NCHRP 534

Guidelines for Inspection and
Strength Evaluation of Suspension

Bridge Parallel Wire Cables
NATIONAL
COOPERATIVE
HIGHWAY
RESEARCH
PROGRAM NCHRP
REPORT 534
TRANSPORTATION RESEARCH BOARD EXECUTIVE COMMITTEE 2004 (Membership as of July 2004)
OFFICERS
Chair: Michael S. Townes, President and CEO, Hampton Roads Transit, Hampton, VA
Vice Chair: Joseph H. Boardman, Commissioner, New York State DOT
Executive Director: Robert E. Skinner, Jr., Transportation Research Board
MEMBERS
MICHAEL W. BEHRENS, Executive Director, Texas DOT
SARAH C. CAMPBELL, President, TransManagement, Inc., Washington, DC
E. DEAN CARLSON, Director, Carlson Associates, Topeka, KS
JOHN L. CRAIG, Director, Nebraska Department of Roads
DOUGLAS G. DUNCAN, President and CEO, FedEx Freight, Memphis, TN
GENEVIEVE GIULIANO, Director, Metrans Transportation Center and Professor, School of Policy, Planning, and Development, USC,
Los Angeles
BERNARD S. GROSECLOSE, JR., President and CEO, South Carolina State Ports Authority
SUSAN HANSON, Landry University Professor of Geography, Graduate School of Geography, Clark University
JAMES R. HERTWIG, President, CSX Intermodal, Jacksonville, FL
GLORIA J. JEFF, Chief Administrative Officer, Michigan DOT
ADIB K. KANAFANI, Cahill Professor of Civil Engineering, University of California, Berkeley
RONALD F. KIRBY, Director of Transportation Planning, Metropolitan Washington Council of Governments
HERBERT S. LEVINSON, Principal, Herbert S. Levinson Transportation Consultant, New Haven, CT
SUE MCNEIL, Director, Urban Transportation Center and Professor, College of Urban Planning and Public Affairs and Department of
Civil and Material Engineering, University of Illinois, Chicago
MICHAEL D. MEYER, Professor, School of Civil and Environmental Engineering, Georgia Institute of Technology
CAROL A. MURRAY, Commissioner, New Hampshire DOT
JOHN E. NJORD, Executive Director, Utah DOT
DAVID PLAVIN, President, Airports Council International, Washington, DC
JOHN H. REBENSDORF, Vice President, Network Planning and Operations, Union Pacic Railroad Co., Omaha, NE
PHILIP A. SHUCET, Commissioner, Virginia DOT
C. MICHAEL WALTON, Ernest H. Cockrell Centennial Chair in Engineering, University of Texas, Austin
LINDA S. WATSON, Executive Director, LYNXCentral Florida Regional Transportation Authority, Orlando, FL
MARION C. BLAKEY, Federal Aviation Administrator, U.S.DOT (ex officio)
SAMUEL G. BONASSO, Acting Administrator, Research and Special Programs Administration, U.S.DOT (ex officio)
REBECCA M. BREWSTER, President and COO, American Transportation Research Institute, Smyrna, GA (ex officio)
GEORGE BUGLIARELLO, Chancellor, Polytechnic University and Foreign Secretary, National Academy of Engineering (ex officio)
THOMAS H. COLLINS (Adm., U.S. Coast Guard), Commandant, U.S. Coast Guard (ex officio)
JENNIFER L. DORN, Federal Transit Administrator, U.S.DOT (ex officio)
EDWARD R. HAMBERGER, President and CEO, Association of American Railroads (ex officio)
JOHN C. HORSLEY, Executive Director, American Association of State Highway and Transportation Officials (ex officio)
RICK KOWALEWSKI, Deputy Director, Bureau of Transportation Statistics, U.S.DOT (ex officio)
WILLIAM W. MILLAR, President, American Public Transportation Association (ex officio)
BETTY MONRO, Acting Administrator, Federal Railroad Administration, U.S.DOT (ex officio)
MARY E. PETERS, Federal Highway Administrator, U.S.DOT (ex officio)
SUZANNE RUDZINSKI, Director, Transportation and Regional Programs, U.S. Environmental Protection Agency (ex officio)
JEFFREY W. RUNGE, National Highway Traffic Safety Administrator, U.S.DOT (ex officio)
ANNETTE M. SANDBERG, Federal Motor Carrier Safety Administrator, U.S.DOT (ex officio)
WILLIAM G. SCHUBERT, Maritime Administrator, U.S.DOT (ex officio)
JEFFREY N. SHANE, Under Secretary for Policy, U.S.DOT (ex officio)
CARL A. STROCK (Maj. Gen., U.S. Army), Chief of Engineers and Commanding General, U.S. Army Corps of Engineers (ex officio)
ROBERT A. VENEZIA, Program Manager of Public Health Applications, National Aeronautics and Space Administration (ex officio)
NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM
Transportation Research Board Executive Committee Subcommittee for NCHRP
MICHAEL S. TOWNES, Hampton Roads Transit, Hampton, VA
(Chair)
JOSEPH H. BOARDMAN, New York State DOT
GENEVIEVE GIULIANO, University of Southern California,
Los Angeles
JOHN C. HORSLEY, American Association of State Highway
and Transportation Officials
MARY E. PETERS, Federal Highway Administration
ROBERT E. SKINNER, JR., Transportation Research Board
C. MICHAEL WALTON, University of Texas, Austin
TRANSPORTATI ON RESEARCH BOARD
WASHINGTON, D.C.
2004
www.TRB.org
NATI ONAL COOPERATI VE HI GHWAY RESEARCH PROGRAM
NCHRP REPORT 534
Research Sponsored by the American Association of State Highway and Transportation Officials
in Cooperation with the Federal Highway Administration
SUBJECT AREAS
Bridges, Other Structures, Hydraulics and Hydrology
Guidelines for Inspection and
Strength Evaluation of Suspension
Bridge Parallel Wire Cables
R. M. MAYRBAURL
S. CAMO
Weidlinger Associates, Inc.
New York, NY
NATIONAL COOPERATIVE HIGHWAY RESEARCH
PROGRAM
Systematic, well-designed research provides the most effective
approach to the solution of many problems facing highway
administrators and engineers. Often, highway problems are of local
interest and can best be studied by highway departments
individually or in cooperation with their state universities and
others. However, the accelerating growth of highway transportation
develops increasingly complex problems of wide interest to
highway authorities. These problems are best studied through a
coordinated program of cooperative research.
In recognition of these needs, the highway administrators of the
American Association of State Highway and Transportation
Officials initiated in 1962 an objective national highway research
program employing modern scientic techniques. This program is
supported on a continuing basis by funds from participating
member states of the Association and it receives the full cooperation
and support of the Federal Highway Administration, United States
Department of Transportation.
The Transportation Research Board of the National Academies
was requested by the Association to administer the research
program because of the Boards recognized objectivity and
understanding of modern research practices. The Board is uniquely
suited for this purpose as it maintains an extensive committee
structure from which authorities on any highway transportation
subject may be drawn; it possesses avenues of communications and
cooperation with federal, state and local governmental agencies,
universities, and industry; its relationship to the National Research
Council is an insurance of objectivity; it maintains a full-time
research correlation staff of specialists in highway transportation
matters to bring the ndings of research directly to those who are in
a position to use them.
The program is developed on the basis of research needs
identied by chief administrators of the highway and transportation
departments and by committees of AASHTO. Each year, specic
areas of research needs to be included in the program are proposed
to the National Research Council and the Board by the American
Association of State Highway and Transportation Officials.
Research projects to fulll these needs are dened by the Board, and
qualied research agencies are selected from those that have
submitted proposals. Administration and surveillance of research
contracts are the responsibilities of the National Research Council
and the Transportation Research Board.
The needs for highway research are many, and the National
Cooperative Highway Research Program can make signicant
contributions to the solution of highway transportation problems of
mutual concern to many responsible groups. The program,
however, is intended to complement rather than to substitute for or
duplicate other highway research programs.
Note: The Transportation Research Board of the National Academies, the
National Research Council, the Federal Highway Administration, the American
Association of State Highway and Transportation Officials, and the individual
states participating in the National Cooperative Highway Research Program do
not endorse products or manufacturers. Trade or manufacturers names appear
herein solely because they are considered essential to the object of this report.
Published reports of the
NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM
are available from:
Transportation Research Board
Business Office
500 Fifth Street, NW
Washington, DC 20001
and can be ordered through the Internet at:
http://www.national-academies.org/trb/bookstore
Printed in the United States of America
NCHRP REPORT 534
Project 10-57 FY00
ISSN 0077-5614
ISBN 0-309-08809-7
Library of Congress Control Number 2004112323
2004 Transportation Research Board
Price $30.00
NOTICE
The project that is the subject of this report was a part of the National Cooperative
Highway Research Program conducted by the Transportation Research Board with the
approval of the Governing Board of the National Research Council. Such approval
reects the Governing Boards judgment that the program concerned is of national
importance and appropriate with respect to both the purposes and resources of the
National Research Council.
The members of the technical committee selected to monitor this project and to review
this report were chosen for recognized scholarly competence and with due
consideration for the balance of disciplines appropriate to the project. The opinions and
conclusions expressed or implied are those of the research agency that performed the
research, and, while they have been accepted as appropriate by the technical committee,
they are not necessarily those of the Transportation Research Board, the National
Research Council, the American Association of State Highway and Transportation
Officials, or the Federal Highway Administration, U.S. Department of Transportation.
Each report is reviewed and accepted for publication by the technical committee
according to procedures established and monitored by the Transportation Research
Board Executive Committee and the Governing Board of the National Research
Council.
To save time and money in disseminating the research findings, the report is
essentially the original text as submitted by the research agency. This report has
not been edited by TRB.
The National Academy of Sciences is a private, nonprot, self-perpetuating society of distinguished schol-
ars engaged in scientic 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 the Congress in
1863, the Academy has a mandate that requires it to advise the federal government on scientic and techni-
cal matters. Dr. Bruce M. Alberts is president of the National Academy of Sciences.
The National Academy of Engineering was established in 1964, under the charter of the National Acad-
emy 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. William A. Wulf 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. Harvey V. Fineberg 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 Academys purposes of furthering knowledge and
advising the federal government. Functioning in accordance with general policies determined by the Acad-
emy, 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
scientic and engineering communities. The Council is administered jointly by both the Academies and
the Institute of Medicine. Dr. Bruce M. Alberts and Dr. William A. Wulf are chair and vice chair,
respectively, of the National Research Council.
The Transportation Research Board is a division of the National Research Council, which serves the
National Academy of Sciences and the National Academy of Engineering. The Boards mission is to promote
innovation and progress in transportation through research. In an objective and interdisciplinary setting,
the Board facilitates the sharing of information on transportation practice and policy by researchers and
practitioners; stimulates research and offers research management services that promote technical
excellence; provides expert advice on transportation policy and programs; and disseminates research
results broadly and encourages their implementation. The Boards varied activities annually engage more
than 5,000 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 Transportation, and other organizations and individuals interested in the
development of transportation. www.TRB.org
www.national-academies.org
AUTHOR ACKNOWLEDGMENTS
The authors wish to acknowledge the contributions of the fol-
lowing people who helped in the successful completion of the
guidelines and report: Robert G. Easterling, for writing Appendix
B of the Report and for his painstaking suggestions regarding sta-
tistics; the several subconsultants who supplied detailed drafts in
their areas of expertise, including Ronald Latanision of Altran on
corrosion mechanisms, Mark Mittelman of Altran on microbially-
inuenced corrosion, and Robert Torbin and Paul Belkus of Foster-
Miller on nondestructive testing; William Moreau and the New
York State Bridge Authority for permission to use information
regarding inspections of the Bear Mountain Bridge and Mid-
Hudson Bridge; Martin Kendall for supplying the actual copies of
the Bear Mountain Bridge inspection reports; Chris Gagnon for
providing reports on Bridge Z wire testing; Bojidar Yanev and the
New York City Department of Transportation for providing the
Williamsburg Bridge inspection reports and for granting permis-
sion to use information regarding their inspection procedures; Par-
sons Transportation Group for providing inspection photographs;
all the bridge owners for responding to the original questionnaires;
and Jeffrey Hu for managing the questionnaire procedure and the
resulting data and Wei Liu for checking the examples. The authors
are especially grateful to Helen Goddard for editing both the
guidelines and the report, a task that almost exceeded the writing
of them.
COOPERATIVE RESEARCH PROGRAMS STAFF FOR NCHRP REPORT 534
ROBERT J. REILLY, Director, Cooperative Research Programs
CRAWFORD F. JENCKS, Manager, NCHRP
DAVID B. BEAL, Senior Program Officer
EILEEN P. DELANEY, Director of Publications
KAMI CABRAL, Associate Editor
NCHRP PROJECT 10-57 PANEL
Field of Materials, Construction, and MaintenanceArea of Materials and
Construction
M. MYINT LWIN, FHWA (Chair)
MACIEJ BIENIEK, Columbia University, NY
JACKSON DURKEE, Bethlehem, PA
CHRISTOPHER HAHIN, Illinois DOT
TIMOTHY M. MOORE, Washington DOT
WALTER PODOLNY, JR., Burke, VA
THOMAS RUT, Moffatt & Nichol, Stockton, CA
BOJIDAR YANEV, New York City DOT
HAROLD BOSCH, FHWA Liaison Representative
STEVEN F. MAHER, TRB Liaison Representative
This report contains recommendations for the inspection and strength evaluation
of suspension bridge parallel-wire cables. A companion CD-ROM provides details of
the research program undertaken to develop the guidelines and explanations of the
guidelines recommendations. The material in this report will be of immediate interest
to suspension bridge owners and suspension bridge main cable inspectors.
There are nearly 50 major suspension bridges in the United States, and more than
half of them are more than 50 years old. These bridges represent major investments and
are essential transportation links for regional and national commerce and lifelines. As
this group of major structures advances in age, the number of in-depth evaluations to
determine their condition and load-carrying capacity is expected to increase. The need
to estimate remaining service life and to take preventative steps to extend service life
will also increase.
There has been no reliable and nationally recognized procedure, either practical or
theoretical, to inspect and evaluate the condition and strength of suspension bridge
parallel-wire cables. At the NCHRP-sponsored Workshop on Safety Appraisal of Sus-
pension Bridge Main Cables held in Newark, New Jersey, in 1998, the highest prior-
ity research needs identied were (1) development of cable inspection, sampling, and
testing guidelines and (2) development of models to predict the strength of deteriorated
cables. Unreliable methods of inspection and evaluation could result in unnecessary
replacement or in unexpected failures.
The objective of this research was to develop the needed guidelines for inspection
and evaluation of suspension bridge parallel-wire cables. These guidelines provide
details of the cable inspection process, including wire sampling and testing. Complete
instructions, illustrated with examples, are provided for using the condition and prop-
erties of the cable wire, determined by inspection and subsequent laboratory testing,
for estimating cable strength. An accompanying CD-ROM (CRP-CD-54) contains a
full account of the research leading to the development of the recommended guidelines
and provides derivations of all equations.
The research was performed by Weidlinger Associates, Inc., with the assistance of
Altran Corporation and Foster-Miller, Inc.
FOREWORD
By David B. Beal
Senior Program Officer
Transportation Research
Board
The Guidelines were written as a general resource for all persons who have respon-
sibility for assessing cable integrity and to clarify and standardize the process of eval-
uating a cable in service for an extended period of time. They will not make a suspen-
sion bridge expert out of a novice, but are intended for use by engineering professionals
who are already versed in the design and analysis of suspension bridges and have some
knowledge of bridge cables. A cable evaluation team should consist of the following:
Chief investigator, a professional engineer with expertise in suspension bridges
who leads all phases of the project
Chief inspector, a professional engineer with experience in bridge inspections,
preferably cable inspections
One or more cable inspectors, graduate engineers with two to three years of design
or inspection experience
Office staff, graduate engineers with experience in computer analysis, spread-
sheets and applying mathematical equations.
Qualied testing laboratory
Metallurgical and corrosion consultants
Statistician (as needed)
The Guidelines present a series of orderly steps that dene a thorough cable eval-
uation from planning inspections through strength estimation. They concern parallel-
wire cables only, although portions of the text may be applied to helical strand cables.
If inspection is limited to only a few panels, the Guidelines cannot provide information
about the strength of the weakest point in the cable. However, they do contain recom-
mendations for when more thorough investigations are needed.
The Guidelines are arranged in two columns, as are other specications and man-
uals published by AASHTO. In general, a description of the required tasks is on the
left, pertinent comments and background information are on the right. The columns are
merged in Section 1, which is introductory. Figures and tables appear at the end of each
section. They are numbered consecutively, prexed by the article numbers in which
they are described (e.g., Figure 1.4.2.1-1).
Section 1 is mostly a general description of bridge cables, including their internal
construction, connections to the bridge structure, and protection systems. Figures are
used to illustrate the various parts of the cable system. The causes of corrosion are dis-
cussed, as well as investigative techniques to locate corrosion. Of special interest are
photographs showing the visual rating scale for corroded wires. The section also
includes a list of persons who should use the Guidelines, a glossary of technical terms,
and information on health and safety requirements.
Section 2 presents three levels of inspection: routine visual inspections by mainte-
nance personnel, biennial inspections, and internal inspections that expose the wires
inside the cable. The section also contains instructions about the data to be recorded
PREFACE
and measured, and the requirements for removal of wire samples. Recommendations
for frequency and locations of internal inspections are based primarily on the data from
a limited number of cable inspections in which a signicant number of panels was
opened. As more cables are inspected, the results can be combined with existing data
to justify or modify these recommendations.
Acoustic monitoring of wire breaks, a recent development, is referred to briey in
Section 2. The sound of a wire as it breaks inside the cable is recorded and the catalog
of recent breaks is used to decide when, and especially where, to perform the next
inspection. This is very helpful with older cables, because timing of inspections and
selection of the best locations are critical.
Section 3 lists the requirements for the physical and chemical tests that are made
on samples removed from the cable, including tensile tests, and tests to determine the
chemical composition of wires and the condition of the zinc coating.
Section 4 focuses on the techniques used to catalog the damage inside the cable
and the statistical analysis of test results. The calculations for obtaining the mean val-
ues and standard deviations of wire properties that are needed for strength estimation
(tensile strength and, in some cases, ultimate elongation) do not require advanced
knowledge of statistics, but can be performed using standard spreadsheet programs, or
even by hand. A slightly more sophisticated statistical analysis is used to estimate the
probable minimum value of these properties in a given length of wire, which is much
the same as estimating the strength of the weakest link in a chain. The graphs that are
provided reduce the complexity of the latter analysis substantially.
It has been noted during internal cable inspections that friction among the wires
introduces tension back into a broken wire as the distance from the break location
increases. A method of estimating the force that is reintroduced as the wire passes
through a cable band is presented in Section 4. In this analysis, the effects of wrapping
wire, which are not negligible, are conservatively ignored.
Three models for estimating cable strength are given in Section 5 based on the fol-
lowing assumptions. All wires are subject to the same elongation between cable bands.
An individual wire breaks at its ultimate elongation and thereafter ceases to share in the
cable tension. Only after some wires have broken does the cable attain its strength,
which is smaller than the product of its area and the mean tensile strength of the wires.
These suppositions are borne out by strand efficiency tests performed during the design
of the Bear Mountain Bridge and Benjamin Franklin Bridge.
Section 5 presents methods for assigning wires to groups that differentiate increas-
ing levels of deterioration. Wherever appropriate, graphs are inserted to assist in esti-
mating the effects of deterioration in panels adjacent to the evaluated panel, including
the effects of broken wires. The statistics in this section are more advanced than in the
previous section. The Weibull distribution of the ultimate strain or tensile strength of
the wires, included in some spreadsheet programs, simplies the calculation. The equa-
tions for the distribution are given in Appendix A, as well as an iterative method for
calculating the parameters of the distribution. An alternative method using Weibull
paper is not presented here, but can be found in statistical texts, Rao [1] for example.
The equations used for estimating the cable strength are also included in Appendix A.
Section 6 lists the points to be covered in written reports for all three levels of
inspection.
Appendix B shows the rationale for calculating the effects of deterioration in adja-
cent panels on cable strength in the evaluated panel. The calculation is extremely
tedious if all the cable panels have been inspected; therefore, it should be reserved for
the worst panel found in an inspection, or eliminated, unless it is essential to take all
sources of cable strength into account. The assumption that all panels are in the same
condition as the evaluated panel leads to a lower estimate of cable strength; graphs are
provided to simplify the calculation.
Three examples of a strength calculation are given in Appendix C, all using the
same inspection data. The rst two examples use the Simplied and Brittle-Wire Mod-
els, with the assumption that all panels are in the same condition. The third example
assumes that all panels have been inspected and employs, in part, a very large spread-
sheet that requires 16 pages to print and is tedious to set up, demonstrating the reasons
for not using it except in extreme cases.
The method that is commonly used to replace broken wires or sample wires is pre-
sented in Appendix D.
REFERENCE
1. Rao, S.S., Reliability-Based Design. 1st ed, ed. R. Hauserman. 1992: McGraw-Hill, Inc. 569.
1-1 SECTION 1 General
1.1 Introduction, 1-3
1.2 Health and Safety Requirements, 1-8
1.3 Suspension Bridges, 1-9
1.4 Causes of Corrosion, 1-15
1.5 Investigative Techniques, 1-18
1.6 Figures for Section 1, 1-19
1.7 References, 1-25
2-1 SECTION 2 Inspection
2.2 Inspection Intervals and Locations, 2-6
2.3 Internal Inspections, 2-12
2.4 Inspection and Sampling, 2-20
3-1 SECTION 3 Laboratory Testing
3.2 Tests of Wire Properties, 3-2
3.3 Zinc Coating Tests, 3-5
3.4 Chemical Analysis, 3-6
3.5 Corrosion Analysis, 3-6
3.6 Figure for Section 3, 3-8
3.7 Reference, 3-8
4-1 SECTION 4 Evaluation of Field and Laboratory Data
4.2 Notation, 4-3
4.3 Mapping and Estimating Damage, 4-5
4.4 Wire Properties, 4-10
4.5 Wire Redevelopment, 4-14
5-1 SECTION 5 Estimation of Cable Strength
5.2 Notation, 5-3
5.3 Estimated Cable Strength, 5-6
6-1 SECTION 6 Inspection Reports
6.2 Maintenance Personnel Inspection, 6-2
6.3 Biennial Inspection, 6-2
6.4 Internal Inspection, 6-3
A-1 APPENDIX A Models for Estimating Cable Strength
B-1 APPENDIX B Effect of Deterioration in Adjacent Panels on Estimated
Cable Strength
C-1 APPENDIX C Illustrative Examples
D-1 APPENDIX D Splicing New Wires
CONTENTS
CONTENTS
SECTION 1 GENERAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-1
1.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-3
1.1.1 Scope of Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-3
1.1.2 Users of Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-3
1.1.3 Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-3
1.1.4 Professional Organizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-7
1.1.5 National Codes and Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-7
1.2 HEALTH AND SAFETY REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-8
1.2.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-8
1.2.1.1 LEAD ON THE CABLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-8
1.2.1.2 OTHER OSHA REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-8
1.2.1.3 PERSONNEL TRAINING PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-8
1.2.1.4 PEDESTRIAN AND VEHICULAR TRAFFIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-8
1.2.1.5 NAVIGATIONAL SAFETY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-9
1.2.1.6 BRIDGE SAFETY PLAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-9
1.3 SUSPENSION BRIDGES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-9
1.3.1 Cables and Cable Wires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-9
1.3.1.1 WIRE MANUFACTURE AND CHEMICAL COMPOSITION . . . . . . . . . . . . . . . . . . . . . .1-9
1.3.1.2 WIRE PROPERTIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-10
1.3.1.3 AERIALLY SPUN CABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-11
1.3.1.4 PARALLEL WIRE STRAND CABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-11
1.3.1.5 HELICAL STRAND CABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-11
1.3.2 Bridge Cable Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-11
1.3.2.1 TOWER SADDLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-12
1.3.2.2 ANCHORAGE AND CABLE BENT SADDLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-12
1.3.2.3 SPLAY CASTINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-12
1.3.2.4 CABLE ANCHORING DEVICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-13
1.3.2.4.1 Strand Shoes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-13
1.3.2.4.2 Parallel Wire Strand Terminations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-13
1.3.2.4.3 Eyebars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-13
1.3.2.5 SUSPENDERS AND CABLE BANDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-13
1.3.2.6 PROTECTIVE APPURTENANCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-14
1.3.3 Cable Protection Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-14
1.3.3.1 ZINC COATING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-14
1.3.3.2 GREASE AND OIL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-14
1.3.3.3 PASTES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-14
1.3.3.4 WRAPPING AND PAINT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-15
1.3.3.4.1 Wire Wrapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-15
1.3.3.4.2 Elastomeric Membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-15
1.3.3.4.3 Fiberglass Reinforced Lucite Composites and Methacrylates . . . . . . . . . . . . . . . . . . . .1-15
1.3.3.5 DEHUMIDIFICATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-15
1.4 Causes of Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-15
1.4.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-15
1.4.1.1 ENVIRONMENTAL FACTORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-16
1.4.1.2 SOURCES OF WATER PENETRATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-16
1.4.1.3 CABLE INSTALLATION PRACTICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-16
1.4.1.4 WIRE SUSCEPTIBILITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-16
1.4.1.5 WIRE STRESSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-16
1-1
1.4.2 Corrosion Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-17
1.4.2.1 ZINC CORROSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-17
1.4.2.2 CORROSION STAGES OF ZINC-COATED WIRES . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-17
1.4.2.3 FRACTURE OF ZINC-COATED WIRES IN THE SPAN . . . . . . . . . . . . . . . . . . . . . . . . .1-17
1.4.2.4 UNCOATED CABLE WIRES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-18
1.4.2.5 ZINC-COATED WIRES IN ANCHORAGES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-18
1.5 INVESTIGATIVE TECHNIQUES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-18
1.5.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-18
1.5.1.1 NDE TECHNOLOGIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-18
1.5.1.1.1 Diagnostic Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-18
1.5.1.1.2 Monitoring Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-18
1.5.1.2 CABLE STRENGTH ANALYSIS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-19
1.6 FIGURES FOR SECTION 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-19
1.7 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-25
FIGURES
Figure 1.3-1. Typical suspension bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-19
Figure 1.3.2.1-1. Tower saddle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-20
Figure 1.3.2.2-1. Cable anchorage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-20
Figure 1.3.2.2-2. Bent strut and saddle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-21
Figure 1.3.2.3-1. Splay casting and collar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-21
Figure 1.3.2.4.1-1. Strand shoe for spun cable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-22
Figure 1.3.2.4.3-1. Eyebars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-22
Figure 1.3.2.5-1. Cable band . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-23
Figure 1.3.2.6-1. Tower saddle opened for cable inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-23
Figure 1.3.2.6-2. Protective sleeve adjacent to tower saddle . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-24
Figure 1.3.3.4.1-1. Two-ply wire wrapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-24
Figure 1.4.2.2-1. Corrosion stages of cable wires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-25
Figure 1.4.2.3-1. Typical wire break . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-25
1-2
1.1 INTRODUCTION
1.1.1 Scope of Guidelines
The purpose of the Guidelines is to provide the following:
Methods of inspecting bridge cables,
Criteria for gathering data to obtain the best statistical sample for the least effort and cost,
Criteria for testing wire, and
Methods of estimating cable strength.
The Guidelines apply to suspension bridge parallel wire cables only. This is especially true of the articles that
describe wedging and interior inspection of the cable and estimation of cable strength. However, much of the con-
tents has relevance in a broad sense to helical strand cables.
1.1.2 Users of Guidelines
The Guidelines are intended for use by the following:
Suspension bridge owners, in planning inspections and preparing specications for scopes of work;
Investigators and subordinate personnel, in planning inspections and evaluating the information gathered in
inspections;
Field inspection personnel, in executing inspections;
Construction crews, in providing access, unwrapping and wedging the cable, and rewrapping or replacing the
protection system; and
Laboratory personnel, in testing wire samples.
The recommended intervals between inspections and locations for inspection are especially useful to bridge
owners, as are the codes and standards and procedures for cable inspection and evaluation. The Guidelines are
not intended, however, to be used as a substitute for hiring experienced staff to inspect or evaluate cables. An engi-
neer who is practiced in the evaluation of cable elements should always be in charge of a cable investigation.
Mathematical prociency at the college graduate level is needed for the estimation of cable strength.
1.1.3 Glossary
A
Adjustment (during cable spinning)Correcting the length of wires in a strand or of strands in a cable so
that all the wires have the same dead load tension.
AnchorageStructure that dead ends the main cable. Also, the material that fastens the cable to it.
B
Bottom of the CableLowest wires in a cross-section.
Bridge WireWire that usually conforms to standard specications of ASTM A586 for suspension bridge cables.
BrightUncoated, or ungalvanized (wire).
SECTION 1
GENERAL
1-3
C
CableMain supporting element of a suspension bridge.
Cable BandsSteel castings that hold the suspenders in place over the cable.
Cable Band BoltsBolts that press the two halves of a cable band against the cable to provide the normal force
that prevents sliding.
Cable CapacitySee Cable Strength.
Cable OpeningLength of cable that is unwrapped for inspection.
Cable StrengthTotal force at which the cable fails.
CastCurvature imparted to a wire during its manufacture.
Cast DiameterDiameter (or radius) of the circle freely adopted by unstressed wire lying on the ground. Also
called Free Coil Radius.
Cold DrawingWire manufacturing method of pulling rods through successively smaller dies at room temperature.
CompactionCompressing the cable to the densest possible circular cross-section.
Continuous WireCable wire that is not broken.
Controlled TensionMethod of maintaining cable wires at a constant tension as they are aerially spun and laid
into strands.
Corroded WireWire that exhibits ferrous corrosion on some of its surfaces (section loss is not implied). See
also Corrosion Stage.
Corrosion StageFour categories of increasing severity that describe deterioration of the zinc coating and fer-
rous corrosion on a bridge wire, based on visual examination.
Cracked WireWire that contains a crack but is not broken, detectable by visual observation of the fracture sur-
faces during laboratory testing of samples (almost impossible to see inside wedged cable).
Crossing WiresCable erection fault in which some of the wires cross over other wires.
D
DehumidicationReduction of ambient relative humidity to prevent condensation.
DistributionMathematical expression used to represent the variation of strength or ultimate elongation of
cable wires.
E
Effective Development LengthLength of a cable outside of which a broken wire will not affect the cable strength
in the panel at the center of this length.
ElongationIncrease of wire length due to changes in stress and/or temperature.
Environmentally Assisted Cracking (EAC)Cracking caused by electrochemical mechanisms associated with
the environment inside the cable.
Error, SystematicRepeating error caused by improper calculation, observation or testing.
Error, RandomError due to the difference between the sample and the real universe being studied.
Evaluated PanelPanel of a cable that has been inspected and for which the cable strength is calculated.
F
Free Coil RadiusSee Cast Diameter.
G
Galvanized WireWire with an applied zinc coating.
Gaussian DistributionNormal probability distribution.
Gross Metallic AreaCross-sectional area of a galvanized wire that includes the zinc coating.
H
Helical StrandStrand composed of wires that are wound around a central straight wire, each layer wound in the
direction opposite the layer below it. Also called Bridge Strand or Structural Strand.
1-4
1-5
Helical Strand CableCable made of helical strands that are placed parallel to one another, generally used for
suspension bridges with shorter spans.
Hydrogen Embrittlement (HE)Brittleness and low strength caused by the penetration of steel by hydrogen.
Hydrogen-Assisted Cracking (HAC)Cracking processes that require the presence of hydrogen.
I
InvestigatorProfessional engineer in charge of performing the inspection and evaluation of a cable.
L
Low Point (of Cable)The panels that are at the lowest elevations in the main and side spans.
M
MacroenvironmentAtmospheric conditions in the general area of a bridge, such as wind, salt spray, and de-icing
salts and other roadway pollutants in the splash zone.
MeanAverage value of a property of cable wire.
MicroenvironmentFactors affecting the wires inside a cable, including water, pollutants on wire surfaces, acid
or alkaline conditions, and radicals such as sulfates, carbonates and nitrates.
Monte Carlo TechniqueStatistical technique for calculating probable variation of minimum cable strength.
N
NDENondestructive evaluation, usually used in reference to devices that detect internal damage without harm-
ing the structure.
NeckingReduction of wire diameter at failure.
Net Steel AreaArea of the steel portion of an uncorroded galvanized wire.
Nominal AreaSpecied area of a bridge wire, either gross metallic or net steel, to be distinguished from the
actual measured area.
Normal DistributionSee Gaussian Distribution.
O
OctantPortion of a cable equivalent to 1 8 of the total cable area lying between two radii and 1 8 of the cir-
cumference, or the wires enclosed in that portion.
OilingIntroduction of oil (generally linseed oil) into the cable to protect cable wires from corrosion.
OwnerPublic or private entity responsible for the operation and maintenance of a bridge.
P
PanelPortion of a bridge or cable that lies between the centers of two adjacent suspenders or cable bands.
Panel LengthLength of a panel measured horizontally.
Panel PointEnd point of a panel.
ParallelismDegree to which cable wires are parallel, with good parallelism marked by very few crossed wires.
Percent ElongationLength change between two gage points, measured under zero load across the break in a
wire in a tension test, expressed as a percent of the pretest unstressed gage length, usually 10 inches.
Ply (of Wrapping Wire)One wire wound around the cable, sometimes in multiple groups next to each other,
forming a single layer. One and two ply wrappings are common, three and four less so; multiple plies do not
imply multiple layers.
PolarizationChange in electrical potential of a component (relative to ground).
Preece TestStandard method for determining the minimum remaining corrosion resistance of a zinc coating that
has deteriorated irregularly.
Protection SystemMethods used to prevent cable corrosion, including wire coating, impregnation with oil, wire
wrapping, painting, neoprene or plastic sheathing, and injection of dried air.
Q
QuadrantPortion of a cable equivalent to 1/4 of the total cable area lying between two radii and 1/4 of the cir-
cumference, or the wires enclosed in that portion.
R
Random NumbersNumbers with no mathematical pattern. See Monte Carlo Technique.
Residual StressStress in a wire with no applied axial or in-plane forces (i.e., lying at on a level oor).
Raw Linseed Oil (RLO)Linseed oil that does not contain a drying agent, used in red lead paste and as a cor-
rosion inhibitor.
RedevelopmentCapacity of a broken wire to regain some or all its prefracture stress at a distance from the fracture.
Redevelopment LengthDistance at which a broken wire regains its full load capacity.
Red Lead PastePaste composed of lead oxide and linseed oil, placed under cable wrapping for additional protection.
Ring (of Cable)Circular portion of a cable that includes all wires at a given distance from the center of the cable.
S
Safety FactorCable strength divided by the maximum cable tension under service loads.
Sag ControlMethod of controlling the position of a wire in an aerially spun cable by controlling its nal drape
during the erection process.
SampleSmall group of wires that are expected to be representative of all such wires with the same charac-
teristics (e.g., the same deterioration stage). See also Wire Sample.
Sector (of Cable)Pie-shaped portion of a cable designated for calculation purposes.
Segment (of Cable)Portion of a cable ring designated for calculation purposes.
ShakeoutInspection of an aerially spun strand by removing all bindings and allowing it to hang free.
Shop-Fabricated Parallel Wire Strand (PWS)Cable strand fabricated offsite that consists of wires placed
parallel to each other.
SpecimenSingle piece of wire, generally cut from a wire sample, on which a test is made.
Spinning (Aerial)Cable erection by repeatedly pulling loops of wire across a bridge until there are suffi-
cient wires to form a strand and sufficient strands to form a cable.
Splash ZoneAlternately wet and dry surfaces caused by stream ow, wave action, or traffic spray.
Standard DeviationStatistical measure of the amount of variation in a property of cable wire.
Straightening StressFlexural stress induced in a wire by decreasing the wire curvature to match the cable
curvature.
StrandIndependent bundle of wires, grouped together systematically to form the main cable.
Strength TestTest in which a cable wire specimen is pulled to failure in a testing machine.
Stress Corrosion Cracking (SCC)Wire cracking due to stress plus corrosion, with or without the assistance
of hydrogen. See Hydrogen Assisted Cracking.
Stress-Strain CurveGraph produced during a strength test showing the relationship between the stress and strain
of a wire specimen.
SuspendersVertical wire ropes or strands connecting the suspended structure to the cable.
T
Tensile StrengthMaximum stress a wire can resist.
Tension ControlSee Controlled Tension.
TestsField or laboratory procedures designed to identify wire properties and forces on cable wires.
Turning PointAny location where a cable changes direction.
U
Ultimate StrainStrain at which a wire fails.
Unloaded Side SpansSide spans with no suspenders (the suspension bridge cable does not support the deck
in the side spans).
UVUltraviolet radiation.
1-6
1-7
V
VarianceStatistical value of a property of cable wire, related to the Standard Deviation.
W
WedgingDriving plastic or oak wedges between cable wires to open a space for observation of internal wires.
Weibull DistributionProbability distribution commonly used for steel strength.
Weight of Coating TestStandard test for determining the average weight of the zinc coating on a bridge wire.
Wire (in cables)Thin longitudinal continuous high-strength steel elements up to 7 mm in diameter that make up
a main cable.
Wire RopeSteel tension element made of 6 or more helical steel strands that are wound around a single central
core of hemp or another wire rope or strand.
Wire SampleLength of wire removed from a cable for testing, from which individual specimens are cut.
Wire StrandA steel tension element that differs from wire rope in that it is fabricated from layers of parallel
longitudinal or helical wires. The latter is also called Bridge Strand.
Wire StrengthForce at which a wire breaks.
Wrapping (Wrapping Wire)Continuous coil of soft steel wire that forms the protective covering on most
cables.
Y
Yield PointFor bridge wire, the point on the stress-strain curve at which the residual strain is 0.2%.
Yield StrengthWire stress at the yield point.
Z
Zinc CoatingZinc layer deposited uniformly on a bridge wire to protect against corrosion, applied according to
galvanizing process codes and standards.
1.1.4 Professional Organizations
The following professional organizations are referred to in the Guidelines with the acronyms listed:
AASHTO American Association of State Highway and Transportation Officials
AISC American Institute of Steel Construction
ANSI American National Standards Institute
ASCE American Society of Civil Engineers
ASTM American Society of Testing and Materials
AWS American Welding Society
EPA Environmental Protection Agency
FHWA Federal Highway Administration
NCHRP National Cooperative Highway Research Program
OSHA Occupational Safety and Health Administration
USCG United States Coast Guard
1.1.5 National Codes and Standards
General PublicationsThe Guidelines recommend experience-based cable inspection frequencies and provide spe-
cic criteria and guidelines for training and qualifying personnel, with reference to the following publications:
AASHTO Manual for Condition Evaluation of Bridges, 2nd Edition, 1994
FHWA Bridge Inspection Training Manual, 1990, 1991
FHWA Inspection of Fracture Critical Members, 1986
NBIS (National Bridge Inspection Standards), Federal Regulations Guide to Bridge Inspection, 1988
Public Safety PublicationsThe federal documents listed below contain codes and standards regarding
public safety.
ARC (American Red Cross) Standard First Aid Workbook, 1991
FHWA Manual on Uniform Traffic Control Devices, 2000
FHWA Guide Design Specifications for Bridge Temporary Works, 1993
OSHA Code of Federal Regulations (CFR), Title 29
OSHA Code of Federal Regulations (CFR), Title 23 (parts dealing with highway safety)
1.2 HEALTH AND SAFETY REQUIREMENTS
1.2.1 General
Cable inspection over trafficked roadways and waterways involves risk to people and the environment. There-
fore, protection of construction workers, inspectors, motorists, pedestrians, and marine traffic is an important con-
sideration in planning and executing the work.
Owners often require that the contractor develop a Health and Safety Plan subject to their nal approval and in
accord with their specic health and safety practices. The plan must incorporate all governmental provisions along
with those of the owner.
1.2.1.1 LEAD ON THE CABLE
Cable unwrapping releases dried lead oxide paste into the environment, which is prohibited by federal reg-
ulations. Protective enclosures must be provided. Monitoring the blood of inspectors and workers and follow-
ing procedures to prevent lead absorption are mandated by both the EPA and OSHA. Work clothes must be
replaced periodically and never taken home.
Smoking accelerates lead absorption and, if permitted, must be accompanied by strict hygiene. Smoking on the
work platform, where there are ammable materials, should be prohibited.
As stipulated by OSHA in Title 29, Subpart D, 1926.62, containment, handling, and disposal procedures must
be approved before the start of work.
1.2.1.2 OTHER OSHA REQUIREMENTS
OSHA requirements also cover work safety in a traffic environment (Title 23) and work under exposed con-
ditions (Title 29 Subparts L and M).
Some owners require that inspectors wear a safety harness with two lanyards attached at all times to two
handropes or other elements independent of the work platform. Other owners require latching onto the handropes
while walking the cable but not while working inside an enclosed work platform.
1.2.1.3 PERSONNEL TRAINING PROGRAM
Personnel should be made familiar with the health hazards related to cable inspection and trained in the use
of the equipment and monitoring procedures associated with health maintenance. They should refer to the
OSHA Compliance Manual Training Requirements.
1.2.1.4 PEDESTRIAN AND VEHICULAR TRAFFIC
To protect traffic on the bridge and on roadways, waterways and areas beneath the bridge, all construction equip-
ment, material, small tools, wedges that are sometimes ejected from the cables by wire pressure, etc., must be retained
within the connes of the work platform. Work platforms should have kick plates and closed oor penetrations around
1-8
1-9
suspenders for this reason. Fine netting or a fall shroud is recommended. It should be secured to the handropes and
enclose the work space completely from one side of the platform to the other.
1.2.1.5 NAVIGATIONAL SAFETY
The USCG must approve all work over navigable waters. Plans for the protection of vessels must be part of the work
specications. The USCG regulates burning (falling sparks), construction over shipping, the quantity of ammable
gases on the catwalk, dropped material, pollution of the waterway, etc. They are expected to impose security regula-
tions, including those concerning communications between crews at the cable site and crews on approaching vessels.
1.2.1.6 BRIDGE SAFETY PLAN
The owner is generally responsible for following the construction-specific regulations of local authorities that
concern work near railroad tracks or pedestrian sidewalks, maintenance of traffic and lane closings, etc. The owner
also determines the various types of insurance needed for the cable work. The contractor and investigator bear
responsibility to the owner for following regulations.
A stand-by boat is often required while inspectors or workers are on the platform above. Painting operations on
cables have been reported to cause damage to vehicles. Removal of suspenders (for access to cable bands) may
require work near active highways or railroad tracks, which may have to be closed for short periods of time to
protect public safety.
1.3 SUSPENSION BRIDGES
Suspension bridges are large, unique structures with two or more cables that carry the immense weight of the
deck and most of the imposed live load to the towers that support them. The suspension system is in tension and
requires substantial anchorage at both ends. The cables are fracture critical and load path nonredundant. Figure 1.3-1
shows an elevation of a typical suspension bridge. The major elements are identied along with the low points of
the cable, which are signicant for cable evaluation and are referred to in Section 2.
1.3.1 Cables and Cable Wires
Most North American bridge cables consist of wire with a 0.192-inch diameter, a 0.002-inch zinc coating, a
6-foot cast (i.e., a 3-foot free coil radius), and approximately 0.8 percent carbon content. The specied minimum
tensile strength varies between 215 ksi and 225 ksi, based on the gross metallic area of the wire, which includes the
cross-sectional area of the zinc coating.
A notable exception is the Williamsburg Bridge in New York, which has cables composed of bright (ungal-
vanized) wire, with an average diameter of 0.1907 inches. Another exception is the Brooklyn Bridge, which has
galvanized wire with a smaller than average diameter of 0.184 inches.
The free coil radius of the wire is the result of the manufacturing (i.e., drawing) methods; it is imparted to the
wire as it is wrapped around the capstans that pull it through the dies. Some wire manufacturers are currently pro-
ducing nearly straight wire with a free coil radius of 25 feet or more, and with a minimum strength of 260 ksi. This
wire may be more crack resistant. The manufacturing process also results in residual stresses in the wire, which may
be about 30 ksi tension at the inside of the cast.
1.3.1.1 WIRE MANUFACTURE AND CHEMICAL COMPOSITION
Cast
Small cast diameters imply high tensile stress on the inside of the coil when the wire is straightened in the cable.
Cracks nearly always originate in this area of the wire, where the tensile bending stress caused by straightening the
wire under service loads is at a maximum.
In contrast to the straightening stress in the span, the stress due to increased curvature around strand shoes
reduces the effect of the residual wire stress in a wire with a cast. A hypothetical trouble-free cable would consist
of straight (no cast) wires that eliminate the exural stress caused by straightening the curved wire. The strands would
be socketed, thus avoiding the introduction of high exural stress as the wire bends around the strand shoe. Also,
with socketed strands the wires would all be exactly the same length, assuming the manufacturer was competent,
reducing the number of crossing wires inside the cable.
Chemistry
The wires alloying elements affect its susceptibility to stress corrosion by governing hydrogen absorption rates.
Cr (chromium) increases and Ni (nickel) decreases the susceptibility to hydrogen embrittlement (HE). Si (sil-
icon) and P (phosphorus) have little or no effect. Silicon is sometimes added to develop ne grain in the wire, while
nickel and chromium are usually present as residual elements in the manufacture of the material.
The following table lists the elements in cable wire and their approximate weight by percent.
Carbon content should be high enough to provide a pearlitic steel structure but not exceed 8%, the maximum when
there is no free cementite. Exceeding 8% results in a structure that is hard but brittle and tends to fracture under cold
drawing. The microcracks produced in cementite reduce toughness and provide traps for hydrogen to do its worst
damage.
Galvanizing
Cold-drawn wires are traditionally acid-cleaned before galvanizing. Care must be taken to remove all hydrogen
from the wire before galvanizing.
A hot-dipped zinc coating isolates the wire from the corrosive media; and, if steel is exposed (at a scratch or hol-
iday in the coating), the zinc-iron cell provides cathodic protection to the exposed steel from the wet environment.
This is a mixed blessing, because polarization will generate hydrogen ions in the electrolyte (the microenviron-
ment) that can enter the steel.
The zinc coating must be applied after cold working and after the wire is treated to remove all possible hydrogen
from the metal. Hydrogen passes through steel but not through zinc.
The zinc/steel interface is a brittle zinc-iron alloy, which may facilitate corrosion of the steel after the zinc coat-
ing has been depleted.
1.3.1.2 WIRE PROPERTIES
Strength
High strength wires are desirable, even though they are more susceptible to corrosion. Cold-drawn wires are more
resistant to stress corrosion and hydrogen-assisted cracking than heat-treated wires having the same mechanical
properties.
Hardness
Hardness is related to carbon content; yield strength increases with increasing hardness. Wires are generally not
tested for this property.
Weight by Percent
Element Minimum Maximum
Carbon 0.65 0.84
Sulfur 0.04
Phosphorus 0.04
Manganese 0.80
Silicon 0.15
Nickel 0.12
Chromium 0.10
1-10
1-11
Toughness
Unless notched by corrosion, bridge wires are not susceptible to fatigue from highway loading. The critical stress
intensity factor, K
IC
, is only meaningful if corrosion causes notches and the material has become embrittled.
The stress intensity factor for stress corrosion cracking, K
ISCC
, which determines time to failure, varies
depending on stress and environmental load intensity. There is no standard test for wire toughness, K
ISCC
or
K
IC
, although Mayrbaurl [1] describes a possible technique of estimating K
IC
from tensile tests of cracked wires.
K
ISCC
is the stress corrosion equivalent of K
I
, which is used for fatigue. It is the stress intensity at which a
crack becomes unstable. The incubation time for corrosion in a constant environment is related to K
ISCC
.
1.3.1.3 AERIALLY SPUN CABLES
Twenty-nine bridges with aerially spun cables in North America were erected before the end of the twenti-
eth century. Making sure that the wires sag equally during erection ensures that all the wire tensions will be
equal. Once all the wires in a strand are erected, that strand is adjusted to the other strands by this sag method,
which is very accurate.
Cutting out or splicing in wire is the method used to correct the few wires found to have incorrect sags dur-
ing the shakeout of a strand. The process occasionally connects a wire to another wire, causing crossing wires
and small voids in the cable that allow water to enter and remain longer than in a well-compacted cable. Cross-
ing wires creates points of contact where the zinc coating can be displaced by the pressure of compaction and
become sites of local deterioration. Broken wires are often found at these points of contact. It is common to
find some voids due to crossing wires during inspection. Modern cables that are aerially spun by a controlled
tension method or made of shop-fabricated parallel wire strands should have few or no crossing wires. They
are expected to be very compact and water-resistant.
1.3.1.4 PARALLEL WIRE STRAND CABLES
Cables that consist of shop-fabricated parallel wire strands may perform better than cables with aerially spun
strands, because the strands are fabricated offsite, where the process can be controlled more easily than in the
field. Moreover, the strands are often composed of nearly straight wires, with minimal residual stress, and have
socketed ends, both distinct advantages in lowering tensile stress.
There are two bridges in the United States with cables that fit this description:
Newport Bridge, Newport, Rhode Island; and
William Preston Lane Jr. Memorial Bridge II, Sandy Point to Kent Island, Maryland.
1.3.1.5 HELICAL STRAND CABLES
In the United States and Canada, there are 21 suspension bridges with cables composed of helical wire
strands. They are not covered by the Guidelines and are primarily bridges with shorter spans.
The spaces between the outer strands are filled with prismatic wood or metal shapes machined to fit between
strands and provide a circular exterior shape. The Dunvegan (Peace River) Bridge in Alberta, Canada, is com-
posed of separated helical strands with wide spaces for strand replacement and no fillers.
1.3.2 Bridge Cable Details
The parts of the cable system other than the cable itself, such as the hardware and fittings at the cable turn-
ing points and the connections of the suspenders to the cable, affect the performance of the wires and their
inspectability. They are discussed below in relation to the bridge component they are part of.
1.3.2.1 TOWER SADDLES
The vertical forces on the tower tops are very large. They are, with few exceptions, greater than the total
dead plus live load of the main span plus both side spans. Virtually the entire weight of the bridge is supported
at the top of the tower where the cable changes direction. This force is transferred from the cable into the tower
by the tower saddles.
The downward force that the anchorages exert on the cables (if the cables slope down toward the anchor-
ages) is far greater than any possible side span reaction. Thus, the tower resists the entire weight of the bridge
plus the downward force of the anchorage on the cable. There is no other component to resist this downward
force.
The normal pressures delivered by the wires to the saddle are inversely proportional to the saddle radius.
Wires in the lower strands touching the bottom surface of the saddle are subjected to the greatest normal pres-
sure. These stresses theoretically reduce the yield capacity of the wire, but no damage will occur if reasonable
saddle proportions are used.
Saddles are usually fixed to the tower tops, which deflect whenever side and center span forces differ. The
deflection reduces the force differential to a negligibly small value that the flexible tower resists.
A typical tower saddle is shown in Figure 1.3.2.1-1. The saddle in Figure 1.3.2.6-1 had a cover plate bolted to
it, which was removed for inspection, and a separate housing that covers the entire saddle. A layer of paraffin
has been taken off the wires to facilitate inspection.
1.3.2.2 ANCHORAGE AND CABLE BENT SADDLES
Anchorages and cable bent saddles redirect the cable so that it comes into alignment with its anchoring mecha-
nisms. Anchorage saddles, called splay saddles, sometimes have a variable vertical radius and a horizontal are, so
that the cable strands can splay directly from the saddle to their anchoring mechanisms. If the anchorage saddle is
not ared, then a splay casting is applied between the saddle and the cable anchoring device. The saddles, splay cast-
ings, strand shoes and eyebars are all located in or near the cable anchorage. Figure 1.3.2.2-1 is an illustration of a
typical anchorage.
Anchorage saddles are sometimes supported on an integral part of the anchorage. They bear on rollers to allow
for changes due to stress and temperature in the length of cable that runs from saddle to anchor.
On several bridges, damaged and broken wires have been observed at the bottom of the cable at both ends of
the saddle. They result from water that has entered the cable outside the anchorage and comes to rest at the upper
end of the saddle or runs through the saddle and wets the wires at the bottom or condenses on the wires at either
end. The condensation is often from humid spring air entering an anchorage that retains the cold of the previous
season.
Cable bent saddles usually bisect the interior angle of the cable. The cable changes direction with a single verti-
cal curve that proceeds from the saddle towards the anchorage without aring to a splay casting.
The saddles are supported on independent struts that may be hinged at the base or, if the struts are exible
enough, xed at the base.
Cable bent saddles are also located on separate cable bents at the ends of side spans whenever the anchorage
structure does not extend to the roadway level. In this case, they do not bisect the interior cable angle, and addi-
tional strands are provided to prevent slippage of the cable on the saddles.
Figure 1.32.2-2 illustrates a typical cable bent saddle.
1.3.2.3 SPLAY CASTINGS
Splay castings are elements that control the direction of strands that are out to their respective anchoring
devices. The castings are designed to resist the outward force exerted by the strands and are anchored against
upward slippage by a cable collar clamped above the splay casting.
Passage of water through the cable may cause corrosion in the wires inside the splay casting. It will take the form
of a white or brown leachate at the lower end of the splay casting.
1-12
1-13
The inspection of wires within the splay casting area is a complex operation that requires the relocation, usually
temporary, of the splay casting. The same tension is maintained in all the splayed strands by rerouting them through
holding frames that move the splay point upward along the cable while both the length and the original stress of
the wires are maintained.
A splay casting detail is illustrated in Figure 1.3.2.3-1.
1.3.2.4 CABLE ANCHORING DEVICES
1.3.2.4.1 Strand Shoes
Traditionally, cable wires loop around strand shoes, which are anchored to two eyebars by a pin. Later versions
use a single eyebar and two strand shoes, one at each end of the pin, splitting the strand into four quarters. Other sys-
tems restrain the strand shoe with high-strength anchoring rods.
If the atmosphere in the anchorages is humid or water enters the anchorage through the deck, damage is often
found in the lower half of the strands, particularly at the interface of the wires and the strand shoe where water
tends to collect. The damage in this area is usually uniform corrosion that causes section loss.
The repair of the strands at the strand shoes depends on the extent of the damage. If there are only a few broken
wires, they are generally repaired individually by cutting and splicing in new wires. Extensive damage is often repaired
by cutting the damaged strand, socketing the end, and then reanchoring it.
A typical strand shoe is shown in Figure 1.3.2.4.1-1.
1.3.2.4.2 Parallel Wire Strand Terminations
Parallel wire strands are terminated in zinc or polyester thermoset resin sockets rather than in strand shoes. The
sockets are connected to anchoring assemblies embedded in the anchorage concrete.
1.3.2.4.3 Eyebars
Eyebars are anchored to a grillage buried deep in the concrete mass of the anchorage. The focus of the eyebars
may not be at the splay casting or cable bent saddle but slightly beyond to provide an inward component for the
strand forces on the ends of the eyebars. To prevent the eyebars from bending, spacers are placed between the eye-
bars of each separate strand so that the eyebars bear against each other. They are often tied transversely with a cen-
ter rod that passes through the holes in all the pins. In humid anchorages, eyebar corrosion is found at the inter-
face with the concrete mass. The damage is often hidden behind pack rust. Figure 1.3.2.4.3-1 illustrates a typical
array of eyebars.
1.3.2.5 SUSPENDERS AND CABLE BANDS
Suspenders are expected to last between 35 and 65 years. The suspenders with longer life spans have been
prevented from rubbing against the structure, particularly at the lower ends.
Suspenders made of wire rope have two legs, but they are actually a single continuous piece that rises from
the deck, loops over the cable band, and returns to the deck, using only two sockets.
If suspenders are designed to ride cross-saddle over the cable bands, then they contribute to the normal force
provided by the cable band bolts and increase resistance to cable band sliding. This is true of most North Ameri-
can bridges.
Suspenders made of wire strands that cannot be bent over the cable require a socket. They are stronger and thus
lighter than rope suspenders, and they simplify cable band detailing somewhat.
Cable bands consist of two cylinder halves bolted together over the circumference of the cable. The number of
bolts per cable band is dependent on the slope of the cable at the suspender attachment point. The friction from
squeezing against the cable prevents it from sliding down the cable. The steeper the cable, the more bolts are needed
to prevent this sliding. If a cable band has to be removed, the suspenders on the band must also be removed.
The tension in the cable band bolt that squeezes the two halves together decreases with time due to creep of the
zinc coating on the wires. The loss may average as much as 65%of the original tension.
Figure 1.3.2.5-1 illustrates a typical cable band.
1.3.2.6 PROTECTIVE APPURTENANCES
Cable wire areas that are exposed to the weather require additional protection. The cable wrapping or protection sys-
tem stops short of the saddle, creating a signicant gap in the system. The cable wires are also exposed at the upper sur-
face of the saddles. A housing should enclose the entire saddle, or a plate should be bolted to the top of the saddle trough
to create a weather-tight space for the cable. These provisions must be made in a manner that allows for inspection.
Figure 1.3.2.6-1 shows a saddle that has both a plate bolted to the saddle and a separate housing.
If tower saddle troughs are hexagonal, then the transition of the cable from its usual circular shape to a hexago-
nal one complicates compaction and long watertight sleeves are used in place of wrapping. The sleeve must allow
for cable slope changes relative to the saddle. A sleeve detail is shown in Figure 1.3.2.6-2.
1.3.3 Cable Protection Systems
1.3.3.1 ZINC COATING
With very few exceptions, cable wires are zinc-coated. Depending on the effectiveness of the exterior protective
system, the coating may last indenitely or could become depleted within 20 years on a large number of the wires.
Complete zinc depletion is equivalent to corrosion at the end of Stage 2 and the beginning of Stage 3 (see Article
1.4.2.2). This condition may describe the entire wire or a local holiday, which is a small discontinuity in the coating
that encourages corrosion. In either case, it marks the start of embrittlement and environmentally assisted cracking.
1.3.3.2 GREASE AND OIL
The wires of some early bridges were greased during spinning or as the cable was being compacted. In some
cases, the greased wires appeared almost new after nearly a century, despite the often deplorable condition of the
exterior wrapping. The Brooklyn Bridge and the Manhattan Bridge are notable examples.
In other cases, despite the presence of grease, wires cracked and failed in localized zinc-depleted regions. It has
been suggested that corrosion may attack holidays in greased or oiled wire for longer periods of time than holidays in
ungreased wire. This could explain why corrosion in greased and oiled wire is limited to a very few small isolated
islands of severe damage.
1.3.3.3 PASTES
In the past, various paste mixtures were used as a layer of protection under new wrapping wires to prevent water
penetration. Red lead paste was the chief ingredient; zinc-based products have been used more recently.
Red lead pasteThe paste used traditionally under the wrapping wire was a mixture of lead oxide powder
(Pb
2
O
3
) and raw linseed oil (RLO), specied in the following proportions by weight:
lead oxide 95%
raw linseed oil 5%
These quantities translate roughly to 70% lead oxide and 30% RLO by volume.
Freshly applied red lead paste releases many inhibitive ions that decrease the potential difference between
cathode and anode. However, in time the oil polymerizes and shrinks and the paste cracks, becomes friable,
and is no longer an impermeable barrier.
Zinc-based pasteDue to the hazardous nature of red lead, zinc-rich pastes are now used in Europe and the
United States.
Moisture-cured urethane with pure zinc pigment was used on the Storebaelt Bridge in Denmark. In the United
States, a material composed of zinc, zinc oxide, corrosion inhibitors, and organic-based oil that resists polymer-
ization was used in several cable rehabilitation projects.
Lead and calcium-based pasteA mixture of calcium plumbate, calcium carbonate, and linseed oil has been
used on the Honshu-Shikoku Bridges in Japan.
1-14
1-15
1.3.3.4 WRAPPING AND PAINT
Several types of exterior cable protection systems have been used on North American bridges, including
Paste, wire wrapping, and a paint system;
An elastomeric membrane;
Fiberglass-reinforced lucite composite; and
Combinations of the above.
1.3.3.4.1 Wire Wrapping
Wrapping typically consists of soft galvanized No. 9 wires with Class A zinc coating. Some newer bridges in
Japan use an S-shaped wire that interlocks with the other wires. Wrapping is installed by power-driven machines
with multiple reels that are capable of placing from one to three wires at a time. Figure 1.3.3.4.1-1 shows a cable
that is wrapped using two reels, which results in a two-ply wrapping. The wires are in a single layer, in side-by-side
helices. In this gure, one of the two wrapping wires has broken, resulting in gaps between wires through which cable
wires are visible.
Paint systems used to cover and seal the wrapping have changed over the years from alkyds to moisture-cured
urethanes and then to water-based paints, which are excellent barriers against water entering the cable. Unfortu-
nately, they also retain water inside the cable.
1.3.3.4.2 Elastomeric Membranes
Elastomeric membranes are also barriers against water penetration, and they also cause water that enters through
the cable bands to be retained inside the cable, unless relief systems are provided.
Overlapped neoprene and hypalon are sometimes applied directly over the entire wrapping as a protective mem-
brane for the wrapping wires. Protective paint is usually hypalon-based, because it is not as sensitive to ultravio-
let (UV) rays as neoprene.
1.3.3.4.3 Fiberglass Reinforced Lucite Composites and Methacrylates
Two North American bridges are protected by a berglass reinforced lucite composite system. The glass ber is used
for reinforcement of the lucite, which is a transparent matrix belonging to the class of compounds known as methacry-
lates. The system is applied over a polyethylene sheet to separate the matrix from the bridge wires, and a pigment is
added to obtain the desired color for the composite material. A well-designed caulking system at the cable bands is
required, because the protective shell in each panel is not subject to the same strain as the cable wires. The performance
of berglass-reinforced lucite cannot be conrmed at this time.
Methacrylates are UV insensitive and fresh methacrylates, unlike thermoset epoxies, will bond to an existing cured
methacrylate. Eliminator, a methacrylate product that is sprayed on and forms a thick exible membrane, was used over
the poorly wire-wrapped cables of the Bosphorous Bridge and was somewhat effective.
1.3.3.5 DEHUMIDIFICATION
Pastes are not used in the most recently developed cable protective system, and the wires are not greased. First
applied to the Akashi Kaikyo Bridge, the new system consists of an S-shaped steel wire wrapping interlocked with
an elastomeric membrane cover. By injecting dehumidied air, the cable is kept at or below 40% relative humidity
[2], which is believed to be the threshold below which corrosion does not occur.
1.4 CAUSES OF CORROSION
1.4.1 General
The fundamental cause of corrosion is the presence of water and its solutes. Without water, corrosion does not
occur. Many conditions affect the quantity of water in the cable, its corrosive quality, and the wires susceptibility
to attack.
1.4.1.1 ENVIRONMENTAL FACTORS
The bridge atmosphere often contains moisture, pollutants, dissolved gases, and salt spray, all of which may
cause corrosion or induce hydrogen attack. De-icing salts and the proximity of the cable to the vehicle splash zone
at deck level are some of the other external factors that are grouped together under the term macroenvironment.
Water enters the cable as a liquid from precipitation or as vapor during periods of high temperature and humid-
ity. Water vapor becomes liquid when the temperature falls, and condensation forms on the surface of the wires.
Conditions inside the cable that can affect the wires are referred to as the microenvironment.
Five types of microenvironments observed in bridge cables cause cable wire to crack and break:
Acid rain chemistry, leading to hydrogen evolution from reaction with zinc;
Carbonate or bicarbonate chemistry, either alkaline or highly acidic;
Nitrate chemistry, either alkaline or acidic;
Alkaline chemistry; and
Seawater or salt spray, moderately alkaline.
1.4.1.2 SOURCES OF WATER PENETRATION
Exterior protection systemWater enters cables that are poorly wrapped or have cracked paint over the
wrapping.
Lower cable band groovesJoints that are often completely or partially open to allow for weeping of inter-
nal water may become points of entry for water streaming along the underside of the cable or from wind-
driven rain.
EnclosuresDamaged or poorly maintained housings and ashing enclosures have been the source of many
wire failures near saddles and anchorages.
CondensationPaint cracks and other entry points for water are also entry points for water vapor. The only
effective way to prevent condensation is dehumidication.
1.4.1.3 CABLE INSTALLATION PRACTICES
Poor cable compaction and crossing wires cause un-usually large voids in the cable that lead water deep into
the cable. Crossing wires also squeezes out zinc at their point of contact, providing opportunities for polarization,
which is activated by the electrolytes in the water.
1.4.1.4. WIRE SUSCEPTIBILITY
The strength of cable wire is due to its high carbon content and the cold working of the steel. Both conditions
render the wire more disposed to corrosion than mild rolled steel. A zinc coating that covers the entire surface of
the wire is benecial as long as it is not breached. Localized zinc depletion encourages the polarization that causes
wire degradation.
1.4.1.5 WIRE STRESSES
Tensile stress reduces the resistance of wires to corrosion. The stress in the wires comes from residual bend-
ing stress due to cast, straightening stress due to removing the cast, and dead and live load tension.
Wires manufactured with small cast radii have a high residual bending stress, estimated to be 30 to 36 ksi by
X-ray diffraction, which adds to the still higher bending stress induced by straightening the wire to the cable cur-
vature. The resulting stresses, which can be greater than the dead load tension, are tensile inside the cast curva-
ture and compressive outside of it. Combined with dead load, the total stress on the inside of the curvature is at
or near the wire yield strength. It is not a coincidence that virtually all cable cracks occur on the inside of the cast
circle.
1-16
1-17
Modern straight wires dont have this stress, but if they are aerially spun, then large diameter strand shoes
must be used to minimize strain. Straight wire strands socketed at the ends are ideal. They have low bending
and residual stress in the span, and the sockets eliminate the need for strand shoes.
1.4.2 Corrosion Mechanisms
1.4.2.1 ZINC CORROSION
Zinc oxidizes in dry air to zinc oxide (ZnO). Further reactions take place in the presence of moisture: e.g.,
atmospheric carbon dioxide and sulfur dioxide dissolved in water react with the zinc and the ZnO to produce
zinc carbonate (ZnCO
3
), zinc hydroxide (Zn (OH)
2
) and zinc sulfate (ZnSO
4
).
ZnO and Zn(OH)
2
are poor passivating films. In contrast, ZnCO
3
is effective. Environmental conditions that
interfere with the formation of ZnCO
3
may lead to rapid degradation of the zinc coating. ZnSO
4
is very solu-
ble and is easily removed by moving water. Hence, acid rain depletes the zinc coating.
Acid rain in contact with zinc generates hydrogen. For example, carbonic acid (H
2
CO
3
) is a weak acid that
produces hydrogen.
The zinc depletion rate is governed by the extent of atmospheric pollution (industrial pollution is more
aggressive than marine environments); the duration, frequency and quantity of moisture that enters the cable;
and the rate of drying.
1.4.2.2 CORROSION STAGES OF ZINC-COATED WIRES
The following discussion applies to galvanized wires in the span. Galvanized wires in the anchorage and all
ungalvanized wires undergo different corrosion mechanisms.
After the zinc coating is depleted, mechanisms that depend on wire stress and probably involve hydrogen
begin to embrittle the wire and to attack it with surface corrosion, transverse cracks, and brittle fracture.
Wire corrosion is categorized visually by corrosion stage. The four corrosion stages are characterized by the
presence of the following:
Stage 1spots of zinc oxidation on the wires;
Stage 2zinc oxidation on the entire wire surface;
Stage 3spots of brown rust covering up to 30% of the surface of a 3-inch to 6-inch length of wire; and
Stage 4brown rust covering more than 30% of the surface of a 3-inch to 6-inch length of wire.
The corrosion stages are illustrated in Figure 1.4.2.2-1. They were adopted from the research performed by
Hopwood and Havens in 1984 [3]. Even if corrosion is highly localized, it is still as destructive as if it were
widespread.
Stage 2 wire may have white surface dust, indicating zinc oxidation, but that does not necessarily imply
depleted zinc. Depletion is indicated by a dull gray color, or a dark gray to black color if sulfur is involved.
Laboratory tests have shown that 5% to 20% of Stage 3 wires and 60% of Stage 4 wires may have cracks.
1.4.2.3 FRACTURE OF ZINC-COATED WIRES IN THE SPAN
Most wires inside a compacted cable display the flat-and-invert type of failure. At the same time, they do
not show the necking (i.e., ductile failure) found in new wires.
The flat-and-invert type of failure is cracking that starts as a transverse fissure on the inside radius of the
wire cast and then continues at an angle to the wire horizontal axis up the final break, which is transverse. The
same type of wire failure was also found in the helical strand cables investigated by Hopwood. Completely
transverse and spiraling failures have also been observed. Figure 1.4.2.3-1 illustrates the most frequently found
failure mechanism.
1.4.2.4 UNCOATED CABLE WIRES
Some cables are composed of uncoated wires, also referred to as bright wires. They are damaged by pitting and
reduction of section. Whenever zinc is present in the wrapping wire, however, the cracking described above is often
present as well, as was the case with the Williamsburg Bridge cables. Very few interior wires were found broken
and no denitive failure type was reported, although one photograph showed an internal wire with a square break,
indicating the possibility that it was cracked.
1.4.2.5 ZINC-COATED WIRES IN ANCHORAGES
Wire failure mechanisms in the anchorages are typically not the same as those in the compacted cables on the
bridge spans. In anchorage strands, which are exposed to an abundance of oxygen, wires display uniform and pit-
ting corrosion with large reductions in cross-section. Failed ends have long needle-like points.
Tension tests made on wires removed from the cables in the anchorage of the Manhattan Bridge showed that the
tensile strength in the steel where corrosion reduced the cross-section was unchanged from the tensile strength in
the original wire; the individual wire strength was proportional to the remaining area. While this data is limited, it
can be assumed that all bridge wires follow this pattern.
1.5 INVESTIGATIVE TECHNIQUES
1.5.1 General
The primary method of assessing cable damage is visual examination of the wires supplemented by laboratory
analysis of representative wire samples removed from the cable.
There are no currently available methods for obtaining significant information based on nondestructive or
remote technologies. Acoustic monitoring, which records wire failure, is effective, but it does not supply data
on the number of failed wires that preceded monitoring or the capacity of degraded wires.
1.5.1.1 NDE TECHNOLOGIES
Nondestructive evaluation (NDE) provides information about conditions inside the cable without the investiga-
tor having to remove the covering or otherwise alter the condition of the cable. See the nal report for NCHRP Proj-
ect 1057, on the accompanying CD, for a more detailed discussion.
1.5.1.1.1 Diagnostic Techniques
Current NDE technologies can be used to detect damage in strands up to 4 inches in diameter. Detectability
in larger strands is theoretically possible but has not been substantially proved. NDE determines the loss of cross-
section, but it will not necessarily identify cracks, and cannot distinguish among corrosion grades.
Many NDE technologies that apply to bridge cables exploit the basic principles of electromagnetism, such as
the following:
Flux leakage reactions,
Magneto-restrictive reactions, and
Eddy current effects.
They require cumbersome magnetic and electric circuitry. Also, calibration is necessary to distinguish between
intact and damaged strands. They are best suited to assessing the condition of suspender ropes or strands outside
the cables.
1.5.1.1.2 Monitoring Techniques
Acoustic Monitoring is a technology that uses sound to detect wire failures as they occur. Accelerometers are
attached to the exterior of the cable at predetermined intervals and monitored continuously. Sonic data provide
1-18
1-19
an inventory of wire failures, along with accurate data on the location of failures, which may be very useful in esti-
mating the rate of cable degradation. Wires that have failed prior to installation cannot be detected.
Acoustic Emissions generally refers to the detection of sound emitted by degrading material at the microscopic
level. Applicability to cables has not been demonstrated.
1.5.1.2 CABLE STRENGTH ANALYSIS
The estimation of cable strength is a probabilistic exercise, due to the impossibility of knowing the properties of
each individual wire in the cable. The models for strength estimation depend on data gathered during internal
inspections and on laboratory testing of samples removed during these inspections.
It may not be necessary to remove a full set of samples from the cable each time it is inspected, providing that
the number of samples taken during earlier inspections is adequate, as described in Section 2.
Sections 3, 4 and 5 present the analytical techniques and strength models that should be used in estimating cable
strength. Section 6 lists the information required in written reports for all levels of inspection.
1.6 FIGURES FOR SECTION 1

