 34.0 356,800 2
15.0 34.0 406,600 2
15.0 34.0 441,000 2
5.0 34.0 506,100 2
5.0 34.0 515,300 2
5.0 29.0 626,000 2
15.0 34.0 645,300 2
15.0 29.0 746,000 2
5.0 29.0 864,500 2
5.0 29.0 920,200 2
15.0 29.0 971,900 2
15.0 29.0 1,232,300 2
15.0 26.0 2,214,500 2
15.0 24.0 3,187,500 2
15.0 25.0 3,496,500 2
15.0 24.0 3,702,400 2
15.0 24.0 8,164,000 2
4.31 39.42 6,250,000 3 Bar A-A15
24.8 22.2 5,200,000 3 Bar A-A15
4.31 34.27 3,782,000 3 Bar A-A15
12.85 31.62 3,375,000 3 Bar A-A15
4.31 38.56 3,142,800 3 Bar A-A15
4.32 38.66 2,934,000 3 Bar A-A15
12.89 34.11 2,342,000 3 Bar A-A15
4.31 42.69 2,037,000 3 Bar A-A15
12.85 36.18 1,598,000 3 Bar A-A15
4.31 39.39 1,316,000 3 Bar A-A15
4.31 38.64 1,060,000 3 Bar A-A15
12.87 35.06 964,000 3 Bar A-A15
4.31 40.2 881,000 3 Bar A-A15
4.32 41.81 750,000 3 Bar A-A15
12.87 35.06 555,000 3 Bar A-A15
4.32 40.21 526,000 3 Bar A-A15
4.31 42.69 450,000 3 Bar A-A15
12.85 34.01 435,000 3 Bar A-A15

504

Concrete (continued)
4.31 42.69 431,000 3 Bar A-A15

4.32 40.22 359,000 3 Bar A-A15
4.32 41.81 281,000 3 Bar A-A15
4.31 42.59 245,500 3 Bar A-A15
4.31 41.75 224,300 3 Bar A-A15
4.31 42.69 183,000 3 Bar A-A15
18.99 44.31 89,200 3 Bar A-A15
5.95 62.75 92,200 3 Bar A-A15
5.93 53.96 113,500 3 Bar A-A15
18.97 47.24 169,500 3 Bar A-A15
5.93 47.61 286,000 3 Bar A-A15
5.93 44.46 317,800 3 Bar A-A15
19.11 42.42 389,200 3 Bar A-A15
19.11 44.6 406,300 3 Bar A-A15
5.95 40.8 432,400 3 Bar A-A440
19.06 37.91 432,600 3 Bar A-A440
19.13 41.23 456,100 3 Bar A-A440
5.93 47.61 505,600 3 Bar A-A440
19.05 37.89 526,800 3 Bar A-A440
5.92 41.46 561,700 3 Bar A-A440
5.94 41.41 590,000 3 Bar A-A440
5.94 38.18 914,700 3 Bar A-A440
18.99 33.44 990,000 3 Bar A-A440
5.95 36.07 1,073,000 3 Bar A-A440
5.94 36.6 1,123,000 3 Bar A-A440
19.04 36.09 1,160,000 3 Bar A-A440
19.02 37.84 1,193,000 3 Bar A-A440
5.93 38.17 1,285,000 3 Bar A-A440
5.92 41.45 1,315,600 3 Bar A-A440
19.07 36.16 1,475,750 3 Bar A-A440
5.94 36.06 1,589,000 3 Bar A-A440
5.94 38.44 2,330,000 3 Bar A-A440
5.92 35.93 2,772,300 3 Bar A-A440
18.98 31.45 2,867,000 3 Bar A-A440
5.95 29.75 3,097,000 3 Bar A-A440
19.03 34.69 3,705,200 3 Bar A-A440
5.93 35.99 3,766,000 3 Bar A-A440
5.96 29.78 4,405,000 3 Bar A-A440

