Integral and Semi-Integral Bridges

Integral and Semi-integral Bridges Martin P. Burke Jr. PE

A John Wiley & Sons, Ltd., Publication

Integral and Semi-integral Bridges

Dedication

This work is affectionately dedicated to the memory of my father, Martin P. Burke Jr. of Pittsburgh, Pennsylvania. After I was born, he became known as Big Mart. But even when I grew a head taller than him, his associates, friends, and acquaintances continued to refer to and speak of him as Big Mart. Both they and I knew that they were thinking and speaking about the right person. And they were using the right name.

Integral and Semi-integral Bridges Martin P. Burke Jr. PE

A John Wiley & Sons, Ltd., Publication

This edition first published 2009 © 2009 Martin P. Burke Jr. Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing programme has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Registered office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom Editorial offices 9600 Garsington Road, Oxford, OX4 2DQ, United Kingdom 2121 State Avenue, Ames, Iowa 50014-8300, USA For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell. The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data Burke, Martin P. Integral and semi-integral bridges / Martin P. Burke. p. cm. Includes bibliographical references and index. ISBN 978-1-4051-9418-1 (hardback : alk. paper) 1. Concrete bridges—Design and construction. 2. Concrete bridges—Joints. I. Title. TG340.B88 2009 624.2—dc22 2009006791 A catalogue record for this book is available from the British Library. Set in 10/12 pt Minion by SNP Best-set Typesetter Ltd., Hong Kong Printed in Singapore 1

2009

Contents

Acknowledgments Introduction

vii xi

Chapter 1

Integral Bridges

Chapter 2

Bridge Damage and the Pavement G/P Phenomenon

21

Chapter 3

Integral Bridges: Attributes and Limitations

41

Chapter 4

Design of Integral Bridges: A Practitioner’s Approach

59

Chapter 5

Genesis of Integral Bridges

71

Chapter 6

Cracking of Concrete Decks and Other Problems with Integral-type Bridges

81

Integral Bridge Design in the Land of No Special Computations

99

Chapter 7

1

Chapter 8

Semi-integral Bridges: Movements and Forces

121

Chapter 9

Emergence of Semi-integral Bridges

139

Chapter 10

Elementalistic and Holistic Views for the Evaluation and Design of Structure Movement Systems

157

Awareness of Reality in Bridge Design

185

Appendix 1 The Pavement Growth/Pressure Phenomenon: The Neglected Aspect of Jointed Pavement Behavior

215

Appendix 2 Glossary

243

Appendix 3 Captions for Photographs

245

Index

247

Chapter 11

v

Acknowledgments

Grateful acknowledgment is made to the publishers of the papers identified below for their permission to use all or parts of them for the various chapters of this book. It should be understood that such permission does not constitute endorsement by the publishers of any statement, method, or practice given or recommended. These have been and must be the full responsibility of the author. However, it should also be understood that those who interpret and implement the opinions and advice given in this book do so with the realization of their complete responsibility for such interpretation and implementation. The author has been diligent in his efforts to avoid errors and to eliminate inconsistencies in this work. However, if and where he has failed, he would appreciate being informed by readers so that this work can be appropriately revised for possible later editions. Chapter 1: Transportation Research Board of the National Academies, Washington, D.C., for “Integral Bridges,” from Transportation Research Record: Journal of the Transportation Research Board (Transportation Research Record) No. 1275, 1990, pp. 53–61. Also, Ohio Department of Transportation, for “Integral Bridges – Development and Design,” Proceedings, Ohio Transportation Engineering Conference, Columbus, Ohio, 1989. Chapter 2: American Concrete Institute, Farmington Hills, Michigan, for “Reducing Bridge Damage Caused by Pavement Forces,” from Concrete International, January, 2004, Vol. 26, No 1, pp, 53–57, and February, 2004, Vol. 26, No. 2, pp. 83–89. Chapter 3: Transportation Research Board of the National Academies, Washington, D.C., for “Integral Bridges: Attributes and Limitations,” from Transportation Research Record, No. 1393, 1993, pp. 1–8. Chapter 4: American Concrete Institute, Farmington Hills, Michigan, for “Design of Integral Concrete Bridges: A Practitioner’s Approach,” from Concrete International, June, 1993, Vol. 15, No. 6, pp. 37–42. Chapter 5: American Concrete Institute, Farmington Hills, Michigan. for “Genesis of Integral Bridges in Ohio,” from Concrete International, July, 1996, Vol. 18, No. 7, pp. 48–51. Chapter 6: Transportation Research Board of the National Academies, Washington, D.C., for “Cracking of Concrete Decks and Other Problems with Integral-Type Bridges,” from Transportation Research Record No. 1688, 1999, pp. 131–138. Chapter 8: Transportation Research Board of the National Academies, Washington, D.C., for “Semi-Integral Bridges: Movements and Forces,” from Transportation vii

viii

Acknowledgments

Research Record No. 1460, 1994, pp. 1–7. Also, Spon Press for “Semi-Integral Bridges: A Concept whose Time Has Come,” from Continuous and Integral Bridges, 1994, pp. 213–224. Chapter 9: Transportation Research Board of the National Academies, Washington, D.C., for “Emergence of Semi-Integral Bridges,” from Transportation Research Record No. 1594, 1997, pp. 179–186. Also, Charles S. Gloyd co-author of “The Emergence of Semi-Integral Bridges.” Chapter 10: Transportation Research Board of the National Academies, Washington, D.C., for “Structure Movement Systems Approach to Effective Bridge Design,” from Transportation Research Record No. 1594, 1997, pp. 147–153. Chapter 11: Engineers’ Society of Western Pennsylvania, Pittsburgh, Pennsylvania, for “Awareness of Reality in Bridge Design,” Proceedings, 11th International Bridge Conference, 1994, pp. 271–283. Appendix 1: Transportation Research Board of the National Academies, Washington, D.C., for “Pavement Pressure Generation: Neglected Aspect of Jointed Pavement Behavior,” from Transportation Research Record No. 1627, 1998, pp. 22–28. This work would not have been possible without the original encouragement and assistance the author received from: the executives and staff members of Burgess and Niple, Engineers and Architects (B&N); the many professional colleagues of the General Structures Committee of the Transportation Research Board of the National Academies who reviewed and critiqued the original papers that formed the basis for the present work; and the many individuals and organizations from throughout the United States and abroad (some of whom are identified below) who graciously responded to the author’s requests by contributing comments, data, drawings, and photographs that were used herein as figure illustrations. The author greatly appreciates the assistance of Anthony Allbery, Dan Landbo, and Chris Carlisle of B&N who helped to prepare some of the drawings used to illustrate this work. The author would also like to take this opportunity to recognize the original structure conceptions and evolutionary contributions made by his many deceased professional mentors and colleagues of the Ohio Department of Transportation, and by the chief engineers of the Washington State and Tennessee Departments of Transportation. Over the years, they and he have shared the same concerns that motivated the origination and development of integral and semi-integral bridges. These tough and durable little bridges are now not only serving many of the nation’s state transportation systems, but also beginning to serve the transportation systems of other nations as well. If it were possible, they would be pleased to learn that their bridge pioneering efforts have had such huge and beneficial consequences. Cover: Edward P. Wasserman and photographer George Hornel, Tennessee Department of Transportation. Introduction: Paul Kinderman, Washington State Department of Transportation (photograph). Chapter 1: Edward P. Wasserman and photographer George Hornel, Tennessee Department of Transportation (photograph and Figure 1.3).

Acknowledgments

ix

Chapter 2: Dennis W. Heckman, Missouri Department of Transportation (photograph); Jack Mecklenborg (Figure 2.1); William Detrict, Indian Department of Transportation (J. F. K. Bridge data); Paul Hasselquist, Ted Barber and Jimmy Camp, New Mexico Department of Transportation (Figures 2.5 and 2.6); Shakir Shatnawi, California Department of Transportation (Figure 2.8d). Chapter 3: Yuqiao Yang, photographer (photograph). Chapter 4: East Nippon Expressway Company Ltd., with all rights reserved (photograph). Chapter 6: Jai Lee, B&N (photograph); I. H. P . J. Taylor, Tarmac Precast Concrete Ltd, UK (Figures 6.5 and 6.6). Chapter 7: I. H. P. J. Taylor, Tarmac Precast Ltd, UK (photograph); Bruce Ford, City of Akron, Ohio, (Figure 7.1); Mitch McCoy, McCoy Associates (Figure 7.3). Chapter 8: Greg Baird, courtesy of Ohio Department of Transportation (photograph and Figure 8.6). Chapter 9: Marc C. Eberhard, University of Washington State (photograph). Chapter 10: Dale Poorman, B&N (Figures 10.11 and 10.12). Chapter 11: Ronald K. Mattox, Gresham, Smith Partners (photograph); Marc C. Eberhard, University of Washington State (Figure 11.8); 1832 Drawing by Charles Graham (Figure 11.12, upper right); Pittsburgh Post Gazette, 2008, Copyright, all rights reserved. Reprinted with permission (Figure 11.12, middle left); Carnegie Library of Pittsburgh (All rights reserved. Unauthorized reproduction or usage prohibited.) (Figure 11.12, middle right); Greg Panza, Mount Washington Community Development Corporation, Pittsburgh (Figure 11.12, bottom). Appendix: 1: John Charman, courtesy of the Highways Agency, UK (photograph); Les Hawker, courtesy of the Highways Agency, UK (Figure A1.4); Toronto Star newspaper and photographer Paul Regan (Figures A1.5 and A1.9). Appendix 2: Bala Tharmabala and Sean Morris, Ontario Ministry of Transportation (photograph).

Introduction The final success of our journey through life depends upon how much we are willing to learn from others. There is not time enough in a lifetime to learn from our own experience alone everything we need to know. A. B. ZuTavern

Captions for the first page photograph of each section of this book are given in Appendix 3.

This book is not a primer on the analysis, design, and construction of continuous bridges, or on the design of many of the common components of integral and semi-integral bridges, components that are typical of all deck-type highway bridges. These subjects are described and discussed in many excellent textbooks that have been developed and published by others especially for that purpose. Rather, this book focuses on those subjects that are of significance for the design and construction of integral and semi-integral bridges, subjects that generally are not described and discussed elsewhere. In brief, these subjects include but are not limited to the following: Chapter 1: The evolution of deck-type highway bridges in the United States is traced from the jointed single-span bridges of the early 1930s to the fully integral bridges of the late 1990s. Also described are a few recent examples of the conversion of existing jointed bridges to integral types of construction. Chapter 2: The uncontrolled G/P (growth/pressure) phenomenon is probably responsible for more pavement and bridge damage than any other cause with xi

xii

Introduction

the exception of de-icing chemical deterioration. Yet its cause and its characteristics are not familiar to most bridge design engineers and bridge administrators. This chapter describes the effect of this phenomenon on three different bridge types. These examples are given as a warning of similar damage that may be sustained by other bridges unless the phenomenon is controlled, or unless more compressive-resistant integral or semi-integral bridges are built instead of the more pressure-vulnerable jointed bridges. Chapter 3: An elaboration about the many attributes and few limitations of integraltype bridges is given that should be considered not only in evaluating the suitability of integral bridges for particular applications but also during their design and construction. Chapter 4: Analysis and design procedures and research findings that form the basis for a pragmatic bridge design engineer’s approach to the design of a limited range of integral bridges are discussed. Some of the primary and secondary stresses that affect these structures are also discussed, including shrinkage, creep, passive pressure, settlement, thermal gradients, buoyancy, earthquakes, etc. Chapter 5: As strange as it may seem, the conceptual background developed for large deck-type, open-spandrel, rib-arch bridges appears to have been the primary inspiration for the design and construction of the 1938 Teens Run Bridge of Gallia County, Ohio, the first fully integral deck-type highway bridge constructed in the United States. This chapter documents the early concerns and design decisions that resulted in the design and construction of this remarkable little historic structure. Chapter 6: Although integral and semi-integral deck-type highway bridges are simple in concept and easy to construct, there are enough structural differences between them and their jointed counterparts that some unique problems arise, especially during construction, which generally are not anticipated by those building them for the first time. This chapter describes some of these problems so that they can be anticipated and prevented. Chapter 7: This chapter describes the “Land of No Special Computations.” It also describes some of the various problem-solving techniques used by pragmatic bridge design engineers to achieve successful bridge designs in the Land of No Special Computations. These techniques become particularly important when considering the significance of secondary stresses during the analysis and design of typical integral bridges. Chapter 8: During the expansion of skewed integral and semi-integral bridges, their superstructures are progressively forced to rotate in a horizontal plane toward their acute corners. This chapter describes this behavior and provides a simplified procedure for estimating the magnitude of the forces involved; it also describes how these forces can be suitably and economically resisted. Chapter 9: In 1977, this author developed the first semi-integral bridge design for the Ohio Department of Transportation. At that time it was presumed that this bridge was the first of its kind constructed in the United States. Subsequently, it was learned that other bridges based on the same general concept had preceded it by many years. Apparently, bridge engineers of several other states had independently devised and built a few of their own versions of this unusual concept. One of them, Willis B. Horne of Washington State, devised his own version more

Introduction

xiii

than a decade before Ohio’s, and his State had been using this basic concept (in place of integral bridges) for most of its typical highway bridge applications. This chapter identifies these other states and provides detail sketches of their semiintegral designs and commentary about their design practices. Chapter 10: The structure movement systems approach to the design of highway bridges is described and illustrated. This is the approach that is used by bridge design engineers with the most experience with integral and semi-integral bridges. It describes how integral and semi-integral bridge applications (or, for that matter, all highway bridge applications) are or can be conceptualized holistically as composite structures, structures that are conceived to be composed of various types of structure movement systems. It also describes how the use of this holistic approach during design can aid design engineers avoid many of the design mistakes that are made by their less experienced and elementalistically oriented and limited colleagues and predecessors. Chapter 11: During a bridge replacement and bridge rehabilitation project for the State of Ohio, a project that consisted of over 1,800 bridges designed by over 80 small consulting engineering firms, it soon became evident that many of the designers working on this project seem to have a predilection towards particular types of design errors. This chapter describes some of these and other errors, and speculates that certain common aspects of an individual’s early education and continued academic conditioning resulted in an orientation that could be described as a general lack of awareness. Appendix 1: This appendix contains a critique of unfortunate concrete pavement recommendations originating from a Midwest State Department of Transportation and a few misguided published recommendations of a certain pavement research “specialist.” These unfortunate recommendations appear to suggest that those making them were not fully knowledgeable about the characteristics of the G/P phenomenon or its destructive potential. The chapter also provides a brief but sufficient description of this destructive phenomenon, and contains rather extensive documentation of some of the pavement and bridge damage associated with the phenomenon. Appendix 2: This is a glossary prepared to aid novice engineers encountering some of the subjects discussed in this book for the first time. Possibly, if others accept these definitions, they may also be found useful for clarifying their own discussions of the same or similar subject matter. Appendix 3: This appendix contains bridge data for and descriptions of the integral and semi-integral bridges shown in photographs that appear on the first page of each section of this book. Typically, these bridges are unique examples of integral and semi-integral bridges (the first of its kind, the longest span, the greatest curvature, etc.) that have been constructed in the United States, United Kingdom, Canada, Japan, and Korea. Hopefully, photographic images of similar bridges constructed in other states and countries will become available for future editions of this book.

Chapter 1

Integral Bridges

We are suspicious of new ideas, however good, if they threaten old ideas however bad. Frank A. Clark

Introduction The first integral bridge in the United States was the Teens Run Bridge. It was built in 1938 near Eureka in Gallia County, Ohio. It consists of five continuous reinforced concrete slab spans supported by capped pile piers and abutments. Since that time construction of integral bridges has spread throughout the United States and abroad. The United Kingdom recently adopted them for routine applications. Japan completed its first two in 1996. South Korea completed its first such bridge in 2002. Integral bridges may be briefly defined as single-span or continuous multiplespan bridges constructed without movable transverse deck joints (movable deck joints) at piers or abutments, or as more generally described in Chapter 10 and Appendix 2, integral bridges may be conceived of as components of composite structure movement systems, systems generally composed of: 1

2

Integral and Semi-integral Bridges

• • • • •

Jointless superstructures constructed integrally with capped pile abutments Abutments supported by embankments and single rows of vertically driven piles Rigid piers with movable bearings, or flexible piers constructed integrally with the superstructure Attached approach slabs that bear on abutments and abutment backfill Cycle control joints, of some sort, for the longer bridges, located between approach slabs and approach pavements.

When multiple-span bridges are constructed without movable deck joints at piers, it is accepted that the continuity achieved by such construction will subject superstructures to secondary stresses, stresses that are induced by the response of continuous superstructures to settlements of substructures, post-tensioning, etc. When continuous bridges are constructed without such joints at the superstructure/ abutment interface, it is likewise accepted that they will, in addition, be subjected to secondary stresses due to superstructure/abutment continuity, and to the resistance of abutment foundations and backfill to cyclic longitudinal superstructure movements. The justification for such construction is based on the growing awareness that, for single- and multiple-span bridges of moderate lengths, significantly more damage and distress have been caused by the use of movable deck joints at piers and abutments than the secondary stresses that these joints were intended to prevent. In addition, elimination of costly joints and bearings and the laborintensive details and construction procedures necessary to permit their use have generally resulted in more cost-effective bridges. Consequently, more and more bridge engineers are now willing to relinquish some of their control of secondary stresses primarily to achieve simpler and more cost-effective bridges and bridges with greater overall integrity and durability. Before continuing this discussion about integral bridges, a pause should be taken here to comment on the use of the unfortunate phrase “integral abutment bridges.” It is this author’s contention that the use of this phrase by members of the bridge engineering profession leaves novice engineers with the incorrect impression that it would be proper and acceptable to provide integral abutments for all bridges including multiple-span non-continuous bridges. Obviously, such construction is totally inappropriate, and especially for those projects that are built in conjunction with jointed concrete approach pavement (see Chapter 2). It therefore should be understood that in this and other chapters of this book, the designations “integral bridges” and “semi-integral bridges” will be used exclusively. The first designation refers to single- or multiple-span continuous bridges without movable deck joints at the superstructure/abutment interface. These are generally supported by embankments with stub-type abutments on flexible piles. The second designation refers to single- or multiple-span continuous bridges without movable deck joints in their superstructures but with movable longitudinal joints between their superstructures and rigidly supported abutments. The piers for such structures may be semi-rigid self-supporting structures generally surmounted by movable bearings, or flexible substructures constructed integrally with superstructures. Approach slabs that span across and are partially supported by structure backfill should be attached to the superstructures of such bridges. Cycle-control joints (see Appendix

Chapter 1

Integral Bridges

3

2) in some form should be provided between their approach slabs and approach pavements.

Continuous superstructures Current design trends (about 1990) received their primary impetus and direction almost six decades ago. In May 1930, a brief 10-page paper on the “Analysis of Continuous Frames by Distributing Fixed End Moments” [1], published in the Proceedings of the American Society of Civil Engineers, generated considerable discussion in academia. Its publication was followed shortly by what could be considered a minor revolution in the design and construction of short- and moderate-span bridges. In that paper, Professor Hardy Cross presented a simple and quick method for the analysis of integral-type structures such as continuous beams and frames. His moment distribution method was quickly adopted by bridge design engineers, and the bridge design and construction practices of many transportation departments began to change. Before Cross’s “Moment Distribution” [1], most multiple-span bridges were generally constructed as a series of simple spans. Following the introduction of moment distribution, bridge design engineers began adopting continuous construction primarily to eliminate troublesome movable deck joints at piers. On the basis of a nation-wide mail survey of state and province transportation departments [2], it appears that the Ohio Highway Department (now the Ohio Department of Transportation, or Ohio DOT) was one of the first agencies to initiate the routine use of continuous construction for the design and construction of multiple-span bridges. Its experience provides an informative background for this movement toward the use of fully integrated construction. To minimize the use of movable deck joints at piers and thus prevent deleterious deck drainage from reaching and saturating the surfaces of vulnerable primary superstructure and pier components, beginning in the late 1920s and early 1930s, Ohio DOT adopted the routine use of continuous construction for multiple-span highway bridges. To make such a practice possible at a time when continuous construction was a rarity, Ohio DOT had to develop and perfect various field-splicing procedures for the bridging materials then available. For the shortest multiple-span bridges and those bridges with spans less than 50 ft. (15 m), continuous reinforced concrete slab bridges were developed and adopted. At first, rolled steel beams were made continuous by the use of riveted field splices at piers (Figure 1.1). To achieve continuous steel girders, field-riveted plate and angle splices were provided at counter flexure points. At about the same time, welding procedures and welder pre-qualification tests were developed for field welding of steel bridge members, and some of the shortest rolled beam bridges were provided with field-welded splices at piers. These initial welded splices consisted of partially butt-welded beam webs supplemented with fillet-welded moment plates. Field-welded splices were constantly being improved by Ohio DOT and, by the mid-1950s, all rolled beam bridges were being made continuous by field butt welding of beam webs and flanges, and by fillet welding of flange moment plates. From the late 1920s to the mid-1950s, steel girder fabrication and girder field splices were of riveted construction. However, in 1954, high-strength bolts were used

4

Integral and Semi-integral Bridges

Figure 1.1 USR 52, Isaacs Creek Bridge, Adams County, Ohio, 1931. This was one of first steel beam bridges in Ohio where riveted field splices at piers were used to achieve superstructure continuity.

in lieu of field-driven rivets for the field splices of the Patterson-Riverside, Great Miami River Bridge of Dayton, Ohio (Figure 1.2). This was one of the first applications of high-strength bolting for highway bridges in the United States. By 1963, high-strength bolting replaced field butt welding in Ohio as the method of choice for integrating multiple-span bridges to achieve full continuity. Consequently, by riveting, field butt welding, and high-strength bolting, Ohio DOT has employed continuous construction for more than 70 years. In conjunction with the development and adoption of continuous construction for all moderate-length highway bridges, Ohio DOT was also the first state to routinely eliminate deck joints at abutments. This was accomplished in the case of continuous reinforced concrete slab bridges by providing embankments and stubtype integral abutments supported by flexible piles in lieu of movable deck joints and wall-type abutments (see Chapter 5). This new abutment type is now designated as an integral abutment and Ohio DOT was the first state transportation department to adopt such construction as a standard practice. A version of this integral abutment design has been used in Ohio for many hundreds of bridges ever since. However, it was not until the early 1960s that the integral concept was first used by Ohio DOT for a steel beam bridge (see Appendix 3, photograph). Since that time, most steel beam and girder bridges with skews 30 ° or less, and lengths not longer than about 300 ft. (91.44 m), were of integral construction (if site geological characteristics and/or embankment heights allowed the use of flexible piles for abutment support). In 1951, Ohio DOT was one of the first transportation organizations in the United States to pioneer the use of prestressed concrete for highway bridges. In fact,

Chapter 1

Integral Bridges

5

Figure 1.2 The Patterson-Riverside, Great Miami River Bridge, Dayton, Ohio, 1954.

Ohio DOT set up its own plant for casting and prestressing concrete T-beams. More than a dozen single-span bridges were constructed using these state-manufactured, prestressed concrete beams. But it was not until the early 1960s that commercially produced, prestressed box beams were adapted to continuous construction. These first continuous, prestressed box beam bridges were also provided with embankments and stub-type abutments on flexible piles but they were not entirely jointless because rotation joints were provided at abutments. Recently, however, some fully integral, continuous, prestressed box beam bridges were built by Ohio DOT. Two design examples serve to illustrate just how strongly Ohio DOT bridge engineers favored the use of integral bridges, and the unusual means that they were willing to consider just to avoid the use of movable deck joints in highway bridges. For instance, at some sites where the depth of overburden was not considered sufficient to provide flexible piles for integral abutment construction, bedrock has been prebored and backfilled to a suitable depth to permit the driving of end-bearing flexible piles for the abutments. At other sites, stream channel alignments have been modified so that integral bridges could be used that would not exceed the 30 ° skew limitation that had been established for such structures. As can be surmised by these examples, and the other practices that have been developed and adopted by Ohio DOT, Ohio bridge design engineers’ primary bridge design goal has always been the avoidance of deck joints whenever practicable. Somewhat paralleling Ohio DOT’s implementation of continuous construction, other state and province transportation departments were also showing interest in similar construction. By 1987, 26 out of 30 mail responses [2, p. 20], or 87 percent of responding transportation departments, indicated that they were using continuous construction for short- and moderate-length bridges.

6

Integral and Semi-integral Bridges

Figure 1.3

Long Island Bridge, Kingsport, Tennessee, 1980.

The Tennessee Department of Transportation (Tennessee DOT) now appears to be leading the way in the construction of continuous bridges. For example, the Long Island Bridge of Kingsport, Tennessee (Figure 1.3) was constructed in 1980 using 29 continuous spans without a single intermediate movable deck joint. The total length of this bridge is about 2,700 ft. (823 m) center to center of abutment bearings. Movable deck joints and movable bearings were furnished, but only at the two abutments. It has aptly been named the “Champ.”

Integral bridges During the last half-century, many bridge engineers have become acutely aware of the relative performance of bridges built with and without movable deck joints. In this respect, bridges without such joints (integral bridges) have performed more effectively because they remain in service for longer periods of time with only moderate maintenance and occasional repairs. Some of this experience was forced upon bridge engineers by circumstances beyond their control. As a result of the growth and pressure generated by jointed rigid pavement (see Chapter 2 and Appendix 1), many bridges built with movable deck joints have been and are being severely damaged. After these joints are closed by pavement growth, the effectively jointless bridge restrains the pavement from further growth, resulting in the generation of longitudinal pavement pressures (compressive forces) against and within the bridge. Over time, these pavement pressures can easily exceed

Chapter 1

Integral Bridges

7

1000 psi (6.89 MPa) or cumulatively the total force due to such pressures can exceed 650 tons (716 tonnes) per lane of approach pavement [3]. When the design of abutments of non-integral-type bridges – bridges with movable deck joints at the superstructure/abutments interface – is considered, forces of these magnitudes are irresistible. Stub abutments subjected to such pressures have routinely been moved, joints closed, and ultimately joints and wingwalls fractured. Wall-type abutments have been split from top to bottom. In longer bridges with intermediate movable deck joints, piers have been cracked and fractured as well (see Chapter 2). In geographical regions of the country that experience low seasonal temperatures and an abundance of snow and freezing rain, the use of de-icing chemicals to maintain dry pavements throughout the winter season has also had a significantly adverse affect on the durability and integrity of bridges built with movable deck joints. Open joints and sliding plate joints of shorter bridges and open finger joints of longer bridges have allowed roadway drainage, contaminated with de-icing chemicals, to penetrate below roadway surfaces and wash over supported beams, bearings, and bridge seats. The resulting corrosion and deterioration have been so serious that some bridges have collapsed while others have had to be closed to traffic to prevent their collapse. Many jointed bridges have required extensive repair. Most of the jointed bridges that have remained in service have required almost continuous maintenance to counteract the adverse effects of contaminated deck drainage. To help minimize or eliminate these maintenance efforts, a whole new industry was born. Beginning in the early 1960s, the first elastomeric compression seals were installed in bridges in the United States to seal movable deck joints. Since these first installations, numerous types of elastomeric joint seals have been developed and improved in an attempt to achieve joint seal designs that would both effective and durable. Most designs have been disappointing. Many leaked. Some required more maintenance than the original bridge built without them. By and large, the many disappointments associated with various types of joint seals have caused bridge engineers to consider other options. Costs of various types of bridges show marked differences. For two bridges built in essentially the same way, except where that one was provided with movable deck joints at the superstructure/abutment interface and the other with integral abutments, the jointed bridge was usually the more expensive. In addition, abutments of integral bridges suffered only minor damage from pavement pressure, were essentially unaffected by de-icing chemicals, and functioned for extended periods of time without appreciable maintenance or repair, whereas jointed bridges suffered major damage from de-icing chemicals and pavement pressure. Consequently, more bridge engineers began to appreciate the merits of integral bridges for short- or moderate-length bridges. Gradually, design changes were made and longer integral bridges were built and evaluated. In 1946, Ohio’s initial length limitation for its continuous concrete slab bridge was 175 ft. (53.3 m). In a 1973 study of integral construction [4], four states reported that they were using integral steel bridges and 15 states were using integral concrete bridges in the 201–300 ft. (61–90.4 m) range. In a 1982 study, even longer bridges were reported. Continuous integral bridges with steel main members have performed successfully for years in the 300-ft. [91.4-m] range in such states as North Dakota, South Dakota

8

Integral and Semi-integral Bridges

and Tennessee. Continuous integral bridges with concrete main members, 500 to 800 ft. (152.4 to 243.8 m) long have been constructed in Kansas, California, Colorado, and Tennessee. [5]

As of 1987, 11 states reported building continuous integral bridges in the 300 ft. (91.4 m) range. Missouri and Tennessee reported even longer lengths. Missouri reported steel and concrete bridges in lengths of 500 and 600 ft. (152.4 and 182.9 m), respectively. Tennessee reported lengths of 400 and 800 ft. (121.9 and 243.8 m) for similar bridges. Actually, 20 of 30 transportation departments, or 60 percent of those departments responding to the 1987 survey, were using integral construction for continuous bridges. The attributes of integral bridges have not been achieved without some concerns about high unit stresses. Parts of these bridges operate at very high stresses levels, levels that cannot easily be quantified. These stresses are significantly above those permitted by current design specifications. In this respect, bridge engineers have become rather pragmatic. They would rather build cheaper integral bridges and tolerate these higher stresses than build the more expensive jointed bridges with their lower stresses and concomitant vulnerability to destructive pressures and deicing-chemical deterioration. This attitude was expressed by Clelland Loveall, then Engineering Director for the Tennessee DOT. At the time he wrote: In Tennessee DOT, a structural engineer can measure his ability by seeing how long a bridge he can design without inserting an expansion joint. … Nearly all our newer (last twenty years) highway bridges up to several hundred ft. have been designed with no joints, even at abutments. If the structure is exceptionally long, we include joints at the abutments but only there. … Joints and bearings are costly to buy and install. Eventually, they are likely to allow water and salt to leak down onto the superstructure and pier caps below. Many of our most costly maintenance problems originated with leaky joints. So we go to great lengths to minimize them. [6]

Tennessee DOT is still leading the bridge engineering profession in the construction of longer and longer integral bridges. Under their present Engineering Director, Edward Wasserman, Tennessee DOT recently completed the Happy Hollow Creek Bridge, a seven-span prestressed concrete curved integral bridge with a total length of over 1,175 ft. (358 m) (see the photograph at start of this chapter). As shown in this photograph, tall flexible twin circular column piers support the superstructure of this outstanding structure. A single row of steel H-piles is used to support each abutment. Although, to some engineers, the length of this structure may seem extreme, it is well within Tennessee DOT’s Bridge Design Policy Statement regarding the length of integral bridges. With respect to expansion joint selection, the policy statement stipulates: When the total anticipated movement at an abutment is less than two (2) inches [50 mm] and the abutment is not restrained against movement, no joint will be required and the superstructure and abutment beam will be constructed integrally. [7]

In 1997, six bridge engineers from the United Kingdom participated in a study tour of integral bridges in North America. This task group visited Ohio, Tennessee,

Chapter 1

Integral Bridges

9

Missouri, Washington State, California, and Ontario. They also visited Construction Technology Laboratories of Skokie, Illinois, where comprehensive integral bridge research was under way. In their report of the study tour, they generalized their opinions about the performance of integral bridges inspected by the task group as follows: Integral bridges were inspected in five States in the USA, and in Ontario, Canada. In all cases these were found to be performing well. It is important to note that, in contrast, the non-integral bridges that were seen all had leaking expansion joints, and several were deteriorating badly. The few minor problems in integral bridges that were found were all considered to be due to poor detailing. Integral construction transfers possible problems from the abutment to the approach slab and pavement. No integral bridges were seen on the tour where the integral concept was considered to have been inappropriate. [8]

Although bridge engineers have conditioned themselves to tolerate higher stress levels in integral bridges, occasionally their design control is not sufficient to prevent these high stresses from resulting in relatively minor structural distress. In this respect, consider some of the responses to survey questions about noticeable structure distress.