LOW POINTS IN MAIN SPAN
PANEL AT EACH SIDE OF
MIDSPAN
LOW POINT IN SIDE SPAN
PANEL ABOVE ANCHORAGE
OR BENT SADDLE
LOCATION OF LOW POINTS
SIDE SPAN MAIN SPAN SIDE SPAN
ANCHORAGE TOWER
CL
STIFFENING TRUSS AND DECK
IDENTIFICATION OF MAJOR ELEMENTS
CABLE SUSPENDERS
Figure 1.3-1. Typical suspension bridge.
1-20
Figure 1.3.2.1-1. Tower saddle.
Figure 1.3.2.2-1. Cable anchorage.
HALF LONGITUDINAL SECTION HALF ELEVATION
END VIEW
CL CABLE
CL CABLE

ANCHORAGE
MASS BLOCK
EYEBAR
ANCHOR
GIRDERS
SPLAY
COLLAR
SPLAY
CASTING
SPLAYED
STRANDS
BENT STRUT
BENT SADDLE
CABLE
1-21
CONCRETE
CABLE OR
SPLAYED
STRANDS
BENT STRUT
CABLE
BENT SADDLE
SEGMENTAL
ROLLERS
Figure 1.3.2.2-2. Bent strut and saddle.
Figure 1.3.2.3-1. Splay casting and collar.
END VIEW ELEVATION SECTION
SHELLS HAVE A GAP SO THAT BOLT
TENSION DEVELOPS FRICTION
BETWEEN COLLAR AND CABLE
SPLAY CASTING COLLAR
END VIEW ELEVATION SECTION
SPLAY CASTING
SHELLS BEAR ON EACH OTHER
TO LIMIT VARIATION IN BOLT
TENSION CAUSED BY LIVE LOAD
1-22
ELEVATION
SECTION
PIN
WIRE
GROOVE
BEARING BLOCK
PIN
SHIMS
Figure 1.3.2.4.1-1. Strand shoe for spun cable.
Figure 1.3.2.4.3-1. Eyebars.
SPLAYED
STRANDS
SHIMS AND BEARING
BLOCK IF NOT PROVIDED
AT STRAND SHOE
MASS BLOCK
PINS
EYEBARS
ANCHOR GIRDER
1-23
SECTION
ELEVATION
CABLE BAND BOLTS
CAULKING
GROOVES
SUSPENDER GUIDES
CL CABLE
BAND BOLT
STANCHION
CAST STEEL
CABLE BAND
HANDRAIL ROPE
CL CABLE
BAND BOLT
CABLE BAND HALVES HAVE A
GAP SO THAT BOLT TENSION
DEVELOPS FRICTION
BETWEEN BAND AND CABLE
SADDLE HOUSING
CABLE SADDLE
CABLE WIRES
Figure 1.3.2.5-1. Cable band.
Figure 1.3.2.6-1. Tower saddle opened for cable
inspection.
1-24
WRAPPING WIRE
ELASTOMERIC
MEMBRANE
COVER
BRIDGE CABLE
SLEEVE
ANCHORING BAND
CABLE SLEEVE
FLASHING COLLAR
TOWER SADDLE
Figure 1.3.2.6-2. Protective sleeve adjacent to tower saddle.
Figure 1.3.3.4.1-1. Two-ply wire wrapping.
1-25
1.7 REFERENCES
1. Mayrbaurl, R. and S. Camo, Cracking and Fracture of Suspension Bridge Wire. ASCE Journal of Bridge Engineering, 2001. 6(6):
p. 645650.
2. Suzumura, K. and S.-I. Nakamura, Environmental Factors Affecting Corrosion of Galvanized Steel Wires (unpublished manuscript).
3. Hopwood, T. and J. H. Havens, Corrosion of Cable Suspension Bridges. 1984, Kentucky Transportation Research Program, University
of Kentucky Lexington.
Stage 1 Stage 1
Stage 2 Stage 2 Stage 3 Stage 3 Stage 4 Stage 4 Stage 1 Stage 1
Stage 2 Stage 2 Stage 3 Stage 3 Stage 4 Stage 4
CRACK SURFACE
FINAL FRACTURE
Figure 1.4.2.2-1. Corrosion stages of cable wires.
Figure 1.4.2.3-1. Typical wire break.
2-1
CONTENTS
SECTION 2 INSPECTION...............................................................................................................................2-1
2.1 INTRODUCTION.......................................................................................................................................2-6
2.2 INSPECTION INTERVALS AND LOCATIONS ...................................................................................2-6
2.2.1 Levels of Inspection ............................................................................................................................2-6
2.2.2 Inspections by Maintenance Personnel .............................................................................................2-6
2.2.3 Biennial Inspections............................................................................................................................2-7
2.2.3.1 CABLES IN SUSPENDED SPANS................................................................................................2-7
2.2.3.2 CABLES INSIDE ANCHORAGES ................................................................................................2-8
2.2.4 Internal Inspections ............................................................................................................................2-9
2.2.5 Locations of Internal Inspections ....................................................................................................2-10
2.2.5.1 FIRST INTERNAL INSPECTION................................................................................................2-10
2.2.5.2 SECOND INTERNAL INSPECTION...........................................................................................2-11
2.2.5.3 ADDITIONAL INTERNAL INSPECTIONS................................................................................2-12
2.2.5.4 INSPECTIONS IF STAGE 4 OR BROKEN WIRES ARE FOUND............................................2-12
2.2.5.5 ACOUSTIC MONITORING.........................................................................................................2-12
2.3 INTERNAL INSPECTIONS....................................................................................................................2-12
2.3.1 Planning and Mobilization...............................................................................................................2-12
2.3.1.1 GENERAL .....................................................................................................................................2-12
2.3.1.2 INSPECTION PLANNING...........................................................................................................2-13
2.3.1.2.1 Review of Available Documents .............................................................................................2-13
2.3.1.2.2 Preliminary Field Observations and Cable Walk.....................................................................2-14
2.3.1.2.3 Interviews of Maintenance Personnel ......................................................................................2-14
2.3.1.2.4 Inspection Forms......................................................................................................................2-15
2.3.1.2.5 Tool Kit....................................................................................................................................2-15
2.3.1.2.6 Inspection QA Plan..................................................................................................................2-17
2.3.1.2.7 Inspection Locations ................................................................................................................2-17
2.3.1.3 CONSTRUCTION PLANNING....................................................................................................2-17
2.3.1.3.1 Design of Work Platform.........................................................................................................2-17
2.3.1.3.2 Construction Equipment ..........................................................................................................2-17
2.3.1.3.2.a Cable Compactors............................................................................................................2-17
2.3.1.3.2.b Steel Straps.......................................................................................................................2-17
2.3.1.3.2.c Wire Wrapping Machines.................................................................................................2-18
2.3.1.3.2.d Wedging Implements ........................................................................................................2-18
2.3.1.3.3 Preparations for Suspender Removal.......................................................................................2-18
2.3.1.3.4 Replacing Wire Wrapping .......................................................................................................2-19
2.3.1.4 NON-DESTRUCTIVE EVALUATION (NDE) TECHNIQUES..................................................2-19
2.3.1.4.1 Monitoring Devices .................................................................................................................2-19
2.4 INSPECTION AND SAMPLING............................................................................................................2-20
2.4.1 Cable Unwrapping............................................................................................................................2-20
2.4.1.1 WRAPPING WIRE TENSION TESTS .........................................................................................2-20
2-2
2.4.1.2 REMOVAL OF WRAPPING WIRE .............................................................................................2-21
2.4.1.3 LEAD PASTE REMOVAL ...........................................................................................................2-21
2.4.1.4 CABLE DIAMETER .....................................................................................................................2-21
2.4.2 Cable Wedging..................................................................................................................................2-21
2.4.2.1 RADIAL WEDGE LOCATIONS..................................................................................................2-21
2.4.2.2 WEDGE INITIATION AND ADVANCEMENT .........................................................................2-23
2.4.3 Wire Inspection and Sampling ........................................................................................................2-24
2.4.3.1 OBSERVATION AND RECORDING OF CORROSION STAGES............................................2-24
2.4.3.2 BROKEN WIRES ..........................................................................................................................2-25
2.4.3.2.1 Wedge Spacing ........................................................................................................................2-25
2.4.3.2.2 Wire Tracing............................................................................................................................2-25
2.4.3.2.3 Failed Wire Ends......................................................................................................................2-26
2.4.3.2.4 Sample Size..............................................................................................................................2-26
2.4.3.2.5 Other Forms of Corrosion........................................................................................................2-26
2.4.3.3 PHOTOGRAPHIC RECORD........................................................................................................2-27
2.4.3.4 MEASUREMENT OF GAPS AT WIRE BREAKS ......................................................................2-27
2.4.3.5 WIRE SAMPLING ........................................................................................................................2-27
2.4.3.5.1 Number of Samples..................................................................................................................2-28
2.4.3.5.2 Sample Location ......................................................................................................................2-29
2.4.3.5.2.a Stage 1 Wires....................................................................................................................2-29
2.4.3.5.2.b Stage 2 Wires....................................................................................................................2-29
2.4.3.5.2.c Stages 3 and Stage 4 Wires ..............................................................................................2-30
2.4.3.5.3 Number of Specimens in Each Sample and Length of Samples..............................................2-30
2.4.4 Identification of Microenvironments ..............................................................................................2-31
2.4.4.1 PH OF INTERSTITIAL WATER..................................................................................................2-31
2.4.4.2 CORROSION PRODUCTS...........................................................................................................2-31
2.4.4.3 PERMANENT PROBES................................................................................................................2-31
2.4.5 Cable Bands and Suspender Removals...........................................................................................2-31
2.4.5.1 CABLE BAND BOLT TENSION.................................................................................................2-31
2.4.5.2 SUSPENDER REMOVAL AND CABLE INSPECTION ............................................................2-32
2.4.5.3 SUSPENDER REINSTALLATION..............................................................................................2-32
2.4.6 Inspection Plan Reevaluation ..........................................................................................................2-33
2.4.7 Reinstallation of the Cable Protection System...............................................................................2-33
2.4.8 Inspection During Cable Rehabilitation.........................................................................................2-33
2.4.8.1 GENERAL .....................................................................................................................................2-33
2.4.8.2 INSPECTION NEEDS VS. OILING OPERATIONS ...................................................................2-34
2.4.9 Inspection and Testing in Anchorage Areas...................................................................................2-34
2.4.9.1 WIRES IN STRANDS ...................................................................................................................2-35
2.4.9.2 WIRES NEAR AND AROUND STRAND SHOES .....................................................................2-35
2.4.9.3 EYEBARS......................................................................................................................................2-35
2.4.9.4 WIRES INSIDE SPLAY CASTINGS ...........................................................................................2-36
2.4.9.5 ANCHORAGE ROOFS.................................................................................................................2-36
2.4.9.6 INSTRUMENTATION OF EYEBARS.........................................................................................2-36
2.4.9.7 DEHUMIDIFICATION.................................................................................................................2-37
2.4.10 Inspection of Cables at Saddles .......................................................................................................2-37
2.4.10.1 TOWER SADDLES.......................................................................................................................2-38
2.4.10.1.1 Tower Top Enclosures .............................................................................................................2-38
2.4.10.1.2 Exposed Saddles with Plate Covers .........................................................................................2-38
2.4.10.2 CABLE-BENT SADDLES ............................................................................................................2-38
2.4.10.2.1 Saddles Inside Anchorages ......................................................................................................2-38
2-3
2.4.10.2.2 Extended Anchorage Housing .................................................................................................2-38
2.4.10.2.3 Exposed Saddles and Plated Roofs ..........................................................................................2-39
2.5 FIGURES FOR SECTION 2....................................................................................................................2-40
2-4
FIGURES

Figure 2.2.3.1-1. Typical cable biennial inspection form. ...........................................................................2-40
Figure 2.2.3.1-2. Typical summary form showing biennial inspection rating system. ................................2-41
Figure 2.2.3.1-3. Typical form for biennial inspection showing detailed ratings........................................2-42
Figure 2.2.3.2-1. Typical form for biennial inspection of cable inside anchorage......................................2-43
Figure 2.2.4-1. Graph of cable condition vs. age at last inspection. ...........................................................2-44
Figure 2.3.1.2.2-1. Damaged caulking and paint at cable band. ...................................................................2-45
Figure 2.3.1.2.2-2. Uneven wrapping. .........................................................................................................2-45
Figure 2.3.1.2.2-3. Ridge indicating crossing wires. ...................................................................................2-46
Figure 2.3.1.2.2-4. Hollow area indicating crossing wires. ........................................................................2-46
Figure 2.3.1.2.2-5. Damage to wrapping caused by vehicular impact. .......................................................2-47
Figure 2.3.1.2.4-1. Form for recording locations of internal cable inspections. ........................................2-48
Figure 2.3.1.2.4-2. Form for recording observed wire damage inside wedged opening.............................2-49
Figure 2.3.1.2.4-3. Form for recording locations of broken wires and samples for testing........................2-50
Figure 2.3.1.3.2.a-1. Cable compactor. .......................................................................................................2-51
Figure 2.3.1.3.2.c-1. Power-driven wrapping machine. ..............................................................................2-52
Figure 2.3.1.3.2.c-2. Manual wrapping machine.........................................................................................2-52
Figure 2.3.1.3.2d-1. Chisels and wedges. ....................................................................................................2-53
Figure 2.3.1.3.2d-2. Hydraulic wedges. .......................................................................................................2-54
Figure 2.4.1-1. Form for recording cable circumference. ...........................................................................2-55
Figure 2.4.1.2-1. Removal of wire wrapping. ..............................................................................................2-56
Figure 2.4.2.1-1. Additional wedges to inspect area with many broken wires. ...........................................2-57
Figure 2.4.2.2-1. Cable wedged for inspection. ...........................................................................................2-58
Figure 2.4.8.1-1. Inspection during cable rehabilitation.............................................................................2-59
Figure 2.4.9.2-1. Deterioration of wires found inside strand shoe..............................................................2-59
2-5
TABLES

Table 2.2.4-1 Interval between internal inspections........................................................................................2-9
Table C2.4.3.5.1-1 Wire characteristics, Bridge X.......................................................................................2-28
Table C2.4.3.5.1-2 Wire characteristics, Bridge Z........................................................................................2-28
Table 2.4.3.5.1-1 Recommended number of wire samples for both cables ..................................................2-29
Table 2.4.3.5.3-1 Sample lengths and number of specimens from each sample..........................................2-30

GUIDELINES COMMENTARY
2-6
2.1 INTRODUCTION
The evaluation of a suspension bridge cable requires
considerable information about the condition and
strength of the cable wires. Obtaining this information
is the goal of an inspection.
Several levels of inspection are performed on a single
structure over time, but only internal inspections
provide data for strength evaluation.
In this section, instructions are given for conducting a
thorough internal inspection, and specific
recommendations are made regarding inspection
frequency and choice of locations. Sample forms are
included, as well as photographs of cables being
inspected and cable defects.
Inspections must be planned in advance. Preparing
forms and hiring a contractor is part of getting ready.
Sometimes the owners staff takes on the contractors
tasks, such as constructing platforms and unwrapping
and rewrapping the cable. The responsible parties,
whoever they are, should be prepared for the removal
of one or more suspenders and cable bands, if
conditions require it. Some cable band bolt tensions
should be checked by the team in either case.
Wedges are driven into the cable to separate the wires
for visual inspection and to take samples from inside
the grooves. Measurements of wire retraction are
used to estimate the capacity of the broken wires to
redevelop their force.
Portions of the cable inside the anchorages, and visible
portions where the cable passes over the saddles, must
be inspected.
C2.1
Information obtained from surveys of U.S. bridge
owners during development of these Guidelines led to
the conclusion that a baseline inspection should be
performed on a bridge when it has been in service for
30 years.
2.2 INSPECTION INTERVALS AND
LOCATIONS