505

18.98 28.28 4,514,000 3 Bar A-A440

8.94 67.1 75,000 3 Bar A-A431
8.94 67.1 101,000 3 Bar A-A431
27.14 58.48 135,000 3 Bar A-A431
26.94 58.04 137,100 3 Bar A-A431
27 48.95 152,000 3 Bar A-A431
8.99 60.72 201,100 3 Bar A-A431
27 48.95 215,000 3 Bar A-A431
26.95 49.09 216,000 3 Bar A-A431
8.94 54.7 225,100 3 Bar A-A431
9 55.03 253,000 3 Bar A-A431
26.9 49.44 301,000 3 Bar A-A431
8.94 45.61 307,600 3 Bar A-A431
26.76 49.18 474,100 3 Bar A-A431
9 45.91 512,000 3 Bar A-A431
27.12 37.18 642,300 3 Bar A-A431
27.15 40.92 702,500 3 Bar A-A431
8.36 40.85 714,200 3 Bar A-A431
26.94 40.61 1,006,000 3 Bar A-A431
26.96 37.06 1,044,000 3 Bar A-A431
8.99 40.53 1,048,000 3 Bar A-A431
26.93 40.59 1,075,000 3 Bar A-A431
26.97 37.07 1,456,000 3 Bar A-A431
26.91 36.99 1,560,000 3 Bar A-A431
8.93 40.26 2,250,000 3 Bar A-A431
8.93 40.23 4,160,000 3 Bar A-A431
27 31.56 6,654,000 3 Bar A-A431
4.85 45.68 127,500 3 Bar B-A15
14.42 38.12 259,000 3 Bar B-A15
4.83 43.09 290,000 3 Bar B-A15
4.85 36 352,000 3 Bar B-A15
14.44 35.81 372,000 3 Bar B-A15
4.83 40.69 411,000 3 Bar B-A15
14.5 33.57 477,200 3 Bar B-A15
4.84 36 504,500 3 Bar B-A15
4.84 43.4 538,200 3 Bar B-A15
14.42 33.38 568,000 3 Bar B-A15
14.46 32.3 646,000 3 Bar B-A15

506

4.84 40.77 661,300 3 Bar B-A15

4.85 36.13 665,000 3 Bar B-A15
4.85 34.73 887,000 3 Bar B-A15
14.44 32.25 890,400 3 Bar B-A15
14.5 31.19 1,157,300 3 Bar B-A15
14.47 29.93 1,478,000 3 Bar B-A15
14.43 31.11 1,664,200 3 Bar B-A15
4.84 34.67 1,900,000 3 Bar B-A15
4.83 33.6 3,012,800 3 Bar B-A15
14.44 29.87 4,819,500 3 Bar B-A15
14.46 28.71 5,350,000 3 Bar B-A15
8.11 38.38 91,500 3 Bar B-A431
8.13 62.66 102,000 3 Bar B-A431
8.13 57.69 110,000 3 Bar B-A431
24.33 52.23 120,200 3 Bar B-A431
8.13 52.55 174,000 3 Bar B-A431
24.39 48.37 188,000 3 Bar B-A431
24.41 44.45 255,300 3 Bar B-A431
8.13 48.37 266,000 3 Bar B-A431
24.26 40.22 313,000 3 Bar B-A431
8.13 44.39 428,000 3 Bar B-A431
24.26 36.08 541,000 3 Bar B-A431
24.38 36.28 604,200 3 Bar B-A431
8.12 40.42 651,000 3 Bar B-A431
24.41 32.14 979,000 3 Bar B-A431
8.13 36.49 1,630,000 3 Bar B-A431
8.11 36.4 1,697,000 3 Bar B-A431
24.2 31.87 3,150,000 3 Bar B-A431
24.39 30.33 4,270,000 3 Bar B-A431
9.32 51.35 134,200 3 Bar C-A431
28.04 46.53 158,000 3 Bar C-A431
9.35 46.73 225,000 3 Bar C-A431
9.31 41.78 257,000 3 Bar C-A431
31.89 38.04 311,000 3 Bar C-A431
9.32 41.81 415,500 3 Bar C-A431
9.32 37.26 428,000 3 Bar C-A431
9.31 41.78 430,000 3 Bar C-A431
9.34 37.36 430,000 3 Bar C-A431