Structural distress Responses to an early survey about continuous integral bridges indicated a rather widespread concern by bridge engineers for the potentially high stresses that would be present in longer integral bridges [4]. This concern, more than any other, appeared to be responsible for the early lack of enthusiasm for using integral construction for the longer continuous bridges. Although most integral bridges perform adequately, many of them operate at high stress levels. For instance, an abutment supported on a single row of piles is considered flexible enough to accommodate thermal cycling of the superstructure and the dynamic end rotations induced by the movement of vehicular traffic. Nevertheless, the steel piles of such an abutment are routinely subjected to axial and flexural stresses approaching, equaling, or exceeding yield stresses [5, 9]. Occasionally, a combination of circumstances results in visible distress. Responding to a 1973 survey, a number of bridge engineers said that some integral bridge abutment wingwalls had minor cracks [4]. This problem was corrected by the use of more generous wingwall reinforcement. Other engineers reported pile cap cracking, cracking that appears to have been eliminated by providing more substantial pile cap connection reinforcement and by rotating steel H-piles to place the weak axis normal to the direction of bridge movement. In a 1984 article in Concrete International, Gamble [10] emphasizes the importance of considering restraint stresses in cast-in-place construction. He discusses cracking that occurred in a continuous concrete frame bridge with footings that were founded in bedrock. Even though the concrete of this structure was considerably below the specified cylinder strength, and shear reinforcement did not meet current requirements, failure of the structure was attributed to its stiffness and

10

Integral and Semi-integral Bridges

resistance to shrinkage and contraction of its bridge deck. Failures of this type emphasize the necessity of achieving suitable flexibility in supporting substructures and conservative reinforcement to withstand the secondary stresses induced by foundation restraint and superstructure shortening. Currently, precast prestressed concrete and prefabricated steel superstructures are generally replacing small cast-in-place bridges in many states and provinces. Consequently, problems associated with initial shrinkage of superstructures are gradually being eliminated. However, where cast-in-place construction continues to be used for substructures, flexibility remains a critical part of bridge design. In this respect, Loveall of Tennessee DOT provides an example of the lack of flexibility in substructure design: Structural analysis of our no joint bridges indicates that we should have encountered problems, but we almost never have. Once we tied the stub of an abutment into rock, and the structure cracked near its end, but we were able to repair the bridge and install [a] joint while the bridge was under traffic. The public never knew about it. That was one of few problems. [6]

Development of new forms of construction will be accompanied by instances of structural distress, and this has certainly been true with continuous integral bridges. However, as indicated by the 1987 mail survey, the application of integral bridges increased exponentially from its beginnings in the 1930s and was beginning to taper off in the 1980s when 20 of 30, or 60 percent, of responding transportation departments were using integral bridge construction in one form or another for longer and longer bridges. Presumably, with continued care and consideration, it appears that the use of integral bridges will continue to see a gradual increase in the numbers of transportation departments adopting the integral bridge concept for routine bridge applications.

Integral bridge details Abutment details of integral bridges used by six transportation departments, as of 1989, when the text that serves as the basis of this chapter was originally prepared, are presented in Figures 1.4 and 1.5. Presumably, the details presently used by these same departments have remained essentially the same except for minor changes in dimensions and reinforcement. What has changed in the intervening years are the numbers of other departments that have adopted standard integral details for their routine bridges. It is probably not an accident that a fair amount of similarity is evident in these designs because structural details from early successful designs are adapted and improved by other bridge engineers for use by their departments. Even though there are similarities, there are also differences that reflect the various types of bridges being built, and the care and concern being given to the conception and development of specific details. It should also be realized that these sketches are mere “bare bones” presentations. They do not reflect other important design aspects and construction procedures that should be considered in the application of these details for specific applications. All of these aspects could not be illustrated and properly described in a chapter as brief as this one. Nevertheless, because these

Chapter 1

(a)

Integral Bridges

11

(b)

(c)

Figure 1.4 Integral abutment: (a) Iowa DOT; (b) Pennsylvania DOT; and (c) North Dakota DOT.

aspects can have a considerable effect on the performance, integrity, and durability of integral designs, it is appropriate to mention something about some of them, especially passive pressure and pile stresses, for those engineers who will be considering such designs for the very first time. Passive pressure To minimize passive pressure development in structure backfill by an elongating integral bridge, bridge design engineers have used a number of controls, devices and procedures. Including but not limited to the following practices, they have:

• • • • •

limited bridge length limited bridge skew limited abutment type to embankment supported stubs provided embankment benches to minimize the length of transverse wingwalls limited the vertical penetration of abutments into the benches

12

Integral and Semi-integral Bridges

(a)

(b)

(c)

Figure 1.5 DOT.

• • • • • •

Integral abutment: (a) Illinois DOT; (b) Tennessee DOT; and (c) Ohio

limited the clearance between the superstructure and embankment benches to make the abutment surfaces exposed to passive pressure as small as possible provided well-drained select granular backfill at abutments provided turn-back wingwalls to minimize total longitudinal pressure on the abutments provided approach slabs to prevent live load surcharging of backfill, and to minimize vehicular compaction of backfill. provided approach slabs with curbs adjacent to curbed bridges to protect backfill from erosion used a semi-integral abutment design to eliminate passive pressure below the bridge seat and permit the use of a semi-rigid foundation design (Figure 1.6).

Chapter 1

Figure 1.6

Integral Bridges

13

Ohio DOT’s first semi-integral abutment (1979).

Pile stresses Knowing that longitudinal forces on superstructures are somewhat directly related to the resistance of abutment pile foundations to longitudinal movement, design engineers have:

• • • • • • • •

provided each abutment foundation with a single row of slender vertical piles provided only those pile types that could tolerate a considerable amount of distortion without failure; in this respect, it has shown that steel H-piles are the most suitable pile type for longer integral bridge applications [11] oriented the weak axis of H-piles normal to the direction of pile flexure provided prebored holes filled with granular material provided an abutment hinge (see Figure 1.5c) to minimize pile flexure limited the length of the structure to minimize pile flexure limited structure skew angle provided semi-integral abutments to minimize restraint on superstructures due to longitudinal movement.

Questionnaires A number of questionnaires about integral bridge practices have been circulated in recent years. The responses reflect the policies, attitudes, and opinions of those engineers responsible for bridge design policies. They also show how some of these attitudes and opinions have changed during the last couple of decades. In 1973, Emanual et al. [4] received responses about the then current design practices from

14

Integral and Semi-integral Bridges

43 transportation departments. In 1982, Wolde-Tinsae et al. [5] used a questionnaire as part of an investigation into non-linear pile behavior. They tabulated the responses that they received from 29 transportation departments. Greimann et al. [12] elicited responses from 30 transportation departments on their pile orientation practices for skewed integral bridges. In 1987, Wolde-Tinsae and Klinger [13] solicited responses from selected transportation departments in the United States, Canada, Australia, and New Zealand. (The reports by Wolde-Tinsae et al. [5], Greimann et al. [12], and Wolde-Tinsae and Klinger [13] also contain valuable bibliographies for those interested in a more in-depth study of available research on the behavior of integral bridges and the performance of abutment pilings.) In addition, 1n 1988, the author received responses from 30 transportation departments describing the limitations that these departments used to control the behavior and performance of the integral bridges designed and constructed by them [2]. Integral conversions Following the trend toward the use of end-jointed continuous construction and the use of jointless continuous construction, transportation departments are also beginning to convert existing multiple-span bridges from simple to continuous spans. This effort began with Wisconsin and Massachusetts DOTs in the 1960s and has gathered strength in the last several decades. Currently, more than 30 percent of the transportation departments have converted one or more bridges from multiple simple spans to continuous spans. Although the 1988 mail survey suggested considerable activity, it was not indicative of the number of bridges that had been converted. For example, positive responses were received from only two departments to the following question: “In recent years, have you converted any bridges from multiple simple spans to continuous spans to eliminate deck joints?” The Ontario Ministry of Transportation and Communications responded: We are modifying a few structures from simple spans to continuous spans, eliminating deck joints in the process. …[2, p. 27]

The Texas Department of Public Transportation (DPT) responded: In recent years, we have eliminated numerous intermediate joints. Generally, this is done while replacing the slab. We simply place the slab continuous across the beams. On a few occasions, we have removed only the joint and surrounding deck area, added reinforcement, and replaced that portion of the deck thus tying the adjacent spans together. [2, p. 27]

Tennessee DOT also has been actively converting simple-span bridges. In a paper on jointless bridges, Edward Wasserman, Engineering Director of Structures, described and illustrated a number of such conversions [14]. To give this movement some direction, in 1980, the Federal Highway Administration (FHWA) issued a technical advisory on the subject [15]. That advisory in part recommends that a study of the bridge layout and existing movable deck joints be made “to determine which joints can be eliminated and what modifications are necessary to

Chapter 1

Integral Bridges

15

revamp those that remain to provide an adequate functional system. …” Further, it recommends: For unrestrained abutments, a fixed integral condition can be developed full length of shorter bridges. An unrestrained abutment is assumed to be one that is free to rotate, such as a stub abutment on one row of piles or an abutment hinged at the footing. … [W]here feasible, develop continuity in the deck slab. Remove concrete as necessary to eliminate existing armoring, and add negative moment steel at the level of existing top-deck steel sufficient to resist transverse cracking. [15]

The detail in Figure 1.7a from the FHWA Technical Advisory mirrors the details used by the Texas Department of Public Transportation (Texas DPT) for its conversion of multiple simple spans to continuous spans. Note that Figure 1.7a shows that only the slab portion of the deck is made continuous. The simply supported beams remain simply supported. For such construction, it is important to ensure that one or both of adjacent bearings supporting the beams at a joint are capable of allowing

(a)

(c)

(b)

(d)

Figure 1.7 (a) Integral conversion at piers, Texas DPT (copied from the FHWA Technical Advisory [15]); (b) integral conversion of existing beams at piers, Utah DOT; (c) integral conversion of precast I-beams at piers during original construction, Wisconsin DOT; and (d) integral conversion of prestressed box beams at piers during original construction, Ohio DOT.

16

Integral and Semi-integral Bridges

horizontal movement. Providing for such movement will prevent large horizontal forces from being imposed on bearings due to rotation of adjacent spans and continuity of the deck slab. Utah DOT has also converted some simple span bridges to continuous spans by using a design similar to the one illustrated in Figure 1.7b. For deck slabs with a bituminous overlay, an elastomeric type of membrane can be used under the overlay to waterproof the new slab section over the piers. With a design like this, it is understood that the deck slab would be exposed to longitudinal flexure due to the rotation of beam ends responding to the movement of vehicular traffic. However, for shortand medium-span bridges, the deck cracking associated with such behavior is preferred by some over the long-term adverse consequences associated with open movable deck joints or a poorly executed joint seals. In new construction, the conversion of simple spans to continuous spans is rather commonplace. Figure 1.7c shows the detail used by Wisconsin DOT for the construction of prestressed concrete I-beam bridges. A substantial concrete diaphragm is provided at piers between the ends of the simply supported beams of adjacent spans. The diaphragm extends transversely for the width of the superstructure. Then a continuous reinforced concrete deck slab is placed to integrate the beams, diaphragms, and slab, thereby providing a fully composite continuous superstructure. This type of prestressed concrete I-beam construction now appears to be standard for many transportation departments. Figure 1.7d illustrates the standard detail used by Ohio DOT to achieve continuous bridges by using simply supported, prestressed concrete box beams with continuity connections at the piers. Boxes are placed side by side and then transversely bolted together. Finally, continuity reinforcement is placed and reinforced concrete closure placements are made. In a 1969 paper, Freyermuth [16] gives a rather complete description of the analysis procedures that can be used to achieve continuity in a bridge composed of a continuously reinforced concrete deck slab on simply supported, precast, prestressed beams. Conversion of existing bridges, by replacing either the deck completely or only portions of the deck adjacent to movable deck joints at piers, can be accomplished by following the procedures developed by Freyermuth for new structures. Obviously, for existing bridges, creep effects will be negligible. Shrinkage effects for other than complete deck slab replacements should also be negligible. Not only does such continuous conversion eliminate troublesome joints, but the continuity achieved also results in a slightly higher bridge load capacity because positive moments due to live load are reduced by continuous rather than simple span behavior. Although too recent to consider in terms of a design trend, conversion of nonintegral abutments to achieve integral bridges or semi-integral bridges for both single- and multiple-span continuous bridges has begun. Figures 1.8–1.10 illustrate design details used for a number of conversions by Ohio DOT. Reconstruction of these abutments was made necessary by the substantial damage caused by pavement growth and pressure, by de-icing chemical deterioration, or both. Instead of replacing backwalls and joints, and in some cases bearings and bridge seats as well, it was decided to pattern reconstruction after the design details used by the department for its new integral and semi-integral bridges. In this way subsequent concern about

Chapter 1

(a)

Integral Bridges

17

(b)

Figure 1.8 Conversion of a very short continuous bridge with movable deck joints at the superstructure/abutment interface (a), into a continuous bridge with integral abutments (b), Ohio DOT.

(a)

(b)

Figure 1.9 Conversion of a continuous bridge with movable deck joints at the superstructure/abutment interface (a), to a continuous bridge with integral abutments (b), Ohio DOT.

the adverse effects of pavement pressure and de-icing chemical deterioration were minimized. When considering the design trends toward integral types of construction, it should not be surprising to learn that a number of transportation departments have also begun to retrofit steel beam and girder bridges constructed with intermediate movable deck joints with hinges into fully continuous structures. Conversion of

18

Integral and Semi-integral Bridges

(a)

(b)

Figure 1.10 Conversion of single- or multiple-span continuous bridges with movable deck joints at the superstructure/wall-type abutment interface (a), into single- or multiple-span continuous semi-integral bridges (b), Ohio DOT.

(a)

(b)

Figure 1.11 Conversion of multiple-span continuous bridges with intermediate deck joints and hinges (a), into continuous bridges with bolted splices (b), Ohio DOT.

these structures is being accomplished by replacing the hinges and leaking joints with bolted splices and continuous deck slabs (Figure 1.11). These joints and hinges were originally intended to accommodate long-term abutment settlement. But as these structures are now more than 20–30 years old, and as embankments are now essentially fully consolidated, the need for these movement systems no longer exists. However, where such labor-intensive conversions are not fully cost-effective, some

Chapter 1

Integral Bridges

19

of these jointed superstructures are being completely replaced with fully continuous superstructures with integral abutments. Finally, within the last two decades, Ohio DOT has converted many of its continuous bridges with movable deck joints at the superstructure/abutment interface by completely replacing independent semi-rigid stub abutments (see Figure 1.8a) with integrated flexible stub abutments (see Chapter 7, Figure 7.7). Presumably, this same rehabilitation technique is now being used by many other transportation departments throughout the United States. However, the number of such retrofitted structures is probably greater in Ohio because most of Ohio’s old multiple-span bridges were originally constructed as continuous bridges. In fact, one would be hard pressed to find a multiple-span bridge in Ohio that was not of continuous construction. These are the bridges that are now being converted in record numbers to fully integrated construction.

Summary As the trends noted above continue, it appears that the use of continuous construction for multiple-span bridges will become standard for all transportation departments in the very near future. It also appears that the use of integral abutments for single- and multiple-span continuous bridges will increase when comprehensive and conservative guidelines for their use become more readily available, and when their long-term performance has been more fully documented. Presumably, the next decade or two will see a burgeoning in the retrofitting of simply supported multiple-span bridges to continuous bridges and from non-integral to integral bridges. When more information on the operating stress levels for these structures is developed and when more fully described design details and construction procedures for integral conversions become available, bridge engineers will be able to more fully justify their consideration. Until then, much intuition and prudent judgment will continue to be used to ensure that integral construction and conversion techniques will provide the structure service life needed to justify their adoption and continued use.

Epilogue As a preliminary to the 2005 FHWA Conference on Integral Abutments and Jointless Bridges, a nation-wide survey was conducted of all major transportation departments of the United States. This survey posed various questions regarding the use of integral and semi-integral bridges. With respect to the number of these structures that have been employed, the following summary statement was made: The survey responses indicate an increase in the number of integral [bridges] of over 200% in the last ten years. As in 1995, Tennessee continues to have over 2000 integral … bridges, but Missouri reports having 4000 integral … bridges, which represents the largest amount of integral bridges. An increase in the number of integral [bridges], since 1995, is most evident in the northern states where Illinois, Kansas and Washington all reported having

20

Integral and Semi-integral Bridges

over 1000 in service. In addition, Michigan, Minnesota, New Hampshire, North Dakota, South Dakota, Oregon, Wyoming and Wisconsin, reported having between 100–500 integral bridges in service. Unlike the northern states, the southern states like Florida, Alabama and Texas do not use integral [bridges] and reported having one or [no] integral [bridges]. [17]

References 1. Cross, H., “Analysis of Continuous Frames by Distributing Fixed End Moments,” ASCE Proceedings, American Society of Civil Engineers, New York, May 1930. 2. Burke, M. P. Jr., National Cooperative Highway Research Program Synthesis 141: Bridge Deck Joints, Transportation Research Board of the National Academies, Washington, D.C., 1989. 3. Burke, M. P. Jr., “Bridge Approach Pavements, Integral Bridges and Cycle-Control Joints,” Transportation Research Record No. 1113, Transportation Research Board of the National Academies, Washington, D.C., 1987. 4. Emanual, J. L., et al., “Current Design Practice for Bridge Superstructures Connected to Flexible Substructures, University of Missouri-Rolla, Rolla, Missouri, 1973. 5. Wolde-Tinsea, A. M., Greimann, L. F., Yang, P. S., Nonlinear Pile Behavior in Integral Abutment Bridges, Iowa State University, Ames, Iowa, 1982. 6. Loveall, C. L., “Jointless Decks,” Civil Engineering, American Society of Civil Engineers, New York, 1985, pp. 64–67. 7. “Expansion Joint Selection,” Tennessee Structures Memorandum, MO 045, Tennessee Department of Transportation, Nashville, Tennessee, 1989. 8. Nicholson, B. A., et al., Integral Bridges: Report of a Study Tour of North America, Concrete Bridge Development Group, Century House, Telford Avenue, Crowthorn, Berkshire, United Kingdom, 1997, pp. 93. 9. Jorgenson, J. L. “Behavior of Abutment Piles in an Integral Abutment Bridge,” Transportation Research Record No. 903, Transportation Research Board of the National Academies, Washington, D.C., 1983. 10. Gamble, W. L., “Bridge Evaluation Yields Valuable Lesson,” Concrete International, American Concrete Institute, Farmington Hills, Michigan, 1984, pp. 68–74. 11. Oesterly, R. G., “Flexible Pile Tests,” Construction Technologies Laboratory, Skokie, Illinois. (unpublished report). 12. Greimann, L. F., Wolde-Tinsea, A. M., Yang, P. S., “Skewed Bridges with Integral Abutments,” Transportation Research Record No. 903, Transportation Research Board of the National Academies, Washington, D.C., 1983. 13. Wolde-Tinsae, D. M., Klinger, J. E., “Integral Abutment Bridge Design and Construction,” Report FHWA/MD-87/04, Maryland Department of Transportation, Annapolis, Maryland, 1987. 14. Wasserman, E., “Jointless Bridges,” Engineering Journal, American Society of Civil Engineers, New York, Vol. 24, No. 3, 1987. 15. FHWA, “Bridge Deck Joint Rehabilitation (Retrofit),” Technical Advisory T1540.16 Federal Highway Administration, Washington, D.C., 1980. 16. Freyermuth, C. L., “Design of Continuous Highway Bridges with Precast Prestressed Concrete Girders,” ACI Journal, American Concrete Institute Farmington Hills, Michigan, Vol. 14, No. 2, 1969. 17. Maruri, R. F., Petro, S. H., “Integral Abutments and Jointless Bridges (IAJB) 2004 Survey Summary,” Presentations and Proceedings, Integral Abutments and Jointless Bridges, 2005, Federal Highway Administration, Baltimore, Maryland, 2005.

Chapter 2

Bridge Damage and the Pavement G/P Phenomenon

If the world has nearly destroyed itself, it is not from lack of knowledge … but is due to the fact that the mass of men have not applied to public policy knowledge they already possess, which is indeed of almost universal possession, deducible from the facts of everyday life. If this is true – and it seems inescapable – then no education which consists mainly in the dissemination of “knowledge” can save us. If men can disregard in their policies the facts they already know, they can just as easily disregard new facts which they do not at present know. What is needed is the development in men of that particular type of skill which will enable them to make social use of knowledge already in their possession; enable them to apply simple, sometimes self-evident truths to the guidance of their common life. Sir Norman Angell

Introduction Innumerable bridges both in the United States and abroad have been and continue to be damaged by the restrained growth of jointed rigid pavement. As a result of such damage, it appears that the pavement growth/pressure (G/P) phenomenon responsible for this damage is not fully appreciated by many pavement research 21

22

Integral and Semi-integral Bridges

specialists, and it appears to be unknown to or not fully appreciated by many pavement and bridge maintenance engineers. This phenomenon is not now described in bridge engineering textbooks, nor is it identified in the American Association of State Highway and Transportation Officials’ (AASHTO’s) Standard Specifications for Highway Bridges [1]. In addition to this apparent somewhat mysterious lack of recognition, it also appears that many bridge design engineers are unaware of the pavement G/P phenomenon. This presumption is based on the fact that these engineers continue to design and construct bridges with movable deck joints in conjunction with jointed concrete pavements. These are the bridges that are vulnerable to pavement growth/pressure-induced damage. In this respect, it also appears that these bridge design engineers are either unfamiliar with or continue to ignore the significant attributes of integral and semi-integral bridges (see Appendix 2), bridges that are highly resistant to such damage. Based on some recent statements and recommendations in national pavement research reports, and in published papers of some state pavement maintenance engineers, it also appears obvious that either the pavement G/P phenomenon is neglected entirely by many pavement engineers, or its significance with respect to the long-term (i.e., ≥10 years) function and durability of both pavements and bridges is not fully understood. As a result, a few misguided pavement design and maintenance practices have recently been advocated and adopted by some state transportation departments, practices that will have a significantly adverse effect on the long-term integrity and durability of both their pavements and their abutting bridges. Concern that such ill-conceived pavement design and maintenance practices might achieve widespread popularity (due primarily to their enticing lower first costs and lower periodic maintenance costs, regardless of the significantly greater long-term pavement and bridge rehabilitation costs) has motivated this author to assemble factual documentation and illustrations and prepare this elaboration of the pavement G/P phenomenon and its destructive potential. As bridge and pavement engineering expertise and practices will continue to change (hopefully for the better), the documentation, illustrations, explanations, and discussion presented herein should, in the author’s opinion, provide a positive direction for those changes. Such changes should help encourage a greater awareness by both pavement and bridge engineers of the pavement G/P phenomenon and its destructive potential, encourage long-term research on this phenomenon, motivate engineers to use more effective pavement and bridge design and maintenance practices, and ultimately help them to achieve more cost-effective, safer, and more durable bridges. The first part of this chapter describes the damaging effect of the pavement G/P phenomenon (see Appendix 2) on three different large bridges containing movable deck joints at both the superstructure/abutment interface and intermediate locations throughout the superstructure. It also provides a brief explanation of the G/P phenomenon. The second part of the chapter describes the effects of this phenomenon on end-jointed continuous bridges and on integral bridges. It also describes the troubling lack of published documentation about this phenomenon and its destructive potential. For those who desire a more complete explanation of this phenomenon, Appendix 1 has been included in this book expressly for that purpose.

Chapter 2 Damage and the Pavement G/P Phenomenon

23

Three bridges Three radically different bridge types – separated by both time and distance – shared a similar fate because each is a multiple-span bridge with movable deck joints at the superstructure/abutment interface, and at intermediate locations throughout the superstructure. In addition, these bridges were built in conjunction with jointed concrete approach pavements. The behavior of these three bridges can be considered somewhat characteristic of many similar smaller bridges located throughout the United States. The three bridges are: the Old Third Street Viaduct of Cincinnati, Ohio; the John F. Kennedy Memorial Bridge of Louisville, Kentucky and Jeffersonville, Indiana; and the Pecos River Bridge of Carlsbad, New Mexico. Old Third Street Viaduct This viaduct consisted of an ugly hodge-podge of structure types, typically continuous steel girders supported by two-legged steel frames (Figure 2.1). But at the shallow western end of the structure, each pier consisted of two short rectangular reinforced concrete columns. In the summer of 1970, after just 11 years of service, bridge maintenance engineers discovered long, essentially vertical cracks in the two columns of Pier 1 (Figure 2.2). The two columns of that pier provided part of the vertical support for the first two continuous spans, and their fixed bolster bearings provided complete longitudinal support for the first of numerous superstructure segments. As the design of these pier columns and their concrete quality were judged to be adequate for the loads to be supported, and because maintenance engineers were confident that pier reinforcement had been provided in accordance with plan

Figure 2.1

Old Third Street Viaduct, Cincinnati, Ohio, 1959–2001.

24

Integral and Semi-integral Bridges

Figure 2.2 Pier 1 of the Old Third Street Viaduct. The vertical cracks in this pier column were induced by the pavement growth/pressure (G/P) phenomenon.

requirements, the integrity of the pier was restored by injecting the cracks with an epoxy adhesive. Periodic inspections of these pier columns made throughout the rest of the year revealed no new or extended cracks, suggesting that the epoxy repair had been successful. During the following summer, however, new and similar cracks began to appear. So instead of further epoxy injections, maintenance personnel responded by apply-

Chapter 2 Damage and the Pavement G/P Phenomenon

25

ing the external hardware visible in Figure 2.2. Subsequently, a broader examination of the structure was made to determine the cause(s) of these unusual cracks so that a more suitable maintenance response could be made. Clues to the cause of pier cracking were revealed when the condition of the adjacent wall-type abutment was considered. For example, the movable deck joint at the west abutment was tightly closed, indicating a permanent 2 in. (50 mm) longitudinal movement from the as-constructed position; the abutment breastwall was tilted toward the superstructure; there were large, essentially vertical cracks between the breastwall and turn-back wingwalls; the approach roadway consisted of jointed concrete pavement; and pier cracking reappearance did not take place until warm summer temperatures were reached. Although there was no visible evidence of pavement or abutment backwall distress, and the approach pavement was only 11 years old, it nevertheless appeared highly likely that the pavement G/P phenomenon was responsible for moving and tilting the west abutment and cracking the first fixed pier of this bridge. More specifically, it appeared that the growing approach pavement jammed the west abutment toward the bridge superstructure, thereby moving and tilting the abutment and closing the movable deck joint at the superstructure/abutment interface. Continual growth of the pavement subsequently moved both the abutment and the first superstructure segment toward the second segment, thereby commencing closure of the movable deck joint between the first and second segments. But the fixed bearings and the two short stiff columns of Pier 1 resisted longitudinal movement of the first superstructure segment. As the forces that can be generated by the pavement G/P phenomenon are so huge, and as they are somewhat proportional to the degree of restraint against pavement growth provided by the bridge, the cracking of the Pier 1 columns was only the first indication of greater forces and more extensive future pier and abutment fracturing, unless pavement pressure relief joints were installed in the bridge approaches. As a first attempt at pressure relief, the western approach pavement was cut transversely so that a 4 in. (102 mm) plank of compressible polyethylene filler could be installed, filler that was manufactured specifically for pavement pressure relief purposes. This installation was made in March 1972. Immediately following the cutting of pavement and the release of longitudinal pavement forces, the wall-type abutment tilted backward ⅝ in. (16 mm) toward the bridge approach, and the slightly tilted Pier 1 columns rotated back to vertical. These responses provided clear visual evidence that the uncut and growing pavement was responsible for the closed movable deck joints and damage to the structure. In March of 1973, just 1 year after its installation, the original 4 in. (102 mm) wide polyethylene pressure relief joint filler was measured and found to be only 2.5 in. (64 mm) wide. This filler thickness change indicated pavement growth and compression of the polyethylene filler of 1.5 in. (38 mm) in just one-year’s time. John F. Kennedy Memorial Bridge Similar trouble for the second of these bridges, the John F. Kennedy Memorial Bridge (Figure 2.3), was brought to the public’s attention when a local newspaper questioned “What’s wrong with the Kennedy Memorial Bridge?” [2]. This bridge consists of continuous through-truss main spans and continuous deck-type steel

26

Integral and Semi-integral Bridges

Figure 2.3 John F. Kennedy Memorial Bridge, Louisville, Kentucky and Jeffersonville, Indiana, 1963.

girder approach spans, stub-type embankment supported abutments, and cap-andround column piers. As described for the Old Third Street Viaduct, the movable deck joints at both ends of the northernmost deck-type superstructure segment were closed, the supporting fixed pier was tilted 5 in. (127 mm) to the south, and the northernmost rocker bearings at the other piers were tilted in the same direction (Figure 2.4). The north abutment had also been moved southward, probably about 7 in. (178 mm) (2 in. [51 mm] joint closure plus 5 in. [127 mm] pier movement). Similar to the Old Third Street Viaduct, the pavement G/P phenomenon was probably responsible for this movement. In 1991, Indiana officials reported that: The northern approach to the bridge has been tilting southward for 8 to 10 years, according to old inspection photographs. The leaning was never enough to cause alarm. [2]

This bridge and its approach pavements were built in 1964. It appears that the approach pavements were about 27 years of age when the growth of pavements had progressed far enough to have closed the movable deck joints and moved all of the northern approach bridge elements (abutment, superstructure, and pier) enough to have provoked public concern about the structure’s stability. Apparently, the fixed pier columns (and presumably their pile-supported foundations) were tall and flexible enough to have tolerated 5 in. (127 mm) of longitudinal superstructure movement and pier tilting without noticeable pier distress. In response to this magnitude of movement, the concrete approach pavements were cut and pressure relief joints installed to eliminate restraint against pavement

Chapter 2 Damage and the Pavement G/P Phenomenon

27

Figure 2.4 A newspaper illustration that appeared in the October 24, 1991, edition of the Louisville Courier Journal showing the condition of the Northernmost approach spans of the John F. Kennedy Memorial Bridge. The illustration was slightly modified by the author to indicate closed deck joints and more representative abutment details.

growth and to minimize pressure transmission from the pavement to the bridge. As embankment and subsoil consolidation and translation at the north approach may have also contributed somewhat to substructure movement, motion detectors were implanted in the north abutment embankment. These detectors were used to monitor possible embankment movements and to ensure that the installation of pavement pressure relief joints was a sufficient response to prevent further longitudinal movement of both superstructure and substructure elements. Subsequent monitoring of the detectors indicated that the embankment was stable and not contributing to substructure movement. Presumably, the pressure relief joints will continue to be monitored and replaced or widened to ensure against further damage from the pavement G/P phenomenon. The movable deck joints have been rebuilt to facilitate anticipated superstructure movements due to live load rotations and longitudinal thermal movements. The rocker bearings will probably be righted when other maintenance repairs make such an effort more cost-effective. Pecos River Bridge Similar trouble for the Pecos River Bridge of Carlsbad, New Mexico (Figure 2.5) came to a head when local maintenance personnel were no longer able to contend with this structure’s unusual behavior and progressive deterioration. This bridge

28

Integral and Semi-integral Bridges

Figure 2.5

Pecos River Bridge, Carlsbad, New Mexico, 1941.

consists of eight deck-type, simply supported, rolled-beam spans with reinforcedconcrete wall-type piers and abutments. An inspection of the structure in 1991 revealed that the three easternmost movable deck joints were closed. The top of the massive wall-type east abutment had been moved and tilted to the west about 5.5 in. (140 mm), the ends of the deck slabs were crushed, and rocker and bolster bearings were shifted and tilted so much that they were edge bearing on the bridge seats (Figure 2.6). Bridge seats of the substructure units supporting the easternmost spans were badly cracked and fractured as well. As a result of the magnitude of the movement at the east abutment, it would appear likely that the adjacent piers were tilted to the west as well. The complete history of the bridge and its concrete approach pavements is uncertain. The bridge was constructed in 1941, but the date of construction of the present approach pavement has not been determined. A report about the bridge stated that, in 1983, the deck was repaired and overlaid, some repair work was done on the substructure, and bridge bearings and structural steel were repainted. Although the precise age of the approach pavement had not been determined, it was apparently old enough for its growth to have been responsible for the bridge’s recent rapid deterioration. From an unpublished inspection report about this bridge, state bridge engineers concluded: “We believe that the [abutment tilting,] deck crushing and bearing misalignments have been caused by the pavement shoving against the [abutment] backwall and bridge deck at the east end…. We have noticed similar problems with pavement shoving at many other bridges. However, the amount of movement and subsequent damage to Bridge No. 1838 is the worst

Chapter 2 Damage and the Pavement G/P Phenomenon

29

Figure 2.6 Rocker and bolster bearings at Pier 7 of the Pecos River Bridge. The rocked bearings of the eighth span were tilted and bearing against the bolster bearings of the seventh span. Bolster-bearing anchors bars were bent and the bearings themselves were horizontally displaced. Not visible are major fractures in pier walls below the bridge seat.

we have seen.” In addition to recommending various structure repairs, the immediate installation of pavement pressure relief joints in the bridge approaches was recommended.