2.2.1 Levels of Inspection
Three levels of inspection are recommended: periodic
routine visual inspections by maintenance personnel of
the cable exterior, biennial hands-on inspections, and
more thorough internal inspections.
C2.2.1
The inspector should walk the entire cable during a
hands-on inspection. For an inspection to qualify as
hands-on, the inspector must be sufficiently close to the
cable to touch it, sound it or inspect it with a
magnifying glass.
2.2.2 Inspections by Maintenance Personnel
During normal maintenance operations such as ice
removal, rinsing of splash zone residue, or touch-up
painting, maintenance personnel should be observant of
C2.2.2
Maintenance personnel, who are familiar with the
cable, may observe changes in condition indicative of
trouble more readily than an engineer or investigator
2-7
changes in the appearance of the cable. This is
especially true of changes that may indicate a potential
problem, including: damage to the paint or wrapping
caused by accidents, weathering of the paint system,
corrosion or severe oxidation of the wrapping wires,
loose caulking, and brown rust stains. Periodic
inspection tours of the cable by maintenance personnel
are recommended. Inspectors should begin by
inspecting the underside of the cable with binoculars,
and then walk the cable for its full length.
The best times of the year for inspection tours are at the
end of winter (March or April) to observe damage due
to frost or deicing salts in the splash zone, and at the
end of summer (September or October) to observe the
effects of extreme heat on paint and caulking.
Additional tours should be scheduled after severe
snow, ice, rain or wind storms. During these
inspections, the underside of the cable should be
examined for evidence of water inside, such as dripping
from the wrapping wire or weep holes in the lower
cable band grooves. Unusually damp areas should also
be noted.
Observations of unusual conditions should be recorded
and documented with color photographs. Both the date
and location of the inspection are noted, along with the
date of the storm that preceded the inspection, if
applicable. This information may be extremely useful
in determining sites for in-depth inspections.
who visits the cable at intervals of two years or more.
On-site observations can be instrumental in
determining where internal inspections should take
place.
While inspection tours are classified as hands-on, in
that they are made by walking along the length of the
cable, they are not intended to replace biennial
inspections. Only if problems are noted need a report
be filed, and a rating system is not required.
A database is highly recommended for summarizing
actual maintenance operations (repair of damage,
repainting, etc.) and inspection tour observations. The
database should include:
report number, part of a consistent reference
system
date of observation or maintenance operation
location of maintenance operation, damage or
repair
description of maintenance operation, damage
or repair (verbal description and numerical
code for rapid searching of the database)
recommended action
reference to report of the action taken
2.2.3 Biennial Inspections
Federally-regulated biennial inspections require that
non-redundant members receive hands-on inspection.
During these inspections, the condition of the items
listed in Articles 2.2.3.1 and 2.2.3.2 should be reported
on and rated.
If a biennial inspection indicates the possible presence
of internal corrosion, and the cable was never inspected
internally in that particular area before, it should be in
the near future.
C2.2.3
Biennial inspections should not be thought of as an
opportunity for internal inspection, because the cable
may be compromised by unwrapping sections of it
every two years.
2.2.3.1 CABLES IN SUSPENDED SPANS
The conditions of the following bridge components
should be reported on and rated as follows:
paint or surface protection, inspected for dried
out, peeling, cracked and crazed paint, or
puncture or tearing of the elastomeric barrier
(rate 3 if localized and 1 if more than 12 inches
long)
C2.2.3.1
Forms used by inspectors in the field, as well as
summary forms, should be prepared for the specific
bridge being inspected.
The ratings used in the Guidelines text and
corresponding figures are specific to New York State.
They progress from 1 (totally deteriorated, or in failed
condition) to 7 (new condition, no deterioration).
Unless the wrapping wire or other components are
2-8
caulking at cable bands, for gaps or cracks (rate
with the cable bands as indicated in Figure 2)
handropes and stanchions, for broken wires,
tightness and corrosion (rate 3 if broken wires
are present or loose, rate 1 if handropes or
stanchions are broken)
wire wrapping, inspected for anomalies,
including:
o unequal tension of wire plies, indicated
by unevenness in wrapping surface
(rate 4)
o bunching below or separating above
the cable bands (rate 4)
o gaps in wrapping, corroded or broken
wrapping wire (rate 3 for small gaps,
rate 1 for broken wrapping)
o surface ridges, indicative of crossing
wires and hollow areas (rate 4)
cable saddles or anchorage penetrations, for
damaged sleeves, bellows or flashing (rate 1 if
cracks that can admit water are present)
bottom of cable or cable bands, for rust stains
or dripping water, indicative of internal
corrosion (rate 1 or 2 and recommend internal
inspection)
Figure 1 shows an example of an inspection form.
Figures 2 and 3 show forms used to report the
conditions found. Figure 2 summarizes the rating
system for various types of defects in wrapping and
cable bands, while Figure 3 is a more detailed listing
for each cable panel or cable band. A bridge plan,
tower elevation, and cable elevation (refer to Figure
2.3.1.2.4-1) should also be included in the inspection
report.
actually new, the highest rating used is 5 (minor
deterioration, but functioning as originally designed).
Other agencies use a scale of 1 to 9, and ratings should
be adjusted proportionally.
Ratings applied to paint are often specified by the
bridge owner. If no such guide exists, the rating system
recommended by the state in which the bridge is
located should be used.
2.2.3.2 CABLES INSIDE ANCHORAGES
The following anchorage features should be reported
on and rated according to state specifications:
strands inside the anchorages, for corrosion or
broken wires, and swelling or bulges at the
strand shoes
anchorage walls and roof, for signs of water
entry
eyebars and strand wires, for signs of
C2.2.3.2
Ratings for these items should follow the system
specified by the state in which the bridge is located.
caulking at cable bands, for gaps or cracks (rate with
the cable bands as indicated in Figure 2.2.3.1-2)
Figure 2.2.3.1-1 shows an example of an inspection
form. Figures 2.2.3.1-2 and 2.2.3.1-3 show forms used
to report the conditions found. Figure 2.2.3.1-2
summarizes the rating system for various types of
defects in wrapping and cable bands, while Figure
2.2.3.1-3 is a more detailed listing for each cable panel
or cable band. A bridge plan, tower elevation, and cable
elevation (refer to Figure 2.3.1.2.4-1) should also be
included in the inspection report.
2-9
condensation
points of contact between eyebars and the
concrete mass, for signs of corrosion
eyebars and anchorage strands, for paint
anomalies
Figure 2.2.3.2-1 shows a typical form for recording the
condition of strands and eyebars inside the anchorages.
2.2.4 Internal Inspections
Internal inspections are necessary at some point during
the life of a cable. The inspection intervals given in
Table 1 are suggested, regardless of the cables external
condition. Access to internal wires requires removal of
the external protective system. The details of
conducting an internal inspection are described in
Article 2.3.
Table 2.2.4-1 Interval between internal inspections
Inspection
Number
Maximum
Corrosion
Stage Found
in Previous
Inspection*
Age of
Bridge at
Last
Inspection
(Years)
Interval
(Years)
First 30
Additional 1-(2) any age 30
2-(3) 40 or more 20
2-(3) 30 10
3-(4) 60 or more 20
3- (4) less than
60
10
4 any age 10
broken wires any age 5

* Each corrosion stage may include up to 25% of
the surface layer of wires in the next higher stage,
indicated by the number in parentheses. Stage 4
may include 5 broken surface layer wires.
At the discretion of the owner and the investigator, the
suggested intervals could be adjusted based on the
history of past internal inspections of the bridge, or
special conditions encountered, e.g., the presence of
dissimilar metals such as copper or bronze in contact
with or in close proximity to the wires, local
C2.2.4
The recommended intervals of inspection reflect the
data taken from condition inspections of 31 bridges of
various ages, described in NCHRP Report 10-57,

the reported maximum corrosion stages may have been
reached well before the cables were inspected. The data
indicate that there is a grace period of about 10 years
after a bridge is completed before deterioration begins.
The bridges have been separated into two groups
according to mean trends of their rates of deterioration
(see NCHRP Report 10-57. Eleven bridges with slowly
deteriorating cables were designated Group A, for
which the interval of time required to advance from one
corrosion stage to the next was about 20 years. The 20
bridges that fall into Group B took only half that time
to advance from stage to stage.
The rate of advancement from one stage to the next is
nearly linear; it increases slightly (i.e., the interval from
one stage to the next is slightly smaller) as the
corrosion becomes worse. Figure 1 shows the linear
rate of deterioration.
The recommended intervals between inspections are
based on these rates of advancement, shown in Figure
1. In all cases, an internal inspection is recommended
when the bridge is 30 years old, based on the
observation that 7 bridges had Stage 3 or worse
corrosion before the age of 40 years. The first
inspection can be used to establish whether the cable is
deteriorating rapidly or slowly.
Once Stage 4 corrosion is present, the interval between
inspections should be shortened to 10 years for all
bridges. Whenever broken wires are found, the interval
to the next inspection should be 5 years. A large
percentage of Stage 4 wires also merits another
inspection in 5 years.
Many of the bridge cables were opened only for a short
distance, and the information is sketchy. In addition,
The recommended intervals of inspection reflect the
data taken from condition inspections of 31 bridges of
various ages.
The bridges have been separated into two groups according to
mean trends of their rates of deterioration (see section 2.3.4 of
the final report for NCHRP Project 10-57, on the accompanying
CD. Eleven bridges with slowly deteriorating cables were
designated Group A, for which the interval of time required to
advance from one corrosion stage to the next was about 20
years. The 20 bridges that fall into Group B took only half that
time to advance from stage to stage.
corrosion becomes worse. Figure 2.2.4-1 shows the
linear rate of deterioration.
Table 2.2.4-1 are suggested, regardless of the cables
external condition. Access to internal wires requires
removal ofthe external protective system. The details
of conducting an internal inspection are described in
Article 2.3.
2-10
deterioration from traffic collisions, or overheating the
wires during a maintenance operation.
The interval between inspections should be shortened
to 5 years when Stage 4 corrosion is found in more than
10% of the wires in the cable.
A description and photographs of the four corrosion
stages can be found in Article 1.4.2.2 and Figure
1.4.2.2-1.
2.2.5 Locations of Internal Inspections
Internal inspections should be made where there are
external signs of internal deterioration. These signs
include:
loose wrapping
dripping water from the cable interior
rust stains
damaged caulking at the cable bands
surface ridges that indicate crossing wires
below the wrapping
hollow sound when the cable surface is tapped
If no external indications of deterioration are found,
then the inspection locations should be selected
according to the method described below.
C2.2.5
No definitive statement can be made about where the
worst conditions in the cable are most likely to be
found. In only one of five bridges, for which data is
available from at least 16 locations along the cable, did
the greatest loss of strength occur at a low point of the
cable. In the other four bridges, the maximum strength
loss occurred near the quarter point of the main span or
near the center of the side span. Furthermore, on one of
the bridges, maximum strength loss above the low
point of the cable was 3.5 times greater than at the low
point.

2.2.5.1 FIRST INTERNAL INSPECTION
The first internal cable inspections should be made at a
minimum of 3 locations along each cable, selected as
follows.
one in each cable at a low point of the main
span
one in each cable at or near a low point of the
side span
one in the first cable in the main span, above
the low point at a distance of from 30% to 70%
of half the main span
one in the other cable in a side span, above the
low point at a distance of from 30% to 70% of
the side span
The cables should be opened for a length of at least 16
feet at each location, and wedged as deeply as possible
at 4 locations around the perimeter. If the corrosion of
the wires exceeds Stage 2, wedging should take place
at 8 locations around the perimeter, and the opening
should be extended to a full panel length. This is to
enable the driving of wedges far enough inside the
cable to determine the depth of Stage 3 or worse
C2.2.5.1
During inspections, Stage 3 corrosion has been found at
one or more of the low points whenever there was
significant strength loss at any location higher up.
There are 4 low points on each cable: in the 2 panels
adjacent to the lowest cable band in the main span, and
in the lowest panel in each side span, usually that panel
on the span side of the anchor or cable bent saddle
where the least slope occurs. In cables that have 2 panel
points at the same midspan elevation, there will be 5
low points, one of which is the entire center panel. It is
recommended that 2 low points be inspected on each
cable in the first inspection, and the other 2 (or 3) in the
second inspection. The inspection of 1 location above
the low point of each cable is recommended in the first
inspection, and not less than 2 such locations in the
second inspection.
When there are more than 2 cables on the bridge, the
same inspection pattern described in Article 2.2.5.1
should be applied to each pair of cables.
It is not likely that wire corrosion on a 30-year-old
bridge will exceed Stage 2, although some Stage 3
corrosion may be present in the exterior wires. Thus,
the cable opening need only be long enough to remove
2-11
corrosion and to remove 16-foot-long samples. a 10-foot-long sample from the outer two layers for
testing. The inspection team should, however, be
prepared to open up a greater length of cable if more
serious corrosion be found.
2.2.5.2 SECOND INTERNAL INSPECTION
When the first internal inspection reveals only Stage 1
or Stage 2 corrosion, the second internal inspection
should be made at not less than 4 locations along
each cable, following the logic of the previous
choices.
The low point location in the main span should be
adjacent to the low point location previously inspected,
but the side span location should be in the side span
opposite the one previously inspected. One location in
the main span and one in a side span above the low
points should also be inspected. A 16-foot-long
opening will suffice, but if either Stage 3 or Stage 4
corrosion is found, opening and wedging should be
increased (follow the instructions in Article 2.2.5.1).
When the first internal inspection reveals Stage 3
corrosion, or Stage 4 corrosion to a depth of 3 wires
or less, each cable should be internally inspected at 6
locations, including any one of the 3 previously
inspected panels that exhibited Stage 2 corrosion or
greater, and 3 additional locations recommended for
the first inspection.
Locations that exhibited only Stage 1 corrosion in the
first inspection need not be reopened at this time, but
additional locations above the low points should be
selected to bring the total locations to 6. All 6 locations
should be inspected for the full length between cable
bands, with wedges driven to the center of the cable, or
as deeply as possible. Whenever Stage 4 corrosion is
present to a depth greater than one wire, and the center
of the cable cannot be reached with a full panel length
unwrapped, one cable band per cable should be
removed to assess the condition of wires at the center
of the cable.
When the first internal inspection reveals Stage 4
corrosion to a depth of more than 3 wires, at least
16% and preferably 20% of the panels in each cable
should be inspected.
Four low points and 2 locations near the towers should
be inspected; the balance of locations should be
selected at random in the remainder of the cable
between the low points and the towers, one each from
contiguous groups of panels that are approximately
C2.2.5.2
If corrosion does not exceed Stage 1 during the first
inspection, a bridge cable could be 60 years old when
the second inspection takes place. The inspection team
should be prepared to open additional locations higher
up on the cable at that time, if Stage 4 corrosion is
found at any of the 4 recommended locations.
In three of the inspection reports mentioned in Article
C2.2.5, little deterioration was found adjacent to the
towers. Therefore, only 2 locations near the towers are
recommended for inspection when Stage 4 corrosion is
found.
Whenever there is any sign of deterioration inside the
cable adjacent to the saddles, or in the saddle housings,
or if the housing or the sleeves at the saddles show
signs of water entry, these locations should be added to
the list of recommended inspection sites.
A method of estimating the minimum strength of a
cable from the findings of an inspection presented in
NCHRP Report 10-57 depends on the adequacy of the
sampling. The estimated error in the minimum strength
is 11% when 16% of the panels in the cable are
inspected, and 8% when 20% of the panels are
inspected. The method and the estimated error are
based on the results from the only bridge for which
sufficient data are available.
A method of estimating the minimum strength of a cable
from the findings of an inspection presented in the final
report for NCHRP Project 10-57, on the accompanying
CD depends on the adequacy of the sampling. The
estimated error in the minimum strength is 11% when
16% of the panels in the cable are inspected, and 8%
when 20% of the panels are inspected. The method and
the estimated error are based on the results from the
only bridge for which sufficient data are available.
2-12
equal in number. The full length of panels between
cable bands should be inspected, with wedges driven to
the center of the cable, or as deeply as possible. At least
2 cable bands should be removed to facilitate
inspection to the center of the cable and under the
bands.
2.2.5.3 ADDITIONAL INTERNAL INSPECTIONS
The number of locations to be opened after the second
inspection depends on the conditions revealed by
previous inspections, and the sites should again be
chosen following the instructions in Article 2.2.5.2.

2.2.5.4 INSPECTIONS IF STAGE 4 OR BROKEN
WIRES ARE FOUND
When more than 10% of the wires in a cable panel are
found to be Stage 4 in any inspection, the cable should
be scheduled for a full interior inspection, and remedial
action, such as the introduction of corrosion inhibitors,
should be taken. Installation of an acoustic monitoring
system is strongly recommended to listen for and locate
continuing wire breaks.
C2.2.5.4
Stage 4 corrosion is usually accompanied by cracked
and broken wires. Whenever more than 10% of the
wires in a cable panel are found to be in this condition,
a full-length inspection of the cable is warranted, along
with some cable band removal to inspect the wires
underneath.
2.2.5.5 ACOUSTIC MONITORING
When Stage 3 wires or worse were found in a previous
inspection, it is recommended that an acoustic
monitoring system be installed and monitored for a
period of 12 to 18 months prior to the next internal
inspection (see Article 2.3.1.4.1). The inspection
locations should be selected to coincide with wire
breaks, if any occur.
At the discretion of the investigator or owner, the same
system could be installed even if the wires found in a
previous inspection were only Stage 1 or Stage 2.
C2.2.5.5
Broken wires in the cable are an indication of active
corrosion, and the sites of breaks are prime locations
for future inspection.
2.3 INTERNAL INSPECTIONS

2.3.1 Planning and Mobilization
Internal inspections require planning, but they also
require the flexibility to respond as the inspection
progresses and to alter initial plans if necessary.
Contractor-assisted inspection must be managed so as
not to compromise accessibility, even to areas not
originally specified.

2.3.1.1 GENERAL
The review of drawings, specifications and documents
from prior inspections is required to provide
background for planning a successful cable

2-13
investigation, defined in part as one that requires
minimum alterations during execution.
2.3.1.2 INSPECTION PLANNING
Before determining inspection locations, the
investigator should perform the series of actions
described in this section.

2.3.1.2.1 Review of Available Documents
The investigator should review bridge design reports,
and become familiar with the design details at cable
bands and cable saddles. If available, the specifications
of wire and eyebar materials should also be checked.
Records of previous inspections, however localized,
may shed light on the causes of previous damage and
suggest locations of potential damage, which could be
important in the eventual assessment of cable capacity.
Review of maintenance records may be of use in
pinpointing areas where caulking or wrapping have
failed in the past, or where water has been observed to
be leaking from the cable or through the anchorage
roof.
The answers to the following questions may be useful
in determining inspection locations, requirements for
laboratory testing, and data needed for reliability
analysis:
Are the wires galvanized or bright? If they are
galvanized, the zinc coating should be tested.
Were the wires originally greased or oiled? If
so, the wires may be less deteriorated, but
corrosion may appear as short black sections
on the wire.
What is the specified wire cast diameter? Small
cast diameters indicate that wire cracking is
more likely.
Are the original wire strengths and other
mechanical properties available from
acceptance testing or from specification
records? If not, the full number of samples of
Stage 1 and Stage 2 wires recommended in
Article 2.4.3.5.1 should be taken during the
first inspection. When original wire strengths
are available, then at least 10 wires from these
stages, for 3 tensile tests per sample, should be
taken during the first inspection, with the
balance of samples required taken during the
C2.3.1.2.1
A highly stressed cable (a safety factor below 2.5) has
less margin for strength loss due to corrosion than a
cable designed for lower stresses. A live to dead load
ratio greater than 0.2 may indicate large deflections,
especially if stiffening trusses or girders are slender.
Such deflections may cause paint to crack or damage at
the sleeves and caulking, and special attention to
inspection of these details is indicated.
2-14
second inspection.
Are test results of wire chemistry available? If
not, chemical tests should be made.
Was the cable aerially spun, or does it have
shop-fabricated parallel wire strands, which
tend to have fewer crossing wires because the
cable is highly compacted?
Is water penetration a possibility at the sleeves
and flashings of the saddle housing, or at deck
penetrations? If so, careful observations of
these areas must be made.
What is the dead load stress? What is the
design live load stress? These values are
needed in preparing the final report. If they are
not available, or if there have been
modifications to the bridge, or if there have
been changes in the traffic load or pattern, the
cable forces will have to be recalculated. High
stress in the cables indicates that wire cracking
is possible.
Which paint systems were used according to
painting records? The information may lead to
a better analysis of the coating performance.
2.3.1.2.2 Preliminary Field Observations and Cable
Walk
Walking on bridge sidewalks or maintenance walkways
allows for observation of the lower portions of the
cable. A cable walk is essential to make close
observations of the external appearance of the cable.
The items listed in Article 2.2.3.1 should be observed,
along with the following:
Paint cracks along wrapping wire valleys
Poor compaction, evidenced by noticeable
angularity of the cable
Broken or torn elastomeric membranes or
cracked Lucite cable shells on newer protection
systems
Figures 1 through 5 illustrate some of these conditions.
C2.3.1.2.2
These observations serve as general background and
may help to establish inspection sites.
Rusted handropes may be a safety hazard and should be
repaired prior to the inspection.
2.3.1.2.3 Interviews of Maintenance Personnel
Interviews of on-site personnel are useful, especially
when they have performed maintenance tasks such as
cleaning and painting the bridge, removing ice, and
rinsing the splash zone.

2-15
The information known to the maintenance staff may
not have been formally recorded, but it could be
valuable in determining the origin of damage and its
time of inception.
2.3.1.2.4 Inspection Forms
A well-prepared inspection form facilitates the
recording of data in the field. Sample forms are given
in Figure 1, Figure 2 and Figure 3. Figure 1 shows the
elevations of each cable, for indicating the locations of
inspected panels.
Figure 2 shows an inspection form for a 9,990-wire
cable. One or more copies are needed for each wedge
line in each inspected panel. The observed conditions
of the wires inside the wedged opening are recorded on
these forms. A typical filled-in form is shown in
Appendix C.
Figure 3 shows a cross-section of the cable that can be
used to record the locations of broken wires near the
surface of the cable, and also to map the internal
conditions observed in the cable.
C2.3.1.2.4
Inspection forms must be tailored to the details of the
cable to be inspected. The number of rings in the cable
shown in Figure 2 and Figure 3 are calculated using
Equation 4.3.1.1-1. If more or less than 9.990 wires are
in the cable, the number of rings shown in the figures
will vary accordingly.

2.3.1.2.5 Tool Kit
Investigators or inspectors should have a simple but
practical set of tools that permit the observation and
recording of all essential data. At minimum, it must
include the following (other items to be added at the
discretion of the investigator):
Clipboard and supply of pens and pencils,
fastened around neck or shoulder, or kept in a
knapsack, but never carried loose in ones
hand.
Adequate number of blank forms. An
inspection of eight wedge lines per panel will
require at least 24 sheets for each panel.
Sturdy pocketknife to test wrapping or scrape
corrosion products from wrapping and exposed
wires. Among other uses, it aids in the
collection of grease samples.
Flashlight, the most powerful one that fits into
a knapsack. Absolutely essential for observing
wires, especially in the lower half of the cable,
it also helps to focus a camera in a dark area
before exposing the film with a flash.
Steel ruler(s) and tapes, for various purposes.
Steel tape, with sufficient rigidity to reach 18
C2.3.1.2.5
Cable environments are always breezy or windy places.
Inspection forms need to be anchored to a clipboard at
both ends (possibly with a rubber band at the bottom)
for taking good notes and preventing the forms from
flying away. Pens have a way of disappearing. The
inspectors should have a supply of pens and pencils
sufficient to ensure an uninterrupted inspection.
Although pens are preferable for filling in inspection
forms, they may be useless on a humid day. HB
graphite pencils can be used instead.
The break location in a wire is not always visible.
Loose wires are evidence of a break nearby, and are
revealed in response to prodding with a rigid
implement. Long implements are required to test deep
wires, in which case a rigid wooden stick can be used if
the wire pick is too short.
A 1/16-inch-thick aluminum or brass sheet can be used
to count the depth of a wire. The sheet should be drilled
down the center with holes the same diameter as the
wires and then split into two, so that notches are
formed by the remaining parts of the holes. Numbers
should be inscribed on the sheet next to the notches to
aid in counting. Since most wires are very nearly 5 mm
in diameter, a metric ruler can be carried to determine
the depth of the wire from the surface. Dividing the
A well-prepared inspection form facilitates the recording
of data in the field. Sample forms are given in Figure
2.3.1.2.4-1, Figure 2.3.1.2.4-2 and Figure 2.3.1.2.4-3.
Figure 2.3.1.2.4-1 shows the elevations of each cable, for
indicating the locations of inspected panels.
Figure 2.3.1.2.4-2 shows an inspection form for a
9,990-wire cable. One or more copies are needed for
each wedge line in each inspected panel. The observed
conditions of the wires inside the wedged opening are
recorded on these forms. A typical filled-in form is
shown in Appendix C.
Figure 2.3.1.2.4-3 shows a cross-section of the cable
that can be used to record the locations of broken wires
near the surface of the cable, and also to map the
internal conditions observed in the cable.
Inspection forms must be tailored to the details of the
cable to be inspected. The number of rings in the cable
shown in Figure 2.3.1.2.4-2 and Figure 2.3.1.2.4-3 are
calculated using Equation 4.3.1.1-1. If more or less than
9.990 wires are in the cable, the number of rings shown
in the figures will vary accordingly.
2-16
inches into the cable. Wires are identified by
counting from the surface of the cable or by
measuring their distance from the surface.
Wire pick, made from a screwdriver with a
round shank and well-rounded edges at the tip,
to prod for loose wires and pry at surface
deteriorated wires to inspect the layer below. It
should not be used on Stage 1 or Stage 2 wires,
to avoid damaging the zinc coating.
Small ruler, for proportioning the damage
recorded in photographs.
Flexible tape, to measure cable diameters for
evaluating cable compaction.
Dial gage caliper, to measure loss of section,
especially in anchorage areas. For loss of
section that is gradual, a micrometer is
sufficient.
A camera (traditional print or digital), required
for a photographic record of conditions
exposed by wedging.
Good lighting (directed flash, ring flash or
externally directed floodlight) for obtaining
adequate photographic records. Conventional
flash units flood the area outside the wedged
opening and cannot illuminate the wedged
cavity.
Tags for identifying sampled wires, typically of
Manila paper, and ballpoint pen or some kind
of permanent marker that wont be smudged by
grease or water. Tags with wires or twine and
reinforced holes for attachment to sampled
wires are preferable.
pH paper with a minimum range of 0.5 pH
units, to determine the acidity of liquid on
cable wires. They are usually available in small
spools, with a calibration chart, from industrial
supply houses.
Several sterile, tightly sealable sample jars, in
case water or corrosion products are found that
require sampling.
Industrial mirror with telescoping arm, 8 inches
square, to observe the underside of the cable as
the inspector walks on it.
Attachment means immediately at hand, for
measured depth by 5, gives the depth into the cable in
terms of wire diameters.
As a backup for identification tags, a strip of duct tape
can be wrapped around the wire and marked with a ball
point pen.
Pharmacies sell small sealable plastic jars that are used
for urine samples but can be used also for collecting
water and corrosion product samples from the cable.
ASA 400 color film is recommended for greater
flexibility. A 50 mm or 28-80 mm zoom lens is
recommended for recording objects at close range,
especially those deep in the cable, which can be seen
only by a camera positioned directly over the wedged
opening. A minimum of 3 megapixels is recommended
for digital cameras to provide sufficient resolution.
2-17
securing all tools, including clipboard, pens
and pencils, to the handrails or platform.
2.3.1.2.6 Inspection QA Plan
Inspection and sampling must be done in a verifiable
manner. For quality control, more than one inspector
should make observations and both parties should agree
on the identification and demarcation of corrosion
stages. Inspectors should be trained by an experienced
investigator.
The QA plan should identify the lead inspector and the
assistant inspector, and describe what steps they will
take to minimize error in data collection.
C2.3.1.2.6
It is crucial that sampled wires be representative of the
damage, so that testing data result in sound statistics for
wire properties.

2.3.1.2.7 Inspection Locations
Preparations for inspection include the selection of
cable panels to be inspected. As the inspection
proceeds, the investigator may alter this plan and
choose to open different panels.

2.3.1.3 CONSTRUCTION PLANNING
Access to the cable is provided by a contractor in most
cases, or maintenance personnel. Panel inspections
follow a predetermined order to some extent. When
construction contracts are used, they should be flexible
enough to provide for unexpected inspection
requirements.

2.3.1.3.1 Design of Work Platform
Work platforms are designed for the safety of
construction and inspection personnel, and for
containment of hazardous material. The platform
should be constructed in a manner that safeguards tools
and wedges, and prevents them from dropping onto the
roadway or the waterway.
C2.3.1.3.1
Wedges have been known to be ejected from the cable
and fall down onto the roadway. Tenting that encloses
the platform should eliminate this potential hazard.
Wedges can also be held in place with straps that wrap
around the cable to protect personnel on the platform.
2.3.1.3.2 Construction Equipment
2.3.1.3.2.a Cable Compactors
It is essential that the compactor have sufficient
capacity to recompact the unwrapped cable, to its
original diameter. This may require a three- or four-
jack assembly, depending on jack and cable size.
Typical compactor details are illustrated in Figure 1.
To prevent cable band slippage, the compactor should
be placed at least 1.5 cable diameters from the edge of
a cable band.
C2.3.1.3.2.a
Placing the compactor immediately adjacent the cable
band on its downward-sloping side could reduce the
cable diameter within the band itself, possibly causing
the band to slip.

2.3.1.3.2.b Steel Straps
It is essential that the compactor have sufficient
capacity to recompact the unwrapped cable, to its
original diameter. This may require a three- or four-jack
assembly, depending on jack and cable size. Typical
compactor details are illustrated in Figure 2.3.1.3.2.a-1.
2-18
In the last stages of compaction, the cable is held
together with steel straps to help it keep its shape; the
straps are removed as the wrapping proceeds. Reusable
synthetic straps are also used for this purpose.
2.3.1.3.2.c Wire Wrapping Machines
Wire wrapping machines may be manually-driven or
power-driven. With some quality supervision, manual
devices can provide as tight a wrapping as power
equipment, but they are recommended only for short
lengths of cable. Figure 1 and Figure 2 show two kinds
of wrapping machines.
Tension is controlled in the wrapping wires by opening
the tensioner jaws of manual machines or by wire spool
friction in power machines. Both tensioner jaws and
spools are calibrated and must be monitored during the
wrapping operation.
C2.3.1.3.2.c
It is usually more efficient to use manual wrapping
devices for small investigations, because of the
relatively high cost of power-driven machines and the
lengthy lead times required to obtain them.
2.3.1.3.2.d Wedging Implements
Several types of wedging implements are employed
during inspection, including wide bronze non-sparking
chisels, and wood, plastic and hydraulic wedges. Chisel
and wedge details are shown in Figure 2.3.1.3.2.d-1.
To prevent wire damage, flat non-sparking chisels, 3 to
4 inches wide and preferably bronze, should be used to
initiate wedging. Long flat screwdrivers with square
shanks are not recommended.
The best wedges for penetrating the cable are made of
oak, rock maple, or high-molecular-weight
polyethylene. The wedges should taper 1 inch for each
5 inches of length. To minimize wedge damage, wedge
tips should be rounded approximately to a 1/8-inch
diameter.
Hydraulic wedges can be used to provide wide
openings with minimum effort. A hydraulic device is
shown in Figure 2.3.1.3.2.d-2.
C2.3.1.3.2.d
A suitable bronze starting chisel can be fabricated
easily from a 3/8-inch x 4-inch flat bronze bar. The tip
should match the one shown in Figure 2.3.1.3.2.d-1.
Wedge tips wear out or become bent out of shape in the
course of wedging, and constant repair is necessary.
2.3.1.3.3 Preparations for Suspender Removal
Suspender removal requires an analysis of the forces
that are necessary to dislocate the suspender from its
anchored position and the forces that are transferred to
adjacent suspenders. Capacity checks of the anchoring
brackets and of the stiffeners and ropes at the adjacent
remaining suspenders are mandatory to ensure safety.
The equipment required for removal of short
suspenders includes framing, jacks, and tension rods to
bring the truss and cable closer together. For long
suspenders, special equipment must be designed to pull
C2.3.1.3.3
The investigator and the contractor usually collaborate
on the design of equipment. The engineer is responsible
for capacity checks.
Suspender brackets on the stiffening trusses are
typically designed to allow for suspender replacement.
Capacity checks are a formality in most cases.
However, increases in dead load unforeseen by the
original designer may require such a check along with
design of modifications.
Wire wrapping machines may be manually-driven or
power-driven. With some quality supervision, manual
devices can provide as tight a wrapping as power
equipment, but they are recommended only for short
lengths of cable. Figure 2.3.1.3.2.c-1 and Figure
2.3.1.3.2.c-2 show two kinds of wrapping machines.
2-19
the suspender against the truss for disconnecting the
two.
Temporary suspenders may be required to carry the
force of the removed suspenders, so as to avoid
overstressing adjacent suspenders or the stiffening truss
and girder.
2.3.1.3.4 Replacing Wire Wrapping
The wire tension should be high enough for the wire to
remain in tight contact with the cable under all
conceivable temperatures and live load conditions.
The wrapping wire loses tension by as much as two-
thirds because of creep in the zinc coating. The
following expression for wire tension during
rewrapping can be used:
P=
where:
= Poissons ratio
LL
= live load stress in the cable
= coefficient of thermal expansion
T = maximum estimated difference in temperature
between the wrapping wire and average cable
temperature
E = Modulus of Elasticity of the wrapping wire
d
w
= diameter of the wire
T will be larger for large cables than for smaller ones.

2.3.1.4 NON-DESTRUCTIVE EVALUATION
(NDE) TECHNIQUES
Many private companies and institutions are marketing
NDE equipment. Investigators should always be aware
of what is available and what can be expected from
NDE devices at the time they plan an investigation.
Technologies are developing at a fast pace. Often,
willingness to try a new technology will lead to its
modification, making it suitable for application to
bridge cables. Public authorities and investigators must
evaluate any proposed or potential technology by
testing it against objective criteria that derive from the
specific characteristics of bridge cables and
information required for strength assessment.
C2.3.1.4
Despite the great need for NDE diagnostic evaluation,
existing devices are of limited value. New and
improved technologies may be developed for
determining the most damaged cable areas, but hands-
on visual inspections are still required for the
foreseeable future. A discussion of NDE techniques is
included in NCHRP Report 10-59.
Acoustic Monitoring is a good example of an existing
technique that has been successfully adapted to cable
work.
2.3.1.4.1 Monitoring Devices
Acoustic monitoring devices should be installed on a
C2.3.1.4.1.a
There are devices that detect changes in the behavior of
3(
LL
+E T)d
w
2
/4
Despite the great need for NDE diagnostic evaluation,
existing devices are of limited value. New and improved
technologies may be developed for determining the most
damaged cable areas, but hands-on visual inspections are
still required for the foreseeable future. A discussion of
NDE techniques is included in the final report for
NCHRP Project 10-57, on the accompanying CD.
2-20
cable that has many Stage 4 or broken wires to
determine whether deterioration is continuing and at
what pace. If wires continue to break, or if the
frequency of breaks accelerates, the inspection
schedule should be revised. If an additional 0.5% of the
wires in a panel break after an inspection, then an
immediate re-inspection and evaluation of the panel is
recommended.
cables or detect wire failures as they occur. Given
known baseline behavior and damage information, a
record of conditions may be constructed over time.
Acoustic Transmission (AT) technology is used to
detect wire failures on large cables, because the sound
waves of a break that travel through the steel can be
detected at receptors placed on the cable surface. Wire
failures have distinctive sound signatures that are easily
differentiated from normal bridge noise. The arrival
times of the sound at several different receptors are
compared to pinpoint the location of the break.
Some U.S. bridges have been fitted with Acoustic
Transmission equipment along both cables, or at least
along parts of them. This service is provided under the
generic technological name of Acoustic Monitoring.
Monitoring devices cannot determine existing
conditions directly, and are not diagnostic. However,
the technology can be used to determine which panels
have wires that are breaking, and hence which panels
are most likely to have damage. This could eliminate
much of the current guesswork involved in selection of
panels to be opened.

2.4 INSPECTION AND SAMPLING
2.4.1 Cable Unwrapping
Prior to unwrapping the cable, the investigator should
record any notable surface defects, such as gaps in the
wrapping wire, damaged paint, and white or brown rust
emanating from the cable.
C2.4.1
The conditions listed apply to cables wrapped with
wire, whereas other protection systems may exhibit
different defects, which should also be noted (e.g.,
cracks in polymeric coverings or tears in neoprene
wrapping).
2.4.1.1 WRAPPING WIRE TENSION TESTS
The wire should be strain-gaged before cutting if the
tension in the existing wire is required. The loss in
stress after cutting, while the wire is held to the cable
circumference, provides the wire tension. The same
procedure is used for single wrapping plies or for all
the plies at once.

C2.4.1.1
If the original tension is known, then the tension in the
unwrapped wire is compared to find out how much
tension has been lost, and whether the existing tension
is adequate.
Measurement of the shortening that occurs when the
wrapping wire extends over several loops is difficult to
accomplish and interpret because of the friction among
the plies. Hence, strain gaging is recommended rather
than measuring the shortening of a wire after cutting.
The circumference of the cable should be measured
prior to unwrapping. If the steel tape is sensitive
totemperature changes, the temperature should be
recorded at the same time. A form that can be used to
record these measurements is shown in Figure 2.4.1-1.
2-21
2.4.1.2 REMOVAL OF WRAPPING WIRE
Removal of the wrapping wire is a health hazard as
well as an environmental one. Workers should wear
masks and handle the wire in such a way that the lead
waste is contained. Workers should also take care not
to generate airborne dust, unless the entire panel is
covered and enclosed, and then a filtering system
should be used to protect workers from the dust that
accumulates inside.

C2.4.1.2
Blood testing for lead toxicity of all personnel involved
in cable unwrapping, inspection, and rewrapping is
mandatory. See Article 1.3.2.
2.4.1.3 LEAD PASTE REMOVAL
The lead paste under the wrapping is usually dry and
friable. Some of it falls onto the platform floor as the
wrapping is removed. Some of it adheres weakly to the
cable wires and may be dispersed by the lightest
breeze. It should be dislodged by striking the cable
with a wooden implement and the remainder removed
with a soft wire brush.
Before wedging the cable, the unwrapped surface
should be vacuumed. Wedging will cause additional
lead dust and waste to be dislodged from the cable.
Spraying or brushing on a light coating of oil will
minimize the production of this additional dust.

C2.4.1.3
Monitoring of airborne dust must comply with
environmental protection guidelines.
2.4.1.4 CABLE DIAMETER
The cable circumference should again be measured
after the wrapping has been removed. Measurements
should be taken immediately adjacent to the cable
bands, 12 inches from the ends of the bands and at the
center of the panel. The cable diameter without
wrapping should be calculated, as well as the solids
ratio of the cable, which is a comparison of the total
metallic area of the steel wires to the area in the cable
circumference without wrapping wire.

C2.4.1.4
The cable diameter, which is calculated from the
circumference measurement, is required so that the
cable can be rewrapped to its original degree of
compaction or more. The solids ratio of a well-
compacted cable usually varies between 0.80 and 0.82.
If the value is significantly smaller, it may indicate that
the reported number of wires in the cable is incorrect.
In this case, the number must be verified by counting
the wires inside the anchorage.
2.4.2 Cable Wedging
The cable is wedged radially at panel locations, as
described in Article 2.4.2.1. In general, wedging should
be done where damage is suspected.

2.4.2.1 RADIAL WEDGE LOCATIONS
When 8 wedge lines are required, they should be
located at every 45 degrees around the cable
circumference. When less than 8 wedge lines are
required, the spacing is adjusted accordingly. The

C2.4.2.1
The upslope is sometimes used as a convenience for
determining direction, because compass directions are
not easily established inside an enclosure. Angular or
other designations may be used instead of clock
To avoid dispersal of lead and lead paint, wrapping
wire should be handled carefully when it is transported
to the temporary storage site. Figure 2.4.1.2-1 shows
wrapping wire in the process of removal.
2-22
wedge lines are usually given clock designations to
describe their angular positions from the top of the
cable:
12:00 at top of cable
1:30 at 45, clockwise from top of cable
Not all cable conditions require wedging on all 8 lines.
On the other hand, some cables may need additional
wedging to improve the statistical data, especially if the
damage is severe. The following guidelines are
recommended for determining the number and position
of the wedge lines:
Always start wedging at one of the bottom
lines (4:30, 6:00, or 7:30), especially if nothing
is known about the condition of the cable, or
unless the plan is to open a minimum of 8
wedge lines.
When only Stage 1 corrosion is found in all
three of the lowest wedge lines, select 1 more
wedge line in the splash zone, if there is one, at
either 1:30, 3:00, 9:00 or 10:30.
When no more than Stage 2 corrosion is found,
a minimum of 6 wedge lines are recommended,
including the bottom 3, plus 9:00 and 3:00 and
either 1:30 or 10:30, preferably in the splash
zone.
When Stage 3 corrosion or worse is found,
open all 8 wedge lines.
When a significant number of broken wires are
found, additional lines should be opened in the
regions where Stage 4 wires are found. Figure
2.4.2.1-1 shows a cable with broken wires
wedged for inspection.
For cables with a diameter greater than 24
inches, 8 additional wedge lines between the 8
recommended above should be opened. They
should extend deep enough to permit
designations.
The bottom of the cable is usually more deteriorated
than the top or sides. By starting at the bottom, it is
likely that the worst condition will be encountered.
When the outer surface of the cable is in Stage 1
condition, wedging in the rest of the cross-section can
be minimized, because the condition is usually worse at
the surface than at the center.
Near the cable low points, the side of the cable facing a
roadway will receive the most splash from passing cars
and will often be more deteriorated than the side facing
away from the roadway.
When many broken wires are found, Stage 4 corrosion
is expected to extend deep inside the cable, and
additional wedge lines are justified to determine its
extent.
On larger cables, additional wedge lines are
recommended in the outer half of the cable radius
where the wedges are at maximum spacing, to avoid
reducing the fraction of wires that are observed. This
helps to minimize the margin of error in the
calculations.
2-23
inspection of the wires at least halfway to the
center of the cable.

2.4.2.2 WEDGE INITIATION AND
ADVANCEMENT
Wedging should be initiated with one of the non-
sparking implements recommended in Article
2.3.1.3.2.d. The wedges are then driven further into the
cable with a sledgehammer. Photographs of cables
wedged for inspection are shown in Figure 2.4.2.1-1.
Start wedging at the middle of the panel (or cable
opening if less than a panel) and insert additional
wedges about every 4 feet, working toward the cable
bands. Drive all the wedges to a uniform depth of about
3 inches, then advance them in sequence for 3 inches at
a time until the center of the cable or the recommended
depth is reached. The following difficulties must be
recognized and overcome to advance wedging.
Gap crossing (i.e., the wire crosses the wedged
opening) occurs whenever the wedge is being
driven along a path on one side of a wire (or
wires) that is opposite to the side where an
adjacent wedge was driven. The condition has
to be recognized quickly, because it impedes
driving and may damage wires. The wedge
should be pulled out and relocated on the same
side of the wire as the prior wedge.
Rejection occurs when a wire that lies in the
middle of the path of the blunt edge of the
wedge will not be pushed aside by the wedge.
The worker experiences a loosening and
regression of the wedge to a position that is
further out than before the hammer blow. The
wire is pushing the wedge back. Driving must
cease to prevent wire damage, and the wire
should be pushed aside on the same side as the
other wedged wires with a non-sparking chisel.
Strewing occurs whenever broken wires are
present. Wedges will often drive the part of a
broken wire nearest to the broken end into
deeper layers of the cable, causing the broken
wire to cross the paths of several intact wires,
and often preventing the wires broken ends
from being seen. This condition cannot always
be prevented, but it can be minimized. Start at
the center of the panel and drive all wedges to
a short and uniform depth. Identify broken
C2.4.2.2
Sometimes lubrication of the wedge with petrolatum or
linseed oil is helpful in driving the wedges. The
investigator should be certain that the lubricant used is
compatible with the corrosion-inhibitive coating, if
there is one.
2-24
wires, and advance wedges first at locations
away from the broken wire ends. Advance
wedges at the ends last, while holding the wires
with additional wedges near the break.
Where broken wires are strewn, as evidenced
by the wires turning inward at a wedge
location, they should be returned to their
original position after the inspection, using
hooked wire picks. Failure to do so will cause
voids inside the cable and crossing wires, both
sources of additional damage.
Wedge tip bending occurs where the cable
offers excessive resistance to penetration. The
wedge tip will select the easiest path, which is
not necessarily radial. The first 3 inches of the
wedge will sometimes bend until driving
becomes impossible, even when the wedge has
penetrated only a small distance into the cable.
The problem can be corrected by pulling out
the wedge and advancing it from the point
where the opening is radial with a non-sparking
implement, and then driving it radially again.
2.4.3 Wire Inspection and Sampling
The purpose of the inspection is to identify corrosion in
the cable wires, and to quantify it according to defined
stages by finding its limits in the cross-section of the
cable.
Sampling of wires determines the physical properties of
the wires in each stage, which are required for
estimating cable capacity.
An example of a filled-in form for a cable cross-section
in a panel is presented in Appendix C.

2.4.3.1 OBSERVATION AND RECORDING OF
CORROSION STAGES
The inspector should identify the corrosion stage of the
wires on both exposed faces of the wedged opening.
This is done visually, using the photographs in Figure
1.4.2.2-1 as a guide. The wire condition is recorded for
at least 3 segments, each approximately 6 feet long,
along the length of a panel. An observed wire should be
first assigned the highest observed stage in the
segment. Each wire is then reassigned the highest
corrosion stage found on that wire in the opening
length, after comparing the recorded data for the wire
in each segment.
Data are entered on inspection forms similar to the one

C2.4.3.1

A corrosion stage may cover only a very short length of
wire, in some cases less than one inch. This is
especially true of wires with black or gray rings of zinc
depletion, which, whenever present, should be counted
separately. Laboratory testing is used to determine the
stage they belong to.
Only the highest stage found along the length of the
wire is used in the analysis of cable capacity in Section
5; it is determined from the data on the forms after the
inspection.
The object of keeping these records is to be able to
2-25
shown in Figure 2.3.1.2.4-2. A typical record of data,
using the form, is presented in Appendix C.
develop a cross-sectional map of the wire stages, as
shown in Appendix C, which is a useful visual tool to
show the extent of damage. The map should represent
the worst condition of all cable segments within the
panel.
2.4.3.2 BROKEN WIRES
Wires found broken in the outer layers of the cable
should be located on the cable cross-section shown in
Figure 2.3.1.2.4-3. Both the tangential location and
depth into the cable should be noted. The maximum
cable depth at which broken wires are seen should also
be noted.
The number and location of broken wires that are
observed inside wedged openings more than a few
layers from the surface of the cable are noted on the
form shown in Figure 2.3.1.2.4-2.
C2.4.3.2
Broken wires can usually be detected in the layer of
wires below the outermost layer. If several adjacent
wires are broken, then it is possible to detect wires up
to 4 layers down. Wires broken at greater depths can
only be seen inside the wedged openings.
2.4.3.2.1 Wedge Spacing
Wedges should be spaced at about 4-foot intervals.
This allows for observing wire conditions deep inside
the wedged cavity. After recording the wire conditions
found, wedges should be removed so that the spacing is
doubled. If there are loose wires, they tend to project
out into the wedged opening or to respond to prodding.
Investigators should experiment to determine the
wedge spacing most suited to detecting loose wires.
Pullout of the intermediate wedges should be partial to
facilitate the observation of loose wires deep inside the
cable; otherwise, loose wires closer to the surface will
hide the deeper ones from view.
C2.4.3.2.1
At a 4-foot spacing of wedges, loose wires are not
visible. Prodding the wires will not give any indication
that they are loose, because they are being tightly held
by the wedges. Experience has shown that loose wires
with a free coil radius of 4 to 6 feet will become
evident at a minimum wedge spacing of 6 feet.