507

9.32 37.26 431,000 3 Bar C-A431

28.05 35.61 462,400 3 Bar C-A431
28.02 35.57 477,700 3 Bar C-A431
31.9 31.5 499,300 3 Bar C-A431
9.28 35.16 503,300 3 Bar C-A431
27.94 30.32 648,400 3 Bar C-A431
27.85 27.86 1,056,000 3 Bar C-A431
9.28 33.15 1,072,000 3 Bar C-A431
9.35 31.81 1,250,000 3 Bar C-A431
28.05 26.06 2,037,000 3 Bar C-A431
9.34 29.82 2,631,000 3 Bar C-A431
5 39 216,400 3 Bar D-A15
5 39 288,000 3 Bar D-A15
5 39 315,600 3 Bar D-A15
5 34 356,800 3 Bar D-A15
15 34 365,200 3 Bar D-A15
15 34 406,600 3 Bar D-A15
5 34 435,000 3 Bar D-A15
15 34 441,000 3 Bar D-A15
5 34 506,100 3 Bar D-A15
5 34 510,000 3 Bar D-A15
5 34 515,300 3 Bar D-A15
5 29 626,600 3 Bar D-A15
15 34 645,300 3 Bar D-A15
15 29 673,000 3 Bar D-A15
15 29 746,000 3 Bar D-A15
5 29 864,500 3 Bar D-A15
15 29 888,400 3 Bar D-A15
5 29 920,200 3 Bar D-A15
15 29 971,000 3 Bar D-A15
15 29 1,030,000 3 Bar D-A15
5 34 1,120,000 3 Bar D-A15
15 29 1,232,000 3 Bar D-A15
15 26 2,214,500 3 Bar D-A15
15 24 3,187,500 3 Bar D-A15
15 25 3,496,500 3 Bar D-A15
15 24 3,702,400 3 Bar D-A15
15 24 8,164,000 3 Bar D-A15

508

4.6 39.7 88,900 4 44CH

4.6 39.7 129,200 4 44CV
4.7 35.5 219,800 4 40CH
4.7 35.5 334,200 4 40CV
4.7 33.5 364,000 4 38CV
4.7 31.6 507,000 4 36CH
4.7 29.5 517,000 4 34CH
4.7 29.5 575,000 4 34CH
4.7 31.5 627,300 4 36CV
4.7 29.5 903,000 4 34CV
4.7 27.5 1,434,000 4 32CH
4.7 27.5 1,941,900 4 32CV
4.7 25.5 2,819,800 4 30CH
4.7 25.5 2,984,600 4 30CV
4.8 23.4 5,237,000 4 28CV
4.7 24.5 5,731,000 4 29CV
4.8 23.4 6,266,500 4 28CH
23.7 51.1 160,000 5 F75-5
22.1 49.4 248,000 5 F75-3
17.1 37.9 350,000 5 F50-7
19.3 47.3 401,000 5 F75-1
15.2 41.2 429,000 5 F60-1
21.3 40.6 604,000 5 F75-4
12.5 34.1 610,000 5 F50-1
13.6 37.8 624,000 5 F50-5
12.3 39.7 672,000 5 F60-3
15 32.4 787,000 5 F40-4
14 33.5 893,000 5 F50-3
13.5 31.9 1,063,000 5 F40-3
14.1 30.7 1,316,000 5 F50-6
16.4 28.6 1,348,000 5 F40-1
12.4 29.6 1,488,000 5 F40-2
21.5 40 1,781,000 5 F75-2
16.1 32.1 1,877,000 5 F60-2
16.4 31.9 3,004,000 5 F60-4
13.9 26.1 3,272,000 5 F50-4
12.3 27.7 3,623,000 5 F50-2