Pavement G/P phenomenon As described above, serious trouble for the Old Third Street Viaduct, the John F. Kennedy Memorial Bridge, and the Pecos River Bridge became evident after these structures and their approach pavements were 11, 27, and less than 30 years old, respectively. As this trouble has been identified with the long-term behavior of jointed concrete approach pavement, it is obviously important that bridge design engineers become familiar with the long-term behavior of such pavements. They also need to understand the causes and consequences of the pavement G/P phenomenon, the phenomenon responsible for such an adverse effect on the performance, integrity, and durability of highway bridges. For this purpose, Appendix 1 has been appended to this book to aid bridge design engineers gain suitable familiarity with this phenomenon. In addition, a brief description of the phenomenon is provided below to introduce the subject to those who are becoming aware of it for the first time. If design had to contend with only the response of structures (pavement and bridges) to ambient temperature ranges, achievement of an efficient and functional

30

Integral and Semi-integral Bridges

transportation system would be relatively easy. However, concrete shrinkage and the less-than-ideal effect of traffic maintenance practices have compounded the problems faced by pavement and bridge maintenance engineers. Pavement contraction joints When pavement segments between transverse contraction joints contract longitudinally in response to lowering temperature and moisture levels, the transverse contraction joints between these segments open wider. Conversely, these joints become narrower due to the segments’ response to rising temperature and moisture levels. However, after the contraction joints open and remain open at low temperatures and moisture levels, compression-resistant fine roadway debris infiltrates the cracks below the saw-cut surfaces of the joints. This infiltration prevents subsequent complete closure to an as-constructed condition. Joint opening, infiltration of debris, and partial closing continue in sequence with daily and yearly temperature and moisture cycles. Of course, debris infiltration is facilitated where de-icing chemicals are used to ensure dry pavements (and consequently, open contraction joints) during low winter temperatures. Infiltration of these joints by compressive-resistant debris begins almost as soon as the cracks form below the saw-cut surfaces. Unsealed joints are infiltrated from the top, sides, and bottom. For “sealed” joints (actually a misnomer because only the upper surfaces of such joints are initially sealed), infiltration begins at the open sides and bottoms. The movement of surface water that penetrates pavement and shoulder joints from above, and ground water that seeps through shoulders and migrates along the sub-base below the joints, facilitate this infiltration. As contraction joint seals begin to fail because of a combination of age degradation, low temperature stiffening, traffic abrasion, neglect, etc., debris infiltration accelerates from both above and below. As a consequence of this contamination of contraction joints by compression-resistant debris, pavements will grow longitudinally in proportion to the amount of compression-resistant debris that infiltrates and accumulates in the joints (if such growth is not resisted). Where restraint against longitudinal growth is present, the pavement will both grow and be partially compressed. But where pavement growth is prevented, pressure generation commences and continues to accumulate until either the weakest pavement joints or the must vulnerable bridge elements are fractured, thereby relieving the built-up restraint stresses. Pressure generation The generation of longitudinal pavement pressures may be visualized as suggested in Figure 2.7, which illustrated an “idealized” chart of the yearly, maximum longitudinally oriented compressive stresses, f ′c, in an extensive length of restrained pavement without movable joints (no pavement pressure relief joints or movable bridge deck joints). Initially, the stress or pressure is insignificant because the joints are relatively clean and joint seals are intact and functioning. However, as the years pass and the joints begin to fill with debris, the yearly maximum pressure increases at a growing rate. As joints continue to fill, the accumulated compressive-resistant

Chapter 2 Damage and the Pavement G/P Phenomenon

31

Figure 2.7 Hypothesized stress- or pressure-generated curve for jointed rigid pavement (brick, stone blocks, concrete).

debris minimizes infiltration of additional material, slowing the rate of joint infiltration and pressure generation. Somewhere along this hypothetical pressure generation curve, the pavement fractures adjacent to a joint, relieving some of the pressure, or the pavement blows up, relieving all of the pressure at the location of the blow-up. A number of different analytical approaches can be used to illustrate the huge compressive stresses associated with the pavement G/P phenomenon. Generally, most reasonable assumptions about measured pavement behavior will yield stresses in the range of 1000 psi (7 MPa). For a 24 ft. × 9 in. (7.3 m × 230 mm) pavement, such pressures could result in a total longitudinal force of about 1300 tons (1200 tonnes) or more than 25 times the force usually assumed in the design of bridge abutments. Blow-ups Blow-ups are unmistakable indications of high pavement pressures. As they impede the movement of vehicular traffic, their occurrence is usually reported. When they occur with considerable regularity, the numbers of them that occur within a certain period are counted. These reports serve as an indication that high pavement pressures are not local peculiarities. Instead, they indicate that high pavement pressures are, or have been before the advent of pressure relief joint use, rather commonplace. The term “blow-up” is generally understood to mean an instantaneous fracture or buckling of pavement or both. Blow-ups may be triggered by the movement of vehicular traffic, but they are caused primarily by high longitudinal compression stresses within the pavement. Compressive stresses are relieved or released by blowups (Figure 2.7a,c). The size of blow-ups has not been quantified. They can consist of minor localized joint fractures and slight buckling, major fractures with little or no buckling, and occasionally minor fracturing with significant buckling. Before 1900, the buckling (or small blow-ups) of stone block streets was probably very commonplace due to the thinness of the pavement and irregularity of block

32

Integral and Semi-integral Bridges

(a)

(c)

(b)

(d)

Figure 2.8 Some pavement blow-ups of the twentieth century: (a) Ohio, about 1920; (b) Wood County, Ohio, 1963; (c) Akron, Ohio, 1975; and (d) Sacramento, California, 1998. For further documentation of these and other blow-ups, refer to Appendix 1, Table A1.

surfaces. However, with the introduction of jointed concrete pavement at the turn of the twentieth century, the G/P phenomenon began generating greater longitudinal pressures and inevitably larger blow-ups. The pavement maintenance engineer, C. D. Buck, had a clear conception of the problem when he reported his experiences with jointed concrete pavement blow-ups in Delaware in 1925 [3]. Similar blow-ups have been occurring periodically throughout the last century, some of which are illustrated in Figure 2.8.

End-jointed continuous bridges Where jointed pavements are constructed together with end-jointed continuous bridges, bridges with movable deck joints at the superstructure/abutment interface, such as the one shown in Figure 2.9, the pavement G/P phenomenon can overwhelm the abutment’s resistance to longitudinal pressures. Pavement growth can compress such bridges and move abutments (supported on other than rigid foundations) until the movable deck joints at the abutment/superstructure interface are permanently closed. Then the abutments begin to fracture. Figure 2.10 shows a close-up view of the west end of the bridge in Figure 2.9 where preliminary fracturing has taken place.

Chapter 2 Damage and the Pavement G/P Phenomenon

33

Figure 2.9 USR 52, Little Scioto River Bridge, Portsmouth, Ohio, 1964. This three-span, continuous steel girder bridge was seriously damaged by the pavement growth/pressure (G/P) phenomenon (see Figure 2.10).

Figure 2.10 Pavement growth/pressure (G/P) phenomenon damage to the Little Scioto River Bridge (see Figure 2.9). The abutment wingwall was fractured, the superstructure was lifted off of its bearings, and the girder webs were buckled. Also notice that the top of the abutment backwall in the roadway area has had prior fractures repaired with asphalt concrete.

34

Integral and Semi-integral Bridges

Two rows of steel H-piles with a battered front row support each abutment of this three-span, continuous welded, steel girder bridge. Even with such stiff foundations, pavement pressures were great enough to have moved both of the bridge’s abutments 3 in. (76 mm) horizontally, permanently closing the movable deck joints at the superstructure/abutment interface. Subsequent pressures generated by the restrained growth of the approach pavements at both ends of this bridge, coupled with the restrained expansion of the bridge itself, lifted the superstructure up and off its abutment bearings, fractured the top of the abutment backwalls (notice the asphalt patches in the roadway over the abutment backwall), and fractured the wingwall. This bridge and its approaches were just 8 years old when generated pavement pressures fractured these abutments and raised the bridge’s superstructure off of its bearings. To protect this bridge from further damage, the approach pavements were cut transversely and 2 ft. (0.30 m) pressure relief joints were installed. Figure 2.11 shows what occurred during the first attempt to cut the pavement. Afternoon temperatures near the bridge site on the day of the cutting were about 80 °F (27 °C). Even with the pressure relief provided by the compressed bridge joints and fractured abutments, pavement pressures were still great enough to squeeze the saw kerf together and freeze the saw blade in the pavement. Seven months after the installation of 2 ft. (0.61 m) pressure relief joints, the east and west joints were compressed ¾ in. (19 mm) and 1⅜ in. (35 mm), respectively.

Figure 2.11 A view of the eastern approach pavement of the Little Scioto River Bridge. Pavement pressures squeezed the saw kerf closed, thereby immobilizing the saw blade during the cutting of pavement in preparation for installation of a pressure-relief joint.

Chapter 2 Damage and the Pavement G/P Phenomenon

35

The damage done by the jointed concrete approach pavement to this continuous bridge with movable deck joints at the superstructure/abutment interface is not unique. In Ohio (and presumably in many other states), there are many hundreds of such end-jointed continuous bridges, mostly built in the 1960s and early 1970s, which have been similarly damaged by jointed approach pavements that were constructed without effective pressure-relief joints. Now, however, in many states, new bridges are being constructed with pavement pressure-relief joints in their bridge approaches. Such construction should protect these new bridges from the pavement G/P phenomenon as long as such joints are properly maintained. Existing bridge approaches are also being retrofitted with pressure-relief joints not only to prevent further bridge and pavement damage, but also to protect vehicular traffic from the hazards created by instantaneous pavement blow-ups.

Integral and semi-integral bridges Where jointed pavements without pressure-relief joints are constructed in conjunction with integral-type bridges, longitudinal pressures generated by the pavement G/P phenomenon are compounded by pressures generated by the restrained expansion of the bridges themselves. Together they generate higher pressures sooner than those caused by the pavement G/P phenomenon alone. However, due to their jointless construction, integral-type bridges are considerably more resistant to longitudinal pressures than the approach pavements. Consequently, when properly designed, these bridges can usually withstand without visible distress the pressures generated by both pavements and bridges. This is the primary reason why aware and pragmatic bridge engineers favor the use of integral or semi-integral bridges where jointed concrete pavements are used. Figure 2.12 illustrates the results of the pavement G/P phenomenon and bridge expansion on the approach slab of a three-span, continuous concrete, slab integral bridge. Pavement growth and bridge expansion compressed and ultimately fractured the bridge approach slab rather than the approach pavement proper because of the unusual geometric shape of the approach slab [3]. This approach slab blowup occurred on June 28, 1971, when the approach pavement was just 15 years old. Notice that black asphalt patches are evidence of prior fractures of both the approach slab and the top leading edge of the bridge slab. As the color of the patches is so dark and uniform, and the white painted roadway edge strip does not cross the patches, the prior fractures probably occurred earlier in the month of the same year or, most likely, not more than a year earlier than the blow-up. Presumably, maintenance engineers were not aware of the reasons for these early fractures. Otherwise, on appearance of the initial fractures, they could have installed pavement pressurerelief joints that would have prevented the blow-up. Also, such a maintenance response would have protected vehicular traffic from such a potential hazard, and it would also have avoided closing the pavement to all traffic while relief joints and new approach slabs were installed. Less than a mile away, on this same highway section, where the pavement G/P phenomenon was not compounded by the restrained expansion of an integral-type

36

Integral and Semi-integral Bridges

Figure 2.12 The northern approach slab of the SR 21, Barberton Reservoir Inlet Bridge, Akron, Ohio, 1956. This approach slab blow-up occurred on June 28, 1971 when this pavement was just 15 years old. The bridge is a continuous concrete slab supported by capped-pile piers and integral-type capped-pile abutments. Notice the initial transverse crack at the apex of the blowup and the asphalt patches of prior fractures (see also Figure 2.13).

bridge, the pavement structure was able to resist generating pressures for a longer period of time. Eventually, the pressures accumulated to such a level that an instantaneous blow-up occurred on June 24, 1975, 4 years after the failure of the integral bridge approach slabs (see Table A1 in Appendix 1, and Figure 2.8c). There is an interesting side aspect to the blow-up shown in Figure 2.12. Transverse cracks or fractures are quite common in the approach slabs of integral-type, continuous concrete, slab bridges as built in Ohio [4]. Numerous such bridge approach slabs exhibit transverse cracks similar to the one located at the apex of the buckled approach slab shown in Figure 2.12. In fact, these cracks are so predictable early in the life of such slabs (or slabs of similar geometric shapes and construction materials) that they could be used as crude indicators of longitudinal pressure generation in almost any type of jointed or continuous pavement. In addition these cracks (Figure 2.13) become so well defined after a brief rain shower that pressure generation in many miles of pavement could by determined by means of aerial photography alone.

Lack of awareness When considering the few representative examples of damaged pavements and bridges described in this chapter, why is the pavement G/P phenomenon, the

Chapter 2 Damage and the Pavement G/P Phenomenon

37

Figure 2.13 Approach slab of USR I-271, Wilson Mills Road Bridge, Cleveland, Ohio, 1963. Notice the transverse crack similar to the one visible at the apex of the blow-up shown in Figure 2.12. Also notice the narrow asphalt patches for prior fractures at the construction joint between the approach slab and the abutting bridge-deck slab. This photograph was taken just 7 years after construction of the approach pavement.

primary cause of this damage, not more widely recognized by bridge design and bridge maintenance engineers? Why are the characteristics of this phenomenon not more thoroughly understood by pavement maintenance and pavement research engineers? This general lack of awareness and understanding within the transportation profession is troubling. Perhaps bridge engineers do not believe that the restrained growth of jointed rigid pavements could cause such pressures, and thus such damage. This should not be considered so surprising because the major textbooks of the bridge engineering profession do not discuss, or even mention, this subject. Even AASHTO’s Standard Specifications for Highway Bridges [1], which mentions almost every conceivable force (including seismic forces that occur in century-long cycles), does not mention the forces generated by the restrained growth of jointed rigid pavement, forces that can be generated in new pavements in less than 10 years. The national Bridge Inspectors’ Training Manual has this to say about undesirable abutment movements: “The most common causes of lateral [abutment] movement are slope failure, seepage, changes in soil characteristics (e. g. frost action and ice) and time consolidation of the original soil” [5, 6]. This quote seems to hint that there may be other causes of abutment movement. However, this author does not hesitate to state that, for bridges with movable deck joints at the superstructure/ abutment interface, constructed abutting jointed concrete pavements, abutment movements due to the pavement G/P phenomenon are more common than the

38

Integral and Semi-integral Bridges

movements due to seepage, slope failure, frost action, and subsoil consolidation combined. Even nationally recognized authorities on pavement maintenance, who are aware of the need for pavement pressure relief, do not appear to be familiar with the characteristics of the pavement G/P phenomenon. This unfamiliarity becomes clear when their statements about “expansion joints” are examined. For example, the following advice is given in a report about pavement pressure relief: “1. Expansion joints near structures may provide [pressure] relief to the main line pavement as far as 2000 ft. (610 m) away” [7] (italics added). Although this quote contains a small element of truth, it can be very misleading if one were to assume that this relief continued beyond a very brief time period. Just glance at the bridge details shown in Figures 2.10 and 2.12. Both these bridges were provided with two standard pavement expansion joints (1 in. [25 mm] joints filled with compressible fillers) in each of their approaches. Obviously, with respect to minimizing the effects of restrained pavement growth, these expansion joints were almost worthless. The pavement G/P phenomenon will permanently compress such joints very quickly. After being compressed, these expansion joints become merely rigid pavement artifacts incapable of minimizing additional generating pavement pressures. There are probably many other reasons for this lack of familiarity with the pavement G/P phenomenon. Probably the primary reason is the lack of specific research. As evidence of this research void, consider that the research report just mentioned above [7]. It uses the term “pressure” over 450 times in the text of the report. Also, the bibliography of the report makes reference to 149 other publications of pertinent pavement research. Yet, astoundingly, this report, which is titled “Pressure Relief and Other Joint Rehabilitation Techniques,” does not contain a single reference in its bibliography to pavement pressure-relief research. Consequently, if pavement research professionals are not familiar with the pavement G/P phenomenon, it is also unlikely that bridge design and bridge maintenance professionals would be familiar with its characteristics. Pavement pressures are obviously recognized in the few research reports that focus on their reduction and control. Also, design details for pavement pressurerelief joints are now appearing in standard construction drawings of many state, county, and city transportation departments. Still, the subject has been mentioned in only two national publications, one of which is the author’s [8], and the other is FHWA’s Bridge Inspector’s Training Manual [5]. In the latter, pavement pressure is recognized as a bridge loading but its severity is significantly understated as follows: On concrete roadways, the pavements tend to migrate toward the bridge, pressing the approach slab against the backwall. Therefore, a pavement pressure relief joint is sometimes used to relieve this additional undesirable loading. (Emphasis added)

When reading this note, while glancing at the fractures shown in Figures 2.8, 2.10, and 2.12, it is easy to recognize that bridge approach pavements do not just tend to migrate towards bridges. They in effect are being relentlessly driven against bridge abutments. Consequently, it appears that the engineers responsible for this note were really not very familiar with the pavement G/P phenomenon, the results of which they were attempting to describe.

Chapter 2 Damage and the Pavement G/P Phenomenon

39

Summary This chapter has attempted to describe the bridge damage that has come to be associated with the pavement G/P phenomenon, the phenomenon that is characteristic of jointed rigid pavements. It describes and illustrates the type of bridge damage that the phenomenon induces in various types of structures. It also documents the fact that the phenomenon is rarely mentioned in bridge design literature. Yet, because it is responsible for so much bridge damage, it is still a mystery how such a destructive phenomenon has remained unknown to so many engineers in the bridge design profession. A few apprentice engineers are made aware of the phenomenon by experienced mentors who have gained awareness and familiarity with it through their own personal experiences with damaged bridges. But without such experienced mentors to guide them, and because the pavement G/P phenomenon is not recognized in bridge design textbooks or in bridge design specifications, unfortunately every bridge design and maintenance apprentice must now learn the same bridge damage lessons that their predecessors did, and unfortunately in the same surprising way. This situation, however, could be changed by recognizing the pavement G/P phenomenon in AASHTO’s Standard Specifications for Highway Bridges [1]. This could easily be done by adding the following brief paragraph: PAVEMENT FORCES: Jointed rigid pavement approaches to bridges shall be provided with effective and long-lasting pressure-relief joints. Otherwise, bridges to be constructed abutting such pavements without pressure-relief joints shall be designed to withstand the potential longitudinal forces generated by the pavement G/P phenomenon.

Such specification recognition will improve awareness of and interest in the pavement G/P phenomenon and its characteristics, and will eventually lead to design and maintenance practices that will blunt or minimize its destructive potential. Hopefully, this specification change can be accomplished before other generations of jointed highway bridges are constructed without such protection. In the meantime, knowledgeable bridge engineers will continue to give preference to the design and construction of integral and semi-integral bridges, the bridge types that can withstand the pressure potential of unrelieved jointed concrete pavement.

References 1. American Association of State Highway and Transportation Officials. Standard Specifications for Highway Bridges, 16th edn. American Association of State Highway and Transportation Officials, Washington, D.C., 1966. 2. “What’s Wrong with the Kennedy Bridge … and How Might it be Fixed?” Louisville Courier Journal, October 24, 1991. 3. Buck, C. D., “Repair of Concrete Road Blow-ups in Delaware,” Engineering News Record, Vol. 95, No. 11, 1925, pp. 432 and 433. 4. Burke, M. P., Jr., “Bridge Approach Pavements, Integral Bridges, and Cycle-Control Joints,” Transportation Research Record No. 1113, Transportation Research Board of the National Academies, Washington, D.C., 1987, pp. 54–65.

40

Integral and Semi-integral Bridges

5. Hartle, R. A., et al., Bridge Inspector’s Training Manual, Federal Highway Administration, McLean, Virginia, revised 1995. 6. Park, S. H., Bridge Inspection and Structural Analysis, 2nd edn, S. H. Park, Trenton, New Jersey, 2000. 7. Smith, K. D., et al., “Pressure Relief and Other Rehabilitation Techniques,” Report No. FHWA/RD-86/xxx, Federal Highway Administration, McLean, Virginia, 1987, p. A-10, 8. Burke, M. P., Jr., “Bridge Deck Joints,” National Cooperative Highway Research Program Synthesis of Highway Practice, 141, Transportation Research Board of the National Academies, Washington, D.C., 1989.

Chapter 3

Integral Bridges: Attributes and Limitations

You can know the name of a bird in all of the languages of the world, but when you’re finished, you’ll know absolutely nothing about the bird. … So let’s look at the bird and see what it’s doing – that’s what counts. Richard Feynman

Introduction Integral structures, or structures without movable joints, are ages old. The most celebrated are the natural arches carved from bedrock by water and wind. The largest such structure is the Rainbow Bridge National Monument in Utah near the Arizona border (see photograph). It is composed of pink sandstone and has a span of 278 ft. (85 m). However, when considering man-made integral bridges, one cannot go much further back in recorded history than the first arch bridges made of unreinforced concrete constructed by the Romans. More recently, most are familiar with the reinforced concrete arch bridges constructed in the early decades of the last century. It began with the substitution of reinforced concrete for stone masonry in the construction of spandrel-filled arch bridges. In these bridges, the pavement and spandrel fill are supported on one or more continuously reinforced arched slabs. Although many of these multiple-span spandrel-filled arches were constructed with movable joints in their spandrel walls and railings, many of the one- or 41

42

Integral and Semi-integral Bridges

two-span bridges of this type can be classified as true integral bridges because they were constructed entirely without any movable transverse joints in their arched slabs. By the third and fourth decades of last century, arch bridge construction culminated in the construction of long-span closed and open-spandrel arch bridges. Although the major supporting elements of these bridges (abutments, piers, and arch ribs) have no movable joints, they are not what could be classified as true integral bridges because they have such joints at each intermediate pier, and occasionally in the deck slabs and spandrel walls within each span. By mid-century, however, many transportation departments were building reinforced concrete rigid frame bridges. These bridges were a standard type of construction for many transportation departments. Those built in Canada by the province of Ontario are good examples. Although vertical movable joints were used between the rigid frame superstructures and their lateral wingwalls, these bridges can be classified as true integral bridges because they have no movable transverse joints in their primary supporting elements. The construction of rigid-frame bridges was paralleled by the construction of multiple-span continuous slab, beam, or girder bridges with embankments and shallow, rigidly supported, stub-type abutments. Movable deck joints were provided but only at the superstructure/abutment interface. Ultimately, the overall economy of multiple-span continuous construction made practical the use of similar bridges but without such joints at abutments. Movable deck joints at the superstructure/ abutments interface were eliminated by the use of integral stub-type abutments supported on single rows of vertically driven flexible piles (Figure 3.1). Although some versions of these bridges are now provided with movable bearings at selfsupporting piers, these are true integral bridges because they have no movable deck joints in their superstructures. Thus, although various types of integral bridges have been constructed for centuries, the designation “integral bridges” is now generally used to refer to single- or multiple-span, continuous, deck-type bridges without movable deck joints. They are generally supported by embankments with stub-type abutments on vertically driven flexible piles and by flexible piers constructed integrally with the superstructure (Figure 3.1a), or by semi-rigid piers provided with fixed and/or movable bearings (Figure 3.1b). For design engineers and engineering administrators who are considering integral bridges for the first time, a review of the brief discussions given herein should help to explain why these short- and moderate-span integral bridges are now being constructed with increasing frequency.

Attributes The popularity of integral bridges has grown with their numbers [1–3]. Originally built as a reaction to the destructive effects of leaking movable deck joints and massive pavement pressures, it soon became evident that these bridges have many more attributes and fewer limitations then their jointed bridge counterparts. Interestingly enough, these attributes not only reduced the first-cost and life-cycle cost of integral bridges, but also reduced the cost of their own future modification (e.g.,

Chapter 3 Integral Bridges: Attributes and Limitations

43

Figure 3.1 Multiple-span continuous integral bridges with stub-type abutments supported by embankments and single rows of vertically driven flexible piles. Piers can be either (a) flexible integral piers or (b) semi-rigid self-supporting piers with fixed and/or movable bearings.

widening) and eventual replacement. Consequently, integral bridges have been found to be ideal structures for state and county secondary road systems. With thoughtful design and crafting, they are becoming popular structures for rural and urban primary and interstate road systems as well. The integral bridge’s jointless construction, significant resistance to the pavement growth/pressure (G/P) phenomenon and its long-term durability has motivated bridge designers to construct such bridges at ever longer and longer lengths. For example, Ohio Department of Transportation (DOT) has recently increased its length limitation of 300 ft. (91 m) for integral bridges with skews of 30 ° or less, to longer than 300 ft. (91 m) for bridges, with lesser skews to up to a maximum of 400 ft. (122 m) for unskewed bridges. As noted in Chapter 1, Tennessee DOT recently shattered integral bridge length records by constructing the 1,175 ft. (358 m) Happy Hollow Greek Bridge of Hickman County, Tennessee (see photograph at start of Chapter 1). As discussions of attributes and limitations of integral bridges would have little significance unless they were considered with respect to those of other bridge types, the discussions that follow should all be considered relative to the attributes and limitations of similar single- and multiple-span continuous bridges with movable deck joints at superstructure/abutment interfaces. Simple design Where abutments and piers of a continuous bridge are each supported by a single row of piles attached to the superstructure (Figure 3.1a), or where self-supporting piers are separated from the superstructure by movable bearings (Figure 3.1b), an integral bridge may, for analysis and design purposes, be considered a continuous frame with a single horizontal member and two or more vertical members. When the stiffness and distribution factors are calculated for such a frame, the vertical members are so flexible compared with the horizontal member that the horizontal member may be assumed to have simple supports. Consequently, except for the design of the continuity connections at abutments, frame action of an integral bridge can be neglected when considering the effects of vertical loads

44

Integral and Semi-integral Bridges

applied to the superstructure (the superstructure can be designed assuming all movable bearings). The design of wide and relatively short integral bridges with flexible piers built integral with superstructures with concrete deck slabs is further simplified because piers and abutments generally need not be designed to resist either lateral or longitudinal loads. This is possible because the laterally and longitudinally rigid concrete deck slabs are rigidly attached to abutments and to one or more piers, and abutments are rigidly restrained by confining embankments. Consequently, essentially all lateral and longitudinal loads applied to the superstructures of such bridges are transmitted directly to abutment embankments. As a result, piers and abutments need not be designed to resist horizontal loads applied to superstructures. The design of abutment/superstructure continuity connections and transverse wingwalls can be standardized for a wide range of bridge applications. A nominal amount of reinforcement will be suitable to resist the slight live loads, dead loads, and secondary effects (shrinkage, creep, passive pressure, etc.) typical of such applications. Also, a nominal amount of reinforcement can be provided for transverse wingwalls to resist the maximum anticipated passive pressure. Once these standard details have been established, each bridge abutment can then be configured and reinforced for the vertical reactions associated with various roadway widths and span lengths. In general, this consists in no more than the determination of appropriate pile loads and spacing and pile cap reinforcement. The design of piers is similarly accomplished. Essentially all horizontal superstructure loads are transmitted to approach embankments. Also, the moments associated with pier/superstructure continuity connections are usually negligible. Therefore, piers of integral bridges (capped-pile or self-supporting types with movable bearings) need be designed only for vertical superstructure and pier loads and for lateral loads that may be applied directly to the piers (stream flow, stream debris, earth pressure, wind). Where these lateral pier loads are small, and this is usually the case, most piers, like abutments, can be designed essentially for vertical loads alone. A word of caution needs to be made with respect to flexible piers. For such piers that receive much of their lateral support from their connection to the superstructure, construction procedures are absolutely necessary to ensure that these piers are not laterally loaded until after they have been connected to the superstructure, and until after superstructure/abutment continuity connections have been completed. For example, bridge embankments must be placed and compacted before pier piles are driven if the piles are to be located in or adjacent to embankments. As the superstructure and abutment embankments resist all primary lateral loads, piers (piles, columns, footings, foundations) of wide and relatively short integral bridges may be reduced to minimum sizes and dimensions. Battered piles are not required. Fixed piers are not required. In general, pier design can be simplified to the extent that standard designs can be developed for a wide range of roadway widths and span lengths. Jointless construction The primary attribute of integral bridges is their jointless construction. To fully appreciate this attribute, one must be somewhat familiar with the performance of

Chapter 3 Integral Bridges: Attributes and Limitations

45

bridges with movable deck joints. For example, open joints permit contaminated roadway drainage to penetrate the joints and cause extensive below-deck deterioration. Closed joints or sealed joints give a measure of protection against bridge deck drainage deterioration. However, all movable deck joints (open, closed, or sealed) are vulnerable to the destructive effects of the pavement G/P phenomenon (see Chapter 2 and Appendix 1). Bridges with such joints constructed together with jointed rigid approach pavement have been inadvertently functioning as elaborate and expensive pavement pressure-relief joints. As approach pavements grow and the movable deck joints accommodate this growth by closing, the joints and the bridge superstructure are squeezed until the joints are fully closed. Thereafter, additional pavement growth and bridge elongation generate sufficient pressure to crush joint seals and fracture abutment backwalls and bridge seats. Consequently, the avoidance of movable deck joints obviates the need for maintenance prone joint seals and the extensive pressure damage repair that has come to be associated with them. As a secondary benefit, the avoidance of damage to movable deck joints also avoids the need for maintenance repair crews to be exposed to the hazards associated with vehicular traffic. In addition it avoids restricted traffic flow and the occasional vehicular accidents that are associated with bridge roadway repair sites. Also the smooth, jointless construction, characteristic of integral bridge roadways, improves vehicular riding quality and diminishes vehicular impact and stress levels. Pressure resistance The solid, jointless construction of an integral bridge distributes longitudinal pavement pressures by the pavement G/P phenomenon over a total superstructure area substantially greater than the approach pavement cross-section. Consequently, the smaller more fragile approach pavements are more likely to fail by localized fracturing or instantaneous buckling than the more pressure-resistant integral bridge superstructure. Unless the approaches to an integral bridge are furnished with cyclecontrol joints that are properly designed – joints that facilitate the thermal cycling of the bridge and approach slabs – they are more likely to experience early distress because the restrained elongation of the bridge also contributes to the generation of approach pavement pressure. As integral bridges are capable of sustaining significant longitudinal compression without distress, almost any pavement pressure-relief joint used specifically by maintenance forces to relieve pavement pressure would be suitable to protect them. However, bridges with movable deck joints need highly efficient pressure-relief joints if pavement pressures are to be reduced low enough to keep these joints functioning. Few such pressure-relief joints are now being used by pavement maintenance forces. Rapid construction Numerous features of integral bridges facilitate their rapid construction, and these features are probably responsible for much of the outstanding economy that has come to be achieved by their construction. Dry construction (most concrete work

46

Integral and Semi-integral Bridges

is done above stream water levels), simple members, broad tolerances, few construction joints, few parts, few materials, avoidance of labor-intensive practices, and many other features combine to make possible the completion of such structures in a single short construction season. Such rapid construction becomes particularly important when collapsed bridges or seriously damaged bridges must be replaced in the shortest possible time. Also, this rapid construction becomes particularly advantageous when bridges carrying high volumes of traffic must be replaced in two or more construction stages. As noted above, such rapid construction is possible because of the many simplifying features that are characteristic of most integral bridges. These include but are not limited by the following. Embankments Embankments can be placed and compacted with large earth-moving and compaction equipment. Only limited use of hand-operated compaction equipment is needed. Cofferdams Integral bridges, especially those constructed with capped-pile or drilled shaft piers, can be constructed with fewer delays resulting from inclement weather and stream flooding. Abutment excavations and pile driving near the top of abutment embankments can be done without cofferdams and generally without the need for dewatering. Foundation construction can progress as fast as pier and abutment piling can be driven. Subsequently, pile cap construction and erection of precast or prefabricated superstructure members can proceed with little regard for stream water levels. Small excavations Abutment excavations need be no more than 2–3 ft. (0.6–0.9 m) deep. Vertical piles At an abutment, vertical piles can be uniformly spaced and driven in a single horizontal row. In contrast, the typical non-integral abutment foundation consists of two or more rows of both vertical and battered piles. Pier piles These can be uniformly spaced and driven in a single horizontal row. This arrangement avoids the need for pile clusters with some battered piles for each column footing, the typical pier foundation for many cap and column piers of jointed bridges. For bridge sites with high water levels, driving piles for pier footings of jointed bridges is more difficult because piles must be driven inside cofferdams.