2.4.3.2.2 Wire Tracing
Wires are identified by their distance from the surface
of the cable. Many times the same wire will not show
up at the same distance from the surface within the
panel being inspected, because of poor parallelism or
the formation of a surface lip during wedging. When
there are several adjacent loose wires, to avoid double
counting, it is necessary to prod the loose wire at one
section and observe its longitudinal movement in other
sections. Tracing should be done for all loose wires

2-26
with unidentified ends.
2.4.3.2.3 Failed Wire Ends
Whenever a wire breaks, two wire ends are formed, one
corresponding to each side of the break.
In the counting of broken wires, both ends of each wire
should be accounted for. Avoid the double counting
that occurs whenever the ends of the same wire are
identified as belonging to two different wires.
Whenever only one end is visible, probe to find the end
of a loose wire nearby that is beneath the surface of the
wedged opening.
C2.4.3.2.3
Not all ends of broken wires are visible. Often, wire
breaks occur near or under cable bands, where wedging
is impossible.
Several inspections have shown that there is roughly
the same number of broken wires in a given length of
cable under or near a cable band as in the observed
portion of the panel before the cable band is removed.
However, this may not be the case for each bridge and
should be further investigated by the removal of one or
more cable bands whenever numerous broken wires are
found. In the process of recording, it is necessary to
count all loose wires and all failed wire ends, and to
eliminate all loose wires ends that have already been
counted.
2.4.3.2.4 Sample Size
Broken or loose wires sometimes project from
underneath the surface exposed by the wedges. This
complicates the estimate of the size of the sample, and
may lead to even larger errors in the estimate of broken
wires in the cable.
To minimize this type of error, the investigator should
count broken wire ends and loose wires that come from
underneath the wedged surface as well as those from
the wedged surface proper, and record them separately.
C2.4.3.2.4
In one experience, the broken wire ends from
underneath the wedged surface were approximately
equal to those on the surface. The origin of the loose
wires could not be properly traced, due to the large
quantity of broken wires. The size of the observed wire
sample, for purposes of counting broken wires, was
conservatively taken to be 3/2 times the wires on the
surface of the exposed cavity. The assumption was that
only surface loose wires were visible.
2.4.3.2.5 Other Forms of Corrosion
The inspector should look for and record all other
significant types of corrosion on the same form that is
used for reporting corrosion stages. Significant
corrosion includes
Pitting
General corrosion, that causes a reduction in
the diameter of the wire (report the diameter).
Crevice corrosion, in which the attack and
corrosion product are primarily along the
contact surface of adjacent wires
When these conditions are prevalent (i.e., observed in
more than one or two wires in a wedged opening),
additional samples for testing should be removed to
establish whether additional corrosion stages need to be
created for strength evaluation. When these conditions
are not prevalent, the recommended number of samples
for each corrosion stage should include a number of

2-27
these wires with the condition noted, proportional to
their incidence.
2.4.3.3 PHOTOGRAPHIC RECORD
Typical as well as singular or atypical conditions
should be photographed in color. For each photograph
taken, the number given to the roll of film and the
number of the exposure should be recorded on the
inspection form. The direction of the view and the
target wire or area should be noted.
C2.4.3.3
The condition of wires is better described with
photographs than with words.
Cameras that date the exposure facilitate identification.
Often, taking a photograph of an object not associated
with wire inspection (e.g., an adjacent cable, the tower,
or the roadway), and noting it on the form, is an aid to
establishing the date and general location of the
photographic record.
2.4.3.4 MEASUREMENT OF GAPS AT WIRE
BREAKS
As many ends of broken wires as possible should be
brought into alignment and the gap between the ends
measured. The record of information should include the
depth into the cable of the measured separation and the
panel in which it took place. It should also be noted
whether the measurement occurred with the cable band
or wrapping in adjacent panels removed.
Whenever a sample wire is removed from the cable for
testing, the gap that forms after the first cut should be
measured. A scratch is made on each side of the cut
prior to making the cut. The distance between the two
scratches is measured before and after cutting the wire.
C2.4.3.4

Friction among the wires, especially under wide cable
bands, can redevelop the force in a broken wire.
From measured gaps in broken wires, the capacity of
the cable band to develop wire tension can be estimated
on a statistical basis, based on known dead load and
postulated live load at the time of measurement. Live
load error will not create a large error in the wire
tension estimate, because the live load is usually a
small percentage of the total load.

2.4.3.5 WIRE SAMPLING
Sample locations should be recorded on the cable
cross-section inspection form shown in Figure
2.3.1.2.4-3.
Samples from broken wires should not be used for
strength testing or for determining the fraction of wires
that are cracked.
A new wire should be spliced to the cut ends of a
sampled wire to restore continuity. The new wire is
tightened to the same tension as the other wires in the
cable. The complete procedure for replacing a cut wire
is given in Appendix D. Since sampling requires the
removal of wires from the cable, it should be kept to a
minimum because the splice between the replacement
wire and the end of the existing wire is never as strong
as the original wire.
C2.4.3.5
The object of sampling is to characterize the physical
properties of wires in each of the various corrosion
stages using laboratory tests. The properties that are of
interest to the investigator are the following:
The extent of and variation in zinc oxidation, to
estimate the remaining usefulness of the protective
coating and to evaluate the susceptibility of the
wires to initiation of stress corrosion at the holidays
in the zinc.
The strength of the corroded wires, because
degraded wire may be embrittled, or have surface
corrosion, corrosion pits or propagating cracks. All
these conditions, implying loss of strength, only
start at Stage 3, with embrittlement. Stage 3 wires
may contain pits and a few cracks; Stage 4 wires
usually contain many cracks.
As the inspection proceeds, investigators have to
become aware of all the possible anomalies that may
reduce wire strength, for which wires will require
2-28
testing.
Sampling of broken wires to test for strength may
overestimate the capacity of continuous cracked wires,
because whenever a wire breaks in the cable, the force
in the wire drops to zero, possibly halting crack growth
in the remainder of the wire. Also, broken wires are
known to contain cracks, so that samples from these
wires cannot be used to determine the number of Stage
4 wires that are cracked. Therefore, only unbroken
wires should be sampled for strength testing. Broken
wires removed from the cable during repairs, however,
should be saved for testing corrosion products.
2.4.3.5.1 Number of Samples
A full set of Stage 1 samples and at least half the
recommended number of Stage 2 samples (if Stage 2 is
present) should be removed during the first inspection.
These samples may be combined with additional
samples removed during the second inspection to bring
the total number of samples to the recommended
number. No further samples of these stages are required
in future inspections.
C2.4.3.5.1
The zinc coating tests recommended for Stage 2 wires
should be performed on Stage 1 wires from the first
inspection for use as a baseline in future inspections.
The proposed numbers of samples are calculated on the
basis of test results that indicate a reduction in tensile
strength and an increased variation of strength
properties as corrosion advances.
The wire characteristics given in Table C2.4.3.5.1-1
and Table C2.4.3.5.1-2 were used to estimate the errors
in cable strength at a 97.5% confidence level. These
characteristics are from laboratory tests on the wires of
two bridges, identified as Bridge X and Bridge Z.
Table C2.4.3.5.1-1 Wire characteristics, Bridge X
% Loss of
Strength
Coefficient
%Variation
% of Wires
Cracked
Stage 2 0% 2% 0%
Stage 3 1% 3% 5%
Stage 4 3% 4% 64%
Cracked 16% 13% N/A

Table C2.4.3.5.1-2 Wire characteristics, Bridge Z
% Loss of
Strength
Coefficient
%Variation
% of Wires
Cracked
Stage 2 1% 3% 0%
Stage 3 5% 4% 5%
Stage 4 6% 4% 64%
Cracked 30% 26% N/A
The recommended number of samples to be taken for
each corrosion stage is given in Table 2.4.3.5.1-1. The
number of proposed samples has been selected to
minimize the error in estimated cable strength.
2-29
Whenever adjacent panels are assumed to be perfect
(no deterioration), application of these wire
characteristics result in the estimated strength losses
given in Table 2.4.3.5.1-1. Whenever deteriorated
wires are present in the adjacent panels, strength losses
will be greater.

Table 2.4.3.5.1-1 Recommended number of wire samples for both cables
Estimated
Error (97.5%
confidence)
Estimated
Cable Strength
Loss
Corrosion Stages Present in Worst
Panel Observed
Total Number of Samples
Bridge Bridge
Stage
1
Stage
2
Stage
3
Stage 4 Stage 1 Stage 2 Stage 3 Stage 4 X Z X Z
100% 0% 0% 0% 10 -- -- -- 3% 5% 0 0
>0% 0% 0% 10 15 -- -- 3% 5% 0% 0%
>0% 10% 0% 10 15 35 -- 3% 5% 1% 2%
>0% 20% 10% 10 15 35 60 4% 5% 9% 10%
>0% 40% 20% 10 15 35 60 4% 6% 16% 18%

2.4.3.5.2 Sample Location

2.4.3.5.2.a Stage 1 Wires
The first panels to be unwrapped should be at the low
points: 2 samples of Stage 1 wires should be removed
from each panel, one at 6:00 and one at 3:00 or 9:00,
on the side facing the roadway, or splash zone. One
sample should be removed at random locations from
each of the two remaining panels opened during the
first inspection.

2.4.3.5.2.b Stage 2 Wires
From 1 to 3 samples of Stage 2 wires should be taken
from each panel. At the low points, one should be
removed at 6:00 and one at 3:00 or 9:00 on the side
facing the roadway. A third sample location may be
randomly selected. If no Stage 2 wires are found at the
first location, then the samples should be taken from
another location where Stage 2 is found, and the
number of samples in each panel increased to provide

2-30
the recommended number.
2.4.3.5.2.c Stages 3 and Stage 4 Wires
The recommended number of samples given in Table
2.4.3.5.1-1 should be divided randomly among the
planned number of inspection locations, with at least
0.5 and at most 1.5 times the average number of
samples taken in any one panel. If no wires of the
required stage are found in a panel, than the number
assigned to that panel should be randomly added to the
remaining uninspected panels. Not more than 10
samples of Stage 3 and 10 samples of Stage 4 should be
taken in any one panel.
For planning purposes, the following percentages of
Stage 3 and Stage 4 wires may be assumed in applying
table 2.4.3.5.1-1:
First inspection 10% Stage 3, 0% Stage 4
Second inspection 20% Stage 3, 10% Stage 4
Later inspections percentages estimated by
the investigator from previous inspections
Samples should be selected at random in each
inspected panel, using tables of random sample
locations prepared in advance for several different
groupings of Stage 3 or Stage 4 wires.
C2.4.3.5.2c
It is possible that in early inspections the total number
of Stage 3 and/or Stage 4 wire samples will be smaller
than recommended, especially if a large number of
higher stage wires is found in the last panel opened. In
this case, the error in estimated strength may be greater
than is assumed in Table 2.4.3.5.1-1. The next
inspection, however, should take place in 5 to 10 years,
and the greater number of panels inspected at that time
will provide enough samples.
2.4.3.5.3 Number of Specimens in Each Sample and
Length of Samples
Table 2.4.3.5.3-1 Sample lengths and number of
specimens from each sample
Minimum Number of
Specimens from Each Sample
Corrosion
Stage of
Sample
Strength
Tests
Weight
of Zinc
Tests
Preece
Tests
Sample
Length
(feet)
1 4 1 4 12
2 4 1 4 12
3 10 0 0 16 to 20
4 10 0 0 16 to 20

C2.4.3.5.3

When the corrosion stage varies along the length of a
sample wire, the specimens to be tested for strength
should be cut from the worst areas of the wire.
When Stage 3 is found during the first inspection, or
Stage 4 to a depth not greater than one wire, remove
12-foot-long samples of Stage 3 and Stage 4 and test 4
specimens. When Stage 4 is found to a depth greater
than one wire during the first inspection, extend the
length of cable that is unwrapped to remove 16-foot-
long samples of Stage 4 wires.

The recommended number of specimens to be cut from
each sample for tensile and zinc loss testing, along with the
recommended sample length, are found in Table 2.4.3.5.3-1.
2-31
2.4.4 Identification of Microenvironments
The field or laboratory tests described in this
subsection may be useful in identifying the conditions
inside the cables that are causing the observed
deterioration.
C2.4.4
Researchers and engineers are trying to identify the
microenvironments that produce wire corrosion. It is
possible that many different types of environments
attack bridges, causing wire deterioration at different
rates. Therefore, it is useful to characterize the nature
of the environment that is affecting the cable under
inspection.
2.4.4.1 pH OF INTERSTITIAL WATER
During inspection, condensed droplets of moisture or
even moving water may be seen on cable wires. The
water should be tested with pH test paper strips. If
possible, samples of the water should be collected and
placed in new tightly-sealed inert containers and
refrigerated to prevent gas loss. The samples should be
sent to a laboratory to detect the presence of dissolved
gasses and salts that are known pollutants, such as
chlorides, sulfates and nitrates.

2.4.4.2 CORROSION PRODUCTS
Corrosion products can be removed from the bridge. It
is useful to select specimens from these samples for the
study of corrosion products. This can also be done for
zinc compounds scraped from the wrapping wire

2.4.4.3 PERMANENT PROBES
Although not currently available, permanent probes
could be inserted at critical locations inside the cable.
They might be used to identify time-dependent wet-dry
cycles and indicate the presence and pH of water, the
availability of oxygen, and other factors.
C2.4.4.3
The presence of water is evidence that corrosion of the
wires may be taking place. The pH of the water
indicates the aggressiveness of the environment inside
the cable. A low pH indicates high acidity, which could
be responsible for rapid depletion of the zinc coating.
2.4.5 Cable Bands and Suspender Removals
Cable bands are an integral part of the suspension
system. They also play a role in maintaining cable
capacity once wires begin to fail. Cable band bolt
tensions affect the capacity of the cable band to
develop cable force in broken wires, as well as their
ability to transfer the tangential component of cable
force to the cable without slipping.
Cable bands have to be removed to compare the level
of deterioration in areas of the cable that are near the
cable band with areas in the middle of the panel.
C2.4.5
While wire damage in or near the cable bands has been
found to approximately equal wire damage in the rest
of the panel, this may not always be the case. It should
be checked whenever Stage 4 or broken wires are
present.
2.4.5.1 CABLE BAND BOLT TENSION
During wire inspections, it is customary to inspect
cable band bolt tensions. This is accomplished by
measuring bolt length, while the bolt is both tight and
loose. The measurements are made with an
C2.4.5.1
During cable inspections, the cable band bolt tension
may have to be determined for assessing reliability
against band slippage. While this aspect of the work is
not directly related to cable capacity, it is useful to
2-32
extensometer, which has a sensitivity of 10
-4
inches.
Extensometers are provided with spherical tips that
bear inside conical center holes provided at the bolt
ends. One bolt is inspected at a time, and retightened
immediately thereafter to the originally specified
tension. All bolts of the cable band should be tested.
The caulking between cable band halves is removed
prior to measuring bolt tension, and the center holes in
the bolt cleaned of paint and debris.
A zero reading of the extensometer is made prior to
loosening a bolt. Three to four readings are averaged to
arrive at the zero reading. The bolt is then loosened,
and three to four readings on the tension-free bolt are
again averaged. The difference in these two sets of
readings is the elongation of the bolt due to tensile
stress. The bolt tension is computed from this
elongation. The bolt length is taken from the underside
of the head to the center of the nut.
If the original bolt tension installation force is known,
the measured bolt tensions should provide an estimate
of the force reduction over time, due to creep in the
zinc on the bridge wires or gradual compaction of the
cable.
know the cable band bolt clamping force and the
friction among wires that it may generate.
There may be as much as a 15% error from the real
average tension of the bolts in a band if 16 cable bands
are inspected, but the safety factors and reliability
against slippage are large enough to encompass the
error.
A long-reach micrometer reading to 10
-4
inch can be
used instead of an extensometer reading, but spherical
anvils are required, and they must be checked for
compatibility with the holes at the bolt ends.
2.4.5.2 SUSPENDER REMOVAL AND CABLE
INSPECTION
Removal of a suspender and its cable band requires
contractor assistance and must be carefully planned and
executed.
Before removal, a suspender should be match-marked
against the cable and its length recorded in such a way
that the suspender and its socket can be reinstalled
easily in its original location and orientation. The exact
location of the cable band on the cable should also be
marked.
Where inspection of two adjacent panels is performed
for deeper access into the cable, the wedge line should
be continuous through both panels and the cable band
area.
A cable wedged to achieve this task is illustrated in
Figure 2.4.2.2-1. The removed suspenders can be seen
where they loop over the cable.
C2.4.5.2

Sections of the cable that can be inspected without
band removal should be differentiated from sections
that cannot. This is necessary to obtain estimates of
unobservable defective wires in locations where the
cable bands are not removed. Then, the worst damage
for the entire panel can be mapped in the cable cross-
section.

2.4.5.3 SUSPENDER REINSTALLATION
The cable must be recompacted before the band is
reinstalled, and the band placed at the exact location on
the cable it was in prior to removal. Failure to do so
may result in a change of suspender tension. Bolts

2-33
should be retightened to the bolt tension originally
specified. The band grooves should be aligned in a
vertical plane.
The reinstalled suspender should be aligned with match
marks on the cable. The suspender should be jacked
down so that the suspender sockets can be placed in the
anchoring brackets while the original orientation of the
entire suspender is maintained. This procedure should
ensure that the tension in the suspender legs before
removal is recovered.
2.4.6 Inspection Plan Reevaluation
After initial inspection of what are assumed to be the
worst panels, the investigator should compare the
predicted and actual damage. This may lead to a
reduction of wedging or an increase in the number of
panels to be inspected, or other alterations of the
inspection plan.

2.4.7 Reinstallation of the Cable Protection
System
Upon completion of inspection and sample removal,
the cable must be recompacted and the cable protection
replaced. Recompaction to the level of original
compaction, based on previous measurements of the
cable diameter, can be accomplished using a hydraulic
compactor (see Article 2.3.1.3.2a). The cable diameter
should be no more than that measured before removal
of the wrapping less two times the wrapping wire
diameter. Steel or stiff plastic (e.g., Kevlar) binding
straps should be applied around the cable at intervals of
12 to 18 inches to hold the compaction until the
wrapping is applied. A protective paste should be
applied just ahead of the wrapping machine, and the
cable rewrapped using a machine that can apply a
tension of at least 300 pounds to each ply of the
wrapping wire. The completed wrapping should then
be painted using a paint system specified by the bridge
owner, and the grooves at the cable bands caulked to
exclude water.

C2.4.7
Generally, the protection system is replaced to meet
current standards. Water-based paint systems or
membranes provide excellent protection and prevent
water from entering the cable through the wire-
wrapped area. However, for partial protection that
replaces only limited inspection areas, this may not be
an advantage. If water has had the opportunity to enter
an aging protection system in areas that have not been
inspected, a good paint system on the newly wrapped
area may cause the water to be retained inside the
cable. It is important that at low cable points, the
replacement protection system be provided with
sufficient weep holes to allow the water to escape.
Many authorities are reexamining the use of red lead
pastes and are experimenting with substitutes or even
no paste at all, placing their confidence in the quality of
the exterior protection system.
Many cable bands are caulked today with polyurethane
or polysulfide caulking rather than the caulked lead
wool that may have been used when the cable was first
constructed.
For more information on protection systems, refer to
Article 1.4.2. and Article 1.4.3.
2.4.8 Inspection During Cable Rehabilitation
2.4.8.1 GENERAL
When cables are damaged to a level of Stage 3
C2.4.8.1
The larger part of inspection work requires access to
2-34
corrosion or worse, or when the wrapping system is
deteriorated and requires replacement, some authorities
opt for oiling the cables to arrest corrosion, albeit for a
limited time, and replacing the exterior protection
system, because it has to be removed to perform the
oiling operation.
These conditions favor inspection of the wires in all
cable panels, which can be done after removal of the
wrapping. Figure 2.4.8.1-1 shows a cable wedged for a
rehabilitation operation. While the wedge alignment is
not radial as normally is the case, the openings
provided can still be used successfully for internal
inspection.
cable wires and replacement of the protection system. It
would be wasteful to replace the protection system,
which is associated with cable damage, and miss the
opportunity of inspecting the wires for very little
additional cost.
2.4.8.2 INSPECTION NEEDS VS. OILING
OPERATIONS
Cable oiling starts at high cable locations and proceeds
downward. To conduct inspections that are not affected
by the oiling operation, wedging must be done several
panels ahead of the oiling, so that the wedge locations
are inspected before the oil arrives.
Wedging for oiling does not always offer adequate
openings for assessing the worst cable damage.
Therefore, auxiliary wedging at the bottom of the cable
is necessary for inspection purposes. Since the wedging
must be done before the oil gets there, wrapping should
be removed far ahead of the oiling work, and bottom-
of-cable wedging conducted independent of wedging
for oiling.

2.4.9 Inspection and Testing in Anchorage Areas
The cables in the anchorages splay outward after
passing over a splay saddle or through a splay casting.
The wires are not wrapped in the splay, and thus are
easier to inspect.
Cable strands are connected to the anchorage with
strand shoes and eyebars, or with sockets and rods. The
inspection of these components is discussed below.
Recently constructed suspension bridges often use
shop-fabricated parallel wire strands (PWS) in place of
aerially-spun strands. Anchorages on these bridges will
look different from those described, and the inspection
forms will also look different from the examples
presented in these Guidelines, but the principles of
inspection are the same. The inspector must prepare
forms specific to each bridge, and the inspection should
cover all items relevant to aerially-spun cables,
substituting, for example, strand sockets for strand
C2.4.9
Many anchorages are susceptible to water
condensation, because their concrete masses serve as
heat sinks. Furthermore, many anchorage roofs
function as roadway decks with construction joints at
or near the curb line. This allows surface water to
penetrate the anchorage and drip on the splayed strands
and eyebars closest to the roadway. Makeshift
diversions for roof water have met with little success.
Field experience has demonstrated that corrosion
mechanisms inside the anchorages are significantly
different from the ones in the protected cable in the
main span and side spans from bent saddle to bent
saddle.
This is evidenced by the differences in appearance of
corroded wires. In wet anchorages, there is
considerable surface corrosion and wire failure occurs
after significant section loss. In contrast, wire failure in
the protected cable occurs after embrittlement and
2-35
shoes and anchorage system for eyebars.
The primary difference will be that PWS construction
always uses a socket at the anchorage point of the
strand instead of a strand shoe, and usually uses a steel
framework instead of eyebars to anchor the sockets.
pitting with flat and invert breaks and little or no
section loss.

2.4.9.1 WIRES IN STRANDS
The strands inside the anchorage should be wedged
open on at least one transverse and one vertical line.
The investigator may add more wedging at locations
that show the worst damage.
The minimum diameter of corroded wires should be
measured and recorded.
C2.4.9.1
In general, anchorage strands have more damage at the
lower end of the strand (i.e., near their anchorage
points), than at higher locations, although upper
portions of the strands may also be damaged where
water runs through the splay casting or drips into the
cable just below the splay casting.
2.4.9.2 WIRES NEAR AND AROUND STRAND
SHOES
In wet anchorages, strand wires deteriorate the most
where water collects over time. This generally occurs
in the bottom half-strand, adjacent to or in contact with
The area is difficult to access, and wire damage can
only be guessed at from the surface. This incomplete
information is insufficient to estimate the total wire
damage. Experience on some bridges indicates that if
there are several broken wires at the edge of the strand
shoe accompanied by section loss in surface wires that
have not yet failed, then a worse condition is likely to
be present in an inaccessible area of the strand. The
investigator must exercise engineering judgment about
the potential capacity of the strand, and whether the
strand should be rehabilitated, in the context of overall
strand group conditions and of the need to restore cable
strength in the anchorage area.
C2.4.9.2

Strand cutting and re-anchoring allows for estimating
the condition and capacity of strands of similar exterior
appearance. NDE devices, in use or in development, do
not permit a reasonable estimate of damage in the wires
in the area around the strand shoe.
2.4.9.3 EYEBARS
All eyebars should be carefully examined during
biennial inspections, and the presence of corrosion
products, exfoliating rust and loose paint reported. The
corrosion product on some of the more accessible
eyebars should be removed with hammer and chisel to
a degree sufficient for determining section loss.
Where section loss on the eyebars is suspected, all
paint and corrosion products on the eyebars should be
removed by shot blasting. The extent of section loss is
measured using specially designed calipers on a
minimum of five equally spaced locations along the
width of the eyebar. The loss on the narrow faces of the
C2.4.9.3
In wet anchorages, a heavy deposit of corrosion
products is often found on the surface of the eyebars
just above the point where they enter the concrete
mass. The extent of the damage is visually deceiving,
because the corrosion products may consist of a dense
black oxide that adheres tightly to the metal, often not
removed in preparation for painting. The corrosion
normally does not extend below the surface of the
concrete.
In some anchorages, access for visual observation may
be so limited that a video camera is required.
the strand shoe, as illustrated in Figure 2.4.9.2-1 (the
end of the steel tape indicates the location of the front
edge of the strand shoe).
2-36
eyebars should also be determined.
The surface of the corroded eyebar is usually too rough
to allow the use of ultrasonic thickness-measuring
devices, but a test should be conducted to determine
their suitability.
The remaining area of the eyebars should be calculated
from these measurements and used in turn to calculate
the capacity of the eyebars to anchor the strands.
Removal of corrosion products manually is not
advisable, regardless of accessibility, because the
products are strongly adherent to the base metal and
cannot be dislodged with a hammer.
Rust is best removed with shot blasting. Pack rust
removal should be executed by a contractor.
The capacity of both strands and eyebars is estimated
with a technique that is equivalent to calculating the
cable capacity with the Limited Ductility Model,
described in Section 5. All strands may be considered
clamped at the splay casting or splay saddle, and this
point is mathematically moved to strain the assembly
of eyebars and strands. The force in each strand is
calculated from the elongation. As an assembly of an
eyebar plus attached strand reaches its strength, the
eyebar yields or the strand fails, and part or all of the
force in the assembly is distributed to the other
elements. The sum of strand forces is the force in the
cable; and the maximum force reached is the strength,
which may be less than the sum of the individual
strengths.
2.4.9.4 WIRES INSIDE SPLAY CASTINGS
Inspection is required if there is any indication of wire
damage inside the splay castings. Engineering and
construction planning are necessary for temporary
upward relocation of the splay casting, which permits
separation of the strands and provides access. Only
competent contractors with experience in bridge wire
and cable inspection should be engaged for this work.
The primary purpose of such inspection is to determine
the condition of the wires. Wires that have significant
section loss or are broken should be replaced. New
zinc-coated wires are spliced to a sound point on the
damaged wires. All ferrules should be outside the final
splay casting location. The access that is provided
should permit inspection of all wires, facilitating
estimation of cable capacity in the splay casting area.
C2.4.9.4
The engineering work will most likely require
installation of guide frames to maintain all the strands
at the same length. Failure to adjust the strand
alignment may result in undesirable movement of the
cable and suspended structure. The procedure also
seeks to maintain the same tension in all the wires at all
times.
2.4.9.5 ANCHORAGE ROOFS
When the strands display significant damage,
anchorage roofs should be inspected to identify sources
of moisture.

2.4.9.6 INSTRUMENTATION OF EYEBARS
Whenever live load stress ranges coupled with
temperature change effects on the cable are required by
the investigator, it may be possible to instrument the
eyebars to obtain the needed data. The effects on the
C2.4.9.6
AASHTO design loads and load factors are often not
applicable to long span suspension bridges. Their use
may lead to low safety indices, unrepresentative of real
conditions.
2-37
eyebars can be translated to effects anywhere in the
suspended span area, providing that the anchorage and
tower saddles are free to move.
If necessary, temperature effects on the cable can be
determined separately by installing instrumentation for
temperature on the cable and calculating changes of
stress on the cable by analytic means.
To eliminate the effect of temperature changes on the
eyebars, one full bridge circuit per eyebar is
recommended. Two gages should be placed on each
eyebar at opposite ends of the horizontal centerline to
determine the flexural stress component in the eyebar.
A line of strands chosen diagonally across the strand
group is a sufficiently large enough sample to
guarantee good averages. The force in the outer strands
should be corrected by multiplying by the cosine of the
angle that the outer strand makes with the center strand.
Instrumentation and data acquisition is an effective
means of estimating bridge loads, especially fatigue
loading stress ranges and histograms.
More accurate live loads and stress ranges than those
obtained from instrumenting eyebars can be obtained
by instrumenting the cable wires directly while the
cable is unwrapped. This should be done in the span
beyond anchorages or cable-bent saddles. At least 8
strain gages should be attached to wires at 45-degree
intervals around the cable.
2.4.9.7 DEHUMIDIFICATION
Whenever a dehumidification system is installed in
the anchorage, the system should be inspected in
the presence of maintenance personnel or a
mechanical engineer familiar with its operation,
according to the following procedure:
Measure the relative humidity immediately
upon opening the chamber.
Measure the relative humidity for a 2-hour
period on a humid or rainy day.
Test the operation of the equipment to
verify that it starts when the relative
humidity is raised to the level the
equipment is set for (an electric pot can be
used to boil water in the chamber to
accomplish this), and that it turns off when
the humidity is reduced to normal levels by
removing the source.
Inspect gaskets at doors and other openings
for leakage.

2.4.10 Inspection of Cables at Saddles
Cable wires inside saddles have not been inspected on
any bridge to date. Observation of wires is possible
only from the top of the saddle and at its ends, where
the cable is visible all around its surface, but not inside.

No currently available or soon to be available NDE
devices can assist in estimating wire damage in saddle
areas.
C2.4.10
A saddle is a turning point of the cable; contact
between the saddle and the cable is necessary to
support the bridge. Based on past inspection
experience, the surface wires in the saddle itself are
usually in good condition, especially if they are
covered with wax. There are exceptions: on one bridge,
the top of the cable in the tower saddles was covered
with bird excrement, despite the presence of a tower
2-38
The wires at the top of the saddle and inside protective
sleeves should be inspected, starting with the second
time the cable is inspected, unless signs of water entry
are observed earlier. Half the saddle wires should be
observed during the second inspection and half during
the third.
enclosure. There was no protective wax or other
coating on the wires, and the upper layer of wires
experienced section loss.
It is not necessary to observe inside the saddle cover or
housing during the first inspection. The second
inspection may occur as early as 35 or 40 years, or as
late as 60 years, after completion of the bridge. In
either event, these areas should be inspected starting
with the second inspection.
2.4.10.1 TOWER SADDLES
There are two types of enclosures for tower saddles,
requiring different access routes for observing wires.

2.4.10.1.1 Tower Top Enclosures
Tower top enclosures are extensions of the tower. They
may consist of a penthouse covering the entire top
surface of the tower, or of a series of separate
enclosures, one for each saddle. The enclosures cover
the entire saddle and permit observation of saddles and
exposed wires by the mere opening of an access door.

2.4.10.1.2 Exposed Saddles with Plate Covers
The saddles are exposed at the tops of towers, but have
top plates cap-bolted onto the sides of the saddle
retainers, thus protecting the cable. Flashings are
similarly mounted. To access the wires, plates and
flashings must be removed temporarily.
Cables within saddles have often been protected with a
layer of wax (beeswax and paraffin are used for this
purpose). This protection should be replaced after
inspection, either in kind or with another type of
waterproof coating.

2.4.10.2 CABLE-BENT SADDLES
Cable-bent saddles are placed on top of a rigid frame
structure or on separate columns to accommodate a
change in the direction of a cable as it deflects
downward into the anchorage at the end of the side
spans. The enclosures for these saddles are of several
types.

2.4.10.2.1 Saddles Inside Anchorages
Bent saddles reside inside the anchorage structure,
where saddle and surface wires are exposed and
observable.

2.4.10.2.2 Extended Anchorage Housing
Steel or concrete housings extend from the anchorage
structure above the roadway and contain the bent strut

2-39
and saddle. Within these housings, saddle and surface
wires are observable.
2.4.10.2.3 Exposed Saddles and Plated Roofs
Exposed saddles with plated roofs create conditions
similar to the ones described in Article 2.4.10.1.2.
Cable bents with exposed saddles extend through the
anchorage roof. Wires are protected by plated roofs
bolted to the sides of the saddles. To access the wires,
plates and flashings must be temporarily removed.

Figure 2.2.3.1-1. Typical cable biennial inspection form.
2-40
2-41
Figure 2.2.3.1-2. Typical summary form showing biennial inspection rating system.
2-42
BD 188 (1/96)
BIN NYS DEPT. OF TRANSPORTATION
BRIDGE INSPECTION REPORT
SHEET OF .
TEAM ASST. TEAM
LEADER: L EADER: DATE
Feature Carried:
Feature Crossed:
NEW PREV PAINT

5 5 5 BMS/10 81 band
5 5 3 BMS/10 81-82 wrap
5 5 5 BMS/10 82 band
5 5 4 MMS/10 81-82 wrap
5 5 5 MMS/10 81 band
5 5 3 MMS/10 80-81 wrap
5 5 5 MMS/10 80 band
4 4 3 89S MMS/10 79-80 wrap
4 5 5 MMS/10 79 band
5 5 3 MMS/10 78-79 wrap
5 5 5 MMS/10 78 band
5 5 3 MMS/10 77-78 wrap
4 5 5 MMS/10 77 band
4 4 3 89S MMS/10 76-77 wrap
4 5 5 MMS/10 76 band
5 5 3 MMS/10 75-76 wrap
5 5 5 MMS/10 75 band
5 5 3 MMS/10 74-75 wrap
5 5 5 MMS/10 74 band
5 5 3 MMS/10 73-74 wrap
5 5 5 MMS/10 73 band
5 5 3 MMS/10 72-73 wrap
5 5 5 MMS/10 72 band
5 5 3 MMS/10 71-72 wrap
5 5 5 MMS/10 71 band
5 5 3 MMS/10 70-71 wrap
5 5 5 MMS/10 70 band
5 5 3 MMS/10 69-70 wrap
5 5 5 MMS/10 69 band
5 5 3 MMS/10 68-69 wrap
5 5 5 MMS/10 68 band
5 5 5 MMS/10 67-68 wrap
5 5 5 MMS/10 67 band
The seal between the cable band and the bottom portion of the
cable is missing.
There are overlapping wires near PP 80.
cable is missing.
cable is missing.
There are overlapping wires near PP 80.
LOC. &
SPAN
TP 350 - [28] PRIMARY MEMBERS
RATINGS
CABLE A
PHOTO
NO.
PP MEMBER REMARKS
Figure 2.2.3.1-3. Typical form for biennial inspection showing detailed ratings.
2-43
Figure 2.2.3.2-1. Typical form for biennial inspection of cable inside anchorage.
2-44
Condition of Cables on 31 Bridges
vs. Age at Last Inspection
0
1
2
3
4
5
6
0 40 80 120 160
Age at Last Inspection (years)
C
o
n
d
i
t
i
o
n
GROUP A
GROUP B
Linear (ALL)
Linear (GROUP A)
Linear (GROUP B)
Condition = presence of corrosion stage of
same number (e.g., Condition 1 =
presence of Stage 1, any amount). See
Figures 2.2-3 to 2.2-6 for illustrations of
Corrosion Stages 1 to 4
Stage 0 = new wire
Stage 1 = start of zinc deterioration
(very slight)
Stage 2 = Wires covered with "white
rust"
Stage 3 = 0 to 30% of surface with
ferrous corrosion
Stage 4 = over 30% of surface with
ferrous corrosion
Stage 5 = broken wires present
Group A = Bridges whose
cables deteriorate slower
than the average of all 31
bridges
Group B = Bridges whose
cables deteriorate faster than
the average of all 31 bridges
Note "A": Where two data points
coincide, the second point is
shown directly below the first.
See note "A"
Figure 2.2.4-1. Graph of cable condition vs. age at last inspection.
2-45
Figure 2.3.1.2.2-1 Damaged caulking and paint at cable band.
Figure 2.3.1.2.2-2. Uneven wrapping.
2-46
Figure 2.3.1.2.2-3. Ridge indicating crossing wires.
Figure 2.3.1.2.2-4. Hollow area indicating crossing wires.
2-47
Figure 2.3.1.2.2-5. Damage to wrapping caused by vehicular impact.
2-48
Figure 2.3.1.2.4-1. Form for recording locations of internal cable inspections.
2-49
Figure 2.3.1.2.4-2. Form for recording observed wire damage inside wedged opening.
2-50

Figure 2.3.1.2.4-3. Form for recording locations of broken wires and samples for testing.
SAMPLE FOR TESTING
2-51
Figure 2.3.1.3.2.a-1. Cable compactor.
2-52
Figure 2.3.1.3.2.c-1. Power-driven wrapping machine.
Figure 2.3.1.3.2.c-2. Manual wrapping machine.
2-53
1.5
R

+

4
4
R/10 or 3/4
whichever is less
3
TIP DETAIL
1/16 dia
Round all edges
1/16 radius
Round corners 1/16 dia
STARTING TOOL
Material: Bronze
See tip
detail
(
R

=

C
a
b
l
e

R
a
d
i
u
s
)
1.5
R

+

4
4
R/10 or 3/4
whichever is less
3
TIP DETAIL
1/16 dia
Round all edges
1/16 radius
STARTING TOOL
Material: Bronze
See tip
detail
(
R

=

C
a
b
l
e

R
a
d
i
u
s
)
1/16 dia
(See tip
detail)
R

+

4
4
1/2
3
Round all edges

1/16 radius
ALTERNATE STARTING TOOL
Material: Bronze
(
R

=

C
a
b
l
e

R
a
d
i
u
s
)
1/16 dia
(See tip
detail)
R

+

4
4
1/2
3
Round all edges

1/16 radius
ALTERNATE STARTING TOOL
Material: Bronze
(
R

=

C
a
b
l
e

R
a
d
i
u
s
)
4
R

=

C
a
b
l
e

R
a
d
i
u
s
4
WEDGE
Material: Rock Maple or Oak
See tip
detail
1
10
TIP DETAIL
1/8 dia
4
R

=

C
a
b
l
e

R
a
d
i
u
s
4
WEDGE
See tip
detail
1
10
TIP DETAIL
1/8 dia
R

=

C
a
b
l
e

R
a
d
i
u
s
4
WEDGE
See tip
detail
1
10
TIP DETAIL
1/8 dia
Figure 2.3.1.3.2.d-1. Chisels and wedges.
Note: For cables with a radius greater than 12 inches, wedges can have smaller slope, resulting in a maximum
thickness of about 2.5 inches.
2-54
Figure 2.3.1.3.2.d-2. Hydraulic wedges.
2-55
Figure 2.4.1-1. Form for recording cable circumference.
2-56
Figure 2.4.1.2-1. Removal of wire wrapping.
2-57
WEDGES

Figure 2.4.2.1-1. Additional wedges to inspect area with many broken wires.
2-58
Figure 2.4.2.2-1. Cable wedged for inspection.
2-59
Figure 2.4.8.1-1. Inspection during cable rehabilitation.
Figure 2.4.9.2-1. Deterioration of wires found inside strand shoe.
CONTENTS
SECTION 3 LABORATORY TESTING...............................................................................................................3-1
3.1 INTRODUCTION........................................................................................................................................3-2
3.2 TESTS OF WIRE PROPERTIES...............................................................................................................3-2
3.2.1 Specimen Preparation ........................................................................................................................3-2
3.2.2 Tensile Tests ........................................................................................................................................3-3
3.2.3 Obtaining Data for Force vs. Strain Curves.....................................................................................3-3
3.2.4 Fractographic Examination of Suspect Wires .................................................................................3-3
3.2.5 Examination of Fracture Surface for Preexisting Cracks...............................................................3-4
3.3 ZINC COATING TESTS.............................................................................................................................3-5
3.3.1 Weight of Zinc Test ............................................................................................................................3-5
3.3.2 Preece Test for Uniformity.................................................................................................................3-5
3.4 CHEMICAL ANALYSIS.............................................................................................................................3-6
3.5 CORROSION ANALYSIS ..........................................................................................................................3-6
3.6 FIGURES FOR SECTION 3 .......................................................................................................................3-8
3.7 REFERENCE................................................................................................................................................3-8
3-1
3-2
3.1 INTRODUCTION
Laboratory testing is an integral part of cable
inspection. The test results are used to estimate the
strength of the wires and their stress vs. strain
relationships, which are used in turn to evaluate
cable strength. The same tests are used to determine
the ultimate strain of the wires for the Limited
Ductility Model. Other tests assess the remaining life
of the zinc coating.
Additional tests are performed on cable wires to
study the causes of corrosion. Although they are
referred to in this section, they are irrelevant to the
assessment of structural safety.

3.2 TESTS OF WIRE PROPERTIES
Strength testing is the most essential type of testing
for the evaluation of cable capacity.

3.2.1 Specimen Preparation
A sample wire is a length of wire that has been
removed from a cable for testing. A specimen is a
piece of wire cut from the sample on which a
specific test is performed. Sample wires obtained in
the field should be long enough to provide the
number of specimens recommended in Table
2.4.3.5.3-1. All of the specimens from a given
sample should be at the same stage of corrosion, but
it is understood this is not always possible.
The cast diameter should be determined prior to
cutting specimens from the sample wires. If the
sample is of sufficient length to form a complete
circle as it lies on a flat surface, measure the cast
diameter in two perpendicular directions and average
the results. If the sample is not long enough to form
a complete circle, measure the rise of the arc on each
of two convenient chords of the curve, calculate the
resulting diameters geometrically, and average the
results.
The diameter d is given by
b
c b
d
8
4
2
2 2
+
= 3.2.1-1
where
b = offset between chord and arc
c = chord length
C3.2.1
A typical Stage 4 sample with no cracks has a standard
deviation that is approximately 1% to 2% of the mean
tensile strength. Ten specimens are sufficient to
determine the sample mean tensile strength within 3% of
the true mean with a 97.5% confidence level.
This number of specimens cannot be obtained during the
first internal cable inspection, if the recommended 16
feet of cable are unwrapped. Longer lengths of cable
should be unwrapped whenever corrosion is found to
exceed Stage 3, so that Stage 4 wire samples that are at
least 16 feet long can be removed.
Whenever corrosion is found not to exceed Stage 3,
cracks are not likely to be present, and 12-foot-long
samples providing 8 Stage 3 specimens are adequate.
Wires must be cleaned and all corrosion product must be
removed prior to using a dye penetrant to find cracks.
Even then, shallow pitting may obscure cracks or be
confused with them. The most reliable method of
identifying cracks is to inspect the fracture surface
visually after testing. Optical microscopy prior to testing
is the alternative to using dye or magnetic flux leakage. It
is extremely time-consuming, and the human eye can fail
to spot many cracks, even with a microscope.
3-3
Before sample wires are cut into specimens of
suitable length for testing, they should be inspected
and assigned to the appropriate corrosion stage. If
possible, NDE testing to locate preexisting cracks
should be performed on individual wires before they
are cut, so that the worst cracks can be arranged to
appear near the center of the specimen. Among the
techniques that may be used to identify cracks are
the application of a dye penetrant on cleaned wires
and magnetic flux leakage inspection.
3.2.2 Tensile Tests
Wire strength derived from tensile tests is used to
estimate cable strength. Tensile tests should be
performed in accordance with ASTM A586 and
ASTM A370 to determine the following wire
properties:
breaking load in the wire
yield strength (0.2% offset method)
tensile strength
elongation in 10-inch-gage length
reduction of area
modulus of elasticity
The tensile strength should be based on the nominal
area of the wire.
C3.2.2
Many engineers prefer to use wire tensile strength rather
than wire strength (a force) in calculations, and therefore
strength equations in the Guidelines are derived using
tensile strength, which is multiplied by the nominal area
of the wire to calculate cable strength. For this reason,
tensile strength test results should be based on the
nominal area as well. Either gross metallic area (including
the area of the zinc coating) or net steel area (not
including the area of the zinc coating, which is equivalent
to the nominal area of the uncoated wire) may be used, as
long as the same area is used consistently in all cable
strength calculations. The zinc coating is often degraded
in the samples removed from the cable and the net steel
area is preferred for the calculations, because the actual
diameter of the wire without galvanizing can be measured
more accurately in the laboratory.
Whenever section loss is observed in the specimen, the
stress in the actual corroded wire cross-section may be of
interest, because it provides the actual tensile strength of
the steel in the corroded area. Tests on wires from the
anchorages of the Manhattan Bridge have shown that this
value does not change when section loss occurs.
Significant variation may indicate hydrogen
embrittlement cracking or pitting of the wires.
3.2.3 Obtaining Data for Force vs. Strain
Curves
In addition to the tests listed above, wire elongation
should be recorded at intervals of tensile force up to
maximum force preceding failure. The data should
be used to construct a full stress-strain curve, or
force vs. strain curve, for each specimen. The
ultimate strain corresponding to tensile strength
should be determined as well.
C.3.2.3
The testing laboratory selects the technique for
determining elongations beyond 2%, because
extensometers can be damaged whenever the wire fails.
One option is to measure the motion of the head of the
testing machine, adjusting the elongation for slippage at
the time the grips are set. The measurement should
include both the elastic and plastic components of the
deformation.
3.2.4 Fractographic Examination of Suspect
Wires
The fracture surface of the wires should be observed
C3.2.4
Techniques that are generally used to study the
microstructure of metal can also be used to study failure
3-4
to detect whether failure is ductile or brittle. A brittle
failure is consistent with pitting or cracking, loss of
ductility, a reduction in elongation and strength, and
little or no reduction in area. Special attention should
be focused on the causes of these phenomena. The
instruments recommended for the task are a
stereoscopic optical microscope and/or a scanning
electron microscope.
It is also recommended that X-ray energy dispersion
spectral analysis be performed on any fracture
surface that displays traces of corrosion or
contamination.
and corrosion morphologies.
Optical (Light) Microscopy
A stereoscopic microscope with 20X magnification is the
most efficient tool for the detection of preexisting cracks
in the fracture surface. Crack depth can be measured
directly, if the microscope is fitted with a reticle, or
indirectly by taking a microphotograph of the fracture
surface.
Longitudinal sections of wire that are microetched may
be studied with an optical microscope at magnifications
of 50X to 200X to identify corrosion morphologies in
pits, both intergranular and intragranular, and to establish
the paths of secondary cracks near the fracture surface.
Scanning Electron Microscopy (SEM)
Any characteristic that can be studied with optical
microscopy is more easily studied with SEM.
SEM allows greater depth of field and better resolution at
higher enlargements than optical microscopy. At the same
time, SEM allows simultaneous close study of the failed
surface.
With SEM, failure surfaces are visible with sufficient
detail to identify the failure mechanism, such as cleavage
or ductile rupture. Corrosion residues are also discernible
on the metal structure or fracture surface.
Machining, Polishing and Etching
Bridge wires should be machined and polished in the
plane of the cast radius. This method generally cuts
across any transverse pits and cracks that initiate at the
inner radius. The polished surface is etched with various
reagents to reveal the steel microstructure in detail.
Image Interpretation
Images of failure morphologies under enlargement should
be interpreted by metallurgists or, if they are unusual, by
corrosion experts. The images may indicate
embrittlement, hydrogen-assisted cracking or other
corrosion mechanisms, recognizable to experts in these
fields.
3.2.5 Examination of Fracture Surface for
Preexisting Cracks
Cracked wires are treated as a separate group in
estimating cable strength. Preexisting cracks are
defined as cracks that are present in the specimen
prior to testing. They are found by examining the
fracture surface of all tension specimens under a
C3.2.5
In Stage 1 and Stage 2 wires, preexisting cracks are
usually due to a manufacturing flaw, while in Stage 3 and
Stage 4 wires, the cracks are most often caused by
hydrogen that results from galvanic action. The surfaces
of the preexisting cracks in the Stage 3 and Stage 4 wires
are usually black, and the cracks themselves are easily
3-5
stereoscopic optical microscope at 20X
magnification. A sample wire is considered to
contain a crack if any of the specimens cut from the
sample contains a preexisting crack.
The outer surfaces of the wire in the vicinity of a
brittle fracture should also be examined under a
stereoscopic optical microscope for the presence of
additional preexisting cracks.
A cracked specimen should be photographed, and
the crack depth should be measured. The wire
diameter at the failure plane, as well as the crack
depth, should be reported in both absolute terms and
as a fraction of wire diameter.
Longitudinal sections of short wire segments in the
vicinity of a brittle fracture should be examined
under either an optical or scanning electron
microscope. In preparation, the surface of the
specimen section should be polished and etched. The
recommended etchant is a 10% solution of nitric
acid in ethyl alcohol.
distinguished from the fracture surfaces caused by the
testing load.