509
References for F.8 Association Research and Development Laboratories, Vol. 6, No. 1,
pp. 65–84.
1. Ople, F. S., and C. L. Hulsbos. 1966. Probable Fatigue Life of Plain 4. Hanson, J. M., K. T. Burton, and E. Hognestad. 1968. Fatigue Tests of
Concrete with Stress Gradient. Journal Proceedings, American Con- Reinforcing Bars: Effect of Deformation Pattern. Journal of the
crete Institute, Vol. 63, No. (1), pp. 59–82. Portland Cement Association Research and Development Laboratories,
2. Fisher, J. W., and I. M. Viest. 1961. Special Report 66: Fatigue Tests of Vol. 10, No. 3, pp. 2–13.
Bridge Materials of the AASHTO Road Test. HRB, National Research 5. Lash, S. 1969. Can High-Strength Reinforcement Be Used in
Council, Washington, D.C., pp. 132–147. Highway Bridges? In First International Symposium on Concrete
3. Pfister, J. F., and E. Hognestad. 1964. High-Strength Bars As Concrete Bridge Design, SP-23, American Concrete Institute, Detroit, Mich.,
Reinforcement, Part 6: Fatigue Tests. Journal of the Portland Cement pp. 283–299.

TRB Oversight Committee for the Strategic Highway Research Program 2*

Chair: Kirk T. Steudle, Director, Michigan Department of Transportation
Members
H. Norman Abramson, Executive Vice President (retired), Southwest Research Institute
Alan C. Clark, MPO Director, Houston–Galveston Area Council
Frank L. Danchetz, Vice President, ARCADIS-US, Inc. (deceased January 2015)
Malcolm Dougherty, Director, California Department of Transportation
Stanley Gee, Executive Deputy Commissioner, New York State Department of Transportation
Mary L. Klein, President and CEO, NatureServe
Michael P. Lewis, Director, Rhode Island Department of Transportation
John R. Njord, Executive Director (retired), Utah Department of Transportation
Charles F. Potts, Chief Executive Officer, Heritage Construction and Materials
Ananth K. Prasad, Secretary, Florida Department of Transportation
Gerald M. Ross, Chief Engineer (retired), Georgia Department of Transportation
George E. Schoener, Executive Director, I-95 Corridor Coalition
Kumares C. Sinha, Olson Distinguished Professor of Civil Engineering, Purdue University
Paul Trombino III, Director, Iowa Department of Transportation
Ex Officio Members
Victor M. Mendez, Administrator, Federal Highway Administration
David L. Strickland, Administrator, National Highway Transportation Safety Administration
Frederick “Bud” Wright, Executive Director, American Association of State Highway and Transportation Officials
Liaisons
Ken Jacoby, Communications and Outreach Team Director, Office of Corporate Research, Technology, and Innovation Management, Federal
Highway Administration
Tony Kane, Director, Engineering and Technical Services, American Association of State Highway and Transportation Officials
Jeffrey F. Paniati, Executive Director, Federal Highway Administration
John Pearson, Program Director, Council of Deputy Ministers Responsible for Transportation and Highway Safety, Canada
Michael F. Trentacoste, Associate Administrator, Research, Development, and Technology, Federal Highway Administration
*Membership as of January 2015.
Renewal Technical Coordinating Committee*

Chair: Daniel D’Angelo, Recovery Acting Manager, Director and Deputy Chief Engineer, Office of Design, New York State Department
of Transportation
Members
Rachel Arulraj, President, InfoInnovation
Michael E. Ayers, Consultant, Technology Services, American Concrete Pavement Association
Thomas E. Baker, State Materials Engineer, Washington State Department of Transportation
John E. Breen, Al-Rashid Chair in Civil Engineering Emeritus, University of Texas at Austin
Steven D. DeWitt, Chief Engineer (retired), North Carolina Turnpike Authority
Tom W. Donovan, Senior Right of Way Agent (retired), California Department of Transportation
Alan D. Fisher, Manager, Construction Structures Group, Cianbro Corporation
Michael Hemmingsen, Davison Transportation Service Center Manager (retired), Michigan Department of Transportation
Bruce Johnson, State Bridge Engineer, Oregon Department of Transportation, Bridge Engineering Section
Leonnie Kavanagh, PhD Candidate, Seasonal Lecturer, Civil Engineering Department, University of Manitoba
Cathy Nelson, Technical Services Manager/Chief Engineer (retired), Oregon Department of Transportation
John J. Robinson, Jr., Assistant Chief Counsel, Pennsylvania Department of Transportation, Governor’s Office of General Counsel
Ted M. Scott II, Director, Engineering, American Trucking Associations, Inc.
Gary D. Taylor, Professional Engineer
Gary C. Whited, Program Manager, Construction and Materials Support Center, University of Wisconsin–Madison
AASHTO Liaison
James T. McDonnell, Program Director for Engineering, American Association of State Highway and Transportation Officials
FHWA Liaisons
Steve Gaj, Leader, System Management and Monitoring Team, Office of Asset Management, Federal Highway Administration
Cheryl Allen Richter, Assistant Director, Pavement Research and Development, Office of Infrastructure Research and Development, Federal
Highway Administration
J. B. “Butch” Wlaschin, Director, Office of Asset Management, Federal Highway Administration
Canada Liaison
Lance Vigfusson, Assistant Deputy Minister of Engineering & Operations, Manitoba Infrastructure and Transportation
*Membership as of July 2014.