Chapter 3 Integral Bridges: Attributes and Limitations

47

Simple forms Pier and abutment pile caps are quickly formed because they are usually composed of simple rectangular shapes. Few joints Few construction joints are used for integral bridges. Consequently few concrete placement and curing days are needed. For example, no more than four concrete placement days are needed for most integral bridges. Only one day each is required for placing pier caps, continuity connections, deck slab, and approach slabs. Actually, single-span integral bridges in some states have been further simplified to the extent that only 2 days are required: the second day is necessary only to place separately cast approach slabs. In contrast, constructing most jointed bridges requires 5 or more placement days and subsequent curing days. Few parts Fixed and movable bearings, armor for bridge deck joints and joint seals are unnecessary. The normal delays usually associated with movable joint installation and adjustment, and anchor bar placement are avoided. Broad tolerances The close construction tolerances usually associated with jointed bridges are not necessary for integral bridges. For example, the elevation, slope and uniformity of bridge seats are not important because only rough surface construction joints are required. Reduced removals Using typical multiple-span integral bridges with embankments and stub-type abutments to replace shorter bridges with wall-type abutments permits new bridges to be constructed without requiring the complete removal of existing substructures. New bridges can be configured to straddle existing foundations (Figure 3.2). Where existing abutments are located in new embankments, most of existing abutments need not be removed. At many sites, significant savings are possible. For example, where normal water levels are high, complete removal of existing substructures could require the building of large cofferdams for this purpose alone. Simple beam seats Some of the labor-intensive practices required for jointed bridge construction are either eliminated or substantially simplified in integral bridge construction. For example, consider the problem of providing appropriate bridge seat surfaces

48

Integral and Semi-integral Bridges

Figure 3.2

New continuous integral bridge straddles old foundations.

for the elastomeric bearings of side-by-side and spread box beam-jointed bridges. Side-by-side box beams of jointed bridges must be canted laterally to match the deck crown and tilted longitudinally to accommodate bridge grade. Also, as the ends of these beams are sloped owing to residual camber, adjustments usually need to be made in beam bottoms, bearings, or bridge seats to compensate for these geometric irregularities and provide parallel loading surfaces for elastomeric bearings. A number of options are available to the designer: 1. 2. 3.

A longitudinally tapered recess can be cast in beam bottoms to match a longitudinally level and laterally crowned bridge seat surface. Bridge seats can be sloped to match the orientation of beam bottoms. Tapered metal laminates can be molded within the bearings to compensate for differences in the longitudinal orientation of beam bottoms and seat surfaces, and bridge seats can be laterally crowned to match the canted beams.

If this is not complicated enough, the specific provisions adopted to compensate for crown, grade, and camber may, in some bridges, have to be unique for each bridge seat because bearing orientation changes from one substructure to the next as a result of changes in grade and span lengths. In addition, poor estimates of residual camber, differences in residual camber from beam to beam, skew effects, errors on computing actual surface orientations, and errors in construction make the attainment of parallel-loading surfaces uncertain. Consequently, even after all of these considerations have been accounted for, occasionally it is necessary to use shims under elastomeric bearings to obtain solid seating of beams on bridge seats. On the other hand, integral bridge construction makes most of these considerations and procedures unnecessary. As box beams of integral bridges need only temporary support until continuity connections between the superstructure and abutments have been casts, a narrow temporary elastomeric erection strip can be used on a temporary construction joint surface to support the beams (Figure 3.3a). Then after continuity connections are cast and cured, because of the relative deformational characteristics of elastomeric erection strips and fully cured continuity connection concrete, all of the beam reactions (dead load, live load, and impact) will be uniformly supported by the rigid cast-in-place continuity connections and not by the compressible elastomeric bearings. These relative rigid connections are,

Chapter 3 Integral Bridges: Attributes and Limitations

49

Figure 3.3 Capped-pile stub-type abutments for integral bridges: (a) for prestressed concrete box beam stringers with temporary elastomeric erection strips; or (b) for steel I-beam stringers with temporary support bolts.

in addition far superior in supporting superstructure loads compared with a series of separate and uncertainly loaded elastomeric bearings that are characteristic of jointed box beam bridges. As concrete or steel I-beams are placed vertically, crown effects need not be considered when appropriate bearings and bridge seats are provided. However, even for I-beam bridges, integral construction with its continuity connections considerably simplifies bridge seat preparation and temporary bearing requirements and improves the distribution of superstructure reactions (Figure 3.3b).

Elimination of bearing anchor bars For typical jointed bridges, superstructures are usually attached at one or more substructure elements, usually at an intermediate pier. For side-by-side prestressed box beam bridges, this attachment is often done by placing anchor bars down through precast holes in the box beams and into field-drilled holes in the bridge seats. As a result of the construction uncertainties regarding beam lengths, and substructure locations, holes in the bridge seat must be field drilled after all of the beams have been placed and laterally compacted together. Considering the dimensional errors that are likely to occur in accurately locating substructures and in placing primary reinforcement in the pier caps, it is reasonable to assume that some of this reinforcement will be cut during field drilling of the anchor bar holes in the bridge seat. As superstructures of short box beam integral bridges receive their lateral and longitudinal support from abutment embankments, only flexible piers not integrally constructed with the superstructure (the type of piers that depend upon the superstructure for lateral and longitudinal support) need to be provided with anchor bars and field-drilled anchor bar holes. All other pier types (flexible integral piers and self-supporting piers with movable bearings) do not need bearing anchors

50

Integral and Semi-integral Bridges

and field-drilled anchor bar holes. Consequently, the damage usually associated with field drilling of anchor bar holes is avoided with these pier types. Of particular significance is the savings in time and costs made possible by eliminating the labor-intensive procedures of drilling and cleaning bar holes and placing grout and anchor bars. This savings in time and costs can be significant where project plans require anchor bars for every beam and for every support. Eliminating field drilling for anchor bar holes becomes particularly important in some bridge modification projects. For example, in replacing only the superstructures of existing bridges with new box beam superstructures (and this type of bridge modification project is occurring with increasing regularity), conversion of jointed bridges to integral bridges with cast-in-place continuity connections at abutments enables designers, when working with self-supporting piers, to provide elastomeric bearing pads without anchor bars. Fixed bearings and anchor bars are not required. By eliminating anchor bars and the need for field drilling of anchor bar holes, designers can not only accelerate construction, but can also ensure against the possibility of damaging existing primary pier cap reinforcement. Broad span ratios The ratio of end span to center span of continuous bridges (Le/Lc) is generally set at or near 0.8 to achieve stable superstructures and a balanced beam design, a design where the maximum positive moments in all spans are approximately the same. This is the ratio that is most often used for stream crossings. Lesser ratios are often used for grade separation structures where short end spans are needed to achieve the shortest practicable bridge length. However, for sites where span ratios of less than 0.6 are required for jointed bridges, provisions usually need to be made to prevent beam uplift during deck placement and uplift due to the movement of vehicular traffic. Such provisions can become complex and expensive when bearings must be provided that will allow horizontal movements of the superstructure but at the same time prevent superstructure uplift. Integral bridges, on the other hand, are more resistant to uplift because the weight of abutments resists uplift. Thus, a span ratio of 0.5 can be used without any change in integral bridge design details. For the smallest span ratios, a procedure for deck slab placement can be used to counteract uplift during construction. Earthquake resistant As the decks of integral bridges are rigidly connected to both abutments and consequently to both embankments, these bridges are in fact part of the earth and will move with the earth during earthquakes. Consequently, when integral bridges are constructed on stable embankments and subsoils, they should have an adequate response to most earthquakes. For a bridge that is to be located across a fault line – a highly unlikely situation – differential lateral movement of the ground at the fault line during an earthquake could seriously stress and possibly collapse the superstructure. However, an integral bridge with a steel beam or steel girder superstructure should be able to survive such unusual ground movements without collapse.

Chapter 3 Integral Bridges: Attributes and Limitations

51

Simplified widening and replacement Many bridges placed on the highway system in the past were designed for immediate needs without much thought being given to future bridge modifications that would be necessary to accommodate substantially higher traffic volumes and vehicular weights. Through arches of reinforced concrete, through trusses of steel, and bridges with wall-type abutments with flared wingwalls are prime examples of such bridges. Most often, through structures have to be completely replaced when increased traffic speeds and traffic weights necessitate building wider roadways and stronger structures. Widening bridges with wall-type abutments and especially those with flared wingwalls is complex and relatively expensive. In contrast, integral bridges with straight capped-pile substructures are convenient to widen and easy to replace if future demands have not been accurately foreseen. Of particular significance is the fact that their substructures (the piling) can be recapped and reused or, if necessary, they can be withdrawn or left in place. Such bridges avoid the necessity of building expensive foundations that interfere with the placement of future foundations. Many of the stream crossings in Ohio – and presumably the same is true for many states and provinces – have been spanned by at least three earlier bridges, with a fourth bridge presently being planned. For many of these earlier removed bridges, the foundations have been left in place. Consequently, when planning today’s replacement structure for small stream crossings, design engineers find that parts of these streambeds are filled with these old foundations. However, with the use of capped-pile integral substructures, the new substructures can usually be placed to clear existing foundations and avoid the expense of removing them. Also, the greater span ratio range makes integral bridges greatly adaptable for foundation-filled bridge sites. Live load distribution Superstructures that are integrally constructed with capped-pile abutments and piers instead of separated from them by numbers of compressible elastomeric bearings give vehicular wheel loads broader distribution than would otherwise be possible. This arrangement reduces superstructure service load stresses. The attributes noted above make integral bridges very desirable structures. However, this desirability has been attained at the expense of a number of limitations, and these limitations should be recognized so that design engineers can evaluate them when considering the use of integral bridges for specific applications.

Limitations Pile stresses Except for abutment piling and wingwalls, the various members of integral bridges are subjected to essentially the same levels of primary stresses (dead load, live load, impact, etc.) and secondary stresses (shrinkage, creep, thermal gradients, etc.) as their jointed bridge counterparts. But because of their flexural resistance, the

52

Integral and Semi-integral Bridges

vertical piling of integral bridge abutments will resist the lengthening and shortening of bridge superstructures responding to temperature changes. Consequently, the piling of long integral bridges can be subjected to flexural stresses considerably greater than those of their jointed bridge counterparts. For long integral bridges, research with abutments supported by steel piles has shown that abutment piling stresses can approach, equal, or even exceed the yield strength of pile material. Such piling stresses, if they are large enough, will result in the formation of plastic hinges that will limit the flexural resistance of the piling to additional superstructure elongation. At the same time, laterally supported piling should retain their capacity to support full vertical loads. As piles of integral bridges may be subjected to high flexural stresses, only suitable pile types should be used for these applications. Such piles should retain sufficient axial load capacity while localized pile transformations occur that will reduce their resistance to bending. For this reason, only steel H-piles or appropriately limited and reinforced prestressed concrete piles should be used to support abutments of the longer (≥300 ft. [≥91 m]) integral bridges. Some very informative pile-test research has recently been conducted to document the structural adequacy of these pile types for integral bridge construction [4, 5], particularly Federal Highway Administration Report [5]. The tests illustrated how common pile types performed successfully even at stress levels considerably beyond what is normally permitted for non-integral types of construction. For short integral bridges (<300 ft. [<91 m]), pile flexural stresses should be well within normal allowable stress levels for the material being used. Consequently, most of the usual steel or well-reinforced prestressed concrete piles can be used for these structures. In addition to the use of the most appropriate pile type, other provisions can be employed to ensure that the chosen pile type will operate well within the usual elastic limits. Steel H-piles can be oriented to place the weak axis parallel to the abutments; the skew of the structure can be limited (typically ≤30 °); piles can be placed in prebored holes filled with granular material; pile/footing connections can be generously reinforced to ensure adequate strength at this high stressed location; and appropriate reinforcement can be placed in concrete piles to facilitate the formation of plastic hinges. For short- and moderate-length bridges provided with the usual cast-in-place concrete, precast concrete, or steel H-piles, pile flexural stresses should be well within the elastic range. So no unusual provisions need to be made in their design. Limited applications The superior economy of integral bridges is due to their ability, within a limited application range, to satisfy all functional requirements with safety, durability and optimal economy. They are not broadly applicable to most bridge applications, as are their jointed bridge counterparts. Integral bridges with stub-type abutments supported by single rows of piles should be limited in a number of ways based on the primary design features that have been incorporated into standard designs. In general, their length should be limited (a) to minimize passive pressure effects and (b) to limit bridge movements

Chapter 3 Integral Bridges: Attributes and Limitations

53

to those that can be accommodated by the movement range of approach-slab/ approach-pavement cycle-control joints and standard approach-guardrail connections. They generally should not be used where curved beams, or beams with horizontal bends, are required, or with extreme skews (>30 °). They should not be used where abutment piles cannot be driven through at least 10–15 ft. (3–4.5 m) of overburden. They should not be used at sites where the stability of subsoils is uncertain or where settlement may be significant (where it cannot effectively be compensated for by added roadway overlays alone). Finally, because of superstructure buoyancy, they should not be used where they can become submerged unless the superstructure is vented, vertically restrained to resist uplift, or both.

Buoyancy As a result of their jointless construction, many types of integral bridges are subject to uplift when they become submerged. This is true for many I-beam bridges and box beam bridges. For single spans, abutments should be heavy enough to give the superstructure adequate resistance to uplift. But for multiple-span integral bridges, some positive provisions must be employed to ensure that a structure, which may become submerged, will have an adequate resistance to uplift. I-beam webs can be provided with 3 in. (76 mm) diameter holes placed near the top flanges and spaced throughout the length of the superstructure, counterweights could be provided, or holddown connections could be provided at piers. Instead of vent holes, added weight, uplift restraints, or integral pier construction, buoyant structures should be used only at those bridge sites where the highest floodwater levels are well below the bridge superstructure.

Construction procedures Embankments Abutments and piers of integral bridges composed primarily of reinforced concrete pile caps supported on single rows of piles have a very limited resistance to lateral loads. So abutment embankments and most major earthwork for such bridges must be placed and compacted first to ensure that lateral consolidation movements of subsoils both below and within the embankments have been allowed to take place before abutment and pier piles are driven. A typical plan note used for this purpose can be phrased as follows: EMBANKMENTS shall be constructed up to the roadway subgrade for a distance of 200 ft. (61 m) back of abutments before excavation is made for abutments, prebored holes placed, and pier and abutment piles driven.

The limitation given above for piers is important. Even though piers may be located beyond the toe of the abutment embankments, they can be adversely affected by subsurface movements if they are placed before completion and consolidation

54

Integral and Semi-integral Bridges

of embankment construction. However, if they must be placed before embankment construction, they also must be of the type that can resist lateral earth pressure without their depending upon attachment to the superstructure for support.

Concrete placement As concrete continuity connections at abutments and concrete approach slabs must be cast integrally with moving superstructures (superstructures that are continuously responding to changing ambient temperatures), concrete placement, especially for long bridges, should be controlled to minimize the effect of such movements on freshly setting abutment and approach slab concrete. It is generally not feasible to restrict concrete placement to those days of the year with the smallest temperature range, and consequently to those periods with the smallest potential for large superstructure movements. But it is practicable to limit concrete placement during daily periods when the superstructure movements are the smallest, generally shortly after the ambient temperature approaches, reaches, and departs from the day’s peak temperature. A plan note to provide such control and some protection for freshly placed concrete can be phrased somewhat as follows: CONCRETE for continuity connections at abutments shall be placed and completed at least 4 hours before the concrete-placement-day’s peak ambient temperature.

Approach slab connections to abutments should similarly be protected from the effects of a superstructure’s response to ambient temperature changes. A plan note somewhat as follows can be used for this purpose: APPROACH SLAB CONCRETE shall be placed toward the superstructure and be completed at least 4 hours before the concrete-placement-day’s peak ambient temperature.

To avoid damaging freshly placed continuity connection concrete during sudden ambient temperature changes, especially for long superstructures, in some states, superstructure beams are mechanically fastened to abutment pile caps before continuity connections are placed (see Figure 1.5b). Thus, after these attachments are completed, the continuity connections can be placed without concern for changes in ambient temperatures because concrete is being placed on pile caps that are moving with the superstructure. However, even for these structures, control of the placement of approach slab concrete is still necessary.

Uplift Deck slab concrete placement on integral bridges with short end spans must be controlled to eliminate uplift of the beams during placement. Such uplift can occur when both deck slabs and continuity connections at abutments are placed simultaneously. To avoid uplift in these applications, continuity connections should be placed first and adequately cured before placement of deck slab concrete.

Chapter 3 Integral Bridges: Attributes and Limitations

55

Approach slabs Full-width approach slabs should be provided for most integral bridges. They should be tied to the bridges to avoid having the slabs shoved off their seats by the constant longitudinal cycling of the bridges as they respond to daily temperature changes. To facilitate approach slab movement, sealed cycle-control joints should be provided between approach slabs and approach pavements to accommodate the longitudinal cycling of the approach slabs. These joints should be designed to prevent roadway drainage from penetrating the joints and flooding the sub-base. To protect these joints, effective pressure relief joints should also be provided between the joints and all rigid approach pavements. Approach slabs have a number of beneficial effects. By spanning between abutments and approach embankments, they help prevent vehicular traffic from consolidating backfill adjacent to abutments, thereby diminishing passive pressure effects. If approach slabs are long enough, they also eliminate live load surcharge on abutment backfill. They help to control bridge deck drainage, especially those slabs that have been provided with curbs, by conducting such drainage to approach pavements and to roadway drains. Such control of deck drainage helps prevent saturation and freezing of abutment backfill. They also help to minimize erosion of the backfill. Finally, they function as ramps from rigidly supported bridge abutments to consolidating approach embankments and thereby serve to help retain smoother riding surfaces and reduce vehicular impacts. In effect, approach slabs minimize the amount of continual maintenance that is necessary adjacent to bridges constructed without them. However, approach slabs tied to integral bridges become part of the bridges. Consequently, they effectively increase overall structure lengths, thereby requiring greater movement ranges for cycle-control joints. To minimize the amount of force necessary to move the approach slabs, they should be cast on smooth, low-friction surfaces (polyethylene sheets, filter fabric, etc.).

Cycle-control joints Integral bridges with attached approach slabs lengthen and shorten in response to temperature and moisture changes. For such structures built adjacent to rigid approach pavements, the joints between approach slabs and approach pavements should be provided with cycle-control joints to facilitate such movements. Otherwise, the longitudinal cycling of both structure and approach slabs can generate pressures sufficient to fracture approach slabs or approach pavement either progressively or abruptly (blow-up). Over time, jointed approach pavement will lengthen progressively (grow). Where such progressive movement is restrained by integral bridges, substantial longitudinal pressure will be generated in the pavements and bridges (see Chapter 2 and Appendix 1). To control such pressures, pavement pressure relief joints should be provided between rigid approach pavements and integral bridges. Consequently, two types of joints are required adjacent to integral bridges built in conjunction with jointed rigid approach pavement. One should be capable of responding to the progressive growth of approach pavement. The other joint should

56

Integral and Semi-integral Bridges

facilitate the longitudinal cycling of the integral bridge and attached approach slabs. Joint designs by four transportation departments are given elsewhere [6]. All of the designs have their limitations. To avoid the maintenance problems associated with complex or unusual cycle-control joints, Ohio DOT has chosen to use a single pavement pressure-relief joint for both purposes. This has been possible because the length of Ohio’s integral bridges has generally been limited to not more than 300 ft. (91 m). For existing bridges, old jointed bridges that have had their deck joints closed by pavement G/P phenomenon, 12 in. (0.3 m) pressure-relief joints have been installed. As pressures against these bridges will be increased as the joints are gradually closed by pavement pressure, these joints have to be periodically enlarged (widened) to control the level of pressure against these bridges. However, for new integral bridges, 4 ft. (1.22 m) pressure-relief joints filled with asphalt concrete and constructed on reinforced concrete sleeper slabs are now being used by Ohio DOT for both purposes. The primary fault with using pressure-relief joints for both types of movement is that these joints will open during cold weather. Such openings will permit roadway drainage to penetrate the joints and saturate the sub-base. Consequently, an effort should be made to ensure that such drainage cannot pond beneath these joints, otherwise a serious pavement pumping problem could result. Consequently, for those who are contemplating the use of pressure-relief joints to serve also as cycle-control joints, the use of sleeper slabs and effective subbase drainage provisions are critical details for the long-term performance of pressure-relief joint applications adjacent to integral bridges. As no other single joint design is available that will effectively compensate for both pavement growth and longitudinal cycling of integral bridges for extended periods of time, Ohio DOT maintenance engineers say that they prefer to use simple pressure-relief joints for both purposes. It is the only type that can be easily maintained by state and county maintenance personnel. Other more complex designs now available do not function well, are difficult to repair, and in many cases have to be replaced.

Research Research on passive pressures is needed to describe both the relationship between the amount of backfill compression and the generation of passive pressure, and the effect of alternating cycles of backfill compression and expansion. Until such research has been accomplished, current integral bridge design will depend on idealizations and simplifications that probably do not accurately predict long-term passive pressure effects. Shrinkage and creep studies are also needed for both integral bridges and their jointed bridge counterparts. Although current research in this area has been illuminating, the numerical procedures presently being recommended do not properly account for the composite behavior of various combinations of beam and slab sizes. Also, the results of recent computer studies have not been verified by comprehensive physical testing or presented in a form suitable for use by practicing design engineers. The lack of comprehensive research on passive pressure characteristics, and the consequent lack of specification provisions to account for the effects of these pres-

Chapter 3 Integral Bridges: Attributes and Limitations

57

sures, is probably the primary reason for the slow development of standard design details and construction procedures for integral bridges.

Summary As the above comments suggest, integral bridges have numerous attributes and few limitations. As design provisions can be made to account for most of these limitations (cycle-control joints, pressure-relief joints, approach slabs, construction procedures, buoyancy countermeasures, etc.) only application limitations (structure length, skew, curvature, overburden depth, subsoil stability, etc.) should negate their use in favor of their jointed bridge counterparts. In many areas of the country, integral bridges are now being used whenever application limitations do not prohibit their use. The high abutment piling stresses and the uncertainties about passive pressure levels are being accepted as the only negative aspects of such designs. In addition, even with these uncertainties, the excellent performance of integral bridges, even those of excessive lengths, skews, etc., have encouraged engineers with the most experience to use them for more and broader applications. Tennessee’s Happy Hollow Creek Bridge (photograph at start of Chapter 1) is a primary example of this tendency.

References 1. Emanual, J. H., et al., “Current Design Practice for Bridge Superstructures Connected to Flexible Substructures,” Civil Engineering Study 77-3, University of Missouri-Rolla, Rolla, Missouri, 1973. 2. Wolde-Tinsae, A. M., Klinger, J. E., “Integral Abutment Bridge Design and Construction,” Report FHWA/MD-87/04, University of Maryland, College Park, Maryland, 1987. 3. Burke, M. P., Jr., “Bridge Deck Joints,” National Cooperative Highway Research Program (NCHRP) Synthesis of Highway Practice 141, Transportation Research Board of National Academies, Washington, D.C., 1989, pp. 20–30. 4. Burdette, E. G., et al., “Behavior of Integral Abutments Supported by Steel H-Piles,” Transportation Research Record No. 1892, Transportation Research Board of the National Academies, Washington D.C., 2004, pp. 24–28. 5. Federal Highway Administration Report, Jointless Bridges, Volume I: Experimental Research and Field Studies, Chapter 7, Part 1, McLean, Virginia. (This reference is based on a prepublication text made available to the author in 2006.) 6. Burke, M. P., Jr., “Bridge Approach Pavements, Integral Bridges, and Cycle-Control Joints,” Transportation Research Record No. 1113, Transportation Research Board of the National Academies, Washington, D.C., 1987.

Chapter 4

Design of Integral Bridges: A Practitioner’s Approach

It is plain there is here but a restricted use for formulas. In this sort of practice, the engineer has need of some transcendental sense. … The rules must be everywhere indeed: but they must everywhere be modified by this transcendental coefficient, everywhere bent to the impressions of the trained eye and the feelings of the engineer. Robert Louis Stevenson

Introduction The analysis and design procedures and research findings that form the background for a simplified approach to the design of integral concrete bridges are described and discussed. The analysis and design of integral bridges with steel beams or girders and concrete deck slabs would be similar. The bridges envisioned for this discussion are wide and moderately long, single- or multiple-span continuous bridges without movable deck joints at the superstructure/abutment interface, and with embankments generally supporting short stub-type abutments on flexible piling. For continuous bridges, piers can be of the flexible integral types attached to superstructures (see Figure 3.1a), or semi-rigid self-supporting types with movable bearings (see Figure 3.1b). 59

60

Integral and Semi-integral Bridges

This simplified approach to the analysis and design of integral bridges will yield not only cost-effective bridges but also very durable ones. In addition, the use of such a simplified approach will enable the members of the bridge engineering profession to modify or replace, in a most expeditious manner, the many thousands of deficient or structurally obsolete bridges that are now burdening the nation’s transportation system. Integral bridges are subjected to the same loading as their jointed bridge counterparts. However, because of their integral construction, they are subjected to a number of secondary effects that are well known but difficult to quantify. Nevertheless, with a proper approach to the configuration of integral bridges and with the use of a number of limitations and simplifications, these secondary effects can be moderated and controlled to such an extent that the resulting designs will provide not only efficient and durable bridges but also cost-effective ones.

Secondary effects Like most of their jointed bridge counterparts, integral bridges are subjected to the same secondary effects, which include shrinkage, creep, thermal gradients, and differential settlement. But they are also subjected to passive pressure when abutment backfill is compressed during superstructure elongation, and to pavement or relief joint pressure when high temperatures and moisture trigger the growth of jointed concrete pavement. The stress levels generated by secondary effects are generally understood but as yet not well quantified. However, they can be controlled and provided for to such an extent that, except for continuity connections at supports, they usually need not be considered when designing single- or multiple-span continuous bridges that are less than 300 ft. (91 m) long. This simplification is possible because design specifications permit higher stresses when secondary stresses (shrinkage, creep, passive pressure, etc.) are combined with primary stresses (dead load, live load, and impact) to determine maximum allowable stresses. In addition, it should be remembered that secondary stresses do not alter the ultimate load capacity of structures.