A cracked wire is shown in Figure 3.2.5-1.
3.3 ZINC COATING TESTS
Two types of tests are performed on the zinc coating
during cable wire evaluation: Weight of Zinc
Coating Tests and Preece Tests. The minimum depth
of the coating determines its condition, not the
average depth.
C.3.3
Wires often display white spots on a shiny silvery field of
sound zinc. If the white spots represent 30% of the
surface area or more, then there may be significant
variations in the depth of the zinc coating.
3.3.1 Weight of Zinc Test
The Weight of Zinc Coating Test, specified in
ASTM A90, is a gravimetric test that measures the
weight of the zinc removed from a unit length of
wire. It is used to determine the average weight of
zinc in that length, separate from variations in
coating thickness.
Weight of Zinc Coating Tests should be conducted
on Stage 1 and Stage 2 specimens that display
uniform zinc or spotted zinc loss.
C3.3.1
The average weight of zinc in a unit length, determined
by testing, can be converted to an average remaining
thickness of zinc coating and used to predict when the
zinc coating will be depleted.

3.3.2 Preece Test for Uniformity
The Preece Test, specified in ASTM A239, is used
to determine the uniformity of the zinc coating on
Stage 1 and Stage 2 wires.
Preece Tests are chemical tests that depend on the
reaction of copper sulfate and zinc. They are used to
confirm whether the coating on the specimen is
depleted uniformly or locally.
Preece Tests should be conducted on Stage 1 and
C3.3.2
Preece Tests are performed in series. Wires are dipped in
a copper sulfate solution for a standard time period. If
sufficient zinc is present, then the wire retains its shiny
surface from the intact zinc. If the zinc is insufficient,
then the copper electroplates the steel, and the wire
surface turns the color of copper.
The tests are terminated after the fourth dip.
3-6
Stage 2 specimens that display uniform zinc or
spotted zinc loss.

3.4 CHEMICAL ANALYSIS
The chemical composition of the steel wire should
be determined under any of these circumstances:
tests were never performed, results from previous
tests are unavailable, or tests reveal unusual
variations in the tensile strength of samples.
Percentages of the following elements should be
obtained:
carbon
silicon
manganese
phosphorous
sulfur
copper
nickel
chromium
aluminum
Five wires should be analyzed to provide a complete
record for future inspections. If the chemistry of the
steel is found to vary significantly, a metallurgist
should be consulted to study the effects on the
properties of the wire.
A chemical analysis of the surface deposits on the
wire samples should be performed if corrosion is
present, to detect harmful contaminants. The
presence or absence of the following salts should be
established:
chloride
sulfates
nitrates
The results should be reported in absolute amounts,
per unit of wire area.
C3.4
Variations in the carbon content of the wires may cause
wider than usual variations in tensile strength. The
Williamsburg Bridge cable wires are an example of this
phenomenon. In ensuing inspections of such bridges, the
pattern of sample taking should differ from the
recommended pattern so that the extent of the variation in
carbon content, the tensile strength and the ultimate strain
throughout the cable can be determined. The proper
procedure to follow is described by Matteo [1].
Aluminum, the last element listed, is not usually present
in bridge wire, unless it has been used as a killing agent in
the production of the steel.

3.5 CORROSION ANALYSIS
In some cases, the investigator may recommend
studying the corrosion product on a wire or
anchorage. Corrosion analysis can be performed on
surface corrosion films, or on the fracture surfaces of
the steel, or on corrosion by-products.
C3.5
Various types of electronic microscopy are used in
corrosion analysis:
X-Ray Photoelectron Spectroscopy (ESCA)
Also referred to as Electron Spectroscopy for
Chemical Analysis or ESCA, X-Ray
3-7
Chlorides from roadway salts; sulfates, and nitrates
from acid rain are some of the causes of corrosion
revealed by the analysis. Remedial measures may be
recommended to eliminate polluting elements.

Photoelectron Spectroscopy is a surface-sensitive
spectroscopic technique that provides information
about the composition and structure of the
outermost atomic layers (2 nm) of a solid
material. ESCA detects all elements except
hydrogen and helium. The element detection limit
is typically about 0.5%. Sometimes it is possible
to determine the chemical state of elements,
including their bonding structure, using this
technique.
Energy Dispersive X-Ray Analysis (EDAX)
The Energy Dispersive X-Ray Spectrometer is an
attachment to the Scanning Electron Microscope
that identifies elements on the surface from X-
rays emitted by the specimen. EDAX can detect
elements as light as boron (atomic number 5). It is
particularly suited to identifying inorganic
elements. The results are only semi-quantitative
without the use of primary standards, which are
recommended. This is due to the complex
combinations of variables, such as sample size,
surface condition, and orientation of the
apparatus. However, the small peak-to-
background ratio encountered in analysis of low
concentrations of elements is an unavoidable
occurrence that makes adequate quantitative
analysis nearly unobtainable.
X-Ray Diffraction (XRD)
X-Ray Diffraction is used to obtain information about the
structure, composition and state of polycrystalline
materials. It can be used to determine the exact
composition and state of the corrosion products. For
instance, if adequate amounts of the product are available,
it can identify various oxides of a particular element (e.g.,
magnetite Fe
3
O
4
and hematite Fe
2
O
3
).
3.7 REFERENCE
1. Matteo, J., G. Deodatis, and D. P. Billington, Safety Analysis of
Suspension-Bridge Cables: Williamsburg Bridge. Journal of
Structural Engineering, 1994. 120(11): p. 31973211.
3-8
CRACK CRACK
Figure 3.2.5-1. Cracked wire.
3.6 FIGURE FOR SECTION 3
4-1
CONTENTS
SECTION 4 EVALUATION OF FIELD AND LABORATORY DATA............................................................4-1
4.1 INTRODUCTION.......................................................................................................................................4-3
4.2 NOTATION.................................................................................................................................................4-3
4.3 MAPPING AND ESTIMATING DAMAGE............................................................................................4-5
4.3.1 Number of Wires at a Specific Depth................................................................................................4-5
4.3.1.1 NUMBER OF RINGS IN THE CABLE.........................................................................................4-5
4.3.1.2 NUMBER OF WIRES IN EACH RING .........................................................................................4-5
4.3.2 Fraction of Cable in Each Stage of Corrosion..................................................................................4-6
4.3.3 Number of Broken Wires ...................................................................................................................4-8
4.3.3.1 BROKEN WIRES IN CABLE INTERIOR.....................................................................................4-8
4.3.3.2 BROKEN WIRES AT THE SURFACE..........................................................................................4-9
4.3.3.3 BROKEN WIRES AT BOTH LOCATIONS ................................................................................4-10
4.4 WIRE PROPERTIES ...............................................................................................................................4-10
4.4.1 General...............................................................................................................................................4-10
4.4.2 Cracked Wires as a Separate Group...............................................................................................4-11
4.4.3 Individual Wires ...............................................................................................................................4-11
4.4.3.1 MEAN PROPERTIES....................................................................................................................4-11
4.4.3.2 MINIMUM PROPERTIES IN A PANEL LENGTH ....................................................................4-12
4.4.4 Means and Standard Deviations for Each Group..........................................................................4-13
4.5 WIRE REDEVELOPMENT....................................................................................................................4-14
4.5.1 Wire Force at a Cable Band.............................................................................................................4-14
4.5.2 Effective Development Length and Redevelopment Coefficient ..................................................4-15

4-2
FIGURES
Figure 4.3.-1. Counting wires in half-sectors. .............................................................................................4-17
Figure 4.3.1-1. Counting wires. ...................................................................................................................4-17
Figure 4.3.1.2-1. Calculation of number of rings and wires in each ring. ...................................................4-18
Figure 4.3.2-1. Spreadsheet for reporting corrosion stages. .......................................................................4-19
Figure 4.3.3.1-1. Spreadsheet for reporting broken wires...........................................................................4-20
Figure 4.4.3.2-1. Inverse of standard Normal cumulative distribution. ......................................................4-21
Figure 4.5.1-1. Redevelopment of wire tension through friction at the cable bands. ..................................4-22
Figure 4.5.2-1. Effective development length...............................................................................................4-22
4-3
4.1 INTRODUCTION
This section covers the tabulation of field observations and laboratory test results and their translation into a
usable form for estimating cable strength. Instructions are given for determining the number of wires in each
stage of corrosion, the effective development length, and the number of cracked wires in adjacent panels when the
evaluated panel is the only panel inspected. If all the panels in the effective development length have been
inspected, then the technique for evaluating cracked wires presented in Appendix B should be used instead.
4.2 NOTATION

a
j
= fraction of a circle corresponding to the width of the sector that contains broken wire j (4.3.3.1)
a
jk
= fraction of a circle corresponding to the width of the half sector that contains observed wire jk;
where all sectors are of equal size, a
jk
is a constant (4.3.2)
a
w
= nominal area of one wire, used in lab analysis (4.5.1) (4.5.2)
B
sm
= number of broken wires observed on the surfaces of the sector m wedged opening (4.3.3.1)
B
um
= number of broken wires observed below the wedged surface layers of sector m (4.3.3.1)
C
d
= redevelopment factor (fraction of 95% of the mean tensile strength of Group 2 wires that is
developed in a broken wire at each cable band) (4.5.2)
d
0
= depth into cable at which no broken wires are found (4.3.3.2)
d
e
= elastic deformation in length L due to force T (4.5.1)
d
jk
= distance from center of surface wire to center of observed wire jk, expressed in the number
of wires with d
jk
= 1 for the surface wire (4.3.2)
d
w
= average of measured gaps between the ends of a wire broken or cut in the cable in a
panel or group of panels with the same length cable bands (4.5.1)
E = Youngs modulus of wire (4.5.1)
e = ultimate strain of wire specimens; replaces x in Equations 4.4.3.1-1 and 4.4.3.1-2 (4.4.3.1)
F = wire force developed at each cable band (4.5.2)
i = panel number (4.3.3.1); number of a specimen (4.4.3.1)
j = identification number of an observed broken wire in sector m (4.3.3.1); number of a sample (4.4.3.1)
(4.4.4)
jk = identification number of an observed wire in Stage k (jk = 1 to J
k
)(4.3.2)
J
bm
= total number of observed wires in sector m (4.3.3.1)
J
k
= total number of observed wires in Stage k (4.3.2)
k = corrosion stage of wires (k = 1, 2, 3 and 4) (4.3.2); corrosion stage of a group of wires
(k=2, 3, 4 and 5) (4.4.4)
L = length of a wire between centers of cable bands (4.4.3.2) (4.5.1)
L
0
= length of the test specimen between grips of the testing machine (4.4.3.2)
L
e
= effective development length (4.5.2)
4-4
m = sector number (4.3.3.1)
M = number of sectors (4 for quadrants, 8 for octants, etc.) (4.3.3.1)
n
x
= number of wires in ring x (4.3.1.2)
N = total number of wires in the cable (4.3.2); actual number of wires in the cable (4.3.1.1) (4.3.1.2)
(4.3.2)
N
B
= number of cable bands required to redevelop the wire (4.5.2)
n
b1,i
= number of broken wires in the outer ring of the cable in panel i (4.3.3.2)
n
bi
= estimated total number of broken wires in cable panel i (4.3.3.1) (4.3.3.2)
n
bm
= unfactored estimate of the number of broken wires in sector m (4.3.3.1)
n
j
= number of wires in the ring that contains broken wire j (4.3.3.1); number of specimens tested from
sample j (4.4.3.1)
n
jk
= number of wires in the ring that contains observed wire jk (4.3.2)
N
jk
= number of wires in cable represented by observed wire jk (4.3.2)
n
k
= number of samples in Group k (4.4.4)
N
sk
= number of Stage k wires in the cable (4.3.2)
N
T
= number of panels on one side of a wire break in which the wire tension is less than T (4.5.1)
p
sk
= fraction of wires in the cable represented by Stage k (4.3.2)
p
c,k
= fraction of Stage k wires that are cracked (4.4.2)
p
c,3
= fraction of Stage 3 wires that are cracked (4.4.2)
s = tensile strength of wire specimens; replaces x in Equations 4.4.3.1-1 and 4.4.3.1-2 (4.4.3.1)
T = tension in wire under service loads (4.5.1)
USF
m
= weighting factor to adjust estimated number of broken wires for the number of layers observed in
sector m (4.3.3.1)
X = number of rings of wires in the cable not including the center wire (4.3.1.1) (4.3.1.2) (4.3.2)
x = number of rings from the center of the cable to a specific ring (4.3.1.1); property of a wire
(i.e., tensile strength or ultimate strain) (4.4.3.1) (4.4.4)
x
1,j
= probable minimum value of x
j
in a length L of the wire from which sample j is removed (4.4.3.2)
(4.4.4)
x
ij
= property of specimen i cut from sample j (4.4.3.1)
x
jk
= number of rings from the center of the cable to center of observed wire jk (4.3.2)
s2
= sample mean tensile strength of Group 2 wires (4.5.2)
sj
= sample mean of the property x for sample j (4.4.3.1) (4.4.4)
sk
= sample mean of property x of Group k wires (4.4.4)
sj
= sample standard deviation of the property x for sample j (4.4.3.1)
sk
= sample standard deviation of property x of Group k wires (4.4.4)
-1
(L
0
/L) = inverse of the standard normal cumulative distribution for the probability L
0
/L (4.4.3.2)
4-5

4.3 MAPPING AND ESTIMATING DAMAGE
An observed wire is assumed to represent all the wires
at the same depth on the same side of the wedged
opening in the half-sector, either left or right of the
opening, where the observation was made (see Figure
C4.3
Spreadsheets similar to the illustrated ones are
recommended for ease of calculation. One spreadsheet
is used to analyze broken wires and one to analyze the
stages of corrosion for each panel.
4.3.1 Number of Wires at a Specific Depth
For the purpose of analyzing data gathered in the field,
it is assumed that the cable is composed of concentric
rings of wires arranged around a central wire, as shown
in Figure 4.3.1-1.
C4.3.1
The wires in a cable do not actually lie in precise rings,
but it is assumed that they do because it facilitates
estimating the number of wires at a specific depth
inside the wedged opening.
It would be convenient to assume that the spacing
from the center of one ring to the next is equivalent to
one wire diameter. In fact, the spacing between rings is
less than one wire diameter, because wires in one ring
partially nest within the wires of the next inner ring.
This more than offsets the spacing between wires in a
specific ring, which is slightly greater than one wire
diameter. The net effect is that the number of rings is
greater than the distance from the surface of the center
wire to the cable surface divided by the wire diameter.
4.3.1.1 NUMBER OF RINGS IN THE CABLE
The number of rings in the cable is estimated by the
following equation
5 . 0 + =
N
X , rounded to the next higher integer

(4.3.1.1-1)
where
X = number of rings in the cable not
including the center wire
N = actual number of wires in the cable

4.3.1.2 NUMBER OF WIRES IN EACH RING
The number of wires in each ring is given by
) X ( X
) N ( x
n
x
1
1 2
+
= (4.3.1.2-1)
where
n
x
= number of wires in ring x
C4.3.1.2
In most cases, wedge lines will be equally spaced
around the cable. Each sector is a pie-shaped portion
of the cable that embraces a wedge line and extends to
a line midway to the next wedge line on either side
(see Figure 4.3.1-1). The sample spreadsheets that
accompany the Guidelines assume 8 wedge lines that
form 8 sectors, but this is a convenience and does not
apply automatically to every inspection. If wedge lines
4.3-1). Field data recorded on inspection forms are
analyzed to estimate the damage in each inspected panel
of the cable. Wires can be either intact or broken and
the calculations differ for each.
4-6
x = number of rings from the center of the
cable to a specific ring
The center wire in the cable lies at x = 0, and n
0
is taken
as 1.
are not equally spaced, each sector still extends from a
line midway between wedges on one side to a line
midway between wedges on the other. In this case,
each observed wire is assumed to represent half the
wires in a ring in the sector rather than the wires on
one side of the wedge. Both half-sectors are assumed
to be of equal size in any given sector, but not all
sectors are the same size.
4.3.2 Fraction of Cable in Each Stage of Corrosion
Each observed wire is assigned a corrosion stage; this is
the highest observed stage in the length of wire exposed
in the wedged opening.
The total number of cable wires in each stage of
corrosion, k, is calculated by adding together the
individual wires in each half-sector represented by an
observed wire at that corrosion stage. The distance from
the center of the cable to the center of each observed
wire should be determined first. In Equation 4.3.1.2-1,
the distance is expressed as x, or the number of rings
from the center of the cable.
After x is determined, the number of wires in the ring,
n
x
, is multiplied by the fraction of the circle represented
by the half-sector that contains the observed wire. For a
cable inspected with four wedge lines (quadrants), the
fraction is 1/8 (half of a quarter-circle); when eight
wedge lines are used (octants), the fraction is 1/16. For
other divisions, or unequal divisions that are the result
of additional wedges being driven into the lower half of
the cable, the appropriate fraction for each inspected
wire should be used.
Let there be J
k
wires observed in a specific corrosion
Stage k. Each observed wire is given an identification
number jk, starting with 1 and ending with J
k
.
Furthermore, the wire lies at a specific depth into the
cable, d
jk
, which is expressed as the number of wires
from the surface wire, which is assigned the depth of 1.
The number of wires in the ring containing wire jk is
given by
) 1 (
) 1 ( 2

=
X X
N x
n
jk
jk
(4.3.2-1)
in which
x
jk
= X + 1 d
jk
(4.3.2-2)
where
x
jk
= number of rings from the center of the
C4.3.2
The most convenient way to calculate the fraction of
the cable in each stage of corrosion is with a
spreadsheet. The spreadsheet is constructed to
calculate quantities of wires in each corrosion stage.
The sum of the number of wires in each ring in each
corrosion stage should equal the total number of wires
in the cable. A sample spreadsheet is shown in Figure
4.3.2-1, which assumes that 8 rows of wedges have
been observed.
The spreadsheet contains one row for each ring of
wires from the surface to the center of the cable, with
two columns for each wedge line inspected. The rings
are numbered, starting with 1 for the outer ring, to fill
out the first column of the spreadsheet. The shaded
areas are filled in for each inspected panel with
corrosion stage entered as an integer. Columns on the
right are filled in automatically with spreadsheet
formulas.
The center wire of the cable lies at x = 0, and n
0
= 1.
The number, n
0
, is not calculated by Equation 4.3.1.2-
1, because the wire does not stand for a ring of wires.
The value of d
jk
for the wire is X +1.
The number of wires in each ring is shown in the
second column of the spreadsheet. The corrosion stage
of each ring of wires on each side of the wedged
opening in each sector is entered on the spreadsheet in
the appropriate column. For each observed wire, the
worst corrosion stage found anywhere in the length of
the panel is entered on the spreadsheet.

The number of wires in each ring is not rounded. Sample
calculations of X and n are given in Figure 4.3.1.2-1.
4-7
cable to the center of observed wire jk
n
jk
= number of wires in the ring that contains
observed wire jk
X = number of rings of wires in the cable not
including the center wire
N = actual number of wires in the cable
d
jk
= distance from the center of the surface wire
(of cable) to center of observed wire jk
expressed in the number of wires with

jk
d = 1 for a surface wire
k = corrosion stage of wires (k=1 ,2, 3 and 4)
jk = identification number of an observed
wire in Stage k (jk = 1 to J
k
)
The number of wires in the cable represented by
observed wire jk is
jk jk jk
a n N = (4.3.2-3)
where
N
jk
= number of wires in cable represented by
observed wire jk
a
jk
= fraction of a circle corresponding to the
width of the half-sector that contains
observed wire jk; whenever all half-sectors
are of equal size, a
jk
is a constant.
and the total number of Stage k wires in the cable is

= =
= =
k k
J
jk
jk jk
J
jk
jk sk
a n N N
1 1
(4.3.2-4)
where
J
k
= total number of observed wires in Stage k
N
sk
= number of Stage k wires in the cable
The sum of all values of N
sk
must equal the total number
of wires in the cable. The fraction of the wires in the
cable in each stage of corrosion is determined by
p
sk
= N
sk
/N (4.3.2-5)
where
p
sk
= fraction of the wires in the cable
represented by Stage k
N = total number of wires in the cable
4-8
4.3.3 Number of Broken Wires
The method of estimating the number of broken wires
in a panel depends on their location. When they are
found beyond the first few surface layers, each broken
wire at a given depth in the cable is assumed to
represent all wires at the same radius from the center of
the cable in that half-sector (half-octant or half-
quadrant, etc.). In Figure 4.3-1, the wires represented by
the observed broken wires at depth d
L
and d
R
are
indicated by the solid areas. This method of counting
the wires is described in Article 4.3.3.1.
Whenever broken wires are found only on the outer
surface and in the first few layers of the cable, they
should be counted using the method described in Article
4.3.3.2.
A combination of the two counting methods is
recommended whenever broken wires are found both on
the surface and inside the cable The investigator should
determine the depth to which each method is applied
(see Article 4.3.3.3).

4.3.3.1 BROKEN WIRES IN CABLE INTERIOR
The number of broken wires is estimated in a manner
similar to the one described for determining corrosion
stages in Article 4.3.2. Interior broken wires are
subdivided into two groups, those broken at the surface
of the wedged opening, and those broken beneath the
surface (a broken wire end is protruding). Broken wires
in each sector, m, are counted separately. The equations
for counting are:

=
=
bm J
j
j
j
bm
a n n
1
5 . 0 (4.3.3.1-1)
where
n
bm
= unfactored estimate of the number of
broken wires in sector m
j = identification number of an observed
broken wire in sector m
m = sector number
J
bm
= total number of observed wires in
sector m
n
j
= number of wires in the ring that contains
broken wire j (see Equation 4.3.2-1)
a
j
= fraction of a circle corresponding to the
width of sector m that contains broken
C4.3.3.1
Using a spreadsheet is also convenient for estimating
the number of broken wires in a panel, as shown in
Figure 4.3.3.1-1.
The shaded areas are filled in with the number of
broken wires on both sides of the wedge at each depth.
The last column is filled in automatically by
spreadsheet formulas. The spreadsheet should contain
a numbered row for each ring of wires from the
surface to the center of the cable, and two columns for
each inspected sector. The number of broken wires
observed on the surface of the wedged opening is
entered in one of the columns. If a wire is observed to
be broken on each side of the wedged opening at the
same depth, the number 2 is entered at the appropriate
ring. If it is observed broken only on one side, the
number 1 is entered. Broken wires observed in the
layer directly beneath the surface layer are entered the
same way in the right column.
Whenever broken wires at the surface (subscript s)
and broken wires beneath the surface (subscript u)
are found in a wedged opening, consideration must be
given to the fact that more than one layer of wires has
been observed. If the number of s and u wires is
equal, then 2 layers have been observed. The
observation of wires at the surface is always more
dependable than the observation of wires beneath the
4-9
wire j
A weighting factor, USF (Under Surface Factor), is
applied to the number of broken wires, to account for
the number of surfaces observed. It is calculated for
each sector, making up for the fact that more wires have
been sampled than the number of wires in contact with
the wedges, protruding broken wires underneath the
exposed layer in particular.
5 . 0
) (

+
=
um sm
m s
m
B B
B
USF (4.3.3.1-2)
where
USF
m
= weighting factor to adjust estimated
number of broken wires for the number of
layers observed in sector m
B
sm
= number of broken wires observed on the
surfaces of the sector m wedged opening
B
um
= number of broken wires observed below the
wedged surface layers of sector m
The estimated total number of broken wires in a panel,
i, is then
bm
M
m
m bi
n USF n =

=1
(4.3.3.1-3)
where
n
bi
= estimated total number of broken wires in
cable panel i
i = panel number
M = total number of sectors (4 for quadrants, 8
for octants, etc.)
surface.
The USF, applied to the total number of wires
estimated to be broken in each sector, is always greater
than or equal to 0.5. If only s wires are observed,
then USF = 1.0. If an equal number of s and u
wires are observed, then USF = 0.50 (i.e., 2 layers of
wires have been observed, each layer corresponds to
half the number of wires at the corresponding radius
from the center of the cable). If more u wires are
observed than s wires, then the number of layers
observed is still 2. USF remains 0.50, because the
surface of the wedged opening is actually seen, and the
lower layer is only assumed to be seen by noting the
broken wires that protrude.

4.3.3.2 BROKEN WIRES AT THE SURFACE
If broken wires are found only near the surface of the
cable, then the depth at which broken wires are no
longer found, d
0
, can be determined by observing the
wedged openings.
The inspector should use additional wedges to ascertain
the depth at which no broken wires are found. The
wedges should be located near surface broken wires,
and driven at least 2 inches beyond the depth of Stage 4
wires.
Depth d
0
is expressed in number of wires from the cable
surface, with the surface wire being number 1. The
number of broken wires in each ring is conservatively
C4.3.3.2
In some cables, broken wires are found only at or near
the outer surface of the cable. To count only wires
found broken at the wedges would distort the assumed
number of broken wires.
When broken wires are observed only in the outer
layer, d
0
will be 2, because no broken wires are found
in the second layer. The number of broken wires using
Equation 4.3.3.2-1 is then equal to n
b1,i
.
4-10
assumed to decrease in a linear manner from the surface
to zero at the depth d
0
. The total number of broken
wires is calculated as the number of broken wires at the
surface of the cable in ring 1 multiplied by d
0
/2,
2
0 , 1
d n n
i b bi
= (4.3.3.2-1)
where
d
0
= depth into cable at which no broken
wires are found
n
b1,i
= number of broken wires in the outer ring
of the cable in panel i
4.3.3.3 BROKEN WIRES AT BOTH LOCATIONS
Whenever broken wires are found at the surface and
inside the cable, the investigator should determine
whether broken wires can be observed in one or two
layers from the surface, and then estimate the number of
additional interior broken wires beyond those layers,
using Article 4.3.3.1. The total number of broken wires,
n
bi
, is the sum of the observed broken wires in the outer
one or two layers and the estimated interior broken
wires.
Additional wedges should be driven between the
recommended wedges halfway to the center of the cable
for a more accurate estimate of the number of broken
wires inside the cable.
C4.3.3.3
When a full panel length is unwrapped, the number of
broken wires in the outer ring between cable bands can
be known with certainty. If an outer wire is broken
under the cable band, it will usually be loose in the
exposed length. Broken wires in the second layer can
often be detected with some prying of the outer layer
of wires.
4.4 WIRE PROPERTIES
4.4.1 General
The properties of a single cable wire can vary from
those of other wires; they can also vary along the length
of that wire. The investigator should obtain the lowest
values of these properties in a panel length for each
wire, as well as the variation among all the wires in the
cable.
As a first step in this process, the laboratory test data for
the sample wires are sorted into these groups:
Group 1 samples exhibiting Stage 1 corrosion, if
determined by the investigator to be needed
Group 2 samples exhibiting Stage 1 or Stage 2
corrosion
Group 3 samples exhibiting Stage 3 corrosion that
are not cracked
Group 4 samples exhibiting Stage 4 corrosion that
are not cracked
C4.4.1
In general, the properties of Stage 1 and Stage 2 wires
are alike, or vary so little that they are considered to
form a single group, called Group 2. Should the
investigator find significant variations between the two
stages, they should be divided into Group 1 and Group
2, respectively.
Wires that are not cracked but that exhibit corrosion
pitting to any depth are assigned to Stage 3 and
Stage 4.
Very few Stage 3 samples are expected to contain
cracks, so that statistical data for the strength or
elongation of any Stage 3 cracked samples should not
be used in estimating cable strength. Cracks usually
occur in Stage 3 wires that are close to being Stage 4,
and therefore, Stage 3 and Stage 4 cracked wire data
can be combined into a single group, called Group 5. It
is more conservative, however, to use only the
4-11
Group 5 samples exhibiting Stage 3 or Stage 4
corrosion that contain one or more cracks

properties of Stage 4 cracked wires to represent all the
cracked wires, a practice that is recommended.
The investigator must decide whether to introduce an
additional group whenever 25% or 8 Stage 3 samples
are found to have cracks. If so, all cracked wires
should form a single group, and the property
distributions for the two stages of cracked wires should
be combined (see Article 5.3.3.1.1).
4.4.2 Cracked Wires as a Separate Group
A sample wire is cracked if any of the specimens cut
from that sample contains a preexisting crack. The
fraction of cracked wires in each stage of corrosion is
given by
p
c,k
= number of cracked Stage k sample wires
total number of Stage k sample wires
(4.4.2-1)
where
p
c,k
= fraction of Stage k wires that are cracked
For Stage 3 wires,
p
c,3
= 0.33 number of cracked Stage 3 sample wires
total number of Stage 3 sample wires
(4.4.2-2)
where
p
c,3
= fraction of Stage 3 wires that are cracked
C4.4.2
The detection of a crack in a specimen is described in
Article 3.2.5. If cracks are found in Stage 3 samples,
they are usually in the outermost layers, with Stage 4
wires close by. The factor, 0.33, adjusts for the fact
that Stage 3 wires found deeper inside the cable rarely
exhibit cracks. If they do, the factor should be
increased accordingly.
4.4.3 Individual Wires
4.4.3.1 MEAN PROPERTIES
The sample mean and sample standard deviation of the
tensile strengths, and the ultimate elongations of the
specimens cut from each sample, are determined as
follows

=
=
j
n
i
ij
j
sj
x
n
1
1
(4.4.3.1-1)
( )
2
1
2 2
1
1
sj
n
i
ij
j
sj
j
x
n

=
(4.4.3.1-2)
where
sj
= sample mean of the property x for sample j

sj
= sample standard deviation of the property x
for sample j
x = property of a wire (i.e., tensile strength or
C4.4.3.1
The symbols and refer to the mean and standard
deviation of a property of the entire population of
wires in the cable. The mean and standard deviation
used in cable strength models are the sample mean and
sample deviation determined from laboratory tests
made on a selection of cable wires removed during
inspection. They are designated by
s
and
s
.
4-12
ultimate strain)
x
ij
= property of specimen i cut from sample j
s = tensile strength of wire specimens,
(replaces x in Equations 4.4.3.1-1 and
4.4.3.1-2)
e = ultimate strain of wire specimens
(replaces x in Equations 4.4.3.1-1 and
4.4.3.1-2)
n
j
= number of specimens tested from sample j
i = number of a specimen
j = number of a sample
4.4.3.2 MINIMUM PROPERTIES IN A PANEL
LENGTH
It is necessary to estimate the weakest point of the wire
between cable bands. To estimate the probable
minimum tensile strength of each sample in a length
equal to the cable band spacing, L, the following
equation is used
sj sj j ,
)
L
L
( x + =
0 1
1
(4.4.3.2-1)
where
x
1,j
j
in a length L
of the wire from which sample j is removed

-1
(L
0
/L) = inverse of the standard normal cumulative
distribution for the probability L
0
/L
L
0
= length of the test specimen between grips of
the testing machine
L = length of a wire between centers of cable
bands
If a negative minimum value of a property results, zero
should be substituted for the minimum value.
C4.4.3.2
A wire will fail at its weakest point within a given
panel. Thus, the estimated minimum strength or
ultimate strain of a wire in a length of panel is required
for estimating total cable strength.
The method for estimating the strength of the weakest
link of a chain of n links is used to find this minimum.
The specimens used to determine the properties in
tensile tests should measure 12 inches between the
jaws of the testing machine. The properties of a
specimen are, by definition, the minimum in this
length, defined as L
0
.
The probability of any 12-inch length of wire in a total
length, L, being the weakest (or having the minimum
ultimate strain) is 1/n, with n being the number of
lengths of wire in the length, L, or n = L/L
0
. The
properties of a single wire will vary according to a
distribution function. The Normal distribution is often
used for this purpose.
A substitution of the Standard Normal distribution for
the Normal distribution is helpful. If the variant z, in
terms of the number of standard deviations, is used,
then the mean will be zero. The function, (z), gives
the probability of an event; thus the inverse,
-1
, will
give the number of standard deviations for a given
probability. The function can be expressed as the
single curve shown in Figure 4.3.3.2-1, which is a
function of L/L
0
= n. The value of
-1
is used in
Equation 1 to determine the minimum value of the
tensile strength. The appropriate number of standard
deviations is subtracted from the mean tensile strength.
The plus sign in the equation is correct, because
-1
is
a negative value. The inverse of the Standard Normal
The value of the term
-1
(L
0
/L) is given in Figure 4.4.3.2-1.
4-13
distribution is included in some spreadsheet programs.
In some extreme cases, especially for ultimate strain,
the Normal distribution may predict a negative value
for the minimum value of a property when the
coefficient of variation, V, is large. In these instances,
4.4.4 Means and Standard Deviations for Each
Group
For each group of samples, calculate the sample mean
and sample standard deviation of the minimum
estimated properties using the following equations:

=
=
k
n
j
j ,
k
sk
x
n
1
1
1
(4.4.4-1)
( )
2
1
2
1
2
1
1
sk
n
j
j ,
k
sk
k
x
n

=

=
(4.4.4-2)
where
sk
= sample mean of property x of Group k
wires

sk
= sample standard deviation of property x
of Group k wires
n
k
= number of samples in Group k
j = number of a sample
k = corrosion stage of a group of wires
(k = 2, 3, 4 and 5)
x
1,j
j
in a length
L of the wire from which sample j is
removed
x = property of a wire (i.e., tensile strength or
ultimate strain)

C4.4.4
In calculating the sample standard deviation, the
number of samples, (n), is usually decreased by 1 as a
conservative measure, especially for small numbers of
samples. The resulting increase in the standard
deviation diminishes as the number of samples
increases. This reduction in n is applied twice in these
Guidelines, first to the individual samples, and again in
calculating the standard deviation of the groups.

zero is sufficient for the minimum value of the property.

A more exact value can be calculated using the Gamma
distribution, which is discussed in Appendix A, on the
accompanying CD.
distribution, which is discussed in Appendix A to the
final report for NCHRP Project 10-57, on the
accompanying CD.
4-14
4.5 WIRE REDEVELOPMENT
A broken wire does not become inactive over the entire
cable length. It redevelops its force as the distance from
the break increases, caused by friction at the cable
bands.
The force in a wire developed at a cable band is
estimated using the measured gap between the ends of a
broken wire, or a wire after the first cut is made for the
purpose of removing a sample.
Wire failure separations used to estimate redevelopment
capacity should come from an area near the evaluated
panel.
Investigators should take measurements of as many
failed wire separations as possible in the area under
study. Refer to Section 2.4.3.4 for further discussion.
C4.5
Wire wrapping has also been found to reduce slippage
in broken wires. The effect is ignored, because it is
unquantifiable and is lost when the cable is
unwrapped, although it is reestablished upon
rewrapping. Whenever a single panel is unwrapped,
the wrapping in adjacent panels remains effective, and
the redevelopment coefficient allows for its effect.
Taking wire failure separations in the area under study
is particularly important on the steeper slopes of the
cable, where the bands are large and exhibit as much
as 2.5 to 3 times the clamping force as at the center of
the span. A smaller value of L
e
may be found at these
locations, which would increase cable strength. If data
from shorter bands elsewhere on the cable were to be
used instead, the redevelopment capacity could be
underestimated.
Wire failure separations predict whether the band can
redevelop dead load plus live load only at the time of
observation. They cannot be used to determine
whether the bands may develop more capacity in the
future (e.g., if the cable band bolts are retightened).
4.5.1 Wire Force at a Cable Band
The effective development length is determined for
each panel from which sample wires are removed, or in
which broken wires are found. In the panels on each
side of the one with the broken wire, the force in that
wire is less than the service tension in an unbroken wire
(Figure 4.5.1-1), The number of these panels is
( ) 1 =
e w T
d d N , rounded to the next higher integer
(4.5.1-1)
in which
E a TL d
w e
/ = (4.5.1-2)
where
a
w
= nominal area of one wire, used in lab
analysis
d
e
= elastic deformation in length L due to
force T
d
w
= average of measured gaps between the ends
of a wire broken or cut in the cable in a
panel or group of panels with the same
length cable bands
E = Youngs modulus of wire
C4.5.1
Whenever a wire breaks or is cut inside the cable, the
ends of the wire separate for a short distance. If the
wires were perfectly clamped at the adjacent cable
bands and could not slip relative to each other at those
points, then the gap would be exactly equivalent to the
elastic stretch in the wire between cable bands under
the action of the cable loads. This is often the case, and
then N
T
= 0 in Equation 4.5.1-1. One cable band is able
to develop at least the working tension in the wire,
conservatively assumed to be exactly the working
tension and no more.
If the solution of Equation 4.5.1-1 is an integer,
including 0, then it is not rounded up to the next higher
integer.
If the gap is greater than d
e
, then it is necessary to
determine the fraction of the working tension in the
wires that one cable band can develop. It is assumed
that the behavior is symmetrical, that the wire tension
in all panels is the same, and that the force
redeveloped in a wire is the same at all the bands near
the wire break.
Bands at the steepest slope will be longer, have more
bolts, and clamp the wires tighter. The number of
locations for which the effective development length is
4-15
L = length of a wire between centers of cable
bands
N
T
= number of panels on one side of a wire
break in which the wire tension is less
than T
T = tension in wire under service loads
calculated can be reduced by grouping panels that have
cable bands of equal length. Then, d
w
is determined by
averaging the measured gaps at wire cuts in all the
panels in the group.
4.5.2 Effective Development Length and
Redevelopment Coefficient
1 2 =
B e
N L (4.5.2-1)
in which
( )

+
+

=
e
w
T
T T w s
B
d
d
N
N N
T
a .
N
1
1
2
1 95 0
2
,
rounded to the next higher integer (4.5.2-2)
and where
L
e
= effective development length
N
B
= number of cable bands required to
redevelop the wire
s2
= sample mean tensile strength of Group 2
wires
When d
w
= d
e

=
B
N
T
a .
w s

2
95 0
,
rounded to the next higher integer (4.5.2-3)
The redevelopment coefficient is the fraction of 95% of
the mean strength of Group 2 wires developed by
friction at each cable band through which the wire
passes, assuming that each cable band in the effective
development length, including the end band, develops
the same force in the wire.
2 1

= =
e B
d
L N
C (4.5.2-4)
where
C4.5.2
Both effective development length, L
e
, and the
redevelopment coefficient, C
d
, are required for
calculating cable strength. Beyond the effective
development length, a broken wire will have no effect
on the calculated panel strength at the center of that
length. The effective development length is a function
of the stress at which the cable reaches its strength,
calculated with one of the various strength models.
Cable efficiency, defined as the cable strength divided
by the mean wire strength multiplied by the number of
wires in the cable, is about 95% for an undeteriorated
cable. Thus, to develop full cable strength, only 95%
or less of the mean wire strength needs to be
redeveloped in the broken wire.
Whenever d
w
= d
e
, Equation 4.5.2-2 contains the
expression N
T
/N
T
= 0/0 and is indeterminate, use
Equation 4.5.2-3 instead.
The redevelopment coefficient is needed to calculate
the force of a wire in the evaluated panel that is broken
in a nearby panel.
Whenever the value of L
e
varies by more than two
panels between areas with short and long cable bands,
the investigator should determine whether different
values for C
d
and L
e
are warranted in cable strength
calculations. If the value varies by only two panels, the
value of C
d
and L
e
calculated from the average of all
the measured gaps at cut and broken wires may be
used for the entire cable.