Geotechnical Solutions for Soil Improvement, Rapid Embankment Construction,

and Stabilization of the Pavement Working Platform (R02)
Innovative Bridge Designs for Rapid Renewal (R04)
Nondestructive Testing to Identify Concrete Bridge Deck Deterioration (R06A)
Mapping Voids, Debonding, Delaminations, Moisture, and Other Defects
Behind or Within Tunnel Linings (R06G)
Bridges for Service Life Beyond 100 Years: Innovative Systems, Subsystems,
and Components (R19A)

The National Academies Press: Bridges For Service Life Beyond 100 Years: Service Limit State Design (2014)

Uploaded by

Copyright:

Available Formats

The National Academies Press: Bridges For Service Life Beyond 100 Years: Service Limit State Design (2014)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The National Academies Press: Bridges For Service Life Beyond 100 Years: Service Limit State Design (2014)

Uploaded by

Copyright:

Available Formats

THE NATIONAL ACADEMIES PRESS

This PDF is available at http://nap.edu/22441 SHARE

Bridges for Service Life Beyond 100 Years: Service Limit

0 pages | 8.5 x 11 | PAPERBACK

GET THIS BOOK Modjeski and Masters, Inc.

FIND RELATED TITLES

National Academies of Sciences, Engineering, and Medicine 2014. Bridges for

– Access to free PDF downloads of thousands of scientiﬁc reports

Copyright © National Academy of Sciences. All rights reserved.

Bridges for Service Life Beyond 100 Years:

Copyright National Academy of Sciences. All rights reserved.

Figure 2.1 Credit

Copyright National Academy of Sciences. All rights reserved.

The Second Strategic Highway SHRP 2 Report S2-R19B-RW-1

More information about SHRP 2:

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Ann M. Brach, Director

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

121 5.4 Development of Statistical Parameters of Fatigue Load

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Purpose of Report and Relation to Scope

Copyright National Academy of Sciences. All rights reserved.

1.2 Scope The INC included appropriate experts from AASHTO, the

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

W. G. Wassef, PhD, PE D. R. Mertz, PhD, PE A. S. Nowak, PhD

Management Limit States Limit States

Figure 1.1. Project R19B organizational chart.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Current State of the Art

2.1 Approach 2.2.1 Serviceability, SLS, Deterioration,

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Table 2.1. Service Life of Expansion Joints

Joint Type Mean Min Max Mean Min Max

Pourable seals 11.5 4 30 5.2, 5.6 1, 0 15, 20

Copyright National Academy of Sciences. All rights reserved.

Table 2.3. Tolerable Movement Criteria for Highway Bridges

Differential Settlement over 100 ft

d ≤ 1 in. d100 ft ≤ 0.75 in. Design and construct

Copyright National Academy of Sciences. All rights reserved.

1. Identify all possible facilities associated with the bridge

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

Copyright National Academy of Sciences. All rights reserved.

2.1 Approach 2.2.1 Serviceability, SLS, Deterioration,

state equation must be developed on the basis of a pre- 3.2 Calibration by