Shrinkage The difference in shrinkage between a new concrete deck slab and partially aged concrete beams creates shear forces at the bonded boundary between these members. These shear forces, being eccentric with respect to the neutral axes of both deck slab and beams, subject both members to axial forces and bending moments. The effects on the composite structure are illustrated in Figure 4.1a. The force PS generated at the beam/deck-slab boundary due to differential shrinkage is conservatively estimated as (bending of members is neglected): PS = ε S ([1 ES AS ] + [1 E b Ab ]) where εS = unrestrained differential shrinkage (in./in. [mm/mm]), ES = elastic modulus of slab (psi [MPa]), Eb = elastic modulus of beam (psi [MPa]), AS =

Chapter 4 Integral Bridge Design: A Practitioner’s Approach

61

Figure 4.1 Shrinkage and creep effects for single- and multiple-span integral bridges: (a) differential shrinkage effects for composite concrete bridges (pavement pressure relief joint effects are similar); and (b) creep effects for composite prestressed concrete bridges (passive pressure effects are similar).

cross-sectional area of slab (in.2 [mm2]), and Ab = cross-sectional area of beam (in.2 [mm2]). The shrinkage moment, MS, associated with this force is: M S = PS ( y + e ) where PS = shear force (lb. [N]), e = distance from top of beam to center of slab (in. [mm]), and y = distance from beam neutral axis to top of beam (in. [mm]). An estimate of the unrestrained differential shrinkage is given by Freyermuth [1]. Shrinkage has the most effect on the positive moments of single spans and the continuity connection moments at abutments of continuous spans. It has a lesser effect on the negative moments at piers of two- and three-span bridges and only a slight effect on the positive or negative moments at the center of continuous spans. As the effects of shrinkage partially compensate for the effects of creep, passive pressure, and thermal gradients (see below), the use of shrinkage-compensating cement for the deck slab should be taken into account when designing continuity connections. Creep Except for the longitudinal compressive force, creep effects in single span beams made continuous are opposite to but greater than those produced by shrinkage (see Figure 4.1b). Again, these effects have been described and estimated by Freyermuth [1]. Interesting and informative half-scale model research by Mattock [2] shows the relationship between shrinkage and creep for a two-span composite prestressed I-beam bridge as both shrinkage and creep change with time. The two-span sketches of Figure 4.1 show support reactions associated with shrinkage and creep moments. The chart of Figure 4.2 shows the change in the pier reaction (and consequently the change in moments and forces) that takes place after forms and curing blankets have been removed from the newly placed concrete deck slab. As external loads on

62

Integral and Semi-integral Bridges

Figure 4.2 Shrinkage and creep studies by Mattock [2]. Clarifying sketches added by author.

the two-span structure remain unchanged, Figure 4.2 shows only the combined effects of shrinkage and creep that take place over the first 18 months. With respect to Figs. 4.1 and 4.2, it can be observed that maximum shrinkage moments occur within 30 days of form stripping with only slight creep effects. Consequently, negative moment reinforcement over piers and positive moment continuity connections at abutments can be designed by combining the effects of live load and superimposed dead load with just the shrinkage that occurs within the first 30 days after form stripping. Subsequently, creep effects begin to counteract shrinkage effects. At 7–8 months, the effects are balanced. After 2 years, continuity moments and vertical reactions are reversed by creep effects. Passive pressure The long-term effects of passive pressure on integral bridges have, as yet, not been thoroughly documented. Consequently, idealized formulas and simplified assumptions are used to account for the effects of passive pressure during design. The ultimate or maximum passive pressure, Ppu, generated in abutment backfill at a depth H, due to bridge elongation and backfill compression, is idealized as follows: Ppu = γ tan 2 (45 + ∅ 2) H + 2c tan 2 (45 + ∅ 2) where Ppu = ultimate passive pressure (psf [N/m2]), γ = unit soil weight (pcf [N/mm3]), Ø = angle of internal friction (degrees), H = depth of calculated pressure below approach slab (ft. [m]), and c = soil cohesion (psf [Pa]). This relationship has been used to guide the design of abutments and backfill of integral bridges. A granular backfill is specified to minimize Ø values and eliminate cohesive effects, the depth of abutment, H, is made as small as is practicable, and

Chapter 4 Integral Bridge Design: A Practitioner’s Approach

Figure 4.3

63

Simplified passive pressure distribution.

approach slabs are used to minimize consolidation of backfill and eliminate live load surcharge on backfill. As total passive pressure on the structure will depend upon the area of the abutments exposed to pressure, abutments should be reduced to the smallest practicable dimensions by the use of embankment benches located close to the superstructure, spill-around 2 : 1 embankment slopes, and shallow penetration of abutments below bench surfaces. In addition, it has been shown [3, 4] that the magnitude of passive pressure is related to the magnitude of soil compression, with a soil compression of about 5 percent of the abutment depth necessary to achieve maximum pressure. Consequently, passive pressure effects can usually be neglected for single-span and short two- and three-span bridges. For most modest length integral bridges, only twothirds of the maximum value needs to be considered during design. Passive pressure distribution on an integral abutment due to superstructure elongation and backfill compression is illustrated in Figure 4.3. To reduce flexural resistance of piles to lateral movement (parallel to the longitudinal axis of the structure), the piles are oriented for weak axis bending and, if necessary, they can be placed in prebored holes filled with pea gravel. The resultant passive pressure, Pp, is idealized as follows: Pp = ( 1 3 ,

2

3

or

3

3

) Σ γ tan 2 (45 + ∅ 2) H 2 2 .

The depth of abutment, H, can be assumed to be slightly greater than the actual depth to account for passive pressure on the piles and pile resistance to lateral movement. This pressure resultant acts eccentric to the composite neutral axis of the superstructure and produces axial loads and bending moments on the superstructure. The moment diagrams of Figure 4.1b show that bending moments induced by passive pressure counteract dead and live load-bending moments in simple spans (increases the service load capacity). Except for negative moments at abutments of continuous spans, passive pressure reduces the negative moments at piers for twoand three-span bridges, and slightly increases center span-positive moments of three-span bridges. For bridges of four or more spans, continuity moments induced by passive pressure are similar.

64

Integral and Semi-integral Bridges

Relief joint pressure The restrained growth of all jointed rigid pavement can generate almost irresistible longitudinal forces. To protect bridge superstructures adjacent to such pavements, pressure-relief joints should be provided in bridge approaches. As integral bridges can sustain substantial longitudinal compression, almost any pressure-relief joint used by maintenance forces should be adequate. Nevertheless, pavement pressure transmitted through these joints should be considered in the design of superstructures. Casual observations of pavement growth at pressure-relief joints reveal that such growth can be rapid and incremental rather than slow and progressive. Consequently, the pressure used in design should be determined from laboratory testing when such joints are instantaneously closed ¼ in. (6.4 mm) or more. As this longitudinal compressive force will be transmitted to the bridge through the approach slabs, it too will be eccentric with respect to the neutral axis of the superstructure and will induce bending moments throughout the superstructure similar to those due to shrinkage. Settlements and deflections Settlement of abutments of multiple-span integral bridges induces bending moments similar to those induced by shrinkage (Figure 4.4a). Settlement of piers induces moments similar to those induced by creep (Figure 4.4b). In both cases, moments at piers are opposite to and substantially greater than those at abutments. The magnitude of these moments is proportional to the amount of settlement. As it is generally not possible to reduce settlement-induced negative moments at piers by raising superstructures at integral abutments, integral bridges should not be used unless the probability of appreciable abutment settlement is remote. Differential vertical deflections and concomitant bending moments at piers and abutments, caused by the placement of superimposed dead loads, can be minimized by equalizing pier and abutment pile loading. Differential vertical deflections due to the movement of live loads along the structure, and the concomitant compression

Figure 4.4 Moments induced by differential settlements: (a) abutments with respect to the pier and (b) pier with respect to abutments.

Chapter 4 Integral Bridge Design: A Practitioner’s Approach

65

of substructures and foundations, are similar to those produced by settlement. However, the deflections are usually so small in comparison to the length and flexibility of the spans that the moments induced are considered negligible and generally ignored. Thermal gradients Depending upon the relative top-to-bottom temperatures of the superstructure, the moments induced by thermal gradients are similar to those illustrated in Figure 4.1. However, where structures are exposed to sustained high temperatures, the moments induced by thermal gradients (modified to account for the compensating effects of concomitant decreases in moisture content along with increases in temperature levels) should be combined with those resulting from other secondary effects to ensure adequacy of continuity connections. As illustrated in Figure 4.1, continuity moments due to shrinkage, relief joint pressure, creep, and passive pressure will have little effect on the positive moments at mid-span, a modest effect on the moments over the piers, and most effect on the continuity connections at abutments. The simultaneous axial forces, however, are all compressive. The moments induced by thermal gradients are similar to those due to shrinkage or creep, although the simultaneous axial load due to thermal effects could be either tension or compression induced by the resistance of abutments to longitudinal bridge movements and extreme low or high temperatures, respectively. An examination of a number of trial designs has shown that, except for the design of single spans and the continuity connections of continuous spans, the sum of these secondary effects is so small in comparison with the usual dead and live load effects that, after establishing a nominal amount of continuity reinforcement, secondary effects can usually be ignored for most bridges less than 300 ft. (91 m) long (Figure 4.5). However, for longer bridges, especially for those bridges where passive pressure effects are greatest and for bridges exposed to sustained high temperatures, a summary of these secondary effects should be evaluated and accounted for during design.

Simplifying assumptions Superstructures Where abutments and piers are supported by single rows of flexible piles, or where semi-rigid self-supported piers are separated from the superstructure by movable bearings, an integral bridge may, for analysis purposes, be considered to be a continuous frame with a single horizontal member and two or more vertical members. When calculating the stiffness and distribution factors for such a frame, it is found that the vertical members are so flexible in comparison to the horizontal member that the horizontal member (the continuous superstructure) may be assumed to have simple or hinged supports. Consequently, except for the design of the continuity connections at abutments and piers, frame action can be ignored when

66

Integral and Semi-integral Bridges

Figure 4.5 Moment comparisons for composite prestressed concrete beams: (a) primary moments and (b) secondary moments.

analyzing the superstructure for superimposed dead and live loads. This assumption yields slightly conservative positive moments and slightly unconservative (±5 percent) negative moments. Substructures The design of integral bridges can be further simplified because piers and abutments generally need not be designed to resist lateral and longitudinal loads applied to the

Chapter 4 Integral Bridge Design: A Practitioner’s Approach

67

superstructure. This is made possible because the laterally and longitudinally rigid concrete deck is rigidly attached to both abutments, and the abutments are rigidly restrained by the confining embankments. Consequently, essentially all lateral and longitudinal loads applied to superstructures of integral bridges are distributed directly to embankments.

Design comments Superstructures In addition to designing the superstructure for the usual primary stresses (dead load and vehicular live load) and making allowances for secondary effects (shrinkage, creep, thermal gradients, etc.) in the design of the continuity connections, provisions should also be made to protect the structure from the effects of buoyancy and snow plows.

Buoyancy Due to their jointless construction, many types of integral bridge superstructures are subjected to uplift when they become submerged. This is true for many I-beam bridges and some box beam bridges. For comments about this critical aspect of integral bridge construction, refer to buoyancy comments in Chapter 3.

Snow plows As the deck surfaces of integral bridges are constructed without movable deck joints and joint seals, the destructive effects of snow plows on such surfaces are minimal. However, edge armor used on approach slab joints should be recessed slightly below the surface (⅛ in. ± ⅛ in. [3 mm ± 3 mm]) to ensure against snagging the plows.

Earthquakes As decks of integral bridges are rigidly connected to both abutments and consequently to both embankments, these bridges are in fact part of the earth and will move with the earth during earthquakes. Consequently, when integral bridges are constructed on stable embankments and subsoils, they should have an adequate response to most earthquakes.

Substructures Abutments The design of abutment/superstructure continuity connections and transverse wingwalls can be standardized for a wide range of bridge applications. It will be found that a nominal amount of reinforcement will be suitable to resist the slight

68

Integral and Semi-integral Bridges

live and dead loads typical of such applications plus a wide range of secondary effects (shrinkage, creep, passive pressure, etc.). Also a nominal amount of reinforcement can be provided for transverse wingwalls to resist the maximum anticipated passive pressure. Once these standard details have been established, each bridge abutment can be configured and reinforced for the vertical reactions associated with various roadway widths and span lengths. In general, this usually consists of no more than determining appropriate pile loads and spacing, and pile cap reinforcement.

Piers As essentially all horizontal superstructure loads are distributed to approach embankments, and as the moments due to pier/superstructure continuity are negligible, typical piers of integral bridges (flexible capped-pile piers or the semi-rigid self-supporting piers with movable bearings) need to be designed only for vertical superstructure and pier loads, and the lateral loads that may be applied directly to the piers (stream flow, stream debris, earth pressure, wind). Where the lateral pier loads are of little consequence, and this is usually the case, most piers, like abutments, can be designed for vertical loads alone. As primary lateral loads are resisted by superstructures and abutment embankments, piers (piles, columns, footings, foundations) of integral bridges may be reduced to minimum sizes and dimensions. Battered piles are not required. Fixed piers are not required. In general, pier design can be simplified to the extent that standard designs can be developed for a wide range of roadway widths and span lengths. For piers that receive much of their support from their connection with superstructures (i.e., capped-pile piers), construction procedures are necessary to ensure that these piers are not laterally loaded until after they have been connected to the superstructures. This is usually accomplished by requiring all major embankment earthwork to be completed before the placement of pier piles.

Summary and recommendations As described above, the structural continuity typical of integral bridges alters the distribution of secondary effects (shrinkage, creep, thermal gradients, etc.), the lateral and longitudinal loads applied to superstructures, and superimposed dead and live loads. In addition, this same continuity induces secondary effects due to the generation of passive pressure in abutment backfill. As illustrated in Figure 4.5, this design complexity is real but not significant. Consequently, bridge design engineers concerned about project schedules should guard against being diverted away from the business of constructing durable bridges to an intellectual study of secondary effects for every bridge design. Instead, the design of integral bridges should be simplified and standardized. Three primary steps are necessary. First, the application range for such bridges should be limited until after the initial design standards have been proven by design applications, construction

Chapter 4 Integral Bridge Design: A Practitioner’s Approach

69

experience, and bridge performance throughout several seasonal cycles. A number of primary limitations should be considered:

• • • • • • •

Limit bridge lengths to not greater than 300 ft. (91 m) Limit bridge skews and curvature to not greater than 30 ° and 5 °, respectively Provide continuous construction for multiple-span superstructures Provide embankments and stub-type abutments supported by single rows of vertically driven piles Provide abutments with flexible piles not shorter than 10 ft. (3 m) and preferably not shorter than 15 ft. (4.6 m); orient piles for weak axis bending Provide approach slabs with 6 in. (150 mm) high curbs adjacent to bridge decks with curbs or parapets. Tie approach slabs to the abutments. Provide pavement joints for rigid pavements to allow longitudinal bridge cycling and pavement growth [5, p. 63].

Second, the basic bridge design should be crafted to minimize secondary effects by one or more of the following:

• • • • • •

Provide abutments with well-drained granular backfill Provide prebored holes for the upper portion of piles to be driven in dense or cohesive soils Provide pea gravel as a void filler Provide embankment benches to help limit abutment heights and wingwall lengths Provide generous reinforcement in wingwalls to resist maximum passive pressure Provide adequate continuity reinforcement between abutments and superstructures.

Finally, specify that embankments be constructed before driving either abutment or pier piles. Also, require a waiting period for embankment and surcharged subsoil consolidation before driving piles to minimize post-construction settlement of piers and abutments. After these initial design steps have been taken, all subsequent design should have been simplified to the extent that superstructures could be treated as continuous members on simple supports, and most piers and abutments could be designed for vertical loads alone. Such an approach to integral bridge design will yield not only the most economical bridges, but also the most durable ones. It will thus also enable the members of the bridge engineering profession to modify or replace, in the most expeditious manner, the many thousands of small, deficient, or structurally obsolete bridges now burdening the nation’s transportation system.

References 1. Freyermuth, C. L., “Design of Continuous Highway Bridges with Prestressed Concrete Girders,” PCI Journal, Precast/Prestressed Concrete Institute, Skokie, Illinois, April, 1969, Vol. 14, No. 2.

70

Integral and Semi-integral Bridges

2. Mattock, A. H., “Precast-Prestressed Concrete Bridges – 5, Creep and Shrinkage Studies,” Journal of the PCA Research and Development Laboratories, Portland Cement Association, Skokie, Illinois, May, 1961, Vol. 3, No. 2. 3. Wu, T. H., Soil Mechanics, T. H. Wu, Worthington, Ohio, 1982, pp. 276–280. 4. “Foundations and Earth Structures,” Design Manual 7.2, Department of the Navy, Naval Facilities Engineering Command, Alexandria, Virginia, 1982, p. 7.2–60. 5. Burke, M. P., Jr., “Bridge Approach Pavements, Integral Bridges and Cycle-Control Joints,” Transportation Research Record No. 1113, Transportation Research Board of the National Academies, Washington, D.C., 1987.

Chapter 5

Genesis of Integral Bridges

Creative thinking may mean simply the realization that there is no particular virtue in doing things the way they have always been done. Rudolf Flesh

Introduction Separated by six decades in time and more than 8,000 miles (12,900 km) in distance, the SR 7 Teens Run Bridge at Eureka in southeastern Ohio (see photograph), and the new Naibekoshinai River Bridge of Hokkaido, Japan (see photograph at start of Chapter 4), have one thing in common – they can both attribute their significant design feature to a third bridge, the open-spandrel rib-arch Ashtabula Viaduct of Ashtabula, Ohio. The relationship of these three bridges is based on conjecture because there is only circumstantial evidence linking them, but that evidence is clear and compelling. These three bridges have major differences. For example, the Teens Run Bridge is a continuous reinforced concrete slab structure, the Naibekoshinai River Bridge is a continuous prestressed concrete box beam structure, and the Ashtabula Viaduct is an open-spandrel, reinforced concrete, rib-arch structure. But they all share one primary attribute: each span of the Ashtabula Viaduct and the multiple spans of both the Teens Run Bridge and the Naibekoshinai River Bridge can be classified as 71

72

Integral and Semi-integral Bridges

integral because they have no movable deck joints in their superstructures, or between their superstructures and flexible abutment supports.

Flexibility The integral type of design originated with the Ashtabula Viaduct (Figure 5.1) and was adopted for the design and construction of all subsequent major arch bridges built by Ohio Department of Transportation (DOT) in the early 1930s and 1940s. It has since been adopted for other structure types including continuous concrete slab bridges and continuous concrete or steel stringer-type bridges. In Ohio, it is also being adapted to the design and construction of continuous, prestressed concrete, I-beam and box-beam bridges. Over the last several decades, many other states and Canadian provinces have adopted this type of design. Now the Naibekoshinai River Bridge is the first integral bridge constructed in Japan, the prototype for the many such bridges that are certain to follow. This design innovation, among others, had its beginning in 1925 with the appointment of J. R. Burkey as the Engineer of Bridges for the Ohio Department of Highways. Under his direction, Ohio became the first state to adopt the “routine” use of continuous construction for all multiple-span bridges [1]. That this design innovation was a significant departure from contemporary bridge design practice is attested to by the fact that even today, more than 80 years later, a few state transportation departments (now less than 10 percent of the nation’s state transportation departments) have not as yet adopted continuous construction as the method of choice for the design and construction of multiple-span bridges.

Figure 5.1

USR 20, Ashtabula River Valley Viaduct, Ashtabula, Ohio, 1928.

Chapter 5 Genesis of Integral Bridges

73

As early as 1926, Burkey mentioned his concern about the design and construction of movable deck joints. In a paper about recent developments in highway bridges, given at a national conference of state highway bridge engineering officials, Burkey [2] summarized the many concerns and new design changes that were being considered by the bridge engineers of Ohio DOT. With respect to these joints he stated: There are several details of design which I am sure have been receiving more attention in recent years than they did formerly. Perhaps the foremost of these in concrete work is the expansion joint [movable deck joint], and devices designed to give it freedom of action. … A little later a slide will show how we [in Ohio] are proposing to take care of this problem on an important structure [presumably, the Ashtabula Viaduct] by the use of flexible columns which eliminate any sliding action. [2]

The following year, in a local magazine article about bridges being built by Ohio DOT, Burkey [3] singled out the Ashtabula Viaduct for comment: Among the more notable structures under construction this year is the Ashtabula Viaduct crossing the valley of the Ashtabula River on the Chicago-Buffalo Road, U.S. Route No. 20, at Ashtabula, Ohio. … With the roadway 100 feet [30.5 m] above the foundation of the piers, this structure when completed will present a striking appearance. The tall, slender columns, designed with a special slenderness to permit them to yield to temperature changes in the deck, afford the most modern solution to this vexing problem, and at the same time add a boldness to the appearance of the structure that is pleasing. By this type of construction, which avoids all sliding expansion joints, it is believed that the structure will be free from the local disintegration and failure which has so frequently attended the type of … joint commonly used in the past. [3]

In a 1928 report of Ohio DOT for the years 1917–1928 [4], a short description of the major bridges designed and constructed by the Bureau of Bridges is given. With respect to several paragraphs about the Ashtabula Viaduct, elimination of deck joints is prominently mentioned again: The outstanding feature of design is the complete separation of the superstructure at the end of each span using only abutting expansion joints. To accommodate deck movement [within each span], long columns were used with a slenderness especially adapted to flexure. [4]

The decision to use flexible columns to support a continuous deck slab, rather than rigid columns and slabs segmented by many movable deck joints, raised concern about two aspects of the design. Would the columns supporting a continuous deck slab be flexible enough to minimize column flexural stresses, and would composite action between the arch ribs and continuous deck slab have an adverse effect on the structure’s response to moving loads? To help alleviate these concerns, field observations and model studies were undertaken. The Department’s 1928 report [4] mentions these efforts: … observations were made during construction on the Conneaut Viaduct, the Miami River Bridge at Piqua, and the Ashtabula Viaduct to assist in studying the behavior of

74

Integral and Semi-integral Bridges

arches and their supporting piers under varying temperatures, loads, and foundation conditions. This Bureau has also conducted extensive tests on the action of arch bridges by means of elastic models and Beggs Deformeter Gages. Cardboard models were made of the Ashtabula Viaduct, both with and without the spandrel columns and the deck, and microscopic deflections were measured to determine the effect of the [continuous] floor system on the arch ribs when acting monolithically. These observations indicate that the [continuous] floor system materially affects the behavior of the arch ribs on such a structure. … The models were made and tested by Mr. D. H. Overman, design engineer in this Bureau. [4]

Throughout the 14-year period beginning with the design of the Ashtabula ribarch Viaduct of Ashtabula, Ohio in 1926 and concluding with the construction of the US Route 40, Big Darby Creek rib-arch Bridge of West Jefferson, Ohio in 1940, the staff of the Bureau of Bridges was engaged in the design of eight major rib-arch bridges. With respect to the largest of these bridges, the Brookpark Viaduct of Cleveland, Ohio, W. H. Rabe and D. H. Overman [5] coauthored a 1934 national magazine article about its design features. Again, the innovation of eliminating movable deck joints by providing flexible supports is mentioned: The practice of omitting expansion joints in the arch deck except at piers, was followed on this structure because there is always the risk of water leakage and concrete deterioration at a joint and because tests conducted by the designers on paper and celluloid models of previously designed structures indicated that intermediate joints are unnecessary. The spandrel columns and the unusually slender pier columns were designed for flexure from expansion and contraction of the deck, assuming the columns as fixed top and bottom. … [5]

Continuous construction As has been stated, the bridge design staff of the Bureau of Bridges pioneered the development of continuous construction of highway bridges by avoiding the use of movable deck joints at piers of all multiple-span structures as early as 1926. This decision to employ continuous construction made the analysis and design of these structures more complicated, but it also gave impetuous to the conception and development of the technological innovations necessary to make such a goal practicable. These innovations included the use of more secure foundations to minimize settlement of supports (friction piles, end-bearing piles, and concrete pedestals to bedrock) and development of riveted and field-welded splices for prefabricated steel beam and girder structures. In this respect, Ohio was the first transportation department to adopt the routine use of field butt welding to achieve continuity for steel stringer bridges. During the 1930s, while some members of Ohio’s Bridge Bureau staff were designing and constructing a series of major rib-arch bridges, other members of that staff were engaged in the design and construction of small continuous concrete slabs and continuous concrete or steel stringer-type bridges. With respect to the design of continuous concrete slab bridges, the typical substructures being used for these structures consisted of various types of supports depending upon site char-

Chapter 5 Genesis of Integral Bridges

75

acteristics. Wall-type piers and flexible capped-pile piers were common. However, abutments consisted of walls or stubs supported on two rows of piles or reinforced concrete pedestals that extended down to bedrock. Slabs were keyed or fixed to one or two substructure elements, with the other substructure elements provided with sliding bearings. For some bridges, rigid abutments were huge by comparison to capped-pile piers. These abutments were also relatively complex to design and construct, and they were expensive. As the continuous deck slabs of arch bridges and continuous slab bridges were being designed simultaneously by members of the same design staff, one or more of these designers must have observed the huge differences in the supports that were being employed for these similar structures. As both slabs serve the same purpose and are exposed to the same live loads, shrinkage, and thermal changes, they must have questioned the justification of using flexible supports for rib-arch deck slabs and both rigid and flexible supports for continuous concrete slab bridges, especially as deck joints had to be provided at one or more of the rigid abutments of slab bridges. Consequently, these designers appear to have decided to use the same design concept for continuous concrete slab bridges that they were successfully using for the continuous concrete slabs of their open spandrel rib-arch structures. All that was necessary to make this transition was to employ flexible capped piles for both piers and abutments of slab bridges for those site locations where an adequate depth was available for providing pile foundations (Figures 5.2 and 5.3).

Figure 5.2 A 1995 photograph of the 1939 Teens Run Bridge. Note the use of precast reinforced concrete piles and cast-in-place, reinforced concrete, pile caps.

76

Integral and Semi-integral Bridges

Figure 5.3

Stub-type integral abutment details, Teens Run Bridge.

Thus it appears that the concept of continuous bridges with flexible integral abutments was born. As the Teens Run Bridge is the longest of these early integral bridges, it can be considered the primary prototype of hundreds of similar bridges that have followed. Initially, these were square structures with the same geometric configuration as the deck slab of rib-arch bridges. But, as the bridge staff observed the performance of these early prototype designs, they apparently then decided to use similar structures even at site locations where up to 30 ° skews were necessary.

Teens Run Bridge As evidenced by the initials on design drawings prepared by the staff of the Bureau of Bridges of the Ohio Department of Highways (now Ohio DOT), the Teens Run Bridge was designed and checked by Virgil Eberly and Josephine E. Powers, respectively. Plans for this bridge were reviewed by both William S. Hindman and William H. Rabe. Usually, plans for the department’s small bridges were reviewed only by Rabe, the bureau’s chief design engineer. For other bridges, especially larger or more complex bridges, the plan review effort was usually shared by one of the more experienced staff designers. Consequently, it is of significance to observe that the other reviewer who shared the review of the tiny Teens Run Bridge, and the primary reviewer, was Hindman, who was the designer of the 1930 Brecksville rib-arch Viaduct of Brecksville, Ohio and the 1931 Broad Road rib-arch Viaduct of Bedford, Ohio, two of the largest rib-arch bridges designed by Ohio DOT. This shared

Chapter 5 Genesis of Integral Bridges

77

reviewed responsibility for a rather small concrete slab structure is indicative of its uniqueness and relative importance. Consequently, it appears that one of the individuals responsible for the adaptation of Ohio’s rib-arch design techniques to continuous concrete slab bridges was Hindman. The author presumes that D. H. Overman, the designing engineer of bridges at the time, and the engineer primarily responsible for most of the design innovations adopted by the Bureau of Bridges, was also partly responsible for what has come to be known as the integral bridge concept. The Teens Run Bridge is significant in another respect. Although it is the longest of these early integral structures, it has, unlike most of its contemporaries, survived intact without the usual concrete deterioration. Although this structure, including its concrete railing (Figure 5.4), had been exposed to yearly applications of de-icing chemicals for almost six decades, it looks in these 1995 photographs as if it had been constructed within the last decade. Consequently, it is a truly remarkable structure in several respects.

Figure 5.4 Original railing of the Teens Run Bridge. As a result of winter applications of roadway de-icing chemicals, the railings of this bridge have periodically been enveloped by chemically contaminated roadway spray. Nevertheless, these railings must be considered remarkably durable because they were more than a half-century old when this photograph was taken. Except for slight deterioration of post tops, the panels and posts appear to be in almost new condition. It is a pity that the engineering profession has been unable to achieve this kind of durability for all reinforced concrete bridges.

78

Integral and Semi-integral Bridges

Japanese connection In October of 1994, the author was honored by the opportunity to give a lecture on integral bridge design and construction to the bridge engineering staff of Nihon Doro Kodan, the Japan Highway Public Corporation. After that lecture, the author was introduced to Shoichi Takahushi, one of the corporation’s bridge design staff members who stated that he had just completed the designs for two of Japan’s first integral bridges. In later correspondence, he stated that the staff ’s interest in the integral bridge concept was motivated by the successes of engineers in California, Ohio, and Tennessee with this concept, and upon the numerous papers about integral bridges by the author who documented the success of Ohio’s integral bridges, bridges that owed their existence to both the Teens Run Bridge and the Ashtabula Viaduct. Thus, although they are separated by about seven decades and 8,000 miles (12,900 km), the Naibekoshinai River Bridge of Hokkaido, Japan (see photograph at start of Chapter 4) appears to be one of the recent descendants of the celebrated Ashtabula Viaduct and a distant cousin of the innovated but obscure little Teens Run Integral Bridge of southeastern Ohio.

References 1. Burke, M. P. Jr., “Bridge Deck Joints,” NCHRP Synthesis 141, Transportation Research Board of the National Academies. Washington, D.C., 1989, pp. 20–21. 2. Burkey, J. R., “Recent Developments in Highway Bridges,” Highway Topics, Ohio Good Roads Federation, Columbus, Ohio, March 1926, p. 11. 3. Burkey, J. R., “Bridges Built by the Department of Highways,” Highway Topics, Good Roads Federation, Columbus, Ohio, November 1927, p. 13. 4. Ohio Department of Highways, Report of the Department of Highways (1917–1928 inclusive), Ohio Department of Highways, Columbus, Ohio, 1928, pp. 156–157, 160–161. 5. Rabe W. H., Overman, D. H., “Long Concrete Arch Viaduct Built Near Cleveland,” Engineering News Record, McGraw-Hill, Inc., October 1934, p. 468.

Addendum: Basic bridge data Ashtabula Viaduct Built in 1928, this is a multiple-span, open-spandrel, reinforced concrete, rib-arch structure with continuous reinforced concrete deck slabs supported by flexible reinforced concrete columns. It has twelve spans consisting of three at 45 ft. (13.72 m), seven at 135 ft. (41.15 m), and two at 45 ft. (13.72 m). It also has a 32 ft. (9.75 m) roadway face-to-face of 5 ft. (1.68 m) sidewalks (see Figure 5.1).

Naibekoshinai River Bridge Built in 1996, this is a continuous, prestressed concrete, box-beam structure with reinforced concrete T-type piers and capped-pile abutments. It has three spans

Chapter 5 Genesis of Integral Bridges

79

consisting of one at 36 m (118.1 ft.), one at 38.5 m (126.3 ft.) and one at 34.5 m (113.2 ft.). It also has a 10.5 m (34.5 ft.) roadway toe-to-toe of parapets (see photograph at start of Chapter 4).

Teens Run Bridge Built in 1939, this is a continuous reinforced concrete slab on capped-pile piers and abutments. Pier and abutment piles are precast reinforced concrete. It has five spans consisting of one at 24′ 11″ (7.59 m), three at 30′ 11″ (9.42 m) and one at 24′ 11″ (7.59 m). It also has a 26 ft. (7.92 m) roadway face-to-face of railings (see photograph at start of chapter).

Chapter 6

Cracking of Concrete Decks and Other Problems with Integral-type Bridges

Our achievements speak for themselves. What we have to keep track of are our failures, discouragements, and doubts. We tend to forget past difficulties, the many false starts, and the painful groping. We see our past achievements as the end result of a clean forward thrust and our present difficulties as signs of decline and decay.

Eric Hoffer

Introduction This chapter is intended to help focus attention on some of the more significant problems encountered during the evolution and construction of integral and semiintegral bridges, two bridge types that are constructed without movable deck joints in their superstructures. Although integral and semi-integral bridges are simple in concept (Figure 6.1), there are enough differences between them and their jointed bridge counterparts (bridges with movable deck joints) that some problems arise, especially during construction, that are not expected by those considering the construction of integral-type bridges for the first time. Also, integral-type bridges are 81

82

Integral and Semi-integral Bridges

Figure 6.1 Representative abutment details are shown at left for integral bridges supported by embankments, stub-type abutments, and flexible (pile) foundations. At right are abutment details for semi-integral bridges supported by embankments, stub-type abutments with pile supported semi-rigid foundations. A movable longitudinal joint at the superstructure/abutment interface is also provided. Not shown are other types of semi-integral bridge abutments including stub- or wall-type abutments on more rigid piling, pedestals, bedrock, etc.

composed of the same materials and built with similar construction practices as jointed bridges. Consequently, their construction is prone to the same or similar problems that have plagued construction and maintenance of jointed bridges for decades (early age transverse bridge deck cracking, embankment consolidation and erosion, marginal performance of structure movement systems [see Chapter 10], etc.). Therefore, although adoption of integral-type bridges will eliminate some of the more troublesome problems associated with jointed bridges and yield significantly more durable structures, they will not eliminate endemic highway bridge construction problems that are somewhat related to accelerated construction, allweather construction, marginal construction supervision, etc. However, it is hoped that the following discussions, explanations, and suggestions will aid in achieving more successful applications of integral-type bridges and thus encourage their more widespread adoption and construction.