As shown in Figure 4.5.2-1, the effective development

length is expressed as the number of cable panels, always
odd, that is twice the number of panels in which the wire
is redeveloped plus the panel in which the wire is
evaluated, or twice the number of cable bands required
to redevelop the wire minus 1, or
4-16
C
d
= redevelopment factor (fraction of 95% of
the mean tensile strength of Group 2
wires that is developed in a broken wire
at each cable band)
The wire force, F, developed by friction at each cable
band is then
w s d
a . C F =
2
95 0 (4.5.2-5)
where
F = wire force developed at each cable band
4-17
Wedge Line
L
R
d
L
Observed Wires
d
R
Perimeter of Cable
Wires in cable
represented by the
observed wires
Figure 4.3.-1. Counting wires in half-sectors.
Right half sector
Wedge
x = 1
x x = 3
Number of wires in cable
represented by the
observed wire at x
X
X
Center of cable
Circumference
of cable X + 0.5 Wires
= Half cabledia.
Observed wire at x
Left half sector
Adjacent wedge
Line midway
between wedges
Sector
Line midway
between wedges
Figure 4.3.1-1. Counting wires.
4-18
Figure 4.3.1.2-1. Calculation of number of rings and wires in each ring.
4-19
Figure 4.3.2-1. Spreadsheet for reporting corrosion stages.
4-20
Figure 4.3.3.1-1. Spreadsheet for reporting broken wires.
4-21
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0 10 20 30 40 50 60 70 80 90 100
L/L
0
-
1
Figure 4.4.3.2-1. Inverse of standard Normal cumulative distribution.
4-22
Figure 4.5.1-1. Redevelopment of wire tension through friction at the cable bands.
Figure 4.5.2-1. Effective development length.
Wire
Cable Bands
Wire Break
Outside L
e
(Force = 0)
Friction
Force

F F
N
T
= Number of panels required to develop
95% of mean strength of Group 2 wires
F
95% of mean strength
of Group 2 wires is
redeveloped in
evaluated panel when
wire breaks outside L
e
Evaluated
panel
Effective development length, L
e
(symmetrical about evaluated panel)
WIRE BREAKS OUTSIDE EFFECTIVE DEVELOPMENT LENGTH
Wire
p Cable Bands
between wire break
and evaluated panel
Wire Break
Inside L
e
(Force = 0)
Friction
Force

F
Force = pF is
redeveloped in
evaluated panel when
wire breaks inside L
e
Evaluated
panel
F
WIRE BREAKS INSIDE EFFECTIVE DEVELOPMENT LENGTH
F
Number of bands = N
B
= N
T
+ 1
Wire
Wire Tension
=X
T
Wire Tension
=nX

Cable Bands
Wire Break
(Force = 0)
Friction Force

Panel Length, L
X X
Measured
Gap,d
w
Band 1 Band n
Band N
T
Band N
T
+1
Panel 1 Panel n Panel N
T
N
T
panels
Friction force
in last band is
< or = X
X
Wire fully
redeveloped
1 panel
Wire Tension
=N
T
X
5-1
CONTENTS
SECTION 5 ESTIMATION OF CABLE STRENGTH .................................................................................5-1
5.1 INTRODUCTION.......................................................................................................................................5-3
5.2 NOTATION.................................................................................................................................................5-3
5.3 ESTIMATED CABLE STRENGTH.........................................................................................................5-6
5.3.1 General.................................................................................................................................................5-6
5.3.2 Wire Groupings ..................................................................................................................................5-6
5.3.2.1 BROKEN WIRES IN EFFECTIVE DEVELOPMENT LENGTH .................................................5-6
5.3.2.2 REPAIRED WIRES IN EFFECTIVE DEVELOPMENT LENGTH ..............................................5-7
5.3.2.3 UNBROKEN WIRES IN EACH CORROSION STAGE ...............................................................5-7
5.3.2.4 CRACKED WIRES .........................................................................................................................5-8
5.3.2.4.1 Discrete Cracked Wires in Effective Development Length...........................................................5-8
5.3.2.4.2 Redevelopment of Cracked Wires That Fail..................................................................................5-9
5.3.2.5 EFFECTIVE NUMBER OF UNBROKEN WIRES......................................................................5-10
5.3.2.6 FRACTION OF THE CABLE REPRESENTED BY EACH GROUP OF WIRES......................5-11
5.3.3 Strength of Unbroken Wires............................................................................................................5-11
5.3.3.1 SIMPLIFIED STRENGTH MODEL.............................................................................................5-13
5.3.3.1.1 Mean Tensile Strength of Uncracked Wires................................................................................5-13
5.3.3.1.2 Cable Strength Using the Simplified Model................................................................................5-14
5.3.3.2 BRITTLE-WIRE MODEL.............................................................................................................5-14
5.3.3.2.1 Compound Tensile Strength Distribution Curve .........................................................................5-14
5.3.3.2.2 Cable Force at Stress s.................................................................................................................5-15
5.3.3.2.3 Force in Cracked Wires That Break in Adjacent Panels...............................................................5-15
5.3.3.2.4 Strength of Unbroken Wires in the Cable....................................................................................5-16
5.3.3.3 LIMITED DUCTILITY MODEL..................................................................................................5-16
5.3.3.3.1 Compound Ultimate Strain Distribution Curve...........................................................................5-17
5.3.3.3.2 Force in Unbroken Wires at Strain e ...........................................................................................5-17
5.3.3.3.3 Force in Cracked Wires That Break in Adjacent Panels..............................................................5-17
5.3.3.3.4 Strength of Unbroken Wires in the Cable....................................................................................5-18
5.3.4 Redevelopment of Broken Wires.....................................................................................................5-18
5.3.5 Cable Strength ..................................................................................................................................5-19
5-2
FIGURES

Figure 5.3.2.4.1-1. Fraction of discrete cracked wires in Stage k...............................................................5-21
Figure 5.3.2.4.2-1. Fraction of cracked wires in Stage k that are redeveloped...........................................5-22
Figure 5.3.2.6-1. Summary of calculations for number of wires in each group. .........................................5-23
Figure 5.3.3.1.2-1. Strength reduction factor, K..........................................................................................5-24
5-3
5.1 INTRODUCTION
The Equations for estimating the current strength of the cable are presented in this section. Three strength models
of increasing complexity are included, along with a graph that aids in the calculation of the simplest model.
The section also includes information about separating the wires into groups that are in a similar state of
deterioration, or that contain cracks or are broken. Grouping wires is required for applying the strength models.
The strength is calculated at a specific inspected location along the cable, called the evaluated panel. The concept
of effective development length, covered previously in Article 4.5.2, is crucial for counting the number of broken
and cracked wires and calculating their contribution to cable strength.
5.2 NOTATION
a
w
= nominal area of one wire used in lab analysis (5.3.3.1.2) (5.3.3.2.2) (5.3.3.3.2) (5.3.7.3.3) (5.3.4)
C
d
= fraction of 95% of the mean tensile strength of Group 2 wires that is developed in a broken
wire at each cable band (5.3.4)
C
di
= wire redevelopment factor for Panel i = C
d
multiplied by the number of cable bands between Panel i
and the evaluated panel (5.3.2.4.2)
e = strain in the unbroken wires of the cable (5.3.3.3.1)
F3
5
(e) = fraction of cracked wires that are broken at strain e = Weibull cumulative distribution of ultimate
strain of Group 5 (cracked) wires at strain e (5.3.3.3.3)
F3
5
(s) = fraction of cracked wires that are broken at stress s = Weibull cumulative distribution of tensile
strength of Group 5 (cracked) wires at stress s (5.3.3.2.3)
F3
k
(e) = Weibull cumulative distribution of ultimate strain of Group k wires (5.3.3.3.1)
F3
k
(s) = Weibull cumulative distribution of tensile strength of Group k wires (5.3.3.2.1) (5.3.3.2.2)
F
c
(e) = compound cumulative distribution of ultimate strain (5.3.3.3.1) (5.3.3.3.2)
F
c
(s) = compound cumulative distribution of tensile strength (5.3.3.2.1) (5.3.3.2.2)
i = number of a panel (5.3.2.1) (5.3.2.4.1) (5.3.2.4.2) (5.3.4)

K = reduction factor (given in Figure 5.3.3.1.2-1 as a function of the coefficient of variation,
s
/
s
)
(5.3.3.1.2)
k = corrosion stage of wires (k = 1, 2, 3 and 4) (5.3.2.3) (5.3.2.4.1)
k = corrosion stage of a group of wires (k = 2, 3, 4 and 5) (5.3.2.5) (5.3.2.6) (5.3.3.2.1) (5.3.3.3.1);
(k=2,3,&4) (5.3.3.1.1)
K = reduction factor (given in Figure 5.3.3.1.2-1 as a function of the coefficient of variation,
s
/
s
)
(5.3.3.1.2)
L
e
= number of panels in effective development length (5.3.2.1) (5.3.2.4.1) (5.3.2.4.2) (5.3.4)
5-4
max = maximum value of the expression inside the brackets (5.3.3.2.4) (5.3.3.3.4)
N
0k
= number of unbroken Stage k wires in the evaluated panel (5.3.2.3) (5.3.2.4.1) (5.3.2.4.2) (5.3.2.5)
N
5
= number of discrete cracked wires in the effective development length (5.3.2.5) (5.3.3.1.1) (5.3.3.1.2)
N
b
= number of broken wires in the effective development length (5.3.2.1) (5.3.2.3)
n
b1
= number of broken wires in the evaluated panel (5.3.2.1) (5.3.4)
n
bi
= number of broken wires in panel i (5.3.2.1) (5.3.4)
N
c,k
= total number of discrete cracked wires in the effective development length that are Stage k in the
evaluated panel (5.3.2.4.1) (5.3.2.5)
N
cr
= effective number of redeveloped cracked wires in the effective development length (5.3.2.4.2)
(5.3.3.2.3) (5.3.3.3.3)
N
cr,k =
effective number of broken cracked wires that are Stage k in the evaluated panel and can be
redeveloped (5.3.2.4.2)
N
eff
= effective number of unbroken wires in the evaluated panel (5.3.2.5) (5.3.2.6) (5.3.3.1.1) (5.3.3.1.2)
(5.3.3.2.2) (5.3.3.3.2)
N
k
= number of Group k wires in the evaluated panel (5.3.2.5) (5.3.2.6) (5.3.3.1.1)
N
r
= number of broken wires that are repaired in the effective development length (5.3.2.2) (5.3.2.3)
n
r1
= number of broken wires that are repaired in the evaluated panel (i=1) (5.3.2.2)
n
ri
= number of broken wires in panel i that are repaired (5.3.2.2) (5.3.4)
N
sk
= number of Stage k wires in the evaluated panel (5.3.2.3)
p
i
= number of cable bands between the evaluated panel and a wire break in panel i (5.3.4)
p
c,k
= fraction of Stage k wires that are cracked (5.3.2.4.1) (5.3.2.4.2)
p
k
= fraction of unbroken wires in the evaluated panel represented by Group k (5.3.2.6) (5.3.3.2.1)
(5.3.3.3.1)
p
uk
= fraction of unbroken and uncracked wires in the cable represented by Group k (5.3.3.1.1)
R = estimated cable strength (5.3.3.1.2) (5.3.5)
R
b
= cable strength attributable to broken wires in adjacent panels (5.3.4) (5.3.5)
R
u
= cable strength attributable to unbroken wires (5.3.3.2.4) (5.3.3.3.4) (5.3.5)
5-5
s = stress in unbroken wires of the cable (5.3.3.2.1) (5.3.3.2.2); stress in wires corresponding to the
estimated cable strength calculated in Article 5.3.3 (5.3.4)
s(e) = stress in wires determined from the average stress-strain curve for all wires at strain e (5.3.3.3.2)
s
d
= redeveloped stress in the evaluated panel for a broken wire in an adjacent panel (5.3.4)
T
cr
(e) = maximum force in broken cracked wires that can be redeveloped in the evaluated panel at strain e
(5.3.3.3.3)
T
cr
(s) = maximum force in broken cracked wires that can be redeveloped in the evaluated panel at stress s
(5.3.3.2.3)
T
u
(e) = force in unbroken wires in the evaluated panel at strain e (5.3.3.3.2)
T
u
(s) = force in unbroken wires in the evaluated panel at stress s (5.3.3.2.2)
s
= sample mean tensile strength of the combined groups of wires excluding cracked wires (5.3.3.1.1)
s2
= sample mean tensile strength of Group 2 wires (5.3.3.2.3) (5.3.3.3.3) (5.3.4)
sk
= sample mean tensile strength of Group k wires (5.3.3.1.1)
= sample standard deviation of tensile strength of the combined groups of wires excluding cracked
wires (5.3.3.1.1)
= sample standard deviation of the tensile strength of Group k wires (5.3.3.1.1)
sk
s
5-6

5.3 ESTIMATED CABLE STRENGTH
5.3.1 General
The strength of a cable at the evaluated panel is the
sum of the strengths of wires in three categories:
all wires in the evaluated panel minus broken
wires in that panel and nearby panels
wires that are cracked in nearby panels,
affecting the strength of the same wires in the
evaluated panel
wires that are broken in nearby panels,
affecting the strength of the same wires in the
evaluated panel
Methods for evaluating each of the three categories
are described below in Articles 5.3.3 and 5.3.4.

5.3.2 Wire Groupings
The wires are assigned to groups that are numbered 2
to 5, corresponding to the corrosion stages they derive
from. Stage 1 and Stage 2 wires are added together to
form Group 2, because their properties are virtually
identical. Stage 3 and Stage 4 wires become Group 3
and Group 4 respectively. All the discrete cracked
wires are subtracted from their corresponding groups
and added together to form Group 5. The number of
discrete cracked wires in the effective development
length is N
5
. Broken wires are treated separately.

5.3.2.1 BROKEN WIRES IN EFFECTIVE
DEVELOPMENT LENGTH
The total number of broken wires in the effective
development length is
=
=
e
i
bi b
L
n N
1
(5.3.2.1-1)
where
N
b
= number of broken wires in the effective
development length
n
bi
= number of broken wires in panel i
L
e
= number of panels in the effective
development length (see Article 4.5.2)
i = number of a panel
If the evaluated panel is the only panel that has been
C5.3.2.1
Wires that are broken in panels other than the panel
being evaluated may affect the cable strength in the
evaluated panel. The force that such a wire can sustain
increases as the wire passes through one or more cable
bands, until the wire is fully redeveloped. The increase
in strength comes from the friction among the wires,
caused by the clamping action of the cable bands and
the cable wrapping.

5-7
inspected, then it is assumed that all panels in the
effective development length are alike, and that all
values of n
bi
are equal to n
b1
, and that
1 b e b
n L N = (5.3.2.1-2)
where
n
b1
= number of broken wires in the evaluated
panel

5.3.2.2 REPAIRED WIRES IN EFFECTIVE
DEVELOPMENT LENGTH
The total number of repaired wires in the effective
development length is
=
=
e
i
ri r
L
n N
1
(5.3.2.2-1)
where
N
r
= number of broken wires that are repaired
in the effective development length
n
ri
= number of broken wires that are repaired in
panel i
inspected, then broken wires in that panel are the only
ones to have been repaired,
1 r r
n N = (5.3.2.2-2)
where
n
r1
= number of broken wires that are repaired in
the evaluated panel (i=1)

C5.3.2.2
inspected, numbering the order of the panels in the
effective development length is not critical.
If adjacent panels have been inspected, and if the
methods specified in Article 5.3.4 and Appendix B are
used to evaluate the cable, then the panels should be
numbered according to the instructions in Appendix
B, Article B.4.1.

5.3.2.3 UNBROKEN WIRES IN EACH
CORROSION STAGE
The number of unbroken wires in each stage of
corrosion is determined by subtracting the unrepaired
broken wires in the effective development length from
Stage 4 wires in the evaluated panel, and when there
are none remaining, from Stage 3 wires in the panel, as
follows:
when
4 s r b
N N N
r b s
N N N N + =
4 04
(5.3.2.3-1)
3 03 s
N N = (5.3.2.3-2)
5-8
1 2 02 s s
N N N + = (5.3.2.3-3)
when
4 s r b
N N N >
0
04
= N (5.3.2.3-4)
r b s s
N N N N N + + =
4 3 03
(5.3.2.3-5)
1 2 02 s s
N N N + = (5.3.2.3-6)
where
N
0k
= number of unbroken Stage k wires in the
evaluated panel
N
sk
= number of Stage k wires in the evaluated
panel (Article 4.3.2)
N
b
= number of broken wires in the effective
development length
N
r
= number of broken wires that are repaired
in the effective development length
k = corrosion stage of wires (k =1, 2, 3 and 4)

5.3.2.4 CRACKED WIRES
The formulas in the following articles apply to
situations in which only the evaluated panel has been
inspected, and are based on the conservative
assumption that all panels in the effective development
length are in the same condition as the evaluated panel.
Of all the cracked wires in the cable, these calculations
are applied to discrete cracked wires only. A discrete
cracked wire is a wire that is cracked in panel i but is
not cracked in all the panels nearer than i to the
evaluated panel.
The effective number of discrete cracked wires that are
assumed to be redeveloped in the evaluated panel due
to friction at the cable bands is also required for the
calculation.

C5.3.2.4
If all panels in the effective development length have
been inspected, the technique presented in Appendix
B should be used. The technique is limited in its
application because it is complex and the data are
usually not available. It is recommended for severely
deteriorated panels, in which case additional adjacent
panels must be opened to obtain these data.
5.3.2.4.1 Discrete Cracked Wires in Effective
Development Length
The number of discrete wires in the effective
development length is calculated separately for each
corrosion stage. The number of discrete cracked wires
in any stage, k, is given by the equation

5-9
( )
1
1
, , 0 ,
1
=

=
i
L
i
k c k c k k c
e
p p N N (5.3.2.4.1-1)
where
N
c,k
= total number of discrete cracked wires in
the effective development length that are
Stage k in the evaluated panel
N
0k
evaluated panel
p
c,k
L
e
= number of panels in the effective
development length
k = corrosion stage of wires (k = 1,2,3 and 4)
As stated above, this calculation is made separately for
each stage. Usually, p
c,1
and p
c,2
will be zero; and p
c,3

may be zero.
The value of the expression
( )
1
1
, ,
1
=

i
L
i
k c k c
e
p p (5.3.2.4.1-2)

5.3.2.4.2 Redevelopment of Cracked Wires That Fail
Cracked wires that are assumed to fail as the cable
stress is increased may redevelop part of their strength
in the evaluated panel. Assuming all of the cracked
wires are broken, redeveloped wires for each stage, k,
are
( )

=
=
e
L
i
di
i
k , c k , c k k , cr
C p p N N
1
1
0
1 (5.3.2.4.2-1)
where
N
cr,k
= effective number of broken cracked wires
that are Stage k in the evaluated panel and
can be redeveloped
C
di
= wire redevelopment factor for Panel i =

in Equation 1 represents the fraction of discrete
cracked wires in each stage in the effective development
length. Values of this expression as a function of the
fraction of cracked wires in each stage are shown
graphically in Figure 5.3.2.4.1-1. The expression is
called N
c,k
/N
0k
.
5-10
C
d
multiplied by the number of cable
bands between Panel i and the evaluated
panel
The total effective number of redeveloped wires is
4 3 2 , cr , cr , cr cr
N N N N + + = (5.3.2.4.2-2)
where
N
cr
= effective number of redeveloped cracked
wires in the effective development length
The value of the expression
( )

e
L
i
di
i
k c k c
C p p
1
1
, ,
1 (5.3.2.4.2-3)

5.3.2.5 EFFECTIVE NUMBER OF UNBROKEN
WIRES
The effective number of unbroken wires in the
cable is
=
=
5
2 k
k eff
N N (5.3.2.5-1)
in which
k c k k
N N N
, 0
= (k = 2, 3, 4) (5.3.2.5-2)
and

=
=
4
2
, 5
k
k c
N N (5.3.2.5-3)
where
N
eff
= effective number of unbroken wires in the
evaluated panel
N
k
= number of Group k wires in the
evaluated panel
N
5
= number of discrete cracked wires in the
effective development length

in Equation 1 represents the effective fraction of
cracked wires that are Stage k in the evaluated panel
but are redeveloped because they are broken at stress, s.
Figure 5.3.2.4.2-1 gives the effective fraction of cracked
wires that will redevelop in the evaluated panel if they
break, to be used in Equation 1. This expression is
called N
cr,k
/N
0k
. Each stage, k, is treated separately, and
then combined using Equation 2.
5-11
N
0k
evaluated panel
N
c,k
= total number of discrete cracked wires in
the effective development length that are
Stage k in the evaluated panel
k = corrosion stage of group of wires (k = 2, 3,
4 and 5)

5.3.2.6 FRACTION OF THE CABLE
REPRESENTED BY EACH GROUP OF
WIRES
The unbroken wires in the cable are separated into four
groups for the purpose of calculating cable strength.
Each group has a different set of tensile strengths
and/or ultimate strain properties. The values of N
k
are
used to calculate the fraction of the cable represented
by each group of wires, k,
eff
k
k
N
N
p = (5.3.2.6-1)
where
p
k
= fraction of unbroken wires in the
evaluated panel represented by Group k
N
k
= number of Group k wires in the evaluated
panel
N
eff
= effective number of unbroken wires in
the evaluated panel
(k =2,3,4 and 5)

C5.3.2.6
5.3.3 Strength of Unbroken Wires
The strength of the unbroken wires in the cable should
be estimated using one of three strength models. The
Limited Ductility Model is the most rigorous; the
others are special cases of this model, using simplified
assumptions. All three recommended models are
further described in Appendix A.
The choice of model depends on the extent of the
deterioration found in the cable:

Use the Simplified Model for cables with no
Stage 4 or cracked wires
C5.3.3
The Limited Ductility and Brittle-Wire models are
used to estimate the strength of a cable composed of
many wires that are subjected to the same strain. The
Simplified Model, which is based on the Brittle-Wire
Model, subtracts all cracked and broken wires and
uses a single distribution curve for the tensile strength
of the remaining unbroken, uncracked wires.
In the Limited Ductility Model, the cable is subjected
to an incremental increase in strain The force in the
cable at any strain is the sum of the forces in the
individual wires at that strain. The wire forces vary in
relation to the individual stress-strain diagrams.
The calculation of the number of wires in each group
is difficult to visualize and is summarized in Figure
5.3.2.6-1.
5-12
Use the Simplified Model (at the discretion of
the owner or investigator) for cables in which
cracked Stage 3 and Stage 4 wires account for
up to 10% of the total wires. It is understood
that the result in cable strength may be up to
10% less than the result using the Brittle Wire
Model.
Use the Brittle-Wire Model if cracks are
present in more than 10% of all the wires in
the cable.
Use the Limited Ductility Model if the wires
display unusual variations in tensile strength
(sometimes due to varying carbon content),
which are reflected in stress-strain curves that
are also unusually varied.
Individual wires share in carrying the cable tension
until one reaches its ultimate strain, at which point it
breaks and no longer resists any force. The total force
in the cable is reduced accordingly by the previous
force carried by the wire. Increasing the strain further
causes the cable force to increase again, until the next
wire breaks. The process continues until the cable
force reaches a maximum value, after which wires
break more rapidly than the force increases in the
individual wires, resulting in decreased cable force
with increased strain. The maximum cable force
achieved is the cable strength.
The Brittle-Wire Model is a special case of the
Limited Ductility Model. In contrast to its parent
model, it is assumed that all the wires are subjected to
the same tensile stress at any given strain; thus, it is
convenient to substitute an increasing stress in the
calculation instead of an increasing strain. Any
individual wire shares in carrying the tension in the
cable until the stress in that wire exceeds its tensile
strength and the wire fails, no longer participating in
the cable force. As with the more general model, the
cable force increases with increasing stress until a
maximum value is reached, which is equivalent to the
cable strength.
The Limited Ductility Model requires determining the
ultimate strain of each wire specimen and developing
a full stress-strain diagram for each wire sample. The
ultimate strain corresponds to the tensile strength of
the wire (see Appendix A). It is also the strain at
failure when there is no reduction of area, for instance
from a crack.
The percentage of elongation in a 10-inch gage length,
determined in accordance with ASTM A370, cannot
be used as the ultimate strain for the Limited Ductility
Model, because it is measured only after necking
down and does not include the elastic component of
the strain.
The Brittle-Wire Model requires knowing only the
tensile strength of each specimen, which can be
obtained by testing in accordance with ASTM A370.
If a cable force vs. strain diagram is wanted by the
investigator, then average stress-strain diagrams of the
cable wires should be developed for each of the
models. Several such diagrams are already needed for
the Limited Ductility Model, whereas the Brittle-Wire
Model requires only one.
5-13

5.3.3.1 SIMPLIFIED STRENGTH MODEL
The Simplified Model should be applied to cables that
have very few cracked wires. The upper limit is no
more than 10% of the total wire population.. The
Brittle-Wire Model is used whenever this limit is
exceeded. The Simplified Model is based on the
Brittle-Wire Model; the estimated number of cracked
wires (Group 5) and broken wires are omitted from the
calculation, and the total number of wires in the cable
is reduced accordingly.
Although the strength may be underestimated by up to
20%, the Simplified Model is useful in locating the
most severely deteriorated panel among those
inspected. Then the more complex models can be
applied to that panel for a more realistic strength
estimate.
C5.3.3.1
A single Weibull distribution that combines Groups 2,
3 and 4 is used in the Simplified Model. It combines
the tensile strength distributions of the individual wire
groups, with the relative size of each group taken into
account.
In order to minimize the computational effort required
by the model, a factor is applied to the mean tensile
strength, which is multiplied by the effective cable
area.

5.3.3.1.1 Mean Tensile Strength of Uncracked Wires
The fraction of the cable represented by Groups 2, 3
and 4 is combined with the sample mean values of
minimum tensile strength of the representative
specimens of each group to determine the sample mean
tensile strength and standard deviation of the entire
unbroken and uncracked wire population, using the
equations

=
=
4
2
) (
k
sk uk s
p (5.3.3.1.1-1)
( )
2 2 2
4
2
s sk sk
k
uk s
p

+ =
=
(5.3.3.1.1-2)
in which
5
N N
N
p
eff
k
uk

= (5.3.3.1.1-3)
where
s
= sample mean tensile strength of the
combined groups of wires, excluding
cracked wires
sk
= sample mean tensile strength of Group k
wires
s
= sample standard deviation of the tensile
strength of the combined groups of wires,
C5.3.3.1.1
The symbols and refer to the mean and standard
deviation of a property of the entire population of
wires in the cable. The mean and standard deviation
used in cable strength models are determined from
laboratory tests on a selection of wires removed from
the cable during inspection and are called the sample
mean and sample standard deviation, designated by
s

and
s
.

5-14
excluding cracked wires
sk
= sample standard deviation of the tensile
strength of Group k wires
p
uk
= fraction of unbroken and uncracked
wires in the cable represented by Group k
(k = 2, 3 and 4)
N
eff
evaluated panel
N
5
= number of discrete cracked wires in the
effective development length
N
k
= number of Group k wires in the evaluated
panel
5.3.3.1.2 Cable Strength Using the Simplified Model
The cable strength is calculated from the equation
K a N N R
s w eff
= ) (
5
(5.3.3.1.2-1)
in which
where
R = estimated cable strength
a
w
= nominal area of one wire used in lab
analysis
C5.3.3.1.2
The strength reduction factor, K, is the brittle-wire
strength of the combined groups of uncracked wires,
divided by the product of mean tensile strength of the
combined groups of uncracked wires and total area of
uncracked wires. The derivation of K is given in
Appendix A.

5.3.3.2 BRITTLE-WIRE MODEL
The tensile strength of each test specimen is used to
determine the minimum tensile strength of each wire
sample. The minima are used to determine the sample
means and standard deviations for each group of wires,
which are combined to construct a compound tensile
strength distribution curve.
Whenever a cable force vs. strain diagram is required,
the test laboratory determines the stress-strain curve up
to the ultimate strain for at least one specimen from
each sample wire. These curves are used to develop an
average stress-strain curve for the entire cable.
C5.3.3.2
The Brittle-Wire Model is recommended with few
exceptions for determining the strength of the cable.
The distribution of the tensile strength of the entire
population of unbroken wires is a compound
distribution curve developed from the distributions of
the individual wire groups, with the relative size of
each group taken into account. The Weibull
distribution is used in the analysis, with the lower
limit of tensile strength, s
0
, assumed to be zero (no
wire can have a negative tensile strength).
5.3.3.2.1 Compound Tensile Strength Distribution
Curve
The fraction of the cable represented by each of the
groups (calculated in Article 5.3.2.6) and the Weibull
distribution curves for tensile strength of the specimens
C5.3.3.2.1
The Weibull distribution is a Type 3 extreme value
distribution function. The function extends from a
minimum value, x
0
, to infinity. The cumulative
distribution function is referred to as F3(x) in the
K = reduction factor (given in Figure 5.3.3.1.2-1
as a function of the coefficient of variation,

s
/
s
)
5-15
that represent each of the groups are combined to
determine the compound distribution curve for the
entire unbroken wire population. The equations used in
the calculation are
) ( 3 ) (
5
2
s F p s F
k
k
k C
=
=
(5.3.3.2.1-1)
where
F
C
(s) = compound cumulative distribution of the
tensile strength
s = stress in unbroken wires of the cable
p
k
= fraction of unbroken wires in the evaluated
panel represented by Group k
(k = 2, 3, 4 and 5)
F3
k
(s) = Weibull cumulative distribution of tensile
strength of Group k wires

Guidelines. The equations for this function, as well as
a method for determining the parameters of the
function, are presented in Appendix A. The term x is
the variable of the distribution and is replaced by s for
tensile strength and by e for ultimate strain.
5.3.3.2.2 Cable Force at Stress s
The total force in the unbroken wires at any value of
cable stress, s, is given by
( ) ( ) ( ) s F s a N s T
C w eff u
= 1 ) ( (5.3.3.2.2-1)
where
T
u
(s) = force in the unbroken wires in the
evaluated panel at stress s
N
eff
evaluated panel
a
w
analysis

C5.3.3.2.2
In Equation 1, the wires with tensile strength less than
s are assumed to have zero force. This is accomplished
mathematically, using the Survivor Function, (1-
F
C
(s)).
5.3.3.2.3 Force in Cracked Wires That Break in
Adjacent Panels
The total force in the evaluated panel at cable stress, s,
in wires that are cracked in other panels within the
effective development length and that have a tensile
strength less than that stress is given by
( ) ) ( 3 95 . 0 ) (
5 2
s F a N s T
s w cr cr
= (5.3.3.2.3-1)
where
T
cr
(s) = maximum force in the broken cracked
wires that can be redeveloped in the
evaluated panel at stress s

5-16
F3
5
(s) = fraction of cracked wires that are broken at
stress s = Weibull cumulative distribution
of tensile strength of Group 5 (cracked)
wires at stress s
N
cr
= effective number of redeveloped cracked
wires in the effective development length
2 s
wires

5.3.3.2.4 Strength of Unbroken Wires in the Cable
The strength of the unbroken wires in the cable is
given by

( ) ( ) ( ) s T s T R
cr u u
+ = max (5.3.3.2.4-1)
where
R
u
= cable strength attributable to unbroken
wires
max = maximum value of the expression inside
the brackets

C5.3.3.2.4
( ) ) ( ) ( max s T s T R
cr u u
= is determined by
calculating the values of the expression for several
values of s at suitably small increments (2 ksi) and
seeking the maximum value. The entire calculation is
best done on a computer spreadsheet program that
incorporates the Weibull distribution functions.
Alternatively, the expression can be evaluated for a
single value of s, and an iterative program can then be
used to determine the value of s that produces the
maximum expression. Should a cable force vs. strain
diagram be required, only the first technique provides
the data to plot it.

5.3.3.3 LIMITED DUCTILITY MODEL
The Limited Ductility Model requires the ultimate
strain of each test specimen based on laboratory results,
as well as a stress-strain curve up to the ultimate strain
for each specimen tested. The data are used to
determine the minimum value of the ultimate strain and
the stress-strain curve of each sample wire. Whenever
the stress-strain curves for the sample wires are
essentially the same, an average curve for all the
samples is constructed, and the method given below of
estimating the cable strength is followed. Whenever the
individual stress-strain curves for the sample wires
vary from one another, the general form of the Limited
Ductility Model, given in Appendix A, must be used.
C5.3.3.3
In the Limited Ductility Model, the ultimate strain of
the wires is used as the variable in the distribution
functions. In the simple form of this model, presented
in the articles below, the distribution of the ultimate
strain for the entire population of unbroken wires is a
compound distribution curve developed from the
distributions of the individual wire groups, with the
relative size of each group taken into account. The
Weibull distribution is used in the analysis, with the
lower limit of ultimate strain, e
0
, assumed to be zero
(no wire can have a negative ultimate strain).
5-17
5.3.3.3.1 Compound Ultimate Strain Distribution
Curve
The fraction of the cable represented by each of the
groups, (calculated in Article 5.3.2.6), and the Weibull
distribution curves for ultimate strain of the specimens
that represent each of the groups, are combined to
determine the compound distribution curve for the
entire unbroken wire population. The equation used for
the calculation is
) ( 3 ) (
5
2
e F p e F
k
k
k C
=

=
(5.3.3.3.1-1)
where
F
C
(e) = compound cumulative distribution of the
ultimate strain
e = strain in the unbroken wires of the cable
p
k
= fraction of the unbroken wires in the
evaluated panel represented by Group k
(k = 2, 3, 4 and 5)
F3
k
(e) = Weibull cumulative distribution of
ultimate strain of Group k wires

C5.3.3.3.1

The compound distribution curve is the sum of several
independent Weibull distributions but is not itself a
Weibull distribution. While there are no explicit
equations for the distribution, it can be described by
Equation 1. The compound distribution curve is used
only when all the groups of wires have the same
stress-strain curve and the cable force in unbroken
wires is calculated by Equation 5.3.3.3.2-1.
5.3.3.3.2 Force in Unbroken Wires at Strain e
If it can be shown that the average stress-strain curves
for all the groups of wires are alike, then the following
equation is used to estimate the force in unbroken wires
in the cable at strain e:
( ) ( ) ( ) ( ) e F e s a N e T
C w eff u
= 1 ) ( (5.3.3.3.2-1)
where
s(e) = stress in wires determined from the
average stress-strain curve for all wires at
strain e
T
u
(e) = force in unbroken wires in the evaluated
panel in strain e
N
eff
evaluated panel
a
w
= nominal area of one wire, used in lab
analysis

5.3.3.3.3 Force in Cracked Wires That Break in
Adjacent Panels

5-18
The total force in the evaluated panel at cable strain, e,
in wires that are cracked in other panels in the effective
development length and that have an ultimate strain
smaller than e is given by
( ) ( ) ( ) e F a N e T
s w cr cr 5 2
3 95 . 0 = (5.3.3.3.3-1)
in which
T
cr
(e) = maximum force in broken cracked
wires that can be redeveloped in the
evaluated panel at strain e
F3
5
(e) = fraction of cracked wires that are
broken at strain e = Weibull cumulative
distribution of ultimate strain of
Group 5 (cracked) wires at strain e
N
cr
= effective number of redeveloped
cracked wires in the effective
development length
s2
= mean sample tensile strength of Group 2
wires
5.3.3.3.4 Strength of Unbroken Wires in the Cable
The strength of the unbroken wires in the cable is given
by

( ) ( ) ( ) e T e T R
cr u u
+ = max (5.3.3.3.4-1)
where
max = maximum value of the expression inside
the brackets
R
u
wires
C5.3.3.3.4
The value of the term
( ) ( ) ( ) e T e T
cr u
+ max
is determined by calculating the values of the
expression for several values of e at suitably small
increments (0.001 inch/inch) and seeking the
maximum value. The entire calculation is best done on
a computer spreadsheet program that incorporates the
Weibull distribution functions.
5.3.4 Redevelopment of Broken Wires
Wires that are broken in panels adjacent to the
evaluated panel share in the cable tension because of
the friction that develops at the intervening cable
bands. If the tension in a wire exceeds the friction in
the cable band adjacent to that panel, the wire will slip,
but it will continue to carry a constant tension as the
cable tension increases. The stress redeveloped in a
broken wire is given by
( )
2
95 0
s d i d
. C p s = , s s
d
(5.3.4-1)
where
s
d
= redeveloped stress in the evaluated panel
C5.3.4
The stress in a broken wire that can be redeveloped is
the number of cable bands between the break and the
panel being evaluated multiplied by the stress
redeveloped at each band. This redeveloped stress is,
however, limited to the stress in the unbroken wires
that corresponds to the cable strength calculated by the
equations in Article 5.3.3. Generally, the stress in the
unbroken wires will be more than 90% of the mean
tensile strength of the Group 2 wires and the limitation
can be disregarded, because the redeveloped stress
with L
e
= 9 is not greater than 76% of the mean tensile
strength of Group 2.
5-19
for a broken wire in an adjacent panel
p
i
= number of cable bands between the
evaluated panel and a wire break in panel i
C
d
= the fraction of 95% of the mean tensile
strength of Group 2 wires that is developed
in a broken wire at each cable band
s2
wires
s = stress in the wires corresponding to the
estimated cable strength calculated in
Article 5.3.3
The contribution of the broken wires to the cable
strength is given by
( )
d ri bi
L
e
i
i s w b
C n n p a R =
=
) ( 95 . 0
2
2
(5.3.4-2)
where
R
b
= cable strength attributable to broken wires
in adjacent panels
a
w
analysis
L
e
= number of panels in effective development
length
n
bi
= number of broken wires in panel i
n
ri
= number of broken wires in panel i that
are repaired
inspected, then the contribution of the broken wires to
the cable strength can be taken as
( ) ( ) 1 5 . 0 95 . 0
1 2
=
e b s w b
L n a R (5.3.4-3)
where
n
b1
= number of broken wires in evaluated panel

5.3.5 Cable Strength
The strength of the cable using either the Brittle-Wire
Model or the Limited Ductility Model in the evaluated
panel is given by
5-20
b u
R R R + = (5.3.5-1)
R = estimated cable strength
R
u
wires
R
b
= cable strength attributable to broken wires
in adjacent panels
5-21
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
Fraction of wires in Group k that contain cracks, p
c,k
F
r
a
c
t
i
o
n

o
f

d
i
s
c
r
e
t
e

w
i
r
e
s

i
n

G
r
o
u
p

k

t
h
a
t

a
r
e

c
r
a
c
k
e
d

i
n

e
f
f
e
c
t
i
v
e

d
e
v
e
l
o
p
m
e
n
t

l
e
n
g
t
h
,

N
c
,
k
/
N
0
k
5
9
7
3 panels = effective development length
Figure 5.3.2.4.1-1. Fraction of discrete cracked wires in Stage k.
5-22
0.00
0.10
0.20
0.30
0.40
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
Percent of wires that contain cracks, p
c,k
F
r
a
c
t
i
o
n

o
f

c
r
a
c
k
e
d

w
i
r
e
s

t
h
a
t

h
a
v
e

f
a
i
l
e
d

i
n

e
f
f

d
e
v

l
g
t
h
t
h
a
t

a
r
e

r
e
d
e
v
e
l
o
p
e
d

i
n

P
a
n
e
l

1
,

N
c
r
,
k
/
(
N
0
k
)
5
9
7
3 panels = effective development length
Figure 5.3.2.4.2-1. Fraction of cracked wires in Stage k that are redeveloped.
5-23
Figure 5.3.2.6-1. Summary of calculations for number of wires in each group.
Type of wires
Eq 5.3.2.6-1
5-24
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Coefficient of Variation
(standard deviation / mean strength)
K
Figure 5.3.3.1.2-1. Strength reduction factor, K.
6-1
CONTENTS
SECTION 6 INSPECTION REPORTS..................................................................................................................6-1
6.1 INTRODUCTION.......................................................................................................................................6-2
6.2 MAINTENANCE PERSONNEL INSPECTION.....................................................................................6-2
6.2.1 Follow-up report .................................................................................................................................6-2
6.3 BIENNIAL INSPECTION .........................................................................................................................6-2
6.4 INTERNAL INSPECTION........................................................................................................................6-3

6-2
6.1 INTRODUCTION
A report is required for each inspection performed.
Over time, the reports form a historical record of the
condition of the cable and help the bridge owner make
informed decisions about maintenance schedules and
budgets.
6.2 MAINTENANCE PERSONNEL
INSPECTION
A report shall be prepared for each periodic inspection
preformed by maintenance personnel. This report shall
include at minimum:
1. date of inspection
2. weather and temperature
3. cable portion inspected (e.g., west main span,
south anchorage, tower saddles and cable housings)
4. list of deficiencies, identified by panel number
5. one-page account of each deficiency with
verbal description (e.g., peeling paint, rust
stains, broken wrapping)
color photograph
recommended action
6. list of recommended actions in order of priority
6.2.1 Follow-up report
A follow-up report shall be prepared for each action
taken, with a description of the action and a photograph
of the completed work.
6.3 BIENNIAL INSPECTION
The basic report for a biennial inspection is described
in specifications provided by the state departments of
transportation; they are not repeated here. The report
should also contain the following information about the
cables and suspension system:
1. separate listings of the ratings applied to each
component (e.g., wrapping, hand ropes, etc.) in
each inspected panel (see Figures 2.2.3.1-2 and
2.2.3.1-3)
2. photographs of deficiencies
3. reasons for ratings lower than 5 (on a scale of 1 to
7)
6-3
4. recommendations for action
5. reasons for recommending an internal inspection, if
applicable

6.4 INTERNAL INSPECTION
The report for an internal inspection shall include all of
the following:
1. executive summary providing a brief synopsis of
the findings of the inspection that incorporates
number of locations opened for inspection
general description of conditions found (e.g.,
severe corrosion with 15 broken wires)
strength of each panel investigated
safety factor of each panel investigated
safety factor using the panel with lowest
strength and maximum cable tension (usually
adjacent to the tower)
recommendations for remedial action
recommendation for date of next inspection
2. table of contents
3. summary addressing executive summary items in
greater detail
4. findings from preliminary cable walk and reasons
for selecting investigated panels
5. plan and elevation of cables showing location of
panels investigated
6. description and photographs of the means of access
to the cable
7. detailed descriptions of each panel opened; cable
cross-sections showing wedge locations,
distribution of stages of corrosion and location of
broken wires
8. summary of laboratory test results; cable cross-
sections showing locations of sample wires
9. verbal description of method used to calculate
cable strength; table of calculated strengths
10. table of cable tensions due to dead load, live load
and temperature; table of cable safety factors
11. investigators estimate of the accuracy of estimated
cable strength
C6.4
The safety factor for an inspected panel is the cable
strength in that panel divided by the cable tension in
that panel, due to dead load, live load, and temperature
effects.
The safety factor of the cable when it is inspected in its
entirety is the lowest value found above.
The safety factor of the cable when only a portion of it
has been inspected is determined by using the
minimum cable strength and the maximum cable
tension, which occurs in the panel with the maximum
slope, usually adjacent to the tower.
The probable error in the strength of a specific panel
(when adjacent panels have been inspected) is
approximately 7% to 8%. The usual safety factors can
accommodate this level of error.
This level of error is reasonably accurate but it applies
only to inspected panels; therefore, it applies to the
entire cable only when all the panels have been
inspected. In all other cases, the error can be greater.
Historically, cables have been designed for a safety
factor of 4; more recently, safety factors of 2.7 (1939)
and 2.4 (1990) have been used. The safety factor that
results from assessing a deteriorated cable will be
lower than the one used in designing the cable. Safety
factors and live loading are not included in the scope of
this project. Nevertheless, whenever the safety factor
falls below 2.15%, it is recommended that:
reductions in traffic be considered
an acoustic monitoring system be installed to
record future wire breaks
a more intensive inspection, of the entire cable
The additional error comes from lack of knowledge about
the uninspected panels. When the results of a full-length
inspection were compared with the results of a hypothetical
inspection of 20% of the panels of the same cable, the error
in the strength estimate of the hypothetical was 8% (See the
final report for NCHRP Project 10-57, on the accompanying
CD). Whenever fewer panels are inspected, the
overestimation is greater, the inverse is also true.
6-4
cable strength
12. conclusions
discussion of cable strengths, safety factors and
possible errors
discussion of probable causes of deterioration
13. recommendations
plan for continued operation of the bridge if the
safety factor is low
general plan for maintenance and repairs
specific plan for time of next inspection and
number of panels to be inspected (The exact
panels should be selected during the
preparatory period for the next inspection,
unless follow-up inspections of specific panels
are recommended.)
14. appendices
laboratory reports for
wire properties from tests; means and
standard deviations of corrosion groups
Weight or Zinc Coating Test and Preece
Test
chemical testing of metal and corrosion
products
metallurgical examinations, including
photographs
sample strength calculations
selected photographs showing condition of
cable exterior (from the cable walk) and cable
interior (from wedging)
selected photographs of inspection and
rewrapping operations
if possible, be scheduled
application of corrosion inhibitors be
considered
A-1
CONTENTS
APPENDIX A MODELS FOR ESTIMATING CABLE STRENGTH........................................................... A-1
A.1 INTRODUCTION......................................................................................................................................A-2
A.2 NOTATION................................................................................................................................................A-2
A.3 STRENGTH MODELS..............................................................................................................................A-4
A.3.1 Behavior of a Single Bridge Wire.....................................................................................................A-4
A.3.2 Strength Models for Wire Bundles...................................................................................................A-5
A.3.2.1 DUCTILE-WIRE MODEL ............................................................................................................. A-5
A.3.2.2 BRITTLE-WIRE MODELS............................................................................................................ A-5
A.3.2.2.1 Limited Ductility Model ...........................................................................................................A-6
A.3.2.2.2 Brittle-Wire Model (A Special Case) .......................................................................................A-6
A.3.2.3 BRITTLE-DUCTILE MODEL....................................................................................................... A-6
A.3.2.4 SIMPLIFIED MODEL....................................................................................................................A-6
A.4 STATISTICS FOR CABLE STRENGTH ANALYSIS.......................................................................... A-7
A.4.1 Extreme Value Distributions ............................................................................................................A-8
A.4.2 Weibull Distribution..........................................................................................................................A-8
A.5

A.5.1 Limited Ductility Model Equations................................................................................................A-10
A.5.1.1 DIFFERENT STRESS-STRAIN CURVES AVAILABLE FOR WIRES....................................A-11
A.5.1.2 ALL GROUPS OF WIRES HAVE THE SAME STRESS-STRAIN CURVE.............................A-12
A.5.2 Brittle-Wire Model Equations ........................................................................................................A-12
A.5.3 Simplified Model Equations............................................................................................................A-13
A.5.3.1 SINGLE DISTRIBUTION CURVE FOR TENSILE STRENGTH..............................................A-13
A.5.3.2 CABLE STRENGTH USING THE SIMPLIFIED MODEL........................................................A-14
A.6 REFERENCES .....................................................................................................................
...................................................................................................................
....................A-14
EQUATIONS FOR ESTIMATING CABLE STRENGTH USING THE
WEIBULL DISTRIBUITON A-10
A-2
A.1 INTRODUCTION
A parallel wire cable on a suspension bridge can be defined as a bundle of parallel structural elements in tension. A
statistical approach to the problem of estimating the strength of a parallel element system is described in Rao [1] He
presents two models based on different assumptions, either that the elements are ductile (they continue to carry load
after reaching maximum capacity) or that they are brittle (they break when they reach maximum capacity). Engineers
currently depend on several variations of these models to estimate the strength of the unbroken wires in a cable that has
experienced deterioration. In practice, the models that include all the wires in a cable are more complex than Raos
models would suggest. The series of tasks required for a proper assessment includes:
developing stress-strain diagrams for the wires
finding the minimum strength of a given length of wire
calculating distribution functions to describe how the wires vary
determining the effectiveness of the cable bands in redeveloping the strength of a wire broken at some distance
from the point at which the strength is being determined
estimating the effect on the cable of deterioration in panels near the one being evaluated
estimating the strength of the cable in a given panel based on this data
The following pages describe the basic models found in the literature for parallel systems and their application to
estimating cable strength.
A.2 NOTATION
a
w
= nominal area of one wire, used in lab analysis (A.5.1.1) (A.5.1.2) (A.5.2) (A.5.3.2)
e = specific value of strain (A.5.1.1)
exp(x) = e (2.7183) to the power (x), the exponential of (x) (A.4.2)
F3
k
(e) = Weibull cumulative distribution for ultimate strain of Group k wires (A.5.1.1) (A.5.1.2)
F3
k
(s) = Weibull cumulative distribution for tensile strength of Group k wires (A.5.2.)
f3
X1
(x) = Type III extreme value probability density distribution for the smallest values of random variable X
(A.4.2.)
F3
X1
(x) = Weibull extreme value cumulative distribution for the smallest values of random variable X (A.4.2)
F3(s) = single Weibull distribution of the tensile strength representing all of Group 2 to 4 wires (without cracked
wires), based on the sample mean and sample standard deviation of the combined groups.(A.5.3.2)
F3(s ) = single Weibull distribution of the tensile strength representing all the Group 2 to 4 wires, based on a mean
tensile strength of 1.0 and a standard deviation of the combined groups, divided by
s
(A.5.3.2)
F
C
(e) = compound cumulative distribution of ultimate strain (A.5.1.2)
F
C
(s) = compound cumulative distribution of tensile strength (A.5.2)
k = corrosion stage of a group of wires (k = 2, 3, 4 and 5) (A.5.1.1)
k = corrosion stage of a group of wires (k = 2, 3, and 4) (A.5.3.1)
The text includes the results of tensile strength tests on corroded wires that were removed from three bridges, a
description of the behavior of a single bridge wire, and the statistical equations required for evaluating the models. An
additional source of information is the simulated ten-wire cable demonstration calculations that appear in the final report
for NCHRP Project 10-57, on the accompanying CD.
A-3
km = subgroup of Stage k wires that follow stress-strain curve m (A.5.1.1)
K = reduction factor from Chart 5.2.3.3.2-1 as a function of the coefficient of variation,
s
/
s
. (A.5.3.2)
max = maximum value of the expression inside the brackets (A.5.1.1) (A.5.1.2) (A.5.3.2)
m = parameter of the Type III extreme value distribution for minimum values (A.4.2)
m = number of a discrete stress-strain curve (A.5.1.1)
M = number of different stress-strain curves considered (A.5.1.1)
N
eff
= effective number of unbroken wires in the cable (A.5.1.1) (A.5.1.2) (A.5.2)
N
eff
= effective number of unbroken wires and uncracked wires in the cable (A.5.3.2)
p
k
= fraction of unbroken wires represented by Group k wires (A.5.1.2) (A.5.2)
p
k
= fraction of unbroken wires and uncracked wires in the cable represented by Group k wires (A.5.3.1)
p
km
= fraction of cable represented by Subgroup km wires (A.5.1.1)
R
u
= cable strength attributable wires in the cable that are not broken (A.5.1.1) (A.5.1.2) (A.5.2) (A.5.3.2)
s = stress in the unbroken wires of the cable (A.5.2) (A.5.3.2)
s
1
= s/
s
(A.5.3.2)
s(e) = stress in wires determined from average stress-strain curve for all wires (A.5.1.2)
S
k
(e) = survivor function, or that fraction of the wires in Group k that has an ultimate
elongation greater than e (A.5.1.1)
s
m
(e) = stress at strain e in wires that follow stress-strain curve m (A.5.1.1)
T
km
= force in Subgroup km wires (A.5.1.1)
T
u
(s) = tensile force in unbroken wires in the cable at stress s (A.5.2)
x = value of random variable X or X
1
(A.4.2)
x
0
=

minimum value of x for which the Type III distribution is valid (A..4.2)
X
1
= random variable representing the smallest values of property x of each wire (A.4.2)
( ) = Gamma function of the expression inside the brackets (A.4.2)
s
= sample mean tensile strength of the combined groups of wires, excluding cracked wires (A.5.3.1) (A.5.3.2)
sk
= sample mean tensile strength of Group k wires (A.5.3.1)
sx
= sample mean of property x (A.2.3.1) (A.2.3.2)
sX1
= sample mean of the extreme value distribution of x (A.4.2)
= parameter of the Type III extreme value distribution for minimum values (A.4.2)
s
= sample standard deviation of the tensile strength of the combined groups of wires, excluding cracked wires
(A.5.2.1)
sk
= sample standard deviation of tensile strength of Group k wires (A.5.3.1)
sx
= sample standard deviation of property x (A.2.3.1) (A.2.3.2)
sX1
= sample standard deviation of the extreme value distribution of x, (A.4.2)
A-4
A.3 STRENGTH MODELS
A.3.1 Behavior of a Single Bridge Wire
The modern method of manufacturing bridge wire is to cold draw a carbon steel rod through successively smaller dies
until the specified diameter and tensile strength are reached. The process imparts strength to the wire, along with an
elongated grain structure.