Deck slab cracks Diagonal deck slab cracks located at acute corners of integral-type bridges are occasionally reported. Manufacturers of prefabricated deck joint seals are quick to suggest that such cracks are indicative of the questionable integrity and limited durability of such bridges. With respect to the appearance of curved and essentially diagonal cracks, recently such cracks were observed in the new deck slab of an integral conversion of a threespan continuous deck-type bridge with movable deck joints at the superstructure/ abutment interface. This conversion consisted in the removal of the existing deck

Chapter 6 Concrete Deck Cracking and Other Problems

83

slab and abutments and replacing them with new abutments cast integrally with a new deck slab. This project was constructed in conformance with Ohio Department of Transportation’s (DOT’s) Bridge Design Regulations and Construction and Materials Specifications. In addition to diagonal cracks at acute corners, other equally prominent but relatively straight cracks were also observed. Some of the uniformly spaced straight cracks were located over and perpendicular to previously placed concrete end diaphragms. An example of these cracks are shown in Figure 6.2, a figure that was prepared from measurements made about a year after the deck concrete was placed. Other less pronounced, relatively straight cracks, perpendicular to longitudinal supporting steel stringers (commonly designated as “transverse cracks”), are located throughout the structure. Periodic examinations of this bridge revealed that the cracks shown in Figure 6.2 were lengthening and turning laterally as crack tips projected further beyond the sides of the end diaphragms. At acute corners of the deck slab, cracks are now perpendicular to transverse end diaphragms at one end and perpendicular to longitudinal stringers at the other end. Apparently, these are the “diagonal” cracks that were occasionally observed and reported. Contrary to the captious integral bridge criticisms noted above, most knowledgeable bridge design and construction engineers recognize early age deck slab cracking

Figure 6.2 Early age deck slab cracks in one of twin structures (USR 62, Yankee Creek Bridge, Trumbull County, Ohio, 1952). These cracks occurred when the right structure was reconstructed in 1994 with integral abutments and a new deck slab.

84

Integral and Semi-integral Bridges

as symptomatic of bridge projects where lapses or omissions occurred in the application of one or more of the precautionary measures usually used to prevent or at least minimize such cracking. Prevention of early age cracking of concrete bridge decks (primarily transverse cracks at relatively uniform longitudinal spacing) has been a continuing concern of bridge design and construction engineers. That concern has been reflected in attention being focused on those aspects of concrete technology and concrete placement procedures that can be controlled to prevent or at least minimize early age deck slab cracking. These aspects include:

• • • • •

Type, quality, proportions, and temperature of concrete constituents Mixing, placing, and curing of concrete Differential thermal gradients within fresh composite sections* due to the heat of hydration and the environmental conditions (primarily ambient temperature extremes and variations) that take place during concrete placement, finishing, and curing Flexural stressing of fresh composite sections during concrete placement and machine finishing Initial post-cure differential shrinkage of composite sections.

The practices and precautions that have evolved over many years to control these aspects that contribute to early age deck slab cracking are now usually part of the requirements for every bridge project. The first three items and last item are usually controlled by standard construction and material specifications; the next-to-last item, which depends upon a bridge’s response to deck slab placement, can be minimized by providing specific concrete placement sequences in bridge plan requirements. The success or lack thereof of bridge designers and construction engineers in achieving crack-free bridge decks has been examined. A comprehensive survey of Transverse Cracking in Newly Constructed Bridge Decks [1] has found that such cracking is widespread and not uncommon. This observation is not surprising when one considers that the research title attests to the fact that early age transverse cracking of bridge decks was widespread enough to motivate state transportation departments to fund a national research project specifically designed to study this one phenomenon. In addition to discussing the primary aspects that contribute to this problem, the study also established that: Generally, the risk of transverse deck cracking increases when girder sizes increase, when thinner decks are used, when spans are continuous instead of simply supported … and when steel girders are used instead of concrete girders if spans are simply supported. … Concrete properties generally are the most important factors affecting transverse deck slab cracking. … [1]

Not mentioned in this report were any noticeable differences between the behavior of deck slabs of jointed bridges and those of integral bridges. *References in this chapter to composite sections refer to deck slabs and stringers that are functionally composite and not just to those that are mechanically joined together by shear connectors.

Chapter 6 Concrete Deck Cracking and Other Problems

85

Early age deck slab cracking in Ohio usually occurs during hot weather. It appears to be primarily and directly related (a) to the type and temperature of concrete constituents, and (b) to differential thermal strains of exposed and thermally responsive support stringers with respect to water-cooled and form- and curing blanket-insulated freshly setting deck slab concrete. Initial cracks are widened and extended after curing by differential thermal and shrinkage strains between slabs and support stringers, and integral end diaphragms. Tensile stressing of fresh composite sections (due to their flexural response to continuing concrete placement and machine finishing) has on occasion also contributed to the location and severity of deck slab cracks. For example, the deck slab cracks shown in Figure 6.2 occurred in one of a pair of side-by-side twin integral structures. The deck slab concrete of the first structure was placed during a warm summer day. Following the appearance of these cracks, the deck slab of its essentially identical twin was placed at night when differential temperatures between slab concrete and supporting stringers were minimized. As a result, most deck cracking was avoided in the twin structure. Because of similar experiences with other bridges, some Ohio DOT Districts require summer placements of concrete bridge decks to be made at night. Incidentally, it has been recognized that high-strength concrete is more prone to transverse deck slab cracking than its lower-strength counterparts. In this respect, the use of high-performance concrete, primarily to improve superstructure durability, may be somewhat counterproductive because this less permeable concrete may yield more transverse cracks. This may be the reason why some construction specifications for high-performance concrete contain provisions for sealing deck slab cracks. With respect to the twin integral structures described above, it also appears that portions of the visible deck cracks at abutments (normal to the superstructure end diaphragms) were caused because the designer neglected to provide continuous reinforcement in the deck slabs over the end diaphragms to resist stresses induced by temperature and shrinkage changes. In recent years, the use of deck slab placement sequences was routine. For the usual continuous bridges with deck joints at the superstructure/abutment interface, portions of the deck slab over piers were placed last. This procedure helped to ensure that fresh concrete of composite superstructures, in negative moment regions over piers, would not be subjected to flexural tensile stresses due to deck slab dead load. However, with the advent of self-propelled finishing machines and set-retarding admixtures, it became more cost-effective on any bridges, except the longest, to permit contractors to place deck slab concrete in one rapid continuous placement from one abutment to the other. Occasionally, tensile stressing of fresh deck slabs resulted in slab cracking. To help compensate for the probability of such cracks, crack sealers were developed that could be applied to deck surfaces after curing. For longer, more significant structures, set retarders, placement sequences, night placement, and crack sealers can all be used to help achieve acceptable results. Similar care and concern should be taken with the design and construction of deck slabs of integral bridges. Suitable shrinkage and temperature reinforcement should be provided in end diaphragms, and in deck slabs over end diaphragms at abutments. Set retarders and rapid concrete placement should be required. During

86

Integral and Semi-integral Bridges

adversely hot weather, concrete constituents should be cooled before mixing, and placement at night should be considered. Where flexural stressing of fresh concrete is probable, placement sequences should be used to allow superstructure concrete for negative moment areas to be placed last. Where cracking still persists, cracks should be effectively repaired. Effective crack repair (instead of just sealing) may be important for fatiguesensitive superstructures exposed to heavy truck traffic, especially where fatigue design strengths of primary members were predicated on composite interaction and full participation of deck slab concrete. Continuity connections One of the primary attributes of bridges with movable deck joints at the superstructure/abutment interface is that they allow placement of concrete for individual elements of a bridge (abutments and superstructures) to be made without fresh concrete being adversely affected by differential movements between these separate elements. This attribute of jointed bridges is lost with the adoption of integral-type bridges. Consequently, in the construction of such bridges, special precautions should be taken to compensate for the adverse effects of these movements. When constructing integral-type bridges, stationary abutments and moving superstructures (responding to ambient temperature variations and deck slab dead load changes during concrete placement) must be joined together by cast-in-place continuity connections (Figure 6.3D). Consequently, these fresh connections could be stressed and cracked if a substantial temperature drop were to occur during initial concrete setting, or if concrete-placement sequences were not suitably controlled. To address this problem, several concrete-placement procedures (in addition to the precautionary procedures noted above for minimizing deck slab cracking) are now being used. These procedures are itemized as follows:

Figure 6.3 Integral abutment construction joints A, B, and C, plus superstructure continuity connections D for two different concrete placement procedures.

Chapter 6 Concrete Deck Cracking and Other Problems

1. 2. 3. 4. 5.

87

Placing continuity connections at sunrise Placing deck slabs and continuity connections at night Placing continuity connections after deck slab placement Using crack sealers Using one or more of the above.

Ohio DOT has had reasonable success constructing short single- and multiplespan continuous integral bridges less than 300 ft. (90 m) long while allowing contractors to place deck slab concrete continuously from one abutment to the other. In addition, only the usual specification provisions are used to control deck slab cracking, except that continuous water curing (and cooling) is always used. Items 1, 2, and 4 also are occasionally used on a case-by-case basis. With respect to transverse differential temperature and shrinkage effects between deck slabs and end diaphragms at abutments, and with respect to longitudinal tensile stressing of deck slabs at abutments due to further slab placement, it is informative to note that Tennessee DOT recommends use of item 3. For this procedure, transverse deck slab construction joints (Figure 6.3C, left abutment) are placed adjacent to and parallel with abutments so that the entire deck slab between these two transverse construction joints can be placed first. Subsequently and lastly, superstructure end diaphragms and the remainder of the deck slab concrete at abutments are completed in a single placement (Figure 6.3D, left abutment). This deck-placement sequence not only minimizes transverse differential temperature effects at abutments, but also ensures that most span deflections and end rotations occur before concrete continuity connections are placed, thus preventing premature flexural stressing of fresh concrete. With respect to achieving continuity connections free of early age cracking, item 3 has two primary advantages. As the short-term or “effective” temperature movement coefficient of individual superstructure members is changed (reduced) when these separate members are compositely integrated with the concrete deck slab, giving a single, less responsive, structural cross-section (material solids and air voids and, for steel/concrete composites, lower temperature conductivity), placing deck slabs before continuity connections not only slows the composite superstructure’s response to short-term ambient temperature changes, but also reduces the magnitude of the differential longitudinal movement between superstructures and abutments while continuity connections are being cast. Second, most dead load rotational movements of superstructures at abutments are designed to take place before placing continuity connection concrete, thus eliminating flexural stressing of fresh concrete, one of the primary culprits of early age deck slab cracking. This procedure also has the added advantage of eliminating bending of abutments and abutment piling due to placement of concrete deck slabs. Advantages gained by design choices usually have one or more negative consequences. In this respect, item 3, placing continuity connections last, has at least three that should be recognized and evaluated to ensure that the most suitable results are achieved. First, contractors would be prevented from placing deck slab concrete, machine finishing, and installing curing materials in one continuous direction. Such discontinuous placement might, for longer bridges, actually require two concrete placement days instead of one. Not only will this procedure increase bridge costs,

88

Integral and Semi-integral Bridges

in some instances such as the replacement of a collapsed bridge or the completion of a multistage bridge replacement, the extra concrete placement day might, depending upon the weather, delay reopening a road to regular traffic by many days. Second, unless mechanical hold-down connections are provided as described below, continuous bridges with very short end spans would no longer be a viable integral bridge option because placement of deck slab dead load on a significantly longer adjacent span could, depending upon relative span lengths, result in uplift of the end-span stringers during concrete placement. Third, although occasional random flexural cracking would be prevented, two well-defined “cracks” in each deck slab (the construction joints themselves) would always result. For unusually long bridges, especially those composed of steel stringers, it would be prudent to eliminate the possibility of differential longitudinal temperature movement between stringers and abutments before placing continuity connection concrete. This could be done by mechanically fastening stringers and pile caps together (see Figure 1.5b). Instead of using temporary stringer support bolts that bear on top of abutment pile caps, these bolts could be cast integrally with the cap. Then, during erection, stringer flanges (provided with longitudinally slotted holes to allow for dimensional irregularities) could be fastened directly to support bolts and thus to pile caps. After being fastened together, differential longitudinal temperature movement between stringers and abutments during casting of continuity connections would be essentially prevented. Similar fasteners could also be provided for unusually long concrete stringers, although such stringers are not as sensitive to rapid ambient temperature changes as steel stringers. Similar mechanical connections could be used to expedite replacement of small flood-damaged or collapsed bridges. Instead of fastening stringers to pile caps at abutments, they could be fastened directly to the piles, if the piles have been laterally spaced to coincide with stringer spacing and extended vertically and field modified to provide secure stringer seats. After stringers have been placed and fastened to pile tops, abutment and deck slab forms completed, and reinforcement placed, concrete for the entire structure, except for raised sidewalks and parapets, could be placed during a single concrete placement day. Although transverse differential thermal and shrinkage strains would be minimized by this design approach, crack sealers might be necessary to seal rotationally induced flexural cracking, unless item 3 placement procedure was used to minimize the possibility of such cracking. Some of the precautions and procedures described above may be needed for placing bridge approach slabs, especially slabs of unusually long integral bridges. As slabs are initially stationary structures (anchored to approach subgrade by slab weight and friction) and superstructures are moving in response to ambient temperature variations, continuity connection items 1, 2, 4, and 5 should be considered. In addition, to reduce frictional forces, approach slab concrete for integral bridges could be placed on bond-breaking flat sheets of polyethylene. For unusually long integral bridges, it may be advantageous to use mechanical connectors and concrete closure placements between approach slabs and superstructures to avoid early age cracking. On the other hand, it may be more cost-effective to seal just occasionally occurring approach slab cracks (item 4), rather than impose more costly crackavoidance controls.

Chapter 6 Concrete Deck Cracking and Other Problems

89

Stage construction As a result of their relatively simple structural characteristics, few basic materials, and rapid construction, integral-type bridges are ideally suited for stage reconstruction or replacement of existing bridges. This is especially true for those projects where early construction completion is needed to minimize the period during which restricted traffic flows must be maintained on only portions of existing and proposed structures. However, rotational restraint provided by abutments of integral bridges and by end diaphragms of semi-integral bridges present bridge designers with unique problems not encountered by designers of simply supported jointed bridges. And these problems become particularly significant with longerspan bridges. In stage construction or reconstruction of long-span, deck-type highway bridges, where transverse portions or segments of bridge cross-sections must be removed and replaced in sequence to permit maintenance of vehicular traffic, stage construction is controlled to the extent that placement of deck slab concrete of subsequent stages will be accompanied by the same stringer deflections and rotations as produced by first-stage placements. To achieve such uniform stringer deflections and rotations, dead load distributions, and surface elevations, superstructure stages are kept separated until deck slabs have been placed on adjacent stages. Such separation is accomplished by not connecting or by disconnecting diaphragms or cross-frames between stages and providing very narrow deck slab closure placements between stages. Diaphragms or cross-frames between stages are then connected before deck slab closure placements are made to minimize live load effects on these placements. Such complete separation of superstructure stages during reconstruction has other beneficial effects. For example, separation allows work on main superstructure members to be done without them being detrimentally affected by live load stresses and deflections. It also minimizes live load-induced movements of transverse deck slab reinforcement that protrude from stage construction joints during placement and setting of abutting and enveloping deck slab concrete. Relatively complete separation of superstructure stages and a deck slab closure placement procedure were used during a recent semi-integral bridge conversion of the SR 771, Big Branch Creek Bridge, Highland County, Ohio, an existing endjointed, single-span bridge. However, after placement of second stage deck slab concrete of this bridge (Figure 6.4), it was found that the dead load deflections of the second stage were considerably less than those of the first stage. As a result, the second stage deck slab surface at the center of the span was more than 1 in. (25 mm) above the adjacent first stage surface. The problem with this application occurred because the bridge designer did not require “complete” separation of the two superstructure stages. Instead, the first stage deck slab and the entire end diaphragms at both abutments were completed up to the underside of the proposed deck slab (see Figure 6.3B right abutment) before placement of second-stage deck slab concrete. Consequently, these end diaphragms integrated the stringers of the second stage to the composite superstructure of the first stage. It was therefore the torsional resistance of these end diaphragms and the rotational stiffness of the first stage composite superstructure that prevented stringers of the second-stage superstructure from freely rotating at abutments while the second-stage deck slab concrete was

90

Integral and Semi-integral Bridges

Figure 6.4 SR 771, Big Branch Creek Bridge, Highland County, Ohio, 1997 (major rehabilitation). This bridge was rehabilitated in two superstructure stages without the use of stage construction joints in the superstructure end diaphragms. This mistake resulted in a 1 in. (25 mm) differential deflection of the second-stage superstructure with respect to the first.

being placed. This end diaphragm restraint prevented stringers of the second stage from achieving the same deflections as first-stage stringers. Had the narrow deck lab closure placement between stages been extended down through the end diaphragms, deflections and rotations of both superstructure stages would have been the same, the main elements of the two stages would have been equally stressed, and the deck slabs of the two stages would have matched and achieved the expected surface elevations. As a result, when constructing integral-type bridges in stages, integrity of joined superstructure and abutment elements must be considered (see Chapter 10) so that appropriate construction procedures can be provided. Tennessee DOT’s continuity connection placement procedure (item 3; see Figure 6.3D, left) eliminates such problems. In placing and completing most of the deck slab before placing continuity connection concrete, superstructure deflections and span rotations at abutments occur before these connections are placed. For short-span stage construction projects, where dead load stringer deflections are relatively small, it would be more cost-effective to interconnect all structural elements of adjacent stages, without the use of deck lab closure placements. Although dead load distribution and deflections of stringers adjacent to stage construction joints would not be exactly as envisioned in design computations, the fatigue and ultimate strength of superstructure members would not be adversely affected. To minimize effects of live load on a completed first stage, as deck slab concrete of a cross-frame-connected second stage is being placed, placement of second-stage concrete could be limited to these weekly periods when the least truck traffic would

Chapter 6 Concrete Deck Cracking and Other Problems

91

be expected. Alternately, truck traffic could be briefly detoured while second-stage deck slab placements were being made and partially cured. Avoiding the use of deck slab closure placements for stage reconstruction of short-span bridges will not only eliminate the need for extra concrete placement days, thus shortening reconstruction periods, but also reduce the number of longitudinal construction joints in each bridge deck (probable cracks) from two to one. Lateral rotation of semi-integral bridges For a new two-span, continuous deck-type, semi-integral bridge in Ohio (SR 180, USR 33 Bridge, Hocking County, Ohio, 1994), movable longitudinal joints were provided between the superstructure end diaphragms and abutments to facilitate differential longitudinal movement of the superstructure with respect to the abutments. Upon completion of the structure, the longitudinal joints between the wingwalls and the superstructure end diaphragms were found to have changed during and shortly after construction. Fillers in acute corner joints were compressed to half their original thickness indicating that the joints were closing. However, fillers in obtuse corner joints were loose indicating that these joints were opening. At another location in Ohio, a three-span, continuous, deck-type bridge was being replaced in stages to carry interstate traffic over a local state highway. Project supervisors noticed that the acute corner of the first stage of the new superstructure was moving perceptively towards the void created by removal of the last stage of the old superstructure. In the state of Washington, a concrete abutment wall adjacent to the acute corner of the superstructure of a long two-span continuous deck-type bridge was found to be fractured (Figures 6.5 and 6.6), and in another state anchor bolts of exterior stringer bearings at a fixed pier were found sheared or deformed. All of these bridges had two things in common: Their designs were based on the semi-integral concept (see Figure 6.1, right), and they all had substantial skews. In

Figure 6.5 SR 250, 148th Avenue N. E. Bridge, King County, Washington State, 1969. This heavily skewed semi-integral bridge, constructed without guide bearings, experienced abutment wall fractures at the acute corners of the structure presumably due to the unresisted horizontal rotation of the superstructure (see Figure 6.6).

92

Integral and Semi-integral Bridges

Figure 6.6 A close-up photograph of the abutment wall fractures noted in Figure 6.5 and discussed in Chapter 8.

Chapter 6 Concrete Deck Cracking and Other Problems

93

addition, most of them had not been provided with guide bearings at abutments to resist the tendency of skewed semi-integral superstructures to incrementally and progressively rotate horizontally towards acute corners. A discussion of this behavior is given in Chapter 8. For those agencies contemplating adoption of the semi-integral bridge concept, one of the primary aspects of these structures that must be considered and effectively resolved is the design of guide bearings for the superstructure of skewed bridges. Unfortunately, many of the retention devices currently being used are not fully functional, because friction and binding of retention devices and, consequently the long-term stability of abutments, especially those not supported by rigid foundations, have not been effectively provided for. However, it appears inevitable that this aspect of the semi-integral bridge concept will be improved when bearing manufacturers and other bridge design engineers unite their talents to design and manufacture more functional structure movement systems for these applications (see Chapter 10). Approach slabs Ohio experienced approach slab distress shortly after it adapted the integral concept to continuous steel beam bridges in the early 1960s. Where these bridges were constructed adjacent to compressible asphalt concrete approach pavements, approach slab seats at the ends of bridge superstructures were found to be fractured, approach slabs had settled, and the vertical discontinuity in the roadway surface at the approach-slab/superstructure interface was hindering movement of vehicular traffic. In these first Ohio DOT adaptations of the integral concept to continuous steel beam bridges, approach slabs were not anchored to superstructures. Instead, the weight of the approach slabs together with friction between slabs and aggregate bases tended to anchor slabs and aggregate bases together. Therefore, as these bridges contracted and expanded in response to daily ambient temperature changes, joints between moving superstructures and stationary approach slabs opened and closed with each daily ambient temperature cycle. As these joints were not sealed, compression-resistant roadway debris infiltrated them while they were open. Subsequently, during superstructure expansion, the force of the expanding superstructures compressed the joint debris and pressed against approach slab ends with sufficient pressure to overcome friction at the approach-slab/aggregate-base interface. Such periodic pressure pushed approach slabs toward compressible asphalt concrete pavement in small incremental movements as joints continued to open, fill with debris, and close with each temperature cycle. In a few years’ time, approach slabs were pushed to near the edges of the slab seats. Eventually, due to diminished bearing areas, slabs and slab seats were fractured by the weight of vehicular traffic. Tying approach slabs to slab seats of moving integral bridges with reinforcing bars has prevented the abutting joints from opening and filling with debris. Ohio DOT places such bars diagonally down through slab seats so that they function not only as longitudinal ties but also as hinges to facilitate settlement of the far end of approach slabs supported by consolidating approach embankments. Such slabs thus provide gradual vertical transitions between consolidating approaches and pilesupported bridge superstructures.

94

Integral and Semi-integral Bridges

Other engineers have occasionally and mistakenly ignored the probability of gradual and long-term consolidation of approach embankments and have used straight extensions of top deck slab reinforcement to tie approach slabs to bridges. Where such straight ties were provided in approach slabs on new embankments, settlement of approach slabs caused cracking of slab concrete and yielding of tie steel. For more effective approach to slab designs, cyclic movement, joint infiltration, and embankment consolidation should all be considered and adequately provided for if such designs are to be more generally successful (see Chapter 10).

Erosion Bridges with closed decks, such as those with raised curbs, sidewalks or parapets (even those bridges where deck scuppers have not been intentionally omitted), function to retain and conduct bridge roadway drainage to approach shoulders. Without the protection of full-width approach slabs with curbs or parapets, accumulated deck drainage will erode shoulder-support materials, embankment surfaces, and backfill at abutments. This situation is made worse during cold weather when contraction of bridge superstructures results in transverse open joints between shoulders and abutment backwalls. Deck drainage penetrating such openings can result in rapid and substantiated erosion of the roadway shoulders. This erosion is such a significant hazard to vehicular movements, especially heavily loaded trucks, that almost continuous maintenance is required to keep unprotected shoulders in serviceable condition. To eliminate this problem of shoulder subsidence, and in some cases erosion of abutment foundation soils, closed deck-type integral and semi-integral bridges should be provided with full-width approach slabs with curbs or raised parapets (Figure 6.7). As drainage retention and control is so important to the long-term durability of shoulder structures, approach slab curbs should also be made at least 6 in. (152 mm) high to compensate for the thickness of probable future overlays. Movable joints Expanded polystyrene is usually specified as joint-forming material for movable bridge seat joints between superstructure end diaphragms and abutments of semiintegral bridges. A number of examples of such designs are illustrated in Chapter 9. In these examples, it will be noticed that elastomeric bearings and polystyrene are both specified as part of the movable bridge seat joints. Unless otherwise specified by plan detail notes, such polystyrene will be left in place because it would not be to the contractor’s advantage to remove it. Unfortunately, polystyrene, if left in place, will compromise the function and performance of the movable joints. Seismic research [2] on a bridge with such movement joints found that, although the elastic modulus of elastomeric bearings was 3.6 times greater than the modulus of polystyrene, the area of polystyrene left in place was 12 times greater than the area of bearings. As a result, total abutment reactions were distributed to both bearings and polystyrene, with polystyrene supporting more than 75 percent of the reactions. It should be stated that this seismic research was performed on a structure

Chapter 6 Concrete Deck Cracking and Other Problems

95

Figure 6.7 TR 45, CSXT Railroad Bridge, Pickaway Township, Pickaway County, Ohio, 1996. This well-designed integral bridge with a scupperless deck was provided with approach slabs with 6 in. (152 mm) high curbs. These curbs should provide adequate long-term protection for the approach roadway shoulders and abutment backfill against the erosive effects of bridge roadway drainage.

that was more than 30 years old. Such performance illustrates that joint-forming polystyrene has considerably more integrity and durability than bridge designers would probably expect. To avoid adverse performance from similar movable joints, it is suggested that plan details or notes require joint-forming materials to be removed after enddiaphragm concrete is placed and cured. One approach to solving this problem is for contractors to use form boards supported on compacted sand as bottom forms for end-diaphragm concrete. After end diaphragms are cast and cured and side forms removed, sand under the bottom form boards can be washed out by use of a low-pressure stream of water. Thereafter, released-form boards can easily be removed. Such removals will provide movable joints that will not only function as contemplated by the design, but also generally yield dry bearings and bridge seats, and thus more durable surfaces. Removal of forming materials will also provide easy access so that bearings, bridge seat surfaces, and joint seals can be examined during periodic bridge inspections. Cycle-control joints Probably the most significant unresolved problem with integral and semi-integral bridges is the availability of cost-effective functional and durable cycle-control joints, the movable transverse joints used between approach slabs of integral-type bridges and approach pavements. For the shortest bridges, the usual pavement

96

Integral and Semi-integral Bridges

movement joints composed of preformed fillers (generally designated by the misnomer expansion joints) are currently being used. For the longest bridges, fingerplate joints with easily maintainable curb inlets and drainage troughs have been successfully employed. However, for intermediate-length bridges, development of suitable cycle-control joints is still in the evolutionary stages. Compression seals, strip seals, and other elastomeric devices have been used with marginal success. After considerable experimentation without success, Ohio DOT decided to use, until a more suitable joint is developed, easily maintainable 4 ft. (1.22 m) pressurerelief joints (a joint filled with asphalt concrete). This decision was made with the recognition that, during cold weather, such joints will crack open and allow some surface water to enter. Sleeper slabs are used not only to support adjacent slabs but also to help minimize the adverse consequences of surface drainage penetrating joints while they are open. This particular approach is very cost effective because such joints not only facilitate cyclic movement of the approach slabs, they also function to protect bridges from longitudinal pressures generated by the pavement growth/pressure phenomenon (see Appendix 1). Lateral subsurface drainage provisions adjacent to relief joint sleeper slabs are important to avoid trapping drainage water and promoting pavement pumping. Burke [3] contains sketches and commentary of early designs used by four DOTs, designs intended to provide for both cyclic movement and pavement pressure protection. Pedestrian safety As the magnitudes of longitudinal movement of typical jointed bridges and their integral or semi-integral bridge counterparts are roughly the same, both must be provided with appropriate movable joints to accommodate these movements. Consequently, adoption of integral or semi-integral bridges does not eliminate movable joints. Instead, it eliminates the need for more complex, critical, and troublesome movable deck bridge joints in favor of simpler, more cost-effective, and durableapproach roadway joints. However, where open joints in sidewalks would be a hazard to pedestrian traffic, deck joints with surface covers (sliding plate or suitable elastomeric devices) should be provided either at the bridge/embankment interface or, where full-width approach slabs are to be provided, at the approach-slab/approach-sidewalk interface. Lack of attention to such details may result in pedestrian hazards and problems for maintenance engineers that are not easy to resolve on a case-by-case basis. Hinged joints Failure or marginal performance of elastomeric devices to seal movable deck joints was one of the primary factors that forced bridge administrators to adopt the use of integral and/or semi-integral bridges. However, depending upon specific design details adopted, the problem of providing effective joint seals has been minimized, but not eliminated. In an effort to reduce bending of integral abutment piling, Ohio DOT devised an integral abutment concept more than 40 years ago that contains a hinge located

Chapter 6 Concrete Deck Cracking and Other Problems

97

between superstructure-encased stringers and abutment pile caps (see Figure 1.5c). This hinge facilitates rotation of superstructures at abutments due to concrete deck slab placement and movement of vehicular traffic across a structure. It also facilitates abutment pile-cap rotation due to thermal expansion and contraction of superstructures. To prevent water in backfill from penetrating these joints and corroding hinge reinforcement, a number of different design provisions were adopted. Full-width approach slabs with raised curbs or barriers were adopted to control and conduct surface waters to approach inlets or embankment-side slope flumes. To control subsurface drainage, longitudinal roadway underdrains were terminated on bridge approaches and drained laterally to embankment side slopes. An 18 in. (0.46 m) thickness of porous backfill was placed against abutments, perforated drain pipes were placed to drain porous backfill to embankment side and front slopes, and elastomeric seals (water stops and/or various types of anchored sheet seals) were provided to seal the back of hinged joints. However, even with all of these provisions and precautions, it is now evident that, in certain projects, some of these water control provisions are ignored or neglected, seal placement is not carefully controlled, and joint seal anchors are inadequate for the purpose intended. Consequently, it is not now unusual to find evidence of water penetrating these hinged joints and corroding hinge reinforcement (Figure 6.8). In retrospect, it now appears that, regardless of the numerous provisions that were

Figure 6.8 USR I-77, SR 18 Bridge, Summit County, Ohio, 1957 (This structure was widened and provided with an integral retrofit in 1972.) Rust stains on the face of this retrofitted 4-year-old hinged-type integral abutment appear to have been caused by the inadequate control of bridge and approach roadway drainage and inadequate joint seals.

98

Integral and Semi-integral Bridges

originally made to ensure the success of Ohio DOT’s hinged abutment design, the long-term performance of those structures relative to integral abutments without hinges must be considered marginal at best. Fortunately, most other transportation agencies that have adopted integral bridges have chosen hingeless design concepts that do not depend upon water control provisions and elastomeric seals for their integrity and durability.

Summary The evolution of integral and semi-integral bridges to replace commonplace jointed bridges is continuing and their numbers are now rapidly increasing. Nevertheless, their evolution has not taken place as smoothly as might be expected because construction of these new bridge types has been accompanied by occasional problems that have had an adverse effect on the function and durability of some of these structures. Transverse and diagonal deck slab cracks, stage construction issues, lateral rotation of superstructures, erosion of embankments, marginal quality of structure movement systems, and other problems have appeared to trouble design, construction, and maintenance engineers. Except for early age deck slab cracking, these problems are generally the result of failure of bridge engineers to anticipate and apply typical design and construction provisions to achieve trouble-free construction and more durable structures. It is hoped that the comments and illustrations contained in this chapter will help to motivate researchers to study and achieve a fuller understanding of the phenomenon of early age cracking of concrete, especially as it pertains to the achievement of relatively crack-free concrete bridge decks. Ultimately, such understanding should help guide material manufacturers and construction engineers in developing material and construction specifications to control concrete placement to the extent that relatively crack-free bridge decks can routinely be achieved. This chapter will also serve its purpose if its messages are used as reminders of the precautions that can be made part of bridge project plans to ensure construction of more problem-free bridges. Finally, it is hoped that this chapter will also motivate designers and manufacturers to develop improved design details, manufactured products, and construction practices for integral-type bridges. All such improvements will thus help transportation engineers achieve the economy and outstanding durability that integral-type bridges make possible.

References 1. Krauss, P. D., Rogalla, E. A., Transverse Cracking in Newly Constructed Bridge Decks, NCHRP Report 380, Transportation Research Board of the National Academies, Washington, D.C., 1996. 2. Eberhard, M. O., et al., Lateral Load Response of a Reinforced Concrete Bridge, Washington State Transportation Center, Seattle, Washington, 1963. 3. Burke, M. P. Jr., “Bridge Approach Pavements, Integral Bridges, and Cycle-Control Joints,” Transportation Research Record No. 1113, Transportation Research Board of the National Academies, Washington, D.C., 1987, pp. 54–65.

Chapter 7

Integral Bridge Design in the Land of No Special Computations

The present age almost certainly tends to carry analysis too far, and engineering schools in most cases have favored this tendency. The ultimate objective for engineering is planning and building. The function of analysis is incidental to this, but it serves as a guide to the ultimate carrying through of the plan. Most important in this picture of design is the sense of scale. Some men never seem to get it – the ability to recognize quickly that certain phenomena, certain stresses are important and others not; the significant ability to put first things first; the ability to weigh the consequences of failure and adapt a factor of safety to the probability and to the consequences.

Hardy Cross

Introduction The above quote was taken from Engineers and Ivory Towers, a book containing some of Hardy Cross’s writings and speeches that were compiled and edited by R. C. Goodpasture [1]. Although Professor Cross developed some of the most elegant and practical methods for the analysis and design of indeterminate structures, he never lost sight of the fact that the results of such analyses must in the end be interpreted by design engineers who are familiar with the limitations and uncertainties associated with the basic assumptions upon which the analyses results were based. As implied by Professor Cross, design engineers are the ones who must make realistic assumptions with respect to the loads and forces to which structures will 99

100

Integral and Semi-integral Bridges

be exposed, make the simplifying assumptions with respect to the response of structures to these loads and forces, and choose the appropriate analysis methods and/or standardized formulas that will be used for determining moments and stresses, and who will then be responsible for evaluating the relative significance of these stresses relative to the assumptions and simplified numerical methods upon which they were based, and on their relative significance with respect to the integrity and durability of the structures. This quote, from the works of such a distinguished engineer as Hardy Cross, is given here as some indication of the professional recognition that has been given and acceptance expressed for the widespread practice of designing countless shortand moderate-span highway bridges without the need to establish precise stress levels for the multitude of events that such bridges are likely to experience. It has also been given here as an indication of the perspective that has been chosen for describing and discussing examples of bridges with designs that are located in the land of no special computations.

Problem solving The primary design activity of structural engineers is to use a scientific education and technical training to solve various types of practical problems associated with the design of private and public structures. In this respect, there are four primary goals that need to be achieve, namely structures that are functionally and economical suitable, reasonably durable, and aesthetically appropriate. With experience, they also come to recognize that different responses to problem solving are generally appropriate, namely: (a) to recognize and learn to ignore non-problems; (b) to eliminate problems where practicable; (c) to redefine problems if appropriate; (d) to simplify problems that cannot be eliminated or redefined; and (e) to solve the remaining problems in the most expeditious way possible, bearing in mind that the four primary goals are more important than a rigorous and exhaustive analysis of the stress levels resulting from combinations of all of the possible load applications and environmental changes. Recognizing and ignoring non-problems As an example of a non-problem, the author was at one time sent to investigate a reinforced concrete retaining wall that according to local maintenance officials appeared to be on the verge of collapse. Concern was expressed to the extent that these officials were considering closing the interstate roadway that was being directly supported by this wall. Upon arriving at the scene of concern, he found a 30 ft. (9.1 m) high retaining wall that paralleled an interstate roadway and butted up against a transverse walltype bridge abutment. The concrete of the wall appeared to be in good condition without cracks or deterioration. However, its top appeared to be tilted laterally about 1½ in. (38 mm) with respect to the adjacent wall-type abutment. Also, the longitudinal joints between the roadway pavement and the paved gutter, and between the gutter and the back surface of the wall, were obviously wider than originally constructed.

Chapter 7 The Land of No Special Computations

101

Examination of the construction plans for the wall revealed, however, that the wall was a cantilever rather than a counterfort type of design, a type of design that would usually have been used for walls of comparable heights. Consequently, the 1½ in. (38 mm) apparent tilt of the wall was about equal to the long-term deflection that would be expected for a cantilever wall of that height. Therefore, this situation appeared to be a non-problem needing nothing more than complimenting the maintenance personnel for being so observant and convincing the local officials that the subject retaining wall was performing as should be expected for a wall of that height, and that there was really no cause for further concern. Actually, as this wall provided direct lateral support for a roadway pavement, a more rigid counterforted type of design should have been used because it would have prevented the gradual spreading of the longitudinal roadway joints. It was the unusual spreading of these roadway joints that first attracted the attention of the maintenance inspectors and provoked their concern about the adequacy of the wall supporting the roadway. Eliminating problems In addition to recognizing and ignoring non-problems, bridge engineers are also occasionally able to recognize problems that can be eliminated rather than solved. For example, a bridge design colleague of the author was assigned the design of a severely skewed highway/railroad grade-separation bridge, a bridge that was to be located above a relatively flat grade-crossing site. As was customary at the time in Ohio, the preliminary design engineers, in attempting to minimize the height of the bridge approach embankments (and the amount of expensive borrow that would be required to build the embankments), established the lowest possible grade line for the bridge and its approaches. As a consequence of this grade-line decision, the depth of longitudinal steel stringers necessary for the long-span superstructure, and the required vertical clearance between the top of rail and bottom of transverse steel box-beam pier caps, it would be necessary to thread the stingers through the pier caps. Such structural and geometric complexity would create significant problems in achieving fatigue-resistant connections between the stringers and pier caps, and would thus result in a difficult and expensive steel fabrication contract and an expensive grade-separation project. Rather than accepting such a design challenge and proceeding to solve some interesting and rather difficult structural connection problems associated with the proposed design, the design engineer contacted the district roadway engineers to describe the complexities involved in achieving a suitable bridge design. It also gave him the opportunity to question the necessity for the established grade line and to verify that the grade line could not be raised to help simplify the structural design. To his surprise, he discovered that for this particular project there was a significantly greater volume of cuts than fills, and that the roadway engineers were pleased to have the opportunity to raise the grade at the bridge site to help them balance their cuts and fills for the project. Consequently, for the cost of a single telephone call, a call that should have been made by the preliminary design engineers, this bridge design engineer was able to solve his most difficult structural problems created by a low grade line by raising

102

Integral and Semi-integral Bridges

the grade line sufficiently to provide adequate room for placing the entire superstructure above the transverse pier caps. In the process, he was able not only to simplify the structural design, but also to produce a significantly more durable structure and a less-expensive project. He solved his primary problem (the low grade line) by eliminating it.

Redefining problems Another example of problem solving, which could be characterized as redefining a problem to simplify its solution, has to do with the control and discharge of bridge deck drainage. It had been the long-term practice of the Ohio Department of Transportation (DOT) to provide small, closely spaced scuppers for bridges with closed (curbed) decks. To clarify scupper requirements, Ohio’s Design Specifications for Highway Structures [2] contained the following provision: 31. The bridge surface shall be effectively drained. This shall be accomplished by one or more of the following means: Transverse crown or superelevation, to cause transverse flow. Closely spaced scuppers along the curbs … . By “close spacing” of scuppers is meant a center-to-center spacing of not more than 20 feet [6.1 m] and preferably not more than 15 feet [4.6 m]. The net area of scupper openings shall be not less than one square inch [25.4 mm] for each 24 square feet [3.3 m] of deck area to be drained.

This practice of using closely spaced scuppers for deck drainage purposes was established in Ohio more than a half-century ago when highway bridges were generally used for stream crossings and when bridge roadway widths were only slightly wider (face-to-face of raised curbs) than the paved portions of approach roadways. However, in the 1960s, when grade separation bridges became more commonplace, bridge deck drainage systems became more complex since these scuppers could not be allowed to drain directly on bridged roadways. Downspouts at piers, and horizontal conductors and drainage troughs, became more commonplace. In addition, these drainage systems were burdened with significant roadway debris, debris that helped to clog the systems. As a result of the amount of sediment and roadway debris that had to pass through these systems and because these systems usually had poor hydraulic characteristics, they became major problems for bridge maintenance personnel who had the responsibility to keep these systems functioning. Obviously, freezing winter temperatures compounded their problems. Eventually, more pragmatic engineers realized that the modern wider bridge decks, especially the shorter bridges (<300 ft. [91 m]), no longer needed closely spaced scuppers to keep the traveled lanes clear of accumulated drainage. Scuppers could be omitted from all but the longest bridges. Accumulated deck drainage could then be conducted to the approaches by means of full-width approach slabs with curbs, and approach shoulders with curbs, either to curb inlets or side slope drainage flumes. This one simple change (providing for bridge deck drainage on bridge approaches) eliminated the need for deck scuppers, downspouts, horizontal conduits and troughs, and the significant maintenance problems associated with them. Looking back, one wonders why so many engineers were unwilling to question

Chapter 7 The Land of No Special Computations

103

authority and long-established practices when the primary justifications for these practices no longer existed. Eventually, bridge deck drainage problems were effectively solved when as suggested above scuppers were omitted from all but the longest bridges. The bridge drainage problem was solved by moving drainage provision off bridges and on to bridge approaches. This solution could be classified as redefining the problem as an approach roadway drainage problem rather than a bridge deck drainage problem. This permitted a cost-effective solution that bridge engineers found themselves unable to provide. Simplifying problems Probably one of the most common approaches to problem solving is the practice of simplifying a problem before its solution. In some cases, the simplification process itself is sufficient to diminish the significance of a problem to such an extent that it becomes a non-problem that can be ignored. The Old North Hill Viaduct As an example, consider the stressing effects associated with differential longitudinal shrinkage and contraction of cast-in-place concrete bridge decks with respect to other relatively rigid supporting elements of a structure. Minimizing these stressing effects by simplifying the structural complexity of a design is a problem that bridge design engineers have been grappling with for over a century. The massive Old North Hill Viaduct of Akron, Ohio (Figure 7.1) is a structure where this problem was compounded by the extreme length of the bridge deck. This bridge, the largest bridge in Ohio at the time of its construction in 1922, served vehicular and pedestrian traffic for over 50 years until it had to be prematurely demolished in 1978. It had 16 major consecutive rib-arch spans and an overall bridge length of about 2810 ft. (857 m). This bridge was designed at a time when the significantly adverse effects of roadway drainage contaminated with de-icing chemicals was unknown to most of the members of the bridge design profession. This particular bridge was chosen to serve not only as an example of a simplification practice, but also as a warning that simplification practices are not all equally effective. Actually some practices, such as the one used to simplify this structure, have had rather significantly adverse long-term durability consequences for countless numbers of deck-type highway bridges. To minimize the adverse effects of the differential shrinkage and contraction of its extremely long reinforced concrete deck with respect to the 16 major rib-arch spans, the structural designers provided it with movable deck joints located throughout the length of the structure. One joint was placed at each abutment and at each pier. Other joints were typically placed at the third points of each rib-arch span. Thus, this structure was provided with a total of over 50 movable deck joints. The joints in the deck over the rib-arch spans not only minimized the stressing effects resulting from differential shrinkage and contraction of the deck relative to the arch ribs, they also functioned to minimize composite interaction between the deck and arch ribs, and helped to assure the designers that the total arch-rib stresses were

104

Integral and Semi-integral Bridges

Figure 7.1 Old North Hill Viaduct, Akron Ohio, 1922–1978: This was the largest bridge in Ohio at the time of its construction.

controlled to such an extent that their calculated stresses were a reasonable representation of the actual stresses imposed on the arch ribs by the structure’s dead load and the superimposed vehicular live load. Unfortunately, what these designers had not anticipated was the deleterious effect that roadway drainage contaminated with de-icing chemicals would have on the integrity and durability of the primary reinforced concrete members of the structure. This contaminated drainage penetrated the bridge deck at every one of its movable deck joints and washed over the members located directly below the joints. This exposure resulted in the gradual disintegration and spalling of the primary members of the structure as illustrated in Figure 7.2a,b. This concrete spalling and disintegration were typical of the deterioration and disintegration that occurred below every one of the structure’s numerous movable deck joints. The maintenance effort required to repair the reinforced concrete members of this structure eventually resulted in maintenance costs of about US$100,000 a year. Finally, the decision was made that the structure had to be pulled down to protect not only vehicular traffic on the structure, but also those individuals who were living and moving about below the structure from the hazards of falling concrete. Consequently, the movable deck joints that the designers of this structure used to simplify the complexity of the structure and keep the secondary stresses at tolerable levels were ultimately responsible for significant maintenance expenses and the

Chapter 7 The Land of No Special Computations

105

(a)

(b)

Figure 7.2 Structure deterioration below the more than 50 movable deck joints of the North Hill Viaduct of Akron Ohio. (a) Typical deterioration of reinforced concrete brackets, floorbeams, and columns located below movable deck joints near the crown of the arch spans. (b) Typical deterioration of reinforced concrete arch ribs located below and adjacent to those deck joints.

106

Integral and Semi-integral Bridges

Figure 7.3 SR 82, Cuyahoga River Valley Bridge, Brecksville, Ohio, 1930: The deck of this bridge was replaced in 1989.

premature loss of the structure. Although this structure was well designed in other respects, the use of movable deck joints to simplify the complexity of the structure and minimize the secondary stresses resulted in a shortened life of less than 60 years. In contrast, consider the simplification practice used by the designers of the Ashtabula Viaduct (see Chapter 5), the Brecksville Bridge (Figure 7.3 and see below), and others to minimize secondary stresses without adversely affecting the integrity and long-term durability of the structures.

The Brecksville Bridge During the early decades of the twentieth century, many bridge design engineers were apparently unaware of the significantly adverse consequences associated with the use of movable deck joints in highway bridges. However, the bridge design engineers of the Ohio Highway Department’s Bureau of Bridges were well aware of these consequences. Writing as early as 1926, the Department’s Engineer of Bridges, J. R. Burkey, stated [3]: There are several details of design which I am sure have been receiving more attention in recent years than they did formerly. Perhaps the foremost of these in concrete work is the expansion joint [movable deck joint]. … A little later a slide will show how we [in Ohio] are proposing to take care of this problem on an important structure by the use of flexible columns which eliminate any sliding action.

Chapter 7 The Land of No Special Computations

107

That statement was written about the Ashtabula Viaduct, which was constructed in 1928. The same design approach was also to be used on the other arch bridges to be designed by the department including the beautiful Cuyahoga River Valley Bridge at Brecksville, Ohio (Figure 7.3), a bridge that was completed in 1930 shortly after the Ashtabula Viaduct. Notice in Figure 7.3 how slender the spandrel columns appear to be. Rather than using movable deck joints within each span, such joints were provided only at piers. Then the bridge deck was intentionally separated from the arch ribs throughout each span by the use of very slender and relatively flexible, reinforced concrete columns. Thus, flexibility of supports (the columns) was employed instead of movable deck joints to provide the bridge deck greater freedom of longitudinal movement. Such freedom of movement not only minimizes differential secondary stresses due to shrinkage and contraction of the deck relative to the arch ribs, but also helped to minimize composite interaction between the deck and arch ribs. That this simplification of the interaction between the deck and arch ribs was successful is attested to by the fact that, after a recent replacement of its deteriorated deck and floorbeams in 1989, the Brecksville Bridge, unlike its demolished predecessor, will continue to serve both vehicular and pedestrian traffic well beyond the 100th anniversary of its construction. This practice of employing flexible supports instead of movable deck joints not only served to help simplify the design of arch-type structures such as the Ashtabula Viaduct, the Brecksville Bridge and others, it is the practice that motivated the origination and evolution of integral-type bridges. Frame-type piers The frame-type pier of Figure 7.4a can be used as another example of how the simplification process can aid in the design of what at first appears to be a very difficult and complex design problem. Designers familiar with the design of piers such as the one illustrated in Figure 7.4a would immediately recognize that shrinkage and contraction of the cap of this pier frame would produce combined moments similar to those illustrated in Figure 7.4b. They would also recognize that, for frames exposed to wide temperature variations, it would be impossible to provide enough reinforcing steel in the upper outside corners of the frame to satisfy either the stress or the strength requirements of the design specifications. After struggling with the analyses of the pier frame illustrated in Figure 7.4a, novice engineers would probably try to simplify the design by the introduction movable joints in the pier cap, somewhat like what is shown in Figure 7.4c. However, they would also immediately recognize that the number of frames to be designed has doubled and that one frame is symmetric while the other is asymmetric. So, although segmenting the pier cap simplified the design with respect to secondary effects, it burdened the design process with two dissimilar structures. With some thought, it would soon be realized that the frames of Figure 7.4c were necessitated by the use of two joints in the pier cap. In turn, it would also soon be realized that the two joints were necessitated by the fact that the proposed superstructure stringer spacing was based on the use of an odd number of stringers, a number that resulted in one of them being located at the center of the deck and at the center of the pier cap. Stringer spacings can be changed, however. Supposing a

108

Integral and Semi-integral Bridges

Figure 7.4 (a) A pier frame located in the Land of Special Computations; (b) diagram of the significant shrinkage and contraction moments for the pier frame illustrated in (a); and (c) a preliminary simplification of the pier frame (a) intended to minimize effects of shrinkage and contraction. This results in two different frames for design. However, other concepts for the simplification of this pier are illustrated in Figure 7.5.

Chapter 7 The Land of No Special Computations

109

wider stringer spacing with an even number of panels were proposed. That spacing would make it possible to provide an open joint at the center of the pier cap as illustrated in Figure 7.5a. This simplification results in two similar but asymmetric frames for design. As a result of the continuous footing, these frames would still be burdened with troublesome secondary stresses, although these effects could be minimized by providing a discontinuous footing as illustrated in Figure 7.5b. At this stage of the conceptual design process, designers would notice that two similar but asymmetric frames were achieved that could be designed to satisfy stress or strength requirements, although they would also be aware that further simplification of the pier frame configuration was practicable. As illustrated in Figure 7.5c, designers have come to realize that, by shifting the central pier columns laterally, the two frames can be designed as symmetric structures. Typically, it would be found that the primary and secondary stresses for the frames of Figure 7.5c would be manageable. However, if the aesthetics of the bridge were not of consequence, further simplification would be possible, as illustrated in Fig. 7.5d. By respacing the columns and providing adequate cantilevers for the pier cap, the vertical loads on the cap would be such that the cap could be designed as a continuous beam. As the cap would be so short, secondary effects could be ignored, axial loaded columns would need to be provided only with nominal reinforcing, footers would probably have to be slightly increased in size, and piling would have to have only a slightly greater capacity. In other words, in the simplification process described above, the frames of this hypothetical pier would have been structurally simplified to such an extent that frame analysis would no longer be needed and the actual design time would have been reduced from days using computer programs, or weeks using manual analysis methods, to less than a day’s effort even without the benefit of an analysis program. A bridge design colleague of the author, Thomas A. Bolte of Burgess and Niple, has described this simplification process as moving the design to, or locating the design in, the “Land of No Special Computations.” It has become known as the land where most pragmatic engineers would try to locate their structures, if such a location would not have an adverse effect on the function, economy, aesthetics, or durability of the structure. Actually, when simplifying a structure to locate it in the Land of No Special Computations, bridge design engineers will usually find that they have achieved structures with more well-defined load paths and more accurate estimations of stresses or strengths, and they will usually also have achieved in the process more economical structures. These benefits would have been achieved because designers used the opportunity during the simplification process to fully evaluate and understand the behavior (responses) of the various structure configurations considered with respect to the loadings and environmental changes to which they could be exposed during their effective service lives. D. P. Billington observed this approach in the design methods used by the famous Swiss engineer, Robert Maillart. In evaluating the design procedures used by Maillart in achieving his remarkable innovative reinforced concrete bridges in the early decades of the twentieth century,

110

Integral and Semi-integral Bridges

a

b

c

d

Figure 7.5 (a) Respaced stringers help to simplify the pier frame design; (b) providing discontinuous footings minimizes effects of shrinkage and contraction; (c) a simplified concept to achieve axially loaded frames although these frames still require rigid frame analyses; and (d) finally, a continuous beam pier cap on axially loaded columns, a concept that truly is located in the Land of No Special Computations.

Chapter 7 The Land of No Special Computations

111

including his beautiful Salginatobel Bridge of Schiers, Switzerland, Billington observed [4]: As engineers in public works, we should be reminding the public of the lessons inherent in Maillart: that all essential features of analysis must be kept always under the watchful eye of the designer and that the bulk of design time should be spent on design thinking and not on analysis refinement. [4]

Land of No Special Computations With respect to the Land of No Special Computations (Figure 7.6), this design region has been recognized by various branches or segments of the bridge design profession, although the region has not heretofore been dubbed with a specific name. This region has been tacitly recognized in the American Association of State Highway and Transportation Officials’ (AASHTO’s) Standard Design Specifications for Highway Structures [5]. It has been recognized in the design regulations, design specifications, and standard drawings of the various transportation departments of

Figure 7.6 A visual representation of the Land of No Special Computations with respect to the Land of Special Computations. The Land of No Special Computations is the land where most pragmatic engineers would prefer their decktype bridge designs to be located.

112

Integral and Semi-integral Bridges

the United States. It has also been recognized in the technical publications of industrial organizations, and by individual bridge design engineers as a goal in the bridge design simplification process, as briefly described above. More specifically, consider the examples given below that describe how the officials of AASHTO have provided bridge design specifications that not only unify design requirements and provisions nationwide; these specifications also help to simplify design procedures and in effect help to locate the design of short- and moderate-span highway bridges in the land of no special computations.

AASHTO bridge design specifications Bridge live loads To simplify the design of highway bridges for live loads, AASHTO has evolved what have come to be known as the HS loadings. These are a numerical system of loads that are intended to simulate the effect of multitude truck sizes, weights, numbers, and roadway positions that are characteristic of the nation’s highway systems. Until just recently, this loading consisted of a single three-axle truck or a continuous uniform load with one or two concentrated loads. In effect, by the use of an HS-simplified system of loads, bridge designers no longer needed to concern themselves with the complex of vehicular loads that would actually use their bridges.

Bridge deck slabs The reinforced concrete deck slabs used on deck-type highway bridges are exposed to a complex of loads, forces, movements, structural differences, material variables, and environmental differences that are easy to recognize but difficult to quantify for each particular application. For example, these consist of, in addition to truck wheel loads (both legal and illegal):

• • • • • • • • • • •

Composite interaction of deck slabs with stringers (intended or unintended) Pavement pressure generation (growth/pressure [G/P] phenomenon) Passive pressure (integral and semi-integral bridges) Negative bending moments due to structure continuity Differential deflection of adjacent stringers Structure skew and/or curvature Differential longitudinal shrinkage of deck slabs with respect to stringers Differential lateral shrinkage of deck slabs with respect to braced stringers Differential longitudinal shrinkage, creep, and elastic shortening of prestressed and post-tensioned stringers with respect to deck slabs Various deck slab placement speeds, sequencing, and set times for continuous structures Various environmental conditions during concrete placement, finishing, and curing

Chapter 7 The Land of No Special Computations

• • • • •

113

Repairs, if any, of transverse deck slab cracks Vehicular collisions (grade separation structures) De-icing chemical corrosion and deterioration Defective materials and inadequate reinforcing steel cover Lateral forces (stream flow, centrifugal, and wind), etc.

Yet bridge design engineers need not contend with such a complex of loads, forces, movements, structural differences, and environmental changes because the AASHTO bridge design specifications provide a single simplified numerical equation (based on the stringer spacing and a truck wheel load alone) that can be used to determine the needed slab depth and primary reinforcing steel. In addition, the necessary amount of secondary reinforcing steel perpendicular to the primary steel is given as a percentage of the primary steel. These requirements are consequently so specific that they have been used to develop deck slab design charts that provide various slab designs based upon a particular HS loading and stringer spacing. Considering the complexity of the deck slab situations and local environment differences, what could be simpler.

Load distribution Longitudinal stringers, reinforced concrete deck slabs, transverse braces, and pier and abutment bearings are the primary members of deck-type highway bridge superstructure structural systems that distribute vehicular live loads both laterally to all stringers and longitudinally to all bearings. Consequently, the geometric configuration of such systems, the width of the systems relative to span lengths, the spacing and depth of stringers, the thickness of deck slabs, and the configuration and spacing of transverse braces all function to help distribute vehicular live loads to all stringers and all bearings. To design the individual stringers of such systems, bridge designers must determine how much of the total vehicular loads on a bridge’s superstructure must be used to design individual stringers. Based on extensive load testing of numerous scale models and of actual bridges, finite-element methods and computer programs have been developed for just such a purpose. To avoid the need to use these programs for the design of short- and moderate-span highway bridges, however, for many decades the AASHTO design specifications provided designers with a simple equation with only two variables (wheel load and stringer spacing) to determine the load distribution factor that should be used to determine the percentage of an HS truck load that should be used for the design of a typical stringer. Recently, for certain applications, that equation was expanded to include other variables (span/width ratio, stringer depth, deck slab thickness, etc.) to refine the design. Nevertheless, even with the present relatively simple equations, bridge designers are still able to ignore the actual complexity of the load distribution throughout complex structural framing systems when determining the loads and forces that may be used to design the structural characteristics of individual stringers at all locations throughout each span.

114

Integral and Semi-integral Bridges

Fatigue design procedures Beginning in 1951, engineers designing Ohio bridges with steel framing systems were required by Ohio DOT to use a new fatigue design specification developed by the Department for the design of such systems. Thus, Ohio was the first state of the Union to routinely require the use of a numerical procedure for the fatigue design of steel bridge members. Following this fatigue design recognition, in 1963, the AASHTO organization published different fatigue design specifications for steel bridge members, which were adopted by most states, including Ohio. Although the provisions of this second specification were an improvement over Ohio’s initial effort, this author and many other designers found them difficult to understand and awkward to implement. Finally, after much fatigue testing of small-scale welded details and full-scale testing of typical welded steel connections, Dr. John Fisher of Lehigh University conceived of, and developed, the present fatigue design specifications that have made the fatigue phenomenon easy to understand and relatively simple to accommodate. These latter specifications and numerical design procedures were based on the establishment of fatigue design categories (A, B, C, D, E, and F) for different types of welded connections. Through the use of these categories, the design of simple and complex steel structures and their connections has now been simplified to the extent that the names of the connection categories imply.

Transportation department methods The transportation departments of the United States have each in their own way adopted standard design details, procedures, specifications, and standard drawings that also help to move the design of short- and moderate-span bridges to the land of no special computations. Briefly, these include but are not limited to the following: 1. Developed computer programs for the analysis and design of various bridge members. 2. Developed standard design drawings for typical bridge members such as deck joints, bearings, roadway parapets, railings, scuppers, approach slabs, etc. 3. Adopted standard details for concrete and steel bridges, including cross frames, diaphragms, bolted connections, etc. 4. Performed crash tests of bridge railing simulations to qualify the design of such members. Such testing obviated the need for numerical methods to achieve the same purpose. 5. Developed standard details for the type of substructures favored by the department. 6. Adopted or established standard configurations for a series of prestressed concrete bridge members. 7. Encouraged the use of movable joints or flexible support members to limit or minimize secondary stresses. 8. Provided rigid foundations (piles, pedestals, rilled shafts, etc.) to minimize settlements and secondary stresses in continuous structures.

Chapter 7 The Land of No Special Computations

115

9. Approved the limiting of bridge characteristics (i.e., length skew, etc.) to allow the neglect of some secondary effects.

Industrial organizations Even the nation’s material industries have contributed to the process of bridge design simplification. They have developed construction manuals, design handbooks, design guides, and sample plans to aid bridge design engineers in the preparation of plans for bridges made of steel, reinforced concrete, prestressed concrete, wood, etc.

Bridge design engineers In addition to the design simplifications described above and implemented by AASHTO, state transportation departments, and the materials industries, bridge design engineers can also simplify structure characteristics, as described above for the Old North Hill Viaduct, the Brecksville Bridge, and various types of pier frames. They have also:

• • • • • • •

used simplifying assumptions with respect to the distribution of superstructure forces to substructures used simplifying assumptions with respect to a structure’s response to superimposed loads and environmental changes (i.e., pinned joints or connections that are absolutely fixed, free to rotate, or partially restrained) used simplified numerical methods for columns or walls subjected to both axial loads and bending (i.e., Odd Albert’s method) neglected secondary effects that have only a negligible effect on either stresses of a structure or its long-term performance used suitable computer programs or standard drawings to obviate the need to use manual numerical methods copied the successful design of colleagues as a basis for new designs used a single design for all substructure elements regardless of small variations in loading to which they may be exposed.

It should be clear from the number of examples described and enumerated above how the various organizations and segments of the bridge engineering profession, from the officials of the AASHTO organization down to individual bridge design engineers, have historically and routinely attempted to locate the design of short- and moderate-span highway bridges in the Land of No Special Computations. In these attempts, there is the recognition that, in responding to public needs of providing and maintaining transportation structures, it is not the responsibility of bridge design engineers to become preoccupied with a continual refinement of numerical methods of analysis and design in an academic quest to determine with absolute certainty every pound and psi to which a structure may be exposed. Rather, it is their responsibility to concentrate their attention and efforts to simplify and expedite the bridge plan preparation process and improve the

116

Integral and Semi-integral Bridges

performance, safety, durability, economy, and appearance of the bridges that they build. An examination of the characteristics of the bridges now being built throughout the United States reveals another indication of how the members of the bridge engineering profession are fulfilling their primary responsibilities. Such an examination reveals that bridge engineers are becoming considerably more pragmatic in that they appear to be willing to relinquish absolute control of secondary effects and associated stresses in order to build more durable and more cost-effective bridges, especially various types of integral bridges.

Integral bridge design As described above, the bridge engineering profession of the United States, from the AASHTO organization at the top to individual bridge design engineers at the bottom, have focused much of their attention and efforts on the simplification of bridge design. Not only have they created simple numerical simulations of vehicular bridge loadings, but they also have simplified the design of complete bridge superstructures including railings, reinforced concrete deck slabs, primary structural framing systems, even bolted, welded, and cast-in-place connections – in other words, the bridge engineering profession has purposely located the design of complete deck-type highway bridge superstructures in the Land of No Special Computations. Considering the structural significance of deck-type superstructures and the fact that the design of such critical bridge elements is located in the Land of No Special Computations, why should the design of the typical short- and moderate-span integral bridges not be located in this same design region? There does not appear to be a single valid reason why this should not be so. Consequently, with this design region as a goal, integral bridges can be suitably crafted to adequately resist superimposed loads and facilitate the bridge’s response to environmentally generated changes. To accomplish such an objective, the limitations and simplifications enumerated and described below could be employed. These would be sufficient to limit secondary effects to negligible levels and thus move the design of the typical short- and moderate-span integral bridges into the same design region as bridge superstructures: 1. 2. 3. 4.