Any bridge wire subjected to a tensile test stretches elastically to the proportional limit. It doesnt exhibit a true yield
point as the strain increases, but enters a strain-hardening range instead, immediately after the transition from elastic
behavior. The stress continues to increase with the strain until the tensile strength is reached, at which point the wire
necks down and fails, resulting in a reduction of area and a cup-and-cone fracture surface. There is no yield plateau as
with milder steel materials. The strain at the tensile strength is the ultimate strain.
Failure occurs almost immediately after the tensile strength and the ultimate strain are reached.
Figure A.3.1-1. Typical stress-strain curve for bridge wire. [2]
New wires have a tensile strength that varies little (the coefficient of variation is about 1.5% to 2%). Corroded wires
that exhibit ductile fractures have a more variable tensile strength than new wires. Wires that are cracked to any degree
0
50
100
150
200
250
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
STRAIN (in/in)
S
T
R
E
S
S

(
k
s
i
)
AVERAGE STRESS-STRAIN CURVE FOR 126 WIRES
DATA FROM BEAR MNTN BRIDGE WIRE TESTS
[ROEBLING, 1925]
TENSILE STRENGTH 229 ksi
ULTIMATE STRAIN 6.13 %
NET DIAMETER OF WIRE 0.191 in
NET STEEL AREA 0.02865 sq in
STRESS BASED ON NET STEEL AREA OF WIRE
PROPORTIONAL LIMIT
TENSILE STRENGTH
ULTIMATE STRAIN
YIELD STRENGTH (0.2% OFFSET)
The typical stress-strain curve for new bridge wire is shown in Figure A.3.1-1. The data were taken from Roebling [2]
and represent the average results of tests on 126 wires from the Bear Mountain Bridge. Failure occurs almost
immediately after the ultimate strain is reached. New, corroded and cracked wires all follow the same curve.
A-5
Table A.3.1-1 Mean tensile strength and coefficient of variation for wires from several suspension bridges
Bridge
Corrosion
Stage
Mean Tensile
Strength of
New Wire (ksi)
Number of
Samples
Mean Tensile
Strength (Fraction
of New Wire)
Coefficient of
Variation
New Wire 240 2 1.00 0.004
1-2 30 0.99 0.018
3 18 0.98 0.024
4 10 0.94 0.038
X
Cracked 14 0.84 0.131
New Wire 231 0 1.00 N/A
1-2 19 1.00 0.089
3 12 0.97 0.119
4 7 0.89 0.085
Y
Cracked 15 0.68 0.210
New Wire 236 20 1.00 0.020
1-2 29 0.98 0.031
3 29 0.95 0.038
4 33 0.94 0.041
Z
Cracked 7 0.75 0.260

A.3.2 Strength Models for Wire Bundles
Perry [3] describes 3 models that can be used to estimate the strength of a parallel wire cable. They are presented
below, starting with the least conservative. One is for ductile wires, another is for brittle wires based on strain and the
third, a subset of the second, is for brittle wires based on tensile strength.
In studies of cables, which are bundles of filaments or wires, it is assumed that all the filaments are held together
firmly at the ends and subjected to equal strain. Internal cable inspections of the Williamsburg Bridge [4] and Mid-
Hudson Bridge [5] can be cited as evidence that the gap in a broken cable wire is equal to or slightly greater than the
elastic elongation of the wire between cable bands under the dead load tension of the cable. The conclusion is that all
the wires are effectively clamped at the cable bands, at least for normal working stresses.
A.3.2.1 DUCTILE-WIRE MODEL
In the Ductile-Wire Model, it is assumed that all of the wires in the cable share in the cable force until the entire cable
breaks as a single unit. In other words, all of the wires are assumed to break simultaneously. For this to happen, the
wires need not have equal strength, but they must be ductile (able to stretch under a constant load). Each wire
elongates elastically and then plastically to the same degree as all of the other wires. The cable strength is the sum of
the individual wire strengths, which is equal to the average wire strength multiplied by the number of wires in the
cable. This model should not be applied to bridge cables, because there are always some cable wires that break before
the entire cable does, and the efficiency of a cable of parallel wires has been found to vary between 94% and 96% [2,
6]. Efficiency is defined in these references as the actual breaking strength divided by the calculated strength of the
Ductile-Wire Model.
A.3.2.2 BRITTLE-WIRE MODELS
have reduced and highly variable tensile strength. They exhibit brittle fracture surfaces with no reduction in area, often
referred to as square breaks. The mean tensile strength and the coefficient of variation of wires removed from
several suspension bridges are shown in Table A.3.1-1 for various stages of corrosion. Tensile strength in this table is
based on the net steel area (the area not including the zinc coating).
A-6
The term brittle in the context of the Brittle-Wire Model is used to describe the behavior of an individual wire. It does
not mean that the material in the wire is brittle, but that the wire fails suddenly when the strain or the stress in the wire
reaches a certain level and no longer shares in the tensile force of the cable. Perry [3]describes two separate models
under this name. The first is a general model that is referred to as the Limited Ductility Model in these Guidelines. The
second is a specific case of the Limited Ductility Model, which is used by Perry and others to calculate cable strength
under the general name, Brittle-Wire Model. It is also called the Brittle-Wire Model in these Guidelines.
A.3.2.2.1 Limited Ductility Model
In the Limited Ductility Model, a wire is assumed to fail suddenly when the strain in the wire reaches a certain level.
Each wire in the cable can elongate only to its individual limit, which is called the ultimate strain of the wire. A
specific wire that reaches this elongation will fail by definition. The limit is different for each wire, and so is the
strength. The small amount of strain that occurs at reduced stress after the tensile strength is reached, as shown in
Figure A 3.1-1, is ignored in the analysis. For any specific value of strain, it is assumed that each intact wire is subject
to a tensile stress that corresponds to the strain in the stress-strain diagram for that wire. Whenever a wire breaks, the
force carried by the wire is distributed to all of the unbroken wires in the cable in the same proportion as before the
wire failed. The wires are assumed to break sequentially as each wire reaches its maximum elongation, and the cable
strength is attained only after some of the wires break.
To determine the cable strength, the cable strain is increased incrementally and then the number of wires that have
reached their elongation limit and failed are calculated at each increment. For each subsequent calculation, the number
of newly failed wires is subtracted from the number of still intact wires to determine the number of unbroken wires that
remain. All of the wires in the cable are subjected to the same strain. The cable force is calculated as the sum of the
forces in the intact wires at that strain. Wires will fail faster than the cable force can increase at some value of
elongation. The maximum force attained is the cable strength. This technique uses a statistical method called ordered
statistics. The strength can be estimated either by sorting the wires in order of ultimate elongation or by using the
statistical distribution curve of this property.
A.3.2.2.2 Brittle-Wire Model (A Special Case)
In a special case of the Limited Ductility Model, it is assumed that all of the wires follow the same stress-strain
diagram, and that the stress in all of the intact wires is the same at any specific value of strain. The model may be
simplified by assuming that a wire fails suddenly when the stress in the wire reaches a certain level. In the Brittle-Wire
Model that results, each wire in the cable can resist a stress only up to its specific limit, equivalent to its tensile
strength. A specific wire that reaches this stress will fail by definition. The limit is different for each intact wire, which
is assumed to carry the same tension as all of the other wires; hence, the model is also called the Load-Sharing Model.
After a wire breaks, the force carried by the wire is distributed equally to all of the other unbroken wires in the cable.
The wires are assumed to break sequentially as each wire reaches its tensile strength, and the cable strength is attained
only after some of the wires have broken.
Determining the cable strength requires increasing the cable stress in steps and calculating the number of wires that fail
at each increment as they reach their tensile strength. The number of newly failed wires is subtracted from the number
of previously intact wires to determine the number of unbroken wires. The cable force is calculated as the area of
unbroken wires at a given level of stress multiplied by the wire stress. At some level of stress, wires will fail faster than
the wire force can increase. The maximum force attained is the cable strength. This technique uses the same ordered
statistics method as the Limited Ductility Model. The strength is estimated either by sorting the wires in order of
tensile strength, or by using a statistical distribution curve of this property.
A.3.2.3 BRITTLE-DUCTILE MODEL
The Brittle-Ductile Model [7] takes into account the cable wires that fail at very low strains, which are subtracted from
the total number of wires in the cable. The model assumes that the remaining wires in the cable are ductile. The
strength calculation is then the same as for the Ductile Model, except that the number of wires is reduced.
A.3.2.4 SIMPLIFIED MODEL
A-7
In a variation of the Brittle-Ductile Model, cracked and broken wires are subtracted from the total wires in the cable,
and the Brittle-Wire Model is applied to the remaining intact wires. This alternative to the Brittle-Wire Model is called
the Simplified Model in these Guidelines.
A.4 STATISTICS FOR CABLE STRENGTH ANALYSIS
Sample wires are removed from the cables and tested to provide the data required for estimation of cable strength.
In the calculation of cable strength using sample statistics, data are used to develop statistical distribution curves that
represent the spectrum of wire strengths or ultimate strains present in the cable. The curves are described below, before
presenting their use in estimating cable strength.
The equations that follow employ the results of physical tests performed on a selection of wires removed from an
actual cable. The mean and standard deviations resulting from these tests are called sample means and sample standard
deviations to differentiate them from the population means and population standard deviations that represent all of the
wires in the cables. The terms (mean) and (standard deviation) apply to the entire population, while
s
and
s

apply to the samples.
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
180 200 220 240 260 280
TENSILE STRENGTH, KSI
f(s)
F(s) = Fraction
of cable weaker
(1-F(s)) =Fraction
of cable stronger
F
R
A
C
T
I
O
N

O
F

C
A
B
L
E
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
180 200 220 240 260 280
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
180 200 220 240 260 280
f(s)
F(s) = Fraction
of cable weaker
(1-F(s)) =Fraction
of cable stronger
F
R
A
C
T
I
O
N

O
F

C
A
B
L
E

Figure A.4-1. Typical probability density distribution curve
0.0
0.2
0.4
0.6
0.8
1.0
180 200 220 240 260 280
F
R
A
C
T
I
O
N

O
F

C
A
B
L
E

W
E
A
K
E
R
F(s) = Fraction
of cable weaker
(1-F(s)) =Fraction
of cable stronger
0.0
0.2
0.4
0.6
0.8
1.0
180 200 220 240 260 280
F
R
A
C
T
I
O
N

O
F

C
A
B
L
E

W
E
A
K
E
R
F(s) = Fraction
of cable weaker
(1-F(s)) =Fraction
of cable stronger
Figure A.4-2. Typical cumulative probability curve
Two types of distribution curves are discussed in the following text. The first is the probability density distribution curve
shown in Figure A.4-1. This curve describes the fraction of the entire population that has a specific value of x. It is
commonly known as the Bell Curve for a distribution that is Normal or Gaussian. The area under the curve is always
unity, describing 100% of the population. The area under the curve to the left of a specific value of x is the fraction of th
population that has a value less than or equal to that value of x. When this area is plotted against x, it becomes the
cumulative probability curve (Figure A.4-2), and varies between 0 and 1. The area to the right of a specific value of x in
Figure A.4-1 is the fraction of the population that has a value greater than the value of x.
A-8
The first of the distribution curves is represented by the expression f(x), while the second is represented by F(x). In
these Guidelines, the expressions f3(x) and F3(x) are used to denote the recommended Weibull distribution.
The value that is derived from these distributions is the fraction of the population that has a value of x (i.e., tensile
strength or ultimate strain) greater than a specific value of x. It is also known as the Survivor Function and is given as
(1-F (x)). It is equivalent to the area to the right of x on the probability density distribution. The term s is substituted
for x for the distribution of tensile strength in these Guidelines, and e is substituted for x for ultimate strain.
A.4.1 Extreme Value Distributions
The distribution of the minimum tensile strengths of the wires must be obtained to determine the strength of the cable
in any panel of length L. The extreme value distributions are very useful for this purpose. In the discussions and
equations that follow, the general variable, X, represents either the tensile strength or the ultimate elongation in the
strength models described previously.
The derivation of the extreme value distributions is given in various references [1, 7-9] and will not be repeated here.
The Type I and Type III extreme value distributions are useful in describing the minimum strength of wires. Type I
(also known as the Gumbel) extends from - to +, as does the Normal distribution.
Type III, extending from a minimum value of x
0
to a maximum of +, corresponds to a material with a lower limit of
tensile strength or ultimate elongation. Rao [1]states that the parent function of this extreme value distribution is a
Gamma distribution. The Gamma distribution is valued for x > x
0
, as is the Weibull distribution, and for very small
standard deviations is virtually the same as a Normal distribution. The value of x
0
is assumed to be zero, because the
tensile strength and ultimate elongation of materials cannot be less than zero.
The Type III extreme value distribution is a Weibull distribution, which is discussed in many references, among them
Rao and Weibull. [1, 9] The parameters of the distribution are not implicit, but it is relatively simple to find a solution
through trial-and-error using a spreadsheet. An alternative method is to use Weibull graph paper. [1] Perry [3]argues
that the Weibull distribution is the only correct distribution in this instance. The Weibull distribution is used in these
Guidelines because it allows limiting the minimum value of the variable, which is of particular importance for the
Limited Ductility Model. Ultimate strains less than zero could result from the use of other distribution functions.
A.4.2 Weibull Distribution
The Type III extreme value distribution is the Weibull distribution. As stated above, it extends from a minimum value,
x
0
, to +. This distribution is given in Rao [1] as
( )
=
m
X
x
x x
x F
0
0
exp 1 3
1
(A.4.2-1)
=
m m
X
x
x x
x
x x
x
m
x f
0
0
1
0
0
0
exp ) ( 3
1

(A.4.2-2)
where
f3
X1
(x) = Type III extreme value probability density distribution for the smallest values of random variable X
F3
X1
(x) = Weibull extreme value cumulative distribution for the smallest values of random variable X
exp(x) = e (2.7183) to the power (x), the exponential of (x)
x = a value of random variable X
1
A-9
x
0
=

minimum value of x for which the Type III distribution is valid. It is taken as 0 in these Guidelines.
X
1
= random variable representing the smallest values of property x of each wire
m = parameter of the Type III extreme value distribution for minimum values
= parameter of the Type III extreme value distribution for minimum values
The function F3
X1
(x) is the cumulative probability distribution. Its value at any value of x is the probable fraction of the
entire population of wires for which the property represented by the general variable X
1
is smaller than the specific
value x.
The probability density distribution f3
X1
(x) represents the fraction of the population for which one unit of property
represented by X
1
is equal to x. The unit of x is the basic unit of the property, e.g., ksi or inches/inch. The cumulative
distribution is the integral of the density distribution between x
0
and x.
In these equations and the ones that follow, the term X
1
represents the smallest values of X in a specific length of wire
determined from tests on specimens cut from sample wires.
The mode, mean and variance are given as
( )
m
m
m
x x
1
0 0
1
mode

+ = (A.4.2-3)
( ) + + = =
m
x x
X
1
1 mean
0 0
1
(A.4.2-4)
( ) + + = =
m m
x
X
1
1
2
1 variance
2 2
0
2
1
(A.4.2-5)
In these equations, x
0
is the lower limit of the tensile strength or ultimate strain of the wire (usually taken as zero) and
and m are parameters of the distribution. If the sample mean and standard deviation for the smallest values of x for
each wire sample are given,
X1
becomes
sX1
and
X1
becomes
sX1
.
( )
2
0 1
2 2
1
1
1
1
2
1 x
m m
sX sX
=

+ +

(A.4.2-6)
where
sX1
= sample mean of the extreme value distribution of X
1
sX1
= sample standard deviation of the extreme value distribution of X
1

( ) = Gamma function of the expression inside the brackets
The Weibull function is one of the functions included in the Microsoft Excel spreadsheet program, which makes it
relatively simple to use once the parameters have been determined. The parameters and m are called beta and alpha,

Eliminating the variable ( - x
0
) in Equation 4 and Equation 5 results in the following equation, given by Castillo
[10] that can be solved for m by successive approximation, illustrated in Figure A.4.2-1 (use theSolverfunction of
a spreadsheet program such as Microsoft Excel for this purpose)
A-10
respectively, in Excel. The Weibull function in Excel uses x
0
= 0. If another value is used, (x x
0
) must be substituted
for x, and ( - x
0
) substituted for beta, in the Excel function.
The recommended value of x
0
in these Guidelines is zero. It can be argued that this value should be the dead load stress
in the cable because the wires have been tested in service at this stress and the tensile strength cannot be smaller, but
comparative calculations using the dead load stress for x
0
in one case and zero in the other show that the difference in
cable strength is very small, and Equations 1 to 6 can be simplified by omitting the term x
0
.
STRENGTH TYPE 3 EXTREME VALUE DISTRIBUTION FOR MINIMUM VALUES
MODELS WEIBULL DISTRIBUTION
CALCULATION OF PARAMETERS FROM MEAN AND STANDARD DEVIATION
WEIBULL USING EQUATIONS A.4.2-4 and A.4.2-6
PARAMETERS
(ksi)
THE MEAN AND STANDARD DISTRIBUTION OF THE TENSILE STRENGTH OF EACH GROUP OF WIRES DETERMINED FROM
THE LABORATORY TESTS ARE USED TO DETERMINE THE PARAMETERS OF THE WEIBULL DISTRIBUTIONS. THE METHOD
PRESENTED IN ARTICLE A.4.2 IS USED BELOW. THE VALUE OF x
0
IS TAKEN AS ZERO AND THIS TERM IS
OMITTED IN THE EQUATIONS SHOWN IN THE CALCULATION BELOW
Excel
k = corrosion group 2 3 4 5, CRACKED parameter
TENSILE STRENGTH DISTRIBUTION FOR EACH CORROSION GROUP
mean tensile strength,
s
ksi 239.0 235.9 231.1 200.5
standard deviation,
s
ksi 4.3 5.7 8.7 26.3
CALCULATION OF WEIBULL PARAMETERS
Eq. A.4.2-6: (1+2/m)/
2
(1+1/m) = 1 -
2
/
2
m (assumed, then determined by solver) 70.6 52.4 33.4 9.1 alpha
( = GAMMA function) (1+2/m) 0.9844 0.9793 0.9688 0.9133
(1+1/m) 0.9920 0.9893 0.9836 0.9475
(1+2/m)/
2
(1+1/m) 1.0003 1.0006 1.0014 1.0172
2
18.490 32.490 75.690 691.690
2
57121 55649 53407 40200
2
/
2
3.2E-04 5.8E-04 1.4E-03 1.7E-02
SOLVE FOR m USING SOLVER:
Equation A.4.2-6 is solved for m by making the value of the
expression (1+2/m)/
2
(1+1/m) - 1 -
2
/
2
equal to zero by
varying m, using the "Solver" routine in Excel:
(1+2/m)/
2
(1+1/m) - 1 -
2
/
2
= 0 -4.7E-10 -8.1E-10 -4.9E-10 2.6E-10
CALCULATE :
The value of

is found by solving Equation A.4.2-4 for this
variable and substituting the value of m found above:

=

/

(1+1/m) 240.9 238.4 235.0 211.6 beta
Corrosion Group
Figure A.4.2-1. Calculation of parameters of Weibull distribution by iteration.
A.5 EQUATIONS FOR ESTIMATING CABLE STRENGTH USING THE WEIBULL DISTRIBUITON
For both the Limited Ductility and Brittle-Wire Models, the Weibull distribution curve is used to estimate cable
strength. The use of the expression (1-F3(x)) is equivalent to making the force equal to zero in wires that break as
strain or stress is increased.
A.5.1 Limited Ductility Model Equations

A-11
A.5.1.1 DIFFERENT STRESS-STRAIN CURVES AVAILABLE FOR WIRES
The separation of test samples into these subgroups would necessitate the removal of many more sample wires from
the cable, and is not recommended. Thus, all wires in a specific group are assumed to follow the same distribution of
the ultimate strain. At any value of strain, the fraction of unbroken wires in each subgroup of wires is represented by
the survivor function,
( ) ( ) ( ) e F e S
k k
3 1 = (A.5.1.1-1)
where
S
k
(e) = survivor function; or that fraction of the wires in Group k that has an ultimate elongation greater than e
e = specific value of strain
F3
k
(e) = Weibull cumulative distribution function for ultimate strain of Group k wires
k = corrosion stage of a group of wires (k = 2, 3, 4 and 5)
The total force in the wires of Subgroup km at strain e is
( ) ( ) e F e s p a N T
k m km w eff km
3 1 ) ( = (A.5.1.1-2)
where
T
km
= force in Subgroup km wires
N
eff
= effective number of unbroken wires in the cable
a
w
= nominal area of one wire, used in lab analysis
p
km
= fraction of cable represented by Subgroup km wires
s
m
(e) = stress at strain e in wires that follow stress-strain curve m
m = number of a discrete stress-strain curve
km = subgroup of Stage k wires that follow stress-strain curve m
The total cable force is the sum of the forces in the subgroups of wires, and the cable strength is the maximum force
attained
( ) ( ) ( )

=

= =
M
m k
k m km w eff u
e F e s p a N R
1
5
2
3 1 max (A.5.1.1-3)

The general form of the Limited Ductility Model is used whenever the stress-strain curves for wires vary significantly. This
can occur when the carbon content of the wires varies from wire to wire because of poor manufacturing quality control or
multiple suppliers. The stress-strain curves should be reduced to a limited number, M, of average curves, and each group of
wires subdivided and assigned to the various M curves proportionally. This requires a map of the cable cross-section in each
evaluated panel showing the distribution for each stress-strain curve. Additional samples must be taken to obtain the data to
prepare these maps; a sampling pattern is suggested in the final report for NCHRP Project 10-57, on the accompanying CD.
These additional samples may also be added to the random samples for the purpose of determining the tensile properties of
the wires. It is recommended that a statistician be added to the investigation team when this procedure is followed.
A-12
where
R
u
= cable strength attributable to wires in the cable that are not broken
M = number of different stress-strain curves considered

This strength model requires M times as many calculations as when only a single average stress curve can be used for
all wires, as in the following article.

A.5.1.2 ALL GROUPS OF WIRES HAVE THE SAME STRESS-STRAIN CURVE
Where all groups of wires have the same stress-strain curve, the term s
m
(e) becomes the same for all groups at any
specific strain, e. The individual cumulative distributions may be combined into a single compound distribution ( not a
Weibull distribution) and Equation A.5.1.1-3 simplifies to
( ) ( ) ( ) e F e s a N R
C w eff u
= 1 ) ( max
(A.5.1.2-1)
in which
) ( 3 ) (
5
2
e F p e F
k
k
k C
=
=

(A.5.1.2-2)
where
F
C
(e) = compound cumulative distribution of the ultimate strain
p
k
= fraction of unbroken wires represented by Group k wires
s(e) = stress in wires determined from average stress-strain curve for all wires

A.5.2 Brittle-Wire Model Equations
The compound tensile strength distribution is used in the Brittle-Wire Model (again, not a Weibull distribution). A
single average stress-strain curve is used to represent the entire cable. The fraction of the cable that is essentially
unbroken at any specific stress level s is given by the expression.
( ) ( ) s F
C
1 (A.5.2-1)
in which
( ) ( ) s F p s F
k
k
k C
3
5
2
=

=
(A.5.2-2)
where
F
C
(s) = compound cumulative distribution of tensile strength
A-13
F3
k
(s) = Weibull cumulative distribution function for tensile stress of Group k wires
s = stress in the unbroken wires of the cable
The force in the cable is
( ) ( ) ( ) [ ] s F s a N s T
C w eff u
= 1 (A.5.2-3)
where
T
u
(s) = tensile force in unbroken wires in the cable at stress s
and the cable strength is the maximum value that T
u
(s) attains,
( ) ( ) ( ) s F s max a N R
C w eff u
= 1 (A.5.2-4)
This equation can be solved by trial-and-error using Solver in the Microsoft Excel spreadsheet program, as shown in
Appendix C. If a cable force vs. strain diagram is required, the cable force must be calculated at selected increments of
strain.
A.5.3 Simplified Model Equations
Using the simplified model requires:
Calculation of the effective number of wires, obtained by subtracting the unrepaired broken wires plus the
estimated number of wires in Stages 3 and Stage 4 that contain cracks from the total number of wires in the cable
(there will be no Group 5, because cracked wires are omitted from this calculation)
Calculation of the combined mean and standard deviation of the tensile strengths of the remaining wires that
comprise Groups 2, 3, and 4, using Equation A-5.3.1-1 and Equation A-5.3.1-2)
Applying the Brittle-Wire Model to the wires, using the single distribution curve
A.5.3.1 SINGLE DISTRIBUTION CURVE FOR TENSILE STRENGTH
The fraction of the cable represented by Groups 2, 3 and 4 is combined with the sample mean and standard deviation
values for the minimum tensile strength of the representative specimens of each group. The result is used to determine
the sample mean tensile strength and standard deviation of the entire unbroken and uncracked wire population as
follows:
=
=
4
2
) (
k
sk k s
p (A.5.3.1-1)
( )
2 2 2
4
2
s sk sk
k
k s
p

+ =
=
(A.5.3.1-2)
where
s
= sample mean tensile strength of the combined groups of wires excluding cracked wires
sk
= sample mean tensile strength of Group k wires
s
= sample standard deviation of the tensile strength of the combined groups of wires, excluding cracked wires

A-14
sk
= sample standard deviation of the tensile strength of Group k wires
p
k
= fraction of the unbroken and uncracked wires in the cable represented by Group k wires
k = corrosion stage of a group of wires (k = 2, 3 and 4)
A.5.3.2 CABLE STRENGTH USING THE SIMPLIFIED MODEL
The estimated cable strength calculated according to the Brittle-Wire Model is given by Equation A-5.2-4. In the
Simplified Model, which uses a single tensile strength distribution, the compound distribution F
C
(s) is replaced by the
single Weibull distribution F3(s), resulting in
( ) ( ) ( ) s F s max a N R
w eff u
3 1 = (A.5.3.2-1)
where
F3(s) = single Weibull distribution of the tensile strength representing all of the Group 2 to 4 wires (without
cracked wires), based on the sample mean and sample standard deviation of the combined groups,
calculated using equations A.5.3.1-1 and A.5.3.1-2
The sample mean tensile strength
s
can be factored out of the expression in brackets in Equation A-5.3.2-2 to result in

( ) ( ) ( ) s F s max a N R s
w eff u
= 3 1 (A.5.3.2-2)
in which
s' = s/
s

F3(s') = single Weibull distribution of the tensile strength representing all the Group 2 to 4 wires, based on a mean
tensile strength of 1.0 and a standard deviation of the combined groups, calculated using Equations
A.5.3.1-1 and A.5.3.1-2, divided by
s

Referring to Equation A.4.2-1, F3
x1
(x) does not change if the terms x, x
0
and on the right-hand side of the equation
are all divided by
s
. The value of the distribution functions F3(s) and F3(s) are identical for any specific value of s, if
the distribution F3(s) is based on
s
= 1 and
s
=
s
/
s
. The value of the term K = max(s'(1 - F3(s))) can be
determined as a function of the coefficient of variation,
s
/
s
. The calculation results in the curve of K vs.
s
/
s
shown
in Figure 5.3.3.1.2-1, where K is the reduction factor to be applied in the equation
K a N R s
w eff u
=
(A.5.3.2-3)
where
K = reduction factor from Figure 5.3.3.1.2-1 as a function of the coefficient of variation,
s
/
s
.
A.6 REFERENCES
1. Rao, S.S., Reliability-Based Design. 1st ed, ed. R. Hauserman. 1992: McGraw-Hill, Inc. 569.
2. Roebling, J.A., Parallel Wire Cables for Suspension Bridges. 1925, John A. Roebling's Sons Company:
Trenton, New Jersey.
3. Perry, R.J., Estimating Strength of Williamsburg Bridge Suspension Cables. The American Statistician, 1998.
52(3): p. No. 3.
A-15
4. Steinman et al., Williamsburg Bridge Cable Investigation Program: Final Report. 1988, New York State
Department of Transportation & New York City Department of Transportation: New York, NY.
5. Modjeski and Masters, Mid-Hudson Bridge - Main Cable Investigation (Phase III). 1991, Modjeski and
Masters, Inc.: Harrisburg, Pennsylvania.
6. Board of Engineers, The Bridge Over the Delaware River Connecting Philadelphia, PA and Camden, N.J..
1927, Delaware River Bridge Joint Commission of the State of PA and NJ.
7. Matteo, J., G. Deodatis, and D.P. Billington, Safety Analysis of Suspension-Bridge Cables: Williamsburg
Bridge. Journal of Structural Engineering, 1994. 120(11): p. 3197-3211.
8. Gumbel, E.J. and J. Lieblein, Statistical Theory of Extreme Values and Some Practical Applications. United
State Department of Commerce - National Bureau of Standards, 1954. Series 33: p. 1 -51.
9. Weibull, W., A Statistical Distribution Function of Wide Applicability. Journal of Applied Mechanics, 1951.
73(18): p. 293-297.
10. Castillo, E., Extreme Value Theory in Engineering. 1988, San Diego, CA: Academic Press, Inc.
B-1
CONTENTS
APPENDIX B EFFECT OF DETERIORATION IN ADJACENT PANELS ON ESTIMATED CABLE
STRENGTH................................................................................................................................. B-1
B.1 INTRODUCTION...................................................................................................................................... B-2
B.2 NOTATION................................................................................................................................................ B-4
B.3 ORDER OF FAILURE OF CRACKED WIRES.................................................................................... B-5
B.4 NUMBER OF DISCRETE CRACKED WIRES AND THEIR REDEVELOPMENT........................ B-5
B.4.1 All Panels in Effective Development Length Inspected ...................................................................... B-5
B.4.1.1 DISCRETE CRACKED WIRES IN EFFECTIVE DEVELOPMENT LENGTH.......................... B-6
B.4.1.2 EFFECTIVE NUMBER OF CRACKED WIRES THAT CAN BE REDEVELOPED.................. B-9
B.4.2 Only Panel 1 Inspected .......................................................................................................................... B-9
B.4.2.1 DISCRETE CRACKED WIRES WITHIN EFFECTIVE DEVELOPMENT LENGTH ............... B-9
B.4.2.2 REDEVELOPMENT OF BROKEN CRACKED WIRES ........................................................... B-10
B.5 MAXIMUM CABLE TENSION THAT CAN BE DEVELOPED IN CRACKED WIRES.............. B-11
B.6 CABLE STRENGTH............................................................................................................................... B-11
B-2
B.1 INTRODUCTION
The equations for cable strength given in the Guidelines are based on the assumption that only wires in the specific
panel being evaluated have been inspected and that, wire for wire, all other panels within the effective development
length are at the same stage of corrosion. Of course, this is not usually the case. The physical strength of a wire may
vary in adjacent panels, the wire may have cracks, or it may be at a different corrosion stage, but there is no way to
know for sure without inspecting all the panels in the effective development length. In Article B.4.1 all the panels are
assumed to have been inspected. In Article B.4.2, only one panel is inspected, and the equations presented in Section 5
are derived.
A wire that is deteriorated or cracked in the evaluated panel as well as in adjacent panels may fail in an adjacent panel
rather than in the evaluated panel. Part of the wires capacity will be redeveloped in the evaluated panel, because of the
friction that is developed in the cable bands between the break location and the investigated panel.
The issue of wires that are broken in service in the cable was addressed in Section 5. The broken wires, which are
treated separately from intact wires, are also partially redeveloped as they pass through the cable bands, but this
Appendix addresses only cracked wires.
The calculation of the effect of adjacent panels becomes manageable once certain assumptions are made. The cable
cross-section is divided into segments. Each segment extends over the arc of a sector of the cable and is composed of
wires that are equidistant from the center of the cable (see Figure B.1-1). All wires in a segment are assumed to be at the
same corrosion stage in any one panel, but not in adjacent panels (e.g., a segment may contain Stage 3 wires in one
panel and Stage 4 wires in another panel).
Cracked wires are the major factor in determining cable strength. As an example, the calculated strength for one bridge
cable was reached at a wire stress of 213 ksi. 14.2% of the wires had a tensile strength lower than this stress and were
assumed to have failed. Wires with cracks accounted for 13.8% of this total, uncracked Stage 3 and Stage 4 wires for
only 0.4%. No good (Stage 1 and Stage 2) wires were broken at this stress level.
A fractured wire regains a constant fraction of its tensile strength at each cable band, equal to the redevelopment
coefficient, C
d
, multiplied by 95% of the mean tensile strength of Group 2 wires. Wires that break in adjacent panels
are redeveloped in the evaluated panel to the extent made possible by cable band friction, and the redevelopment force
is added to the cable strength. It is assumed that these wires will not break again in the effective development length.
Usually, the stress in the cable at failure will be greater than the stress that can be redeveloped in a wire that breaks
within the effective development length. The use of a constant for the maximum force that can be redeveloped will not
affect the contribution to cable strength of cracked wires that fail, because the maximum redeveloped stress is
generally smaller than the stress in the cable at failure. The calculated cable force at stresses lower than the stress at
which the cable fails may be slightly too high, but the cable strength will not be affected.
The cable force and wire stress are assumed to be the same in all the panels. Whenever a wire breaks in one panel, the
redistribution of wire force makes an infinitesimal change in the stress in the remaining wires in the other panels, but
this change in stress can be ignored.
B-3

Figure B.1-1. Definitions of segments, panels and discrete cracked wires.
segments are
identified by sector
and ring numbers
and are sequentially
numbered from 1 to
NS
ring
sector
segment
CABLE
e
1
=q
1
r
1
=p
1
e
2
=e
1
*q
2
r
1
*p
2
r
2
=e
1
*p
2
PANEL 1
PANEL 3
PANEL 2
e
3
=e
2
*q
3
(r
1
+r
2
)*p
3
r
3
=e
2
*p
3
SEGMENT "j" IN PANEL 1
Panel 1 is the panel in which
strength is calculated
CABLE
Cable band
Panel 7 Panel 5 Panel 3 Panel 1 Panel 2 Panel 4 Panel 6
CROSS SECTION OF CABLE
DIAGRAMS OF SEGMENT "j" IN THREE ADJACENT PANELS
ELEVATION OF CABLE
1. Wires in Segment j that are cracked in Panel 1
2. Wires in Segment j that are cracked in Panel 2 but not in Panel 1
4. Wires in Segment j that are cracked in Panel 3 but not in Panels 1 or 2
3. Wires in Segment j that are cracked in Panel 2 and in Panel 1
5. Wires in Segment j that are cracked in Panel 3 and in Panel 1 or 2 or both
1, 2 & 4 are "discrete cracked wires"
SEGMENT "j" IN PANEL 2 SEGMENT "j" IN PANEL 3
B-4
B.2 NOTATION
a
w
C
d
= redevelopment factor, (fraction of 95% of the mean tensile strength of Group 2 wires that is developed in a
broken wire at each cable band)
C
di
= wire re-development factor for panel i = C
d
multiplied by the number of cable bands between panel i and
Panel 1
e
ij
= fraction of uncracked wires in segment j, panel i, not cracked in panels less than i
F3
5
(s) = fraction of cracked wires that are broken at stress s = Weibull cumulative distribution of tensile strength
for Group 5 (cracked) wires at stress s
i = number of a cable panel
j = number of a cable segment
jk = a segment j that contains Stage k wires in Panel 1
k
ij
= corrosion stage of wires in segment j, panel i
L
e
= effective development length (number of panels)
N = total number of wires in the cable
N
0k
= number of unbroken Stage k wires in the evaluated panel
N
5
= number of discrete cracked wires in the effective development length
n
c,j
= number of discrete cracked wires in segment j for the effective development length
n
c,jk
= number of discrete cracked wires in a segment, j, that contains Group k wires in Panel 1, in the effective
development length
N
cr
= total number of cracked wires in the effective development length that are redeveloped in Panel 1 after
breaking
N
cr,k
= effective number of broken cracked wires that are Stage k in Panel 1 and can be redeveloped
N
c,k
= total number of discrete cracked wires in the effective development length in segments that are Group k in
Panel 1
N
eff
= effective number of unbroken wires in the evaluated panel
NS = total number of segments in the cross-section of the cable
n
j
= number of wires in segment j
p
5
= fraction of unbroken wires in the evaluated panel represented by Group 5
p
c,ij
= fraction of wires cracked in segment j, panel i
p
c,k
p
k
= fraction of Stage k wires in the cable
p
cr
(s) = fraction of cracked wires that have broken at stress level s and are redeveloped in Panel 1
q
ij
= fraction of uncracked wires in segment j, panel i
R
b
= cable strength attributable to broken wires in adjacent panels.
r
ij
= fraction of wires in segment j that are cracked in panel i but that are not cracked in panels less than i
B-5
r
ki
= fraction of wires cracked in panel i in all segments that are Group k in Panel 1
R
u
= cable strength attributable to unbroken wires
T
cr
(s) = maximum force in the broken cracked wires that can be redeveloped in the evaluated panel at stress s
s2
= sample mean tensile strength of Group 2 wires
Note: p
c,ij
, q
ij
, r
ij
and e
ij
are also called p
i
, q
i
, r
i
and e
i
respectively in the derivation in Article B.4.1.1
B.3 ORDER OF FAILURE OF CRACKED WIRES
It is assumed that wires in a given stage in Panel 1 will break before wires at the same stage or better in an adjacent
panel, and equally that a wire in an adjacent panel will break before a better stage wire in Panel 1. Cracks have not
been found in Stage 1 or Stage 2 wires, and therefore only Stage 3 and Stage 4 cracked wires are considered in this
analysis, and they are assumed to share a common tensile strength distribution curve. If the cracked wires in the two
stages have different curves, they should be treated separately. A determination must be made about where they will
break first. It is suggested that a single tensile strength distribution curve be calculated for the combined groups of
wires, thus simplifying the calculation. The net effect is that a wire with cracks along its length will break as near to
Panel 1 as possible.
B.4 NUMBER OF DISCRETE CRACKED WIRES AND THEIR REDEVELOPMENT
B.4.1 All Panels in Effective Development Length Inspected
The segments in the cable cross-section are numbered from 1 to NS, the total number of segments. The integer j is used
to identify the segments.
Cable panels are numbered from 1 to L
e
, where L
e
is the number of panels in the effective development length. The
panel numbers begin at Panel 1, which is the panel under evaluation, located at the center of the effective development
length. Panel numbers increase as the distance from Panel 1 increases on both sides, as shown in Figure B.1-1. The
integer i is used to identify the panels.
When all panels within the effective development length are inspected, the corrosion stage and thus the wire group
observed in each segment of the cable should be known.
In the following derivation, some shorthand notation is used. The subscript j is omitted because the derivation is for a
single segment, j, of the cable, and the subscript c is omitted for cracked wires.
p
i
= p
c,ij
= fraction of wires in segment j in panel i that are cracked
q
i
= 1-p
i
= fraction of wires in segment j in panel i that are not cracked
e
i
= e
ij
= fraction of wires in segment j in panel i that are not cracked and are also not cracked in
panels less than i
r
i
= r
ij
= fraction of wires in segment j that are cracked in panel i but that are not cracked in panels
less than i
The reduction is calculated for each group of wires, k. The fraction of discrete cracked wires, r
i
, in each panel is then
multiplied by the appropriate factor to arrive at the effective number of cracked wires that fail in the calculation of
strength, and that are redeveloped in Panel 1. This factor is zero for Panel 1; C
d
for panels 2 and 3, 2C
d
for Panels 4
To account for the effect of deteriorated wires in adjacent panels, the number of wires in each corrosion stage in Panel 1
is reduced by the effective number of cracked wires in that stage in all panels in the effective development length. The
number of discrete cracked wires is the total number of cracked wires that have an impact on Panel 1when they break in
the effective development length, and is equal to the number of cracked wires in each panel that are not cracked in lower
numbered panels ( the black areas in Figure B.4.1.1-1). A wire that is cracked in Panel i and also cracked in a panel less
than i will break in that other panel and is already counted there. It is not counted again in Panel i.
B-6
and 5, etc., where C
d
is the fraction of 95% of the mean tensile strength of Group 2 wires that is developed through
friction at each cable band.
B.4.1.1 DISCRETE CRACKED WIRES IN EFFECTIVE DEVELOPMENT LENGTH
The separation of Stage 3 and Stage 4 wires into uncracked and cracked sets in each segment in each panel is
illustrated in Figure 1, and is made as follows. In Panel 1, there is a set of cracked wires, r
1
= p
1
, and a set of uncracked
wires, e
1
= q
1
= (1 p
1
). In Panel 2, the first adjacent panel, a fraction of the cracked wires will probably be the same
wires as wires in each of these Panel 1 sets, in proportion to p
2
. Thus, p
2
x p
1
wires will correspond one-for-one with
cracked wires in Panel 1 and will break in Panel 1. These wires will not break again until they are fully redeveloped,
and they will not affect the strength of Panel 1. In this same Panel (Panel 2), however, there will be r
2
= p
2
x e
1
cracked
wires that do not correspond to cracked wires in Panel 1 and will break in Panel 2. They must be subtracted from Panel
1 wires of the appropriate stage in the calculation of uncracked, unbroken wires in Panel 1.
The calculations for the fraction of discrete cracked wires in each panel are tabulated in Table B.4.1.1-1. The values
of e
i
and r
i
are calculated in numerical panel order, starting from Panel 1. The table assumes an effective development
length of 7 panels and must be expanded or contracted for other values of the effective development length. Note that
panels separated by more than four cable bands are calculated as if no wires could be redeveloped. Wires can obviously
break in another panel more than 4 cable bands away from a break, but because they correspond one-for-one with broke
wires closer to Panel 1 that have already been considered, they can be ignored.
B-7

Figure B.4.1.1-1. Probable location of cracked wires in adjacent panels.
B-8
7 5 3 1 2 4 6
Corrosion Stage
g
7
g
5
g
3
g
1
g
2
g
4
g
6

Fraction Cracked,
p
i

p
7
p
5
p
3
p
1
p
2
p
4
p
6

Fraction Not
Cracked, qi
1-p
7
1-p
5
1-p
3
1-p
1
1-p
2
1-p
4
1-p
6

Net Not Cracked,
ei
Note that
redevelopment at
cable bands is not
considered in this
value
q
1
*
q
2
*
q
3
*
q
4
*
q
5
*
q
6
*
q
7

q
1
*
q
2
*
q
3
*
q
4
*
q
5

q
1
*
q
2
*
q
3

e
1

=

q
1

q
1
*
q
2

q
1
*
q
2
*
q
3
*
q
4

q
1
*
q
2
*
q
3
*
q
4
*
q
5
*
q
6

Additional
cracked wires in
segment, ri
p
7
*e
6
p
5
*e
4
p
3
*e
2
p
2
*e
1
p
4
*e
3
p
6
*e
5

The calculation of discrete cracked wires is made for each of the NS segments in the cable cross-section. For each
group of wires, k, in Panel 1, the total number of discrete cracked wires is required in the effective development length.
The number of discrete wires, regardless of group in Panel 1, that are cracked in a segment j is calculated by adding the
fraction of discrete cracked wires in that segment, r
ij
, for all panels in the effective development length, and
multiplying by the number of wires in segment j.

=
=
e
L
i
ij j j , c
r n n
1
(B.4.1.1-1)
where
n
c,j
= number of discrete cracked wires in segment j in the effective development length
L
e
= number of panels included in the effective development length
n
j
= number of wires in segment j
r
ij
= fraction of wires in segment j that are cracked in panel i but that are not cracked in panels less than i
i = number of a cable panel
j = number of a cable segment
The total number of discrete cracked wires corresponding to Stage k wires in Panel 1 is found by adding N
c
,
jk
for all
segments that contain Stage k wires in Panel 1.

=
=
NS
jk
jk c k c
n N
1
, ,
(B.4.1.1-2)
where
Table B.4.1.1-1 Calculation of discrete cracked wires
Panel Number, i
B-9
N
c,k
= total number of discrete cracked wires in the effective development length in segments that are Group k in
Panel 1
n
c,jk
= number of discrete cracked wires in a segment, j, that contains Group k wires in Panel 1, in the effective
development length
NS = total number of segments in the cross-section of the cable
jk = a segment j that contains Stage k wires in Panel 1
The cracked wires, N
c,k
, are subtracted from the wires in the appropriate stages in the strength calculation for Panel 1.
There may be cracked wires that correspond to Group 2 in Panel 1, N
c,2
, because a higher stage is present in a segment
in an adjacent panel that is Group 2 in Panel 1. The number of cracked Stage 4 wires in adjacent panels must be
adjusted for broken wires, because the number of cracked wires is calculated based on the entire cable. The adjustment
is made in proportion to the number of Stage 4 unbroken wires in Panel 1, N
04
, divided by the total number of Stage 4
wires in Panel 1, N
s4
.
B.4.1.2 EFFECTIVE NUMBER OF CRACKED WIRES THAT CAN BE REDEVELOPED
Also required in the calculation is the effective number of cracked wires that are redeveloped in Panel 1 when all the
cracked wires fail. This is the sum of discrete cracked wires in each segment in each panel multiplied by the
appropriate redevelopment factor, C
di
, for the panel

= =
=
NS
j
L
i
di ij j cr
e
C r n N
1 1
(B.4.1.2-1)
where
N
cr
= total number of cracked wires in the effective development length that can be redeveloped in Panel 1 after
breaking, if all cracked wires break
C
di
= wire redevelopment factor for panel i = C
d
, multiplied by the number of cable bands between panel i and
Panel 1
This number, N
cr
, is independent of the corrosion stage or group.
The number of redeveloped cracked wires, which includes both Stage 3 and Stage 4 wires, must also be adjusted for
broken wires in the cable. This single adjustment is the ratio of all unbroken wires in Panel 1, N
eff
, to the total number
of wires in the cable, N.
The fraction of broken cracked wires that is developed in Panel 1 is calculated as the effective number of redeveloped
cracked wires divided by the total number of discrete cracked wires.
The calculation of cable strength then follows the method given in the Limited Ductility Model or the Brittle-Wire
Model, with all cracked wires treated as another group (Group 5). At each value of the stress, the redeveloped force in
Panel 1 in broken cracked wires is added to the cable force, in the same manner as the redeveloped force in the in-
service broken wires was added. As before, the maximum cable force reached is the cable strength.
B.4.2 Only Panel 1 Inspected
When only selected panels are opened for inspection, and no observations are made in adjacent panels in the effective
development length, it is conservative to assume that the condition of all panels in that length is the same as that of the
evaluated panel. In this case, the calculation is considerably simplified. Only segments with Group 3 and Group 4
wires need be considered, and each of these groups can be combined into a single segment, which is continuous in the
effective development length.
B-10
For a segment that is Group k in Panel 1, the value of p
c,k
is the same in all panels in the effective development length.
Table B.4.1.1-1 shows that the fraction of discrete cracked wires in panel i is equal to the fraction of uncracked wires
in panel i-1, multiplied by the fraction of cracked wires in panel i. The fraction of uncracked wires in panel i-1 is equal
to the fraction of uncracked wires in Panel i-2, multiplied by one minus the fraction of cracked wires in panel i-1, etc.
The number of discrete cracked wires in panel i can be expressed as
1
, ,
) 1 (
=
i
k c k c ki
p p r (B.4.2.1-1)
where
r
ki
= fraction of wires cracked in panel i in all segments that are Group k in Panel 1
p
c,k
The number of discrete cracked wires in the effective development length in segments that are Group k in Panel 1 is

=
=
e
L
i
i
k , c k , c k k , c
) p ( p N N
1
1
0
1 (B.4.2.1-2)
where
N
0k
= number of unbroken Stage k wires in the evaluated panel
All the cracked wires are treated as a separate group, Group 5
=
=
4
2
, 5
k
k c
N N (B.4.2.1-3)
where
N
5
B.4.2.2 REDEVELOPMENT OF BROKEN CRACKED WIRES
Cracked wires that fail as the cable stress is increased in the calculation of cable strength can redevelop a part of their
strength in Panel 1. Assuming all the cracked wires are broken, redeveloped wires for each Stage k are
( )
=

=
e
L
i
di
i
k , c k , c k k , cr
C p p N N
1
1
0
1 (B.4.2.2-1)
where
N
cr,k
= effective number of broken cracked wires that are Stage k in Panel 1 that can be redeveloped after
breaking, if all cracked Stage k wires break
C
di
= wire redevelopment factor for Panel i = C
d
, multiplied by the number of cable bands between panel i and
Panel 1
The total effective number of redeveloped wires is

=
=
4
2 k
k , cr cr
N N (B.4.2.1-2)
The broken wires are subtracted in determining N
0k
, and hence the corrections for broken wires given in Articles
B.4.1.1. and B.4.1.2 are not required in this case.
B.4.2.1 DISCRETE CRACKED WIRES WITHIN EFFECTIVE DEVELOPMENT LENGTH
B-11
B.5 MAXIMUM CABLE TENSION THAT CAN BE DEVELOPED IN CRACKED WIRES
The effective fraction of cracked wires that will be redeveloped in Panel 1 is N
cr
/N
5
. This fraction is applied to the
fraction of cracked wires that have failed at a stress, s, to arrive at the fraction of cracked wires that have broken and
are redeveloped in Panel 1 at stress level s.
) ( 3 ) (
5
5
s F
N
N
s p
cr
cr
= (B.5-1)
where
p
cr
(s) = fraction of cracked wires that have broken at stress level s and are redeveloped in Panel 1.
N
cr
= total number of cracked wires in the effective development length that can be redeveloped in Panel 1 after
breaking
N
5
F3
5
(s) = fraction of cracked wires that are broken at stress s = Weibull cumulative distribution of tensile strength
for Group 5 (cracked) wires at stress s
This is an effective fraction, based on 95% of the mean tensile strength of Stage 2 wire; hence, the maximum force that
can be redeveloped in these wires is
) ( 95 . 0 ) (
2 5
s p a N p s T
cr s w eff cr
= (B.5-2)
where
T
cr
(s) = maximum force in the broken cracked wires that can be redeveloped in the evaluated panel at stress s
p
5
= fraction of unbroken wires in the evaluated panel represented by Group 5
N
eff
= effective number of unbroken wires in the evaluated panel
a
w
s2
= sample mean tensile strength of Group 2 wires
Substituting Equation B.5-1 into Equation B.5-2, and noting that p
5
multiplied by N
eff
= N
5
, results in
( ) ) s ( F . a N ) s ( T
s w cr cr 5 2
3 95 0 = (B.5-3)
This equation is given in Article 5.3.3.2.3 as Equation 5.3.3.2.3-1.
B.6 CABLE STRENGTH
The force redeveloped in broken cracked wires at stress s is added to the cable force at stress s calculated for unbroken
wires, and the maximum value of the sum plus the redeveloped strength of the in-service broken wires is the cable
strength.
( ) [ ]
b cr s c w eff
R s p p s F s a N R + + = ) ( 95 . 0 ) ( 1 max
2 5
(B.5-4)
in which
R = cable strength
R
b
= cable strength attributable to broken wires in adjacent panels.
N
5
F3
5
(s) = fraction of cracked wires that are broken at stress s
B-12
= Weibull cumulative distribution of tensile strength for Group 5 cracked wires at stress s
C-1
CONTENTS
APPENDIX C ILLUSTRATIVE EXAMPLES.....................................................................................C-1
C.1 INTRODUCTION......................................................................................................................................C-2
C.2 EXAMPLE 1 SIMPLIFIED MODEL ......................................................................................................C-3
C.2.1 Simplified model assuming that adjacent panels are perfect.........................................................C-4
C.3 EXAMPLE 2 BRITTLE-WIRE MODEL, ONE PANEL INSPECTED...............................................C-5
C.4 EXAMPLE 2A (EXAMPLE 2 - CONDENSED FORMAT)................................................................... C-6
C.5 EXAMPLE 3 BRITTLE WIRE MODEL, ENTIRE EFFECTIVE LENGTH INSPECTED.............C-7
C.6 EXAMPLE 3A (EXAMPLE 3 - CONDENSED FORMAT)................................................................... C-8
C.7 SUMMARY.................................................................................................................................................C-9
C.8 EXPLANATION OF STEP NUMBERING........................................................................................... C-10
C.8.1 Inspection (I) ....................................................................................................................................C-10
C.8.2 Data Reduction (D). ......................................................................................................................... C-10
C.8.3 Estimation of Cable Strength (S).................................................................................................... C-10
C.8.3.1 Simplified Model (Steps 2 and 3 not used) ...................................................................... C-10
C.8.3.2 Brittle-Wire Model ........................................................................................................... C-10
C-2
C.1 INTRODUCTION
Three different calculations for estimating the strength of a cable panel on the Centennial Bridge are presented in
Appendix C. The bridge and the data are entirely fictional, although the same level of deterioration has been found in
some real inspections. The events are reported as if they had actually occurred (e.g., so many samples taken, so much
deterioration found, etc.). The number of samples was deliberately reduced to limit the size of the tables, but obtaining
less than the recommended number of samples can be defended as reasonable, considering that this is only the first
internal inspection.
Three locations on each cable were selected for inspection, following the recommendations in Article 2.2.5.1. Ten
Stage 1 and 15 Stage 2 wires were sampled. After a significant number of Stage 4 wires were found in two of the
panels, 20 Stage 3 and 30 Stage 4 wires were also sampled. Table 24.3.5.1-1 requires even more samples: 35 Stage 3
and 60 Stage 4, but opening only 6 panels reduced the number of samples that could be removed. It is expected that
more samples will be obtained during the next scheduled inspection, when more panels are opened.
The strength of each opened panel was calculated using the Simplified Model presented in Article 5.3.3.1, which
excludes broken and cracked wires. It was determined from these calculations that the lowest cable strength occurred
at the low point of the south cable in Panel 77-78. In the third sample calculation, it is assumed that the owner was
concerned enough about the apparent low strength of this panel to open up additional panels for inspection. The other
two examples are based on the original plan of 6 locations. Calculations for the estimated strength of this panel are
presented as follows:
Example 1: Simplified Model, with all panels in the effective development length assumed alike, since
adjacent panels were not inspected (Article 5.3.3.1)
Example 2: Brittle-Wire Model, with same assumptions (Articles 5.3.3.2, 5.3.4 and 5.3.2.4)
Example 3: Brittle-Wire Model, with all panels in the effective development length inspected, using a more
elaborate method for determining the effects of deterioration in adjacent panels (Articles 5.3.3.2, 5.3.4 and
Appendix B).
In all three examples, the results are often given to more significant figures than is necessary, implying an accuracy
that does not exist. For instance, the mean tensile strength of wires is shown to 1 decimal place, while the nearest
integer is sufficient for the calculation. This is also true of the standard deviation. The cable strength is shown to the
nearest integer, implying that it is known to 5 significant figures, whereas 3 figures, or the nearest 1%, are sufficient.
Various methods for checking the accuracy of the calculations are part of the reason for retaining significant figures
beyond their apparent effectualness. For example:
The number of wires in each ring, when added together, should result in exactly the total number of wires in
the cable. This is a check on the formulas entered into the cells of the spreadsheet. If the total is not exact, the
error should be found. This is one of the reasons that the number of wires in each ring is not rounded.
On Page EX1-12, the number of wires that are less than Stage 1 or greater than Stage 4 must be exactly zero,
or the data are incorrect.
The total number of wires in each stage must add up exactly to the number of wires in the cable, or there is an
error.
It is primarily for these checks that the calculations are shown with so much seeming accuracy. The results should be
rounded in the final summary. As stated above, 1% is close enough.
C-3
C.2 EXAMPLE 1 SIMPLIFIED MODEL
Each cable of the Centennial Bridge is composed of 9,990 galvanized steel parallel wires.
Each wire is 0.196 inches in diameter, including the zinc coating. The nominal diameter of the wire before galvanizing
is 0.192 inches. Three panels on each cable were selected for inspection, as shown in the Inspection Location Diagram
on Page EX1-02. The condition of the wires in Panel 77-78 was found to be severe enough to warrant removal of the
cable band at panel point 77 to facilitate inspection of the entire cable cross-section. The steps required for this
representative inspection (prefixed I), data reduction and testing (prefixed D), and calculation of cable strength
(prefixed SS) are described on the following pages, and then illustrated. The page numbers in the follow paragraphs
refer to the calculation page numbers that can be found in the upper right corner of the calculation sheets, e.g., EX1-02.
Step I-1 Prepare inspection forms. Prior to the preparation of inspection forms, the number of rings of wire in the
cable and the number of wires in each ring were estimated. This calculation is shown on Page EX1-03. The relevant
forms are shown in the next step, on Pages EX1-04 and EX1-05.
Step I-2 Record the inspection observations. The condition of the wires at one of the inspected locations along the
length of Panel 77-78, inside wedged opening No. 5 (at 6:00), was recorded on the Field Inspection Sheet shown on
Page EX1-04; the conditions at other locations were similarly recorded, but are not shown in the example.
Step I-3 Remove sample wires for testing and measure retraction of the cut ends. The separations of the ends of four
wires, at the location of the first cut for removal of samples, were measured and recorded on another field inspection
form showing the cable cross-section, along with the location of the sample wires and wires found broken in the cable.
This inspection form is shown on Page EX1-05.
Step D-1 Calculate the redevelopment coefficient and the effective development length. The effective development
length is calculated on Pages EX1-06 and EX1-07, using the measurements of wire retraction made during the
inspection.
Step D-2 Test the sample wires and calculate the tensile strength distribution (mean and standard deviation) of each
group of wires. Many sample wires were removed from the cable for testing. Ten Stage 1 and 15 Stage 2 wires were
selected, as well as 18 Stage 3 and 30 Stage 4 wires. Although the deterioration of the cable was observed to be severe
in Panel 77-78, the limited number of panels opened in this inspection was insufficient for a larger sampling of the
wires. The condition of this panel shows that a more intense investigation is needed at the next internal inspection,
during which at least 35 Stage 3 and 60 Stage 4 wires should be sampled.
The results of tensile tests on these samples are shown on Pages EX1-08 to EX1-11; the test results on an individual
specimen from one sample are shown on Page EX1-08, along with the calculation of the estimated minimum tensile
strength in one panel length, or 41 feet. When a crack is present in one or more specimens, the calculation of the
estimated minimum strength using Equation 4.4.3.2-1 is not valid, and the lowest strength found for a cracked
specimen is used instead, as shown on Page EX1-09. These values are calculated for each sample, and the results
carried to Pages EX1-10 and EX1-11, where the sample means and sample standard deviations of the tensile strengths
for each stage of corrosion are calculated.
The fraction of samples in each corrosion stage that contain one or more cracks is calculated on Page EX1-11.
Step D-3 Determine the number of wires in each corrosion stage. The corrosion stage of each wire in the wedged
openings is tabulated on Page EX1-12. Upon reviewing all inspection records for wedge line 5, in Panel 77-78, Stage3
and Stage 4 were found to extend one to two wires deeper into the cable at another panel segment than at the middle
panel for which the example on Page EX1-04 is shown, and this is reflected in the data on Page EX1-12. There is one
line on this spreadsheet for each ring in the cable, and the estimated number of wires is entered for each ring. Two
columns represent each wedge, one for the left-hand side and one for the right-hand side of the opening formed by the
wedge. The fraction of the circle corresponding to the arc subtended by each half-sector is given at the top of each
column. The number of wires in each stage in each ring is calculated by formulas in the appropriate cells, and the totals
for the entire cable given at the bottom of the spreadsheet.
C-4
These numbers are carried to Pages EX1-14 and EX1-16. Page EX1-13 presents a corrosion map of the cable cross-
section in the inspected panel.
All data up to this calculation on Page EX1-14 are the same for all the examples, and these pages will not be repeated
for Example 2 and Example 3.
Step D-4 Determine the number of broken wires in the effective development length and the effective number of
unbroken wires in the cable. In this inspection, broken wires were found only on the periphery of the cable, up to 6
wires from the surface. The spreadsheet for broken wires is therefore not needed.
The number of broken wires and the number of unbroken wires in the cable are estimated directly on Page EX1-14.
Equation 4.3.3.2-1 is used to estimate the number of broken wires, because no wires were found broken beyond the
sixth ring. The depth at which no broken wires are found is 7 wires (i.e., broken wires were found 6 layers into the
cable), and d
0
equals 7. Of the 8 broken wires found, 6 were repaired.
The number of broken wires in all the panels in the effective development length is assumed to be the same as in the
inspected panel, because only that panel was inspected. It follows that only wires in the inspected panel were repaired.
Step D-5 Determine the number of discrete cracked wires in the effective development length. The total fraction (and
total number) of wires in each stage that are cracked in the effective development length is found using the graph in
Figure 5.3.2.4.1-1. These calculations are shown on Pages EX1-15. This information is added to the summary sheet
shown on Page EX1-16.
Step SS-1 Determine the fraction of the cable in each group of wires. The fraction of the cable in each stage of
corrosion, and the fraction represented by each group of wires, is calculated on Page EX1-16. The testing laboratory
reported that 50% of Stage 4 wires and 5% of Stage 3 wires contained preexisting cracks (i.e., they were observed to
be cracked before testing). The cracked wires represented by these fractions are subtracted from the total number of
wires in the appropriate stages to determine the net number of wires per stage that are not broken or cracked. The
cracked wires are added together to form Group 5. In the Simplified Model, broken and cracked wires are ignored, and
thus Group 5 has no wires.
Steps SS-2 (Weibull parameters) and SS-3 (Strength of broken wires) are not used in the Simplified Model.
Step SS-4 Determine the combined distribution of the tensile strength of unbroken and uncracked wires. The mean
and standard deviation of the tensile strength of the combined groups of wires are calculated on Page EX1-17 using
Equation 5.3.3.1.1-1 and Equation 5.3.3.1.1-2; the coefficient of variation (standard deviation divided by the mean) is
also calculated.
Step SS-5 Estimate the cable strength. On Page EX1-18, the strength reduction factor for use in Equation 5.3.3.1.2-1
is found from Figure 5.3.3.1.2-1, and the cable strength is calculated as 38,968 kips. The result is rounded to 3
significant figures, or 39,000 kips.
C.2.1 Simplified model assuming that adjacent panels are perfect
The Simplified Model can be used to quickly and easily identify the worst of the inspected panels for more detailed
analysis, by assuming that only the inspected panel is deteriorated and all other panels are perfect. This calculation is
shown on Page EX1-19. Only broken and cracked wires in the evaluated panel are considered; the entire calculation
can be made on a single sheet. The effective development length is taken to be 1 panel because there are no broken
wires outside this panel to be redeveloped. The number of cracked wires in each corrosion stage is the fraction cracked
multiplied by the number if wires in that stage. The calculations on Pages EX1-14, 16, 17 and 18 are shown on Page
EX1-19; Calculation Page EX1-15 is not required.
C-5
C.3 EXAMPLE 2 BRITTLE-WIRE MODEL, ONE PANEL INSPECTED
In Example 2, the investigator seeks to estimate the lower bound of the cable strength by including the effect of
adjacent panels. The condition in adjacent panels is assumed conservatively to be the same as in the inspected panel,
because only the latter has been inspected. Pages EX1-02 to EX1-14 also apply to this example, and are not repeated.
The example starts with Step D-4.
unbroken wires in the cable. The number of broken wires and the number of unbroken wires in the cable are estimated
directly on Page EX2-02. Equation 4.3.3.2-1 is used to estimate the number of broken wires, because no wires were
found at a depth greater than 6 rings. This page is the same as Page EX1-15, but is included here because some of the
data calculated on this page are required on the following pages. This information is added to the summary sheet
shown on Page EX2-04. Of the 8 broken wires found, 7 were repaired.
Step D-5 Determine the number of discrete cracked wires in the effective development length. The number of
discrete cracked wires is calculated on Page EX1-15, as described previously. This information is added to the
summary sheet shown on Page EX2-04.
Step D-6 Determine the number of discrete cracked wires that can be redeveloped in the evaluated panel when they
break in an adjacent panel. The effective fraction of discrete cracked wires that will be redeveloped in the evaluated
panel if they break is found by using Figure 5.3.2.4.2-1, and their effective number is calculated for each corrosion
stage. The total fraction of wires that can be redeveloped is calculated on Page EX2-03. This fraction is used in the
calculation of the cable tension at a given stress, found on Page EX2-08, and in the calculation of the cable strength on
Page EX2-09.
Step BS-1 Determine the fraction of the cable in each group of wires. The fraction of the cable in each stage of
corrosion and the number of wires represented by each group are calculated on Page EX2-04. The number of cracked
wires is subtracted from the total wires one stage at a time to determine the net number of wires per stage that are not
broken or cracked. Unlike the Simplified Model, broken and cracked wires are included in the calculation, and are
used to form Group 5.
Step BS-2 Determine the Weibull coefficients. Coefficients for the Weibull distribution of each Group of wires are
calculated on Page EX2-05, using the method given in Article A.4.2, Appendix A. The Microsoft Excel Spreadsheet
Program is used for this purpose, and the tool, Solver, is used to determine the values of the parameters.
Step BS-3 Determine the force that can be redeveloped in wires found broken in adjacent panels. The maximum
force that wires found broken in the effective development length can sustain in the evaluated panel because of friction
among wires developed at the cable bands is calculated on Page EX2-06, using Equation 5.3.4-3.
Step BS-4 Determine the cable force at a specific value of stress. This calculation is shown on two pages, and is
divided into two steps as follows.
Step BS-4A Develop the cumulative compound distribution of the tensile strength. The Weibull parameters
calculated on Page EX2-05, along with the fractions for each group of wires, are applied on Page EX2-07, using
Equation 5.3.3.2.1-1 to evaluate the cumulative compound distribution curve for the entire cable at a specific stress.
The calculation on this page is for the value of the distribution at a stress of 220 ksi.
Step BS-4B Determine the cable force at a specific value of stress. The cumulative distribution is calculated on Page
EX2-07 (the calculations are shown again in condensed form). The capacity of redeveloped wires found broken in the
cable from Page EX2-06 and the fraction of Stage 5 wires that have failed at Stress s (F3
5
(s)) from Page EX2-07 are
used on Page EX2-08 to determine the force in the cable when the stress is 220 ksi.
Step BS-5 Determine the estimated cable strength. The estimated cable strength is calculated on Page EX2-09. The
Solver function is used to maximize the cable force by varying the wire stress. The cable strength is found to be
50,824 kips, which rounds to 50,800 kips.
C-6
C.4 EXAMPLE 2A (EXAMPLE 2 - CONDENSED FORMAT)
On Pages EX2A-01 to EX2A-09, the calculations detailed in Example 2 are given in condensed form on spreadsheets.
Steps D-4, D-5, D-6, BS-1 and BS-2. Page EX2A-02 shows the data from the inspection and calculated values (from
Pages EX2-02 to EX2-05) for use in the spreadsheets that follow.
Steps BS-3 to 5. Page EX2A-03 is a calculation of cable strength using the technique on Pages EX2-07 to EX2-09,
while Pages EX2A-04 to EX2A-07 illustrate the steps required to develop both the cable strength and cable force vs.
strain curve. On these 4 pages, each line of the spreadsheet calculates the cable tension at one value of the stress, the
same as on Page EX2A-03. The calculation of redeveloped cracked wires that break as the stress is increased uses
Equation B.5-1 and Equation B.5-2, instead of Equation 5.3.3.2.3-1, for convenience.
The stress-strain curve of the wires and tensile strength distribution curves are shown on Page EX2A-08; Page EX2A-
09 includes curves for cable force vs. stress and strain.
C-7
C.5 EXAMPLE 3 BRITTLE WIRE MODEL, ENTIRE EFFECTIVE LENGTH INSPECTED
When all the panels in the effective development length have been inspected, it is possible to estimate cable strength
more accurately than in Example 2, using the method in Appendix B. The example applies whenever the entire length
of the cable is opened for inspection, for instance during a maintenance operation. In this instance, the panels adjacent
to the worst panel were opened after a low strength estimate was found in one of the panels.
The initial calculations on Pages EX1-02 to EX1-14 are the same, and are not repeated here. The effective
development length, calculated on Page EX1-07, is 7 panels.
unbroken wires in the cable. In this example, since all panels in the effective development length are inspected, the
number of broken wires found in each panel is entered into the summary sheet on Page EX3-02. The number of broken
wires in each panel is estimated separately, and the number of repaired wires is entered for each panel. Then, the
number of unbroken wires in the cable is calculated, taking into account the number of broken wires that were repaired
in each panel.
Step D-5 Determine the number of discrete cracked wires in the effective development length, and
Step D-6 Determine the number of discrete cracked wires that can be redeveloped in the evaluated panel when they
break in an adjacent panel. The number of discrete cracked wires in the effective development length is calculated on
the large spreadsheet shown on Pages EX3-08 to EX3-23. Each line of this spreadsheet compares the corrosion stages
of a single cable segment along the entire effective development length, L
e
, determines where cracks are likely,
assumes the location of a crack, and calculates the probable number of discrete cracks in that segment and how many
will redevelop in the evaluated panel. Totals are made on the last page. A detailed calculation of a single line of the
spreadsheet is shown on Pages EX3-05 to EX3-07.
The number of discrete cracked wires, given at the bottom of Page EX3-23, is used to divide the wires into groups on
Page EX3-24; the effective number of cracked wires that can be redeveloped in the evaluated panel, also calculated on
Page EX3-23, is used on Page EX3-28 to calculate the fraction of wires that are redeveloped at a given stress level.
Step BS(adj)-1 Determine the fraction of the cable in each group of wires. The fraction of the cable in each stage of
corrosion and the number of wires represented by each fraction are calculated on Page EX3-24. The number of cracked
wires is subtracted from the total wires in each stage to determine the net number of wires per stage that are not broken
or cracked. Unlike the Simplified Model, broken and cracked wires are included in the calculation; cracked wires are
used to form Group 5.
Step BS(adj)-2 Determine the Weibull coefficients. The calculation of the parameters of the Weibull distribution is the
same as in Example 2, repeated here on Page EX3-25. These parameters are used on Page EX3-27, along with the
fraction of wires in each group, to develop the compound cumulative distribution of the tensile strength of the wires in
the cable at a stress of 220 ksi.
Step BS(adj)-3 Determine the force that can be redeveloped in wires found broken in adjacent panels. The force that
can be resisted in the evaluated panel by the broken wires is calculated using Equation 5.3.4-2 on Page EX3-26.
Step BS(adj)-4A Develop the cumulative compound distribution of the tensile strength. The Weibull parameters
calculated on Page EX3-25, along with the fractions for each group of wires, are applied on Page EX3-27, using
Equation 5.3.3.2.1-1 to evaluate the cumulative compound distribution curve for the entire cable at a specific stress.
The example calculation on this page is for the distribution at a stress of 220 ksi.
Steps BS(adj)-4B and 5 Determine the cable force at a specific value of stress and calculate the estimated cable
strength. The cable tension at a stress of 220 ksi is calculated on Page EX3-28; the estimated cable strength is
calculated on Page EX3-29 by varying the cable stress until a maximum cable tension is reached, resulting in a cable
strength of 53,092 kips (which should be rounded to 53,100 kips).
C-8
C.6 EXAMPLE 3A (EXAMPLE 3 - CONDENSED FORMAT)
A condensed form of Example 3, using spreadsheets, is given on Pages EX3A-02 to EX3A-07. This is in the same
form as Example 2A. The calculation for the effect of cracked wires in adjacent panels is the same as in Example 3
(Steps D(adj) 5 and 6, Pages EX3-08 to EX3-23) and is not repeated here.
Steps D(adj)-4, BS(adj)-1 and BS(adj)-2. Page EX3A-02 shows the data from the inspection and calculated values
(from Pages EX3-02 and EX3-24 to EX3-26) for use in the spreadsheets that follow.
Steps BS-3 to 5. Page EX3A-03 is a calculation of cable strength using the technique on Pages EX3-27 to EX3-29,
while Pages EX3A-04 to EX3A-07 illustrate the steps required to develop both the cable strength and cable force vs.
strain curve. On these 4 pages, each line of the spreadsheet calculates the cable tension at one value of the stress, the
same as on Page EX3-29.
The stress-strain curve of the wires and tensile strength distribution curves are shown on Page EX3A-08; Page EX3A-
09 includes curves of cable force vs. stress and strain.
C-9
C.7 SUMMARY
The initial strength of a new cable can be estimated using the Simplified Model by assuming that only Group 2 wires
are present. This calculation, which is not shown, results in a cable strength of 66,400 kips, for the cable in these
examples.
A severely corroded condition was assumed in the examples to demonstrate more clearly the differences in the three
calculations.
The Simplified Model, as expected, results in very low cable strength, 39,000 kips. This is the result of assuming that
all cracked and broken wires in the effective development length do not contribute to the cable strength;
redevelopment of force in these wires is assumed to be zero.
When the conservative assumption of all panels alike is made in Example 2, the predicted cable strength is 50,800
kips. The assumption that all panels are alike is reasonable, especially when the entire cable is not inspected and such a
severe condition is found. The difference between the cable strengths predicted by Examples 1 and 2 would be smaller
in a less deteriorated cable. Nevertheless, the Simplified Model is useful as a quick way of finding the worst inspected
panel.
Estimating the strength when all panels in the effective length have been inspected is far more complex, but it results
in a higher strength of 53,100 kips. The conditions assumed for this example approximate those that were found in an
actual cable; the evaluated panel was known to be the worst in the entire cable.
As a quick and easy way to identify the weakest panel of those inspected, the adjacent panels can be assumed to be
perfect in the Simplified Model, resulting in a strength of 53,600 kips. The similarity between this strength and that
found using the Brittle-Wire Model when all panels in the effective development length are inspected (Example 3), is
coincidental. Use of the Simplified Model (i.e., assuming adjacent panels are perfect), however, can provide a quick
indication of the weakest inspected panel. Example 3, which is complex, should be reserved for the worst panel, and
then it is only worth the effort when the cable strength is marginal. All other panels inspected can be evaluated using
the method shown in Example 2, or when the cable is less deteriorated, by the Simplified Model shown in Example 1.
When only a few panels are inspected, as in these examples, it is important to remember that it cannot be known
whether there is a panel in worse condition elsewhere in the cable.
C-10
C.8 EXPLANATION OF STEP NUMBERING
The steps required in the evaluation of a cable are numbered in these examples sequentially. The letters indicate the
phase of the investigation, the numbers indicate the order of the steps in the phase. Only those actions required for
evaluating the strength of the cable are included.
C.8.1 Inspection (I)
Step I-1 Prepare inspection forms
Step I-2 Record the inspection observations
Step I-3 Remove sample wires for testing and measure retraction of the cut ends
C.8.2 Data Reduction (D)
In this phase, sample wires are tested and the data obtained are reduced for use in estimating cable strength. Other
information obtained in the field is also reduced so that it is in the form required by the calculations. Whenever all
panels in the effective development length are inspected, the suffix (adj) is added (e.g., D(adj)-5), signifying that the
techniques given in Appendix B are used.
Step D-1 Calculate the redevelopment coefficient and the effective development length
Step D-2 Test the sample wires and calculate the tensile strength distribution (mean and standard deviation)
of each group of wires
Step D-3 Determine the number of wires in each corrosion stage
Step D-4 Determine the number of broken wires in the effective development length and the effective number
of unbroken wires in the cable
Step D-5 Determine the number of discrete cracked wires in the effective development length
Step D-6 Determine the number of discrete cracked wires that can be redeveloped in the evaluated panel
when they break in an adjacent panel
C.8.3 Estimation of Cable Strength (S)
In this phase, the data obtained in Steps D-1 to 6 are used to estimate the strength of the cable. The steps are prefixed
and suffixed by a letter indicating the model that is used in the analysis. The prefixes used are S for the Simplified
Model and B for the Brittle-Wire Model. When all panels in the effective development length are inspected, the suffix
(adj) is added (e.g., BS(adj)-1), signifying that the techniques given in Appendix B are used.
C.8.3.1 SIMPLIFIED MODEL (STEPS 2 AND 3 NOT USED)
Step SS-1 Determine the fraction of the cable in each group of wires
Step SS-4 Determine the combined distribution of the tensile strength of unbroken and uncracked wires
Step SS-5 Estimate the cable strength
C.8.3.2 BRITTLE-WIRE MODEL
Step BS-1 Determine the fraction of the cable in each group of wires
Step BS-2 Determine the Weibull coefficients
Step BS-3 Determine the force that can be redeveloped in wires found broken in adjacent panels
Step BS-4A Develop the cumulative compound distribution of the tensile strength
Step BS-4B Determine the cable force at a specific value of stress
Step BS-5 Determine the estimated cable strength
CALCULATION PAGE EX1-06
EXAMPLE CALCULATION NO 1
PROJECT: REDEVELOPMENT COEFFICIENT
CENTENNIAL BRIDGE EFFECTIVE DEVELOPMENT LENGTH
CABLE ARTICLES 4.5.1 AND 4.5.2
SPAN
PANEL
STEP D-1
CALCULATION OF REDEVELOPMENT COEFFICIENT
CALCULATION OF EFFECTIVE DEVELOPMENT LENGTH
REDEVELOPMENT OF WIRE TENSION THROUGH FRICTION AT CABLE BANDS
SOUTH
EAST MAIN SPAN
77-78
1 panel
X X X
N
T
panels
Panel Length, L
Friction force
Wire break
(Force = 0)
Panel N
T Panel 1 Panel n
Cable bands
Wire
Wire tension
fully restored
T
Friction force
at last band is
< or = X
Band N
T
+1 Band N
T
Band n Band 1
Wire tension
= nX
Wire tension =
X
Measured
Gap, d
w
Wire tension =
N
T
X
Friction
force
F
F
F
N
T
= number of panels required to develop
95% of mean strength of Group 2 wires
Evaluated
panel
Wire break
outside L
e
(Force = 0)
Cable bands
Wire
95% of mean strength of
Group 2 wires is
redeveloped in evaluated
panel when wire breaks
outside L
e
F
Effective development length, L
e
(symmetric about evaluated panel)
WIRE BREAKS OUTSIDE EFFECTIVE DEVELOPMENT LENGTH
(Number of bands = N
b
= N + 1)
T
D-1
CONTENTS
APPENDIX D SPLICING NEW WIRES........................................................................................................... D-1
D.1 SPLICING NEW WIRES INTO THE CABLE.......................................................................................D-2
D.1.1 Introduction........................................................................................................................................D-2
D.1.2 Typical specification for splicing wires............................................................................................D-2
D.1.2.1 WIRE SPLICING COMPONENTS................................................................................................ D-2
D.1.2.2 VERIFICATION TESTING OF WIRE SPLICING COMPONENTS ...........................................D-2
D.1.3 Wire splicing procedure ....................................................................................................................D-3
D.2 MEASURING THE WIRE TENSION..................................................................................................... D-4
D.2.1 Calibration..........................................................................................................................................D-4
D.2.2 Measuring procedure ........................................................................................................................D-4
D-2
D.1 SPLICING NEW WIRES INTO THE CABLE
D.1.1 Introduction
Whenever a sample wire is removed from the cable, or a broken wire is found there, it is necessary to replace the
removed or corroded section of wire by splicing in a new wire. The usual method is to attach two lengths of new wire
to the cut ends of the original wire with pressed-on, or swaged, ferrules, and then to connect the ends of the two new
wires with a threaded ferrule that acts like a turnbuckle. The pressed-on ferrules can develop at least 90% of the
strength of the wires. The threaded ferrule somewhat less, usually 75% to 85%, because of the section loss in the
threads: rolled threads develop strength near the upper limit, cut threads near the lower limit.
D.1.2 Typical specification for splicing wires
D.1.2.1 WIRE SPLICING COMPONENTS
The material for the new wires shall conform to ASTM A586 specifications for tensile strength, elongation in
a 10-inch-gage length and reduction of area. Single wires of sufficient length for the longest replacement shall
be provided, with a left-hand thread at one end and a right-hand thread at the other.
Threaded ferrules, pressed-on ferrules and equipment for the installation shall be manufactured and certified
subject to the engineers approval.
D.1.2.2 VERIFICATION TESTING OF WIRE SPLICING COMPONENTS
Verification testing shall be performed to confirm that the required specifications for these components have been met,
as follows:
Threaded ferrules shall be tested by fully connecting the ferrule to the threaded ends of two new wires. They
shall develop at least 75% of the ultimate strength of the new wires.
Pressed-on ferrules shall be tested by fully connecting the ferrule to the smooth ends of two new wires using
field equipment. Pressed-on ferrules shall develop at least 90% of the ultimate strength of the new wires.
Wires used in these tests shall have been previously tested and certified by the manufacturer and accepted.
Assemblies of ferrules and wires for testing shall be made by the contractors field personnel, in order to qualify
them to do the work.
Testing shall be performed by an approved testing laboratory.
The wire used in verification testing shall be provided by the contractor from the lot of wire to be used in the
actual work.
The sample for testing each component shall be 1% of each lot or 10 specimens, whichever is greater. A lot is
defined as the number of components contained in a single shipment.
Each component shall be packed separately, and each package shall be clearly labeled, noting the specific
component and the number of pieces. Packages that lack the required labeling may be returned to the place of
manufacture uninspected, untested, and unaccepted. No extra compensation shall be granted, nor extension of time
allowed for any delays attributable to the return of unlabeled or poorly labeled packages.
All of the specimens that constitute the first sample shall be tested before a second sample is taken. If more than
one specimen from the first sample fails the required testing, then the entire lot shall be rejected. If one specimen
of a sample fails the required testing, then an additional random sample, equal in number to the first sample, shall
be taken. If one specimen of the second sample fails the required testing, then the entire lot shall be rejected.
The contractor is notified that the testing procedure will render the tested specimen unfit for further use, and that
all tested specimens shall be discarded after testing has been completed. The contractor is advised to order a
sufficient number of components to allow for testing and discards and still have enough left to do the work.

D-3
D.1.3 Wire splicing procedure
Remove the wire sample or portion of broken wire as directed by the engineer.

Clean the ends of wires #1 and #4 and the unthreaded ends of wires #2 and #3 to bright metal for a length of 2
inches. The (spliced) ends of wires shall be kept clean and free of dirt, oil and any other foreign material
throughout the procedure.
Connect the ends of the existing wires #1 and #4 to splice wires #2 and #3 as follows (see Figure D.1.3-1):
Splice existing wire end #1 to the unthreaded end of splice wire #2 using pressed-on ferrule A. This
operation is shown in Figure D.1.3-2.
Splice the threaded end of splice wire # 2 to threaded end of splice wire #3 using threaded ferrule B.
Tension wires #3 and #4 to a load equal to the dead load tension in the cable using come-alongs equipped with
suitable wire grips. Trim the end of wire #3. A gap of approximately 1/4 to 3/8 of an inch shall be provided
between wires # 3 and # 4.
Disconnect the threaded ferrule between new wires #2 and #4.
Splice the unthreaded ends of wires #3 and #4 using pressed-on ferrule C.
Attach come-alongs and suitable wire grips to wires #2 and #3. Install threaded ferrule B. Tension the wire
assembly to a load equal to the dead load tension in the cable. Use threaded ferrule B as a take-up system. A
suitable dynamometer or load cell shall be used for wire tension control.
Existing wire after
sampling or removal of
corroded section
Pressed-on ferrules
New wires
Threaded ferrule
Existing wire after
corroded section
1 2 3 4
Wire
number
A C B
Existing wire after
corroded section
Pressed-on ferrules
New wires
Threaded ferrule
Existing wire after
corroded section
1 22 33 44
Wire
number
A C B
Figure D.1.3-1. Wire splicing.
Cut one new wire into two pieces, numbered #2 and #3 on Figure D.1.3-1. Piece #2 shall be of a length that places
the threaded ferrule at the desired location. One end of each of these pieces will be threaded, the other plain
D-4

Figure D.1.3-2. Applying ferrules with a hydraulic ram.
D.2 MEASURING THE WIRE TENSION
After the new wire is spliced into the cable, the tension shall be measured and adjusted to within 10% of the dead load
tension in the wires. The tension is measured with the device shown in Figure D.2-1, which shall be fabricated and
calibrated in advance of the work. Calibration is necessary rather than calculating the wire tension indicated by an
induced offset, because the flexural stiffness of the wire as well as its cast will affect the force vs. deflection curve.
D.2.1 Calibration
Calibration shall be performed by an approved testing laboratory. A piece of bridge wire 20 feet long shall be placed in
a testing machine and stressed in increments of 100 pounds. The force required to cause a deflection of the wire
between the ends of the device shall be measured using the same spring balance that will be used in the field. The
recommended offsets are given in Table D.2.1-1; these offsets should result in the required force of approximately 60
pounds.
Table D.2.1-1 Recommended offsets for various wire tensions
Wire Tension
(pounds)
Offset
(inches)
1500 0.70
2000 0.55
2500 0.45
A calibration curve shall be prepared, with applied force as the abscissa and wire tension as the ordinate. The offset
used shall be shown on the calibration curve.
D.2.2 Measuring procedure
The device is held against the wire, which is in contact with the grooves in the end plates. Only enough pressure is
used to achieve contact with the wire; the tube is not bent. The center hook is pulled outward with a spring balance
until the offset is exactly equal to the calibration offset and the applied force is measured. The wire tension is
determined from the calibration curve.
D-5
6 0
Force to deflect wire
measured by spring
balance
2 x 2 x 1/8 alum. tube
1/4 alum. plate
Weld
Cable wire
Ruler attached
to tube
Hole to clear rod
6 0
Force to deflect wire
measured by spring
balance
2 x 2 x 1/8 alum. tube
1/4 alum. plate
Weld
Cable wire
Ruler attached
to tube
Hole to clear rod
Figure D.2-1. Wire tension measuring device.
Abbreviations used without denitions in TRB publications:
AASHO American Association of State Highway Officials
AASHTO American Association of State Highway and Transportation Officials
APTA American Public Transportation Association
ASCE American Society of Civil Engineers
ASME American Society of Mechanical Engineers
ASTM American Society for Testing and Materials
ATA American Trucking Associations
CTAA Community Transportation Association of America
CTBSSP Commercial Truck and Bus Safety Synthesis Program
FAA Federal Aviation Administration
FHWA Federal Highway Administration
FMCSA Federal Motor Carrier Safety Administration
FRA Federal Railroad Administration
FTA Federal Transit Administration
IEEE Institute of Electrical and Electronics Engineers
ITE Institute of Transportation Engineers
NCHRP National Cooperative Highway Research Program
NCTRP National Cooperative Transit Research and Development Program
NHTSA National Highway Traffic Safety Administration
NTSB National Transportation Safety Board
SAE Society of Automotive Engineers
TCRP Transit Cooperative Research Program
TRB Transportation Research Board
U.S.DOT United States Department of Transportation

NCHRP 534

Uploaded by

Copyright:

Available Formats

NCHRP 534

Uploaded by

Document Information

Original Description:

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

NCHRP 534

Uploaded by

Copyright:

Available Formats

Guidelines for Inspection and

Strength Evaluation of Suspension

Round all edges

Round all edges

X , rounded to the next higher integer

zero is sufficient for the minimum value of the property.

As shown in Figure 4.5.2-1, the effective development

You might also like