Limit bridge lengths to not more than 300 ft. (91 m) Multiple-span superstructures should be of continuous construction Bridge skew and curvature should be limited to 30 ° and 5 °, respectively Provide embankments and single rows of vertically driven piles to support stub-type abutments (Figure 7.7) 5. Provide flexible piles (steel-H piles for the longest bridges) with lengths of not less than 10 ft. (3 m) and preferably not less than 15 ft. (4.6 m); piles should be oriented for weak axis bending (webs parallel to the abutment) 6. Provide approach slabs tied to the abutments; use diagonal ties where slabs are partially supported on consolidating embankments and surcharged

Chapter 7 The Land of No Special Computations

117

Figure 7.7 A typical integral stub-abutment supported by a single row of vertically driven piles. These piles constitute the primary movement system of integral bridges and composite structures (see Chapter 8). Abutment details are illustrated for steel bridges. Abutments details for concrete bridges would be similar.

7. 8. 9. 10. 11. 12. 13. 14. 15.

16.

subsoils; provide the slabs with 6 in. (150 mm) curbs for bridges with curbs or parapets Provide prebored holes for the upper portion of piles to be driven in dense or cohesive soils; use pea gravel as a void filler Provide well-drained granular backfill behind abutments Provide an abutment width sufficient to accommodate pile misalignment and longitudinal reinforcing steel (Figure 7.7) Provide generous wingwall reinforcement to resist maximum passive pressure Provide adequate continuity reinforcement between abutments and superstructures Provide continuous upper transverse shrinkage reinforcement in the deck slab immediately above abutments Consider the use of embankment benches to reduce the height of abutments and the total passive pressure on abutments Require embankments to be constructed up to the roadway subgrade before driving either abutment or pier piles To minimize abutment settlements, use a waiting period for the consolidation of embankments and surcharged subsoils before pile driving and construction of abutments Provide pressure relief joints between jointed, rigid pavements (concrete, stone block, brick) and bridge approach slabs.

118

Integral and Semi-integral Bridges

Based on the simplifications enumerated above, the design of integral bridges that conform with these simplifications and limitations would be located in the Land of No Special Computations. For such integral bridges, standard superstructure designs could be used without modifications (the effects of the slight rotational restraint provided by the integral abutments could be ignored). Lateral and longitudinal forces on the superstructure could be ignored. Except for lateral or longitudinal loads applied directly to piers and passive pressures applied directly to abutment wingwalls, piers and abutments could be designed for vertical loads alone.

Summary In 1992 the American Iron and Steel Institute studied the lengths of bridges contained in the National Bridge Inventory. They found that 83 percent of the Nation’s bridges had lengths of 200 ft. (61 m) or less. An extrapolation of the bridge length data of that study reveals that fully 90 percent of the Nation’s Bridges have lengths of 300 ft. (91 m) or less (see Figure 7.8). Consequently, the length limitation of 300 ft. (91 m) suggested above for the design of integral bridges in the Land of No Special Computations suggests that the design of most integral bridges could be easily located in that desirable design region. Nevertheless, bridge design engineers should not feel hampered by the “curse of the single solution.” For example, for bridges longer than 300 ft. (91 m), those skewed more that 30 °, or those with roadway curvatures, special designs could be

Figure 7.8 This chart was prepared by the author to illustrate the National Bridge Inventory bridge length data that was compiled by the American Iron and Steel Institute. As illustrated, in 1992, about 83 percent of the nation’s bridges had overall lengths of 200 ft. (61 m) or less. An extrapolation of this data indicates that about 90 percent of the nation’s bridges had overall lengths of 300 ft. (91 m) or less.

Chapter 7 The Land of No Special Computations

119

prepared in the Land of Special Computations. Consequently, as illustrated by the design of the Happy Hollow Creek Bridge of Hickman County, Tennessee (see photograph at start of Chapter 1) and the West Fork of the Black River Bridge of Reynold County, Missouri (see photograph at start of Chapter 2), the attributes of integral bridges could be achieved, even for integral bridges with designs not located in the land of no special computations. Finally, as all design decisions have both positive and negative consequences, designers should be concerned not only that they have considered the immediate and beneficial aspects of their simplification decisions, but also about the possible long-term negative consequences of these some decisions (i.e., such as those responsible for the early demolition of the Old North Hill Viaduct). Only with such pragmatic evaluations can designers be comforted by the fact that they have, to the best of their ability, provided their clients not only with functional, economical and aesthetically appropriate structures for the sites where they are to be located, but also with reasonably durable structures as well.

References 1. Goodpasture, R. C., Engineers and Ivory Towers, McGraw-Hill Book Company, Inc., New York, 1952, pp. 67, 136. 2. Department of Highways, Ohio, Design Specifications for Highway Structures, State of Ohio, Department of Highways, Columbus, Ohio, 1957, pp. 16, 17. 3. Burkey, J. R., “Recent Developments in Highway Bridges,” Highway Topics, Ohio Good Roads Federation, Columbus, Ohio, March 1926, p. 11. 4. Billington, D. P., “Deck-Stiffened Arch Bridges of Robert Maillart,” Journal of the Structural Division, American Society of Civil Engineers, Vol. 99, No. ST7, 1973, p. 1536. 5. American Association of State Highway and Transportation Officials. Standard Specifications for Highway Bridges, 17th edn, AASHTO, Washington, D.C., 2002.

Chapter 8

Semi-integral Bridges: Movements and Forces

When you can measure what you are speaking about, and express it in numbers, you know something about it; otherwise, your knowledge is of a meager and unsatisfactory kind. It may be the beginning of knowledge, but you have scarcely, in your thoughts, advanced to the stage of science.

Lord Kelvin

Introduction There are three primary characteristics of the semi-integral bridge concept. First, the concept features a single- or multiple-span continuous superstructure without movable deck joints. Second, the abutments are supported on rigid foundations (footings on bedrock, vertical and battered end-bearing piling, end-bearing drilled shafts, pedestals that extend down to bedrock, or equivalent friction piles or drilled shafts). Third, the superstructure moves longitudinally independent of the abutments. As a consequence of this third characteristic (the relative freedom of superstructure movement with respect to abutments), the semi-integral bridge superstructure depends upon other means for its longitudinal and lateral stability. This chapter describes these other means, means that historically have not been used 121

122

Integral and Semi-integral Bridges

Figure 8.1 Semi-integral abutment concept for single- and multiple-span bridges on rigid supports (spread footings on bedrock, two or more rows of piles, pedestals to bedrock, etc.). This design concept is not recommended for abutment footings on yielding foundations (spread footings on compressible subsoils, single rows of flexible piles, etc.).

by the bridge engineering profession for this purpose. It also describes how these other means have mistakenly been neglected by some designers who have proposed using the semi-integral bridge concept for skewed applications without providing for the lateral stability of superstructures. Figures 8.1 and 8.2 illustrate the basic features of the semi-integral bridge concept as it was developed by the Ohio Department of Transportation (DOT). For features of this concept as it has been envisioned by other state transportation departments, see Figures 9.6 and 9.7 in Chapter 9. These features include the absence of movable deck joints, a superstructure that moves longitudinally on elastomeric bearings almost independent of rigidly supported abutments, abutment members (including piling) that can be designed to operate well within the usual allowable stress limits, and abutment and superstructure end-diaphragm configurations that are simple to design, simple to reinforce, and relatively simple to construct. Notice, however, that this design concept does not eliminate the need for movable joints; in fact, it doubles their number. In addition to the longitudinally oriented movable joint between the superstructure and abutments, another movable joint is needed between the ends of each approach slab and the approach pavements. Yet, while doubling their

Chapter 8 Semi-integral Bridges: Movements and Forces

123

Figure 8.2 A typical superstructure cross-section of a semi-integral bridge abutment. Longitudinally oriented movable joints are provided between the superstructure end diaphragms and the abutments.

number, this design concept has reduced a bridge’s vulnerability to substantial maintenance. If these joints fail to function as desired, their failure will not have the significantly adverse bridge maintenance consequences (de-icing chemical corrosion and deterioration, and crushing caused by the pavement growth/pressure [G/P] phenomenon) that have come to be associated with bridges with movable deck joints. However, a few words of caution. Although the semi-integral design concept has expanded the application range of bridges without movable deck joints, it possesses some unusual characteristics that must be recognized and provided for. Otherwise, application of this type of design may result in bridges that do not satisfy all of their functional requirements. A discussion of these characteristics is the primary focus of this chapter. Superstructure restraint Of all the characteristics of the semi-integral bridge concept described in this chapter, the longitudinal and lateral restraint of superstructures are the most important. This type of structure should not be considered for design unless the bridge designer is familiar with these characteristics and makes appropriate design provisions to account for them. Longitudinal restraint As shown in Figure 8.1, the semi-integral bridge superstructure is supported on movable elastomeric bearings and therefore moves longitudinally essentially independent of the abutments. That is the reason why this design concept is adaptable to bridges with various types of rigidly supported abutments. For bridges without fixed piers, it receives longitudinal restraint almost exclusively from sources not

124

Integral and Semi-integral Bridges

normally utilized in bridge design for this purpose. Longitudinal restraint comes from the compressive resistance of structure backfill (passive pressure), friction between the approach slabs and subbase, and the shearing resistance of the elastomeric bearings. However, during cold weather, after the superstructure contracts away from the abutments (and away from the backfill), only active earth pressure, the frictional resistance of the approach slabs moving relative to the subbase, and the shearing resistance of the elastomeric bearings will be immediately available to restrain the superstructure against externally applied longitudinal forces. For this reason, it would be desirable if the granular backfill at abutments could be placed and compacted during cold weather, or at night during hot weather, so that the backfill could initially contribute more restraint to supplement that of the approach slabs and bearings for resisting longitudinal forces. Providing turn-back wingwalls cantilevered from the superstructure, in place of straight transverse wingwalls, can provide additional longitudinal restraint by mobilizing the resistance of the backfill/wingwall friction, or, for wingwalls with irregular surfaces, the shearing resistance of backfill. Generally, however, the longitudinal resistance provided by approach slabs and bearings should be sufficient to satisfy specification requirements respecting the resistance to longitudinal forces. For moderate earthquake forces, the resistance provided by consolidated backfill should provide the additional longitudinal restraint needed for moderate length bridges, even during cold weather. For longer bridges, anchorage to piers can provide the extra longitudinal restraint needed, even for large longitudinal forces. Lateral restraint With respect to Figures 8.1 and 8.2, notice that, although the horizontal bridge seat joint is provided with elastomeric bearings and elastomeric joint seals, the vertical joints between the superstructure and lateral wingwalls are usually provided with only joint fillers and elastomeric joint seals. (It is the author’s opinion that all jointforming materials, including fillers, should be removed after fresh concrete has set sufficiently to permit their removal.) To facilitate longitudinal movement of the superstructure, the vertical joints are oriented parallel to the superstructure. If these joints are not thus oriented, binding of the superstructure and abutment wingwalls will occur. Consequently, the abutments of the semi-integral bridge concept not only provide vertical support for the superstructure, but also function essentially as longitudinal guides for the superstructure. As noted above, the superstructure of a semi-integral bridge is laterally supported by the interaction of the superstructure and backfill, by the interaction of the approach slabs and subbase, and to some extent by the shearing resistance of the elastomeric bearings in the bridge seat joint. For applications where substantial lateral resistance is necessary (such as skewed bridges [described later], or structures exposed to stream flow pressure or earthquake forces), however, elastomeric guide bearings are necessary at abutments and their use in the wingwall joints or on the bridge seat between beams is recommended. For superelevated bridges where bridge seat joints (Figure 8.2) are sloped parallel to the deck surface and elastomeric bearings are also sloped, in addition lateral guide bearings are needed to resist the lateral component of the superstructure

Chapter 8 Semi-integral Bridges: Movements and Forces

125

reaction. Otherwise, primary support bearings should be set on level bridge seat surfaces. When considering most of the application situations that have to be contended with in the design of semi-integral bridges with characteristics similar to the concept described in this chapter, the routine use of elastomeric guide bearings should be considered as standard for most if not all such applications. Rotational restraint Superstructures of some skewed semi-integral bridges will, unless restrained by guide bearings, tend to rotate in a horizontal plane. This tendency will be greater for bridges with greater skews. Horizontal rotation will initiate sooner for longer bridges. The characteristics of this behavior are described below. As superstructures of semi-integral bridges expand in response to rising ambient temperatures, superstructure elongation (ΔL) will be resisted by backfill being compressed at abutments (Figure 8.3). Force is required to compress backfill and this same force will restrain superstructure elongation by producing compressive stresses on the superstructure. When considering the relative compressibility of backfill (even thoroughly consolidated granular backfill) and a reinforced concrete superstructure, it should be clear that almost all the expected superstructure elongation will occur as compression of backfill. Only a slight amount of superstructure compression will occur, as evidenced by a slight reduction in the amount of superstructure elongation that would have been evident had the elongation been unresisted. These compressive stresses are shown summarized in Figure 8.4 as the resultant longitudinal compressive force Pp sec θ. The centralized location of this resultant force is based on the assumption that structure backfill is homogeneous, it is placed

Figure 8.3 Assumed passive pressure distribution for an expanding semiintegral bridge.

126

Integral and Semi-integral Bridges

Figure 8.4 Assumed resultant passive pressure forces on a skewed semi-integral bridge superstructure before horizontal rotation of the superstructure.

symmetrically with respect to the centerline of the structure, and it would be uniformly compressed throughout the width of the superstructure. The components of this resultant force against the backfill are the normal force due to passive pressure, Pp, and the lateral force Pp sec θ sin θ or, in simpler terms, Pp tan θ. If guide bearings for the superstructure are not provided and the force Pp tan θ is not adequately resisted at the structure/backfill interface (by friction of backfill on superstructure end diaphragms [Pp tan δ] or by the shearing resistance of backfill [Pp tan Ø]), differential movement at the end diaphragm/backfill interface will commence. When considering the shearing resistance of backfill (Pp tan Ø) or the frictional resistance of backfill on smooth concrete surfaces (Pp tan δ), the latter force will usually be found to be the smaller of the two and will govern behavior at the structure/backfill interface. As the external forces act “on” both ends of the superstructure of a semi-integral bridge (Figure 8.4), the eccentric longitudinal force component Pp will tend to rotate the superstructure towards the acute corner of the structure or, for the skew shown on Figure 8.4, in a clockwise direction. The lateral force components, on the other hand (Pp tan Ø or Pp tan δ), will tend to resist this rotation. Using the shearing resistance of an idealized granular backfill and the frictional resistance of backfill on the end diaphragm/backfill interface surfaces, it can be shown that superstructures of semi-integral bridges skewed greater than 15 ° will be unstable unless they are provided with guide bearings at both abutments. With respect to Figure 8.4 and the symbols tabulated below, the above statement can be justified by a short series of computations as follows: L = bridge deck length θ = bridge skew angle PP = total passive pressure FS = factor of safety

Chapter 8 Semi-integral Bridges: Movements and Forces

127

∅ = angle of internal friction of backfill δ = angle of end diaphragm backfill interface friction For the superstructure of a skewed semi-integral bridge to be stable, the force couple tending to resist rotation (Pp tan δ L cos θ) must be equal to or greater than the force couple tending to cause rotation (Pp L sin θ), or: PP L sin θ ≤ Pp tan δ cos θ

(1)

Providing an FS against rotation: PP L sin θ ≤ Pp tan δL cos θ FS

(2)

As the weight of attached approach slabs and approach-slab/subbase friction will tend to resist movement, a safety factor of 1.5 seems sufficient for this situation. Inserting this factor in Equation (2) and simplifying yields the following: Sinθ ≤ tan δ cos θ 1.5 Tanθ ≤ tan δ 1.5 θ ≤ arc tan ( tan δ 1.5)

(3)

Assuming that the angle of friction of the structure backfill interface (δ) is 22 ° [1, p. 7.2–63] for granular backfill on a smooth concrete surface, Equation (3) suggests that θ must be equal to or less than 15 ° to be stable. For greater skews, it is likely that rotation will be initiated unless guide bearings are provided at both abutments to resist the forces inducing such movement. Other observations can be made with respect to Equations (1) and (3). Although Equation (1) indicates that some level of passive pressure must be generated to cause rotation, Equation (3) indicates that the skew angle at which rotation will be initiated is independent of passive pressure and bridge length but directly related to structure/backfill interface friction. What would be the result of the restrained expansion and subsequent contraction of the superstructure and differential rotational movement at the end diaphragm/ backfill interface? Figure 8.5 and the speculative analysis described below are offered as an answer for this question. As before, it is assumed that guide bearings have not been provided. Consequently, sliding friction (Pp tan δ) will govern the behavior at the end diaphragm/ backfill interface. As the force due to sliding friction would not be sufficient to resist the lateral force component, Pp tan θ, for bridges with large skews, sliding (rotation) of the superstructure towards the acute corners of the structure will be induced. Rotation of the flat-ended superstructure will alter the earth pressure distribution within the backfill. As rotation commences, the obtuse corners of the superstructure will move into and compress the backfill while the acute corners will move away from and allow the backfill to expand. The amount of movement into and away from the backfill may appear rather insignificant when compared with the original movement into the backfill (ΔL) due to the thermal elongation of the superstructure. However, slight movements of soil retaining structures can have significant effects on soil pressures. Earth pressure research (documented in the literature [1, p. 7.2–60, 2]) indicates that a fair amount of structure movement into the

128

Integral and Semi-integral Bridges

Figure 8.5 Assumed resultant passive pressure forces on a skewed semiintegral bridge superstructure after superstructure expansion and horizontal rotation.

backfill (about 5 percent of its height) is needed to achieve full or ultimate passive pressure. On the other hand, only a very small amount of movement away from the backfill (about 0.1 percent of its height) will result in active pressures. Based on these relationships of movement to pressure, it is assumed that backfill compression due to the rotational movement of the obtuse corners into the backfill would only slightly increase passive pressure because this movement is so small relative to the initial superstructure elongation, ΔL. However, at the acute corners, the slight rotational movement of the superstructure away from the backfill will probably result in a drop of soil pressure from the initial passive pressure caused by ΔL to active pressure. The slight rise in soil pressure at the obtuse corners and the significant drop in pressure at the acute corners of the structure will alter the pressure distributions within the backfill throughout the width of the structure. This change in pressure distribution will be accompanied by a lateral shift of the pressure resultant, identified as Pp in Figure 8.5. The designation Pp in this case is somewhat misleading because this resulting pressure is now intended to represent a summation of a whole spectrum of pressures from active pressure near the acute corner through various levels of pressure to a maximum pressure at the obtuse corners. This shift of the resulting pressure will decrease the lateral distance between the resulting pressures at each end of the structure and consequently the moment couple-inducing rotation will diminish. Figure 8.5 illustrates the condition when the force component at an abutment tending to induce sliding (Pp tan θ) has diminished until it equals the force component due to frictional resistance (Pp tan δ). Similarly, the moment couple tending to induce movement (Pp L sin θ) has diminished to (Pp)(L cos θ tan δ) and is in equilibrium with the force couple resisting rotation (Pp tan δ L cos θ).

Chapter 8 Semi-integral Bridges: Movements and Forces

129

It is presumed that the movements described above were the result of a single increase in the ambient temperature. Subsequently, a temperature drop would be accompanied by a shortening of the superstructure or movement of the end of the structure away from the backfill. In response to this movement, the backfill would expand and soil pressure would drop to active pressure or less, depending upon the composition of backfill and state of its consolidation. Depending upon the amount of backfill reconsolidation that would occur while the superstructure was withdrawn, it is presumed that a similar but more modest superstructure rotation would accompany each cycle of superstructure elongation. Over time, the significant number of thermal cycles that would take place suggests that the superstructures of semi-integral bridges would continue to experience incremental and accumulative rotation until or unless such rotation is terminated by the restraint provided by some other stable part of the structure. In addition to the resistance provided by the interaction of the end diaphragms and backfill, the horizontal rotation described above would be prevented or moderated due to approach-slab/subbase friction, and the shearing resistance of elastomeric support bearings. Depending upon the characteristics of the bridge-seat joint seal, even this device may offer some resistance to horizontal rotation of the superstructure. When considering these supplemental resistance elements, it seems apparent that, for some structures, the most susceptible period for rotational movement would occur during construction when the superstructure would be exposed to “at-rest” placement pressures before approach slabs have been placed. There is one bridge with twin semi-integral superstructures that was constructed in Ohio with a skew of 45 ° (see the photograph at the start of this chapter). It was provided with accessible guide bearings in end-diaphragm recesses (Figure 8.6) so that, some time after completion of the structures, the guide bearings could be removed from one of the structures. This would provide the opportunity to observe the behavior of two essentially similar structures exposed to the same environmental effects with one structure with guide bearings and the other without. The construction of these two superstructures and their early performance under similar service conditions could provide some of the experience needed to determine with greater certainty the effects of skew on the movement of semi-integral bridge superstructures. However, to date, this bridge remains as constructed. Possibly, some time in the near future, it may be used to document the behavior of bridges with and without guide bearings. Early earth pressure measurements at the Forks Bridge of Forks, Maine, appear to provide some support for the analysis described above. The Forks Bridge is a skewed long-span steel rigid-frame structure. According to the report: Earth pressures were measured at 8 pressure cells on each abutment with measurements on both sides of the abutment centerline and at different elevations. … The effect of skew was noticeable during the summer, although the average increase for all cells at El. 583 was 1,200 psf [57.46 kPa] the increase at the obtuse sides was 1,800 psf [86.19 kPa] while the increase at the acute sides was 620 psf … [29.69 kPa]. [4]

Pressure measurements at this structure have recently been completed. The final report for this project provides valuable background that should be beneficial for

130

Integral and Semi-integral Bridges

Figure 8.6 Bridge seat-mounted accessible guide bearings located in enddiaphragm recesses between superstructure stringers (see Figure 8.7).

guiding subsequent earth pressure research on semi-integral bridges [5]. Hopefully, such research will be funded and completed within the not-too-distant future.

Guide bearings The magnitude of guide bearing reactions is another indication of the potential for superstructure rotation. As most thermal movement of a superstructure will be parallel to the longitudinal axis of the bridge, the bearing surface of the guide bearings should be placed parallel to that axis. Then, based on the lateral force components shown in Figure 8.4, the guide bearing reaction, which would be normal to this axis, is given by Pp tan θ cos θ or, in a simpler form, Pp sin θ. Based on this relationship and including a safety factor of 1.5, the required capacity of guide bearings for a structure with a 30 ° skew is equal to 0.5(1.5)Pp or 75 percent of the total passive pressure. For a structure with a 45 ° skew, the required capacity equals at least 100 percent of the total passive pressure. Consequently, it is clearly evident that neither the frictional resistance (Pp tan δ) nor the shearing resistance (Pp tan Ø) of the backfill can resist forces of this magnitude, and one or more guide bearings should be provided at each abutment for skewed structures if the long-term stability of the superstructure is to be achieved (Figure 8.6). Some of the details of the guide bearings provided for the structure in the photograph at the start of this chapter are shown in Figure 8.7.

Chapter 8 Semi-integral Bridges: Movements and Forces

131

Figure 8.7 General details of an accessible guide bearing. Two of these bearings were provided in end-diaphragm recesses at each end of each superstructure of the USR 23, SR 32 Bridge, Pike County, Ohio (see Figure 8.6 and photograph at start of chapter).

132

Integral and Semi-integral Bridges

The 148th Avenue N. E. Bridge over SR 250 in King County, Washington State (see Figure 6.5) provides an indication of the forces that can be generated against the wingwalls of a heavily skewed semi-integral bridge constructed without guide bearings. The superstructure of this bridge rotated toward the acute corners of the structure. Without guide bearings, this rotation resulted in bearing of the end diaphragms against the joint filler, compression of the fillers, and fracturing of the abutment walls (see Figure 6.6).

Vertical restraint Buoyancy Due to their jointless construction, many types of integral bridges are buoyant when they become submerged. This may be true for I-beam-type bridges and probably is true for most box-beam-type bridges as well. The weight of concrete diaphragms provides some resistance to uplift. But, generally, some positive design provisions must be made to ensure that semi-integral bridges have a reasonable factor of safety against flotation. I-beam webs can be pierced near the top flanges by 3 in. (75 mm) diameter vent holes spaced uniformly throughout the length of the beams. Similar holes can be placed in all box-beam webs. However, unless adequate top vent holes and bottom drain holes are provided for box-type members to permit air circulation and water discharge, respectively, and unless interior forming materials are removed, I-beam-type members should be given preference for use at sites where inundation of superstructures is possible. Counterweights could be used to provide resistance to buoyant effects but their weight would have to be taken into account during beam design. Uplift restraints could be provided at pier bearings, or some piers could be constructed integral with the superstructure to add sufficient vertical uplift restraint to counteract buoyancy. In place of top vent holes, bottom drain holes, added weight, uplift restraints, and/or integral piers, etc., the use of the most buoyant structures should be restricted to these bridge sites where the highest floodwater levels are well below the superstructure.

Design aspects Movable joints As has been mentioned, the semi-integral bridge concept has enabled the bridge designer to provide bridges without movable deck joints at site locations where abutments must be provided with rigid foundations. Instead of movable deck joints, however, two other joint types must be provided to accommodate longitudinal superstructure movement. Movable longitudinal joints must be provided between the superstructure and abutments, and movable cycle-control joints should be provided at the approach-slab/approach-pavement interfaces. Although doubling their number, the semi-integral bridge concept has minimized their significance. Less than desirable performance for either of these joint types will not have the significantly adverse consequences that have come to be expected with the failure of

Chapter 8 Semi-integral Bridges: Movements and Forces

133

movable deck joints. In addition, rigid approach pavements also should be provided with effective pressure-relief joints to guard these bridges from the pressures generated by the pavement G/P phenomenon (see Appendix 1).

Bridge seat joints With respect to the design shown in Figures 8.1 and 8.2, corrosion-resistant elastomeric bearings are provided for the bridge seat joint so that the superstructure can move longitudinally almost independently of the rigid abutments. Very durable elastomeric joint seals are also provided. These seals must be very durable because they will be buried in backfill and consequently would be inaccessible for repair or replacement. In addition the joint seal’s most important characteristic is the ability to prevent backfill from being forced into the joints by compressed backfill. It would be desirable but not absolutely necessary for seals to be watertight. They also must permit unrestrained differential longitudinal movement between the abutments and superstructure, even for bridges with large skews, and they must retain these characteristics for many years without the need for repair or replacement. Although elastomeric joint seals are an important aspect of the semi-integral bridge concept, the Ohio DOT and most other state transportation departments have not as yet adopted seal designs that appear to fulfill all of the necessary functional and durability characteristics. At the Ohio DOT, a number of trial designs have been developed and used. Initially, standard compression seals were employed. Then it became apparent that a reinforced elastomeric sheet seal was more functionally suitable for both square and skewed applications. The sheet seal now being used is a 3 32 in. (2.4 mm) thick nylon-reinforced neoprene. It is attached to the bridge by various means including elastomeric anchor rods in formed recesses, steel clamp bars with expansion anchors, masonry nails with metal washers, or bonding adhesives. It remains to be seen which of these attachment methods will be perfected for these critical joints.

Cycle-control joints Semi-integral bridges with attached approach slabs lengthen and shorten in response to temperature and moisture changes. Consequently, for such structures, the interface between approach slabs and approach pavements should be provided with cycle-control joints to facilitate such movements, otherwise, longitudinal cycling of both structure and approach slabs could damage both flexible and rigid approach pavements. At present, standard pavement expansion joints with compressible fillers are being provided for short semi-integral bridges. Longer bridges are being provided with pavement pressure-relief joints that are 4 ft. (1.22 m) wide and filled with asphalt concrete. For semi-integral bridges built adjacent to jointed, rigid-approach pavement, it is imperative that they also be protected from pressures generated by the pavement G/P phenomenon (see Appendix 1). Effective pressure-relief joints should be provided for all semi-integral bridges, even the shortest. Consequently, for bridges

134

Integral and Semi-integral Bridges

adjacent to rigid approach pavement, two types of pavement joints are required, one to facilitate the cyclic movement of the bridge and one to protect the structure and cycle-control joints from the adverse effects of pavement growth. Initial designs by four transportation departments are illustrated in Burke [3]. As noted there, all designs being used have their limitations. For longer integral and semi-integral bridges, the Ohio DOT is using 4 ft. (1.22 m) wide pressure relief joints to serve both purposes. As both types of bridges are such relatively new conceptions, much additional development is needed if the approach pavement joints adjacent to such structures are to be provided with all the necessary characteristics that these joints must have to satisfy structure requirements without continuous maintenance. Backfill Backfill for semi-integral bridges should not be considered a nuisance that has to be contended with, as is the case with integral bridges with flexible abutments. Instead, backfill should be recognized as an integral and important part of the semiintegral bridge concept. As in the case of retaining walls supported by spread footings on subsoils, such walls when properly designed will interact somewhat compositely with the supporting subsoil and be adequately supported both vertically and horizontally. Similarly, the superstructure and backfill at abutments of semi-integral bridges form a partially composite interactive structure. In this context, the backfill performs multiple functions. While rigid abutments provide vertical support for the superstructure and lateral support for backfill, the backfill at the superstructure level of semi-integral structures provide longitudinal support for the superstructure and vertical support for approach slabs. The ultimate success or failure of the semi-integral bridge concept will depend to a great extent upon the methods and procedures that are developed by the bridge engineering profession to enhance the interaction between the superstructure and backfill. As backfill is such an integral part of the semi-integral bridge concept, every effort should be made to ensure that it is properly selected, properly constructed, and properly maintained. Backfill should be selected and designed to have characteristics suitable for superstructure/backfill interaction, be of a composition that protects it from erosion, and be protected from above by full-width approach slabs. For bridge superstructures with closed decks (raised curbs, barriers, parapets, etc.), approach slabs should be provided with curbs with a height of at least 6 in. (150 mm) or more to confine roadway drainage and conduct it along the bridge approaches and away from the backfill. Approach roadway curb inlets should be considered and provided if necessary to ensure effective drainage control. An effective subdrainage system should also be provided in the backfill above impervious embankments to ensure against retention of subsurface water. Provisions should be made to intercept drainage water from approach roadway underdrains and discharge such water away from the bridge and its approach slabs. In this respect, roadway underdrains should be terminated beyond the bridge approach slabs and provided with lateral drains to embankment side slopes. Otherwise, underdrain accumulations should be conducted in closed conduits longitudi-

Chapter 8 Semi-integral Bridges: Movements and Forces

135

nally through the structure backfill and, where necessary, through the abutments to embankment front slopes. Finally, bridge maintenance engineers should become familiar with semi-integral bridge characteristics so that they can properly appreciate the importance of backfill/end diaphragm interaction and provide the corrective maintenance that such structures must have if they are to provide the service life that their design anticipated.

Construction aspects Unlike bridges with movable deck joints, the design peculiarities of the semi-integral bridge concept described here have created concrete construction problems that are unique to this type of design. These problems have to do with the forming, placement, and curing of fresh concrete for second or subsequent stages of construction while differential movements between stages are taking place due to the response of these stages to thermal and moisture changes.

End diaphragms Integral end diaphragms shown in Figures 8.1 and 8.2 are part of the superstructures of semi-integral bridges and, consequently, after they have been constructed, they will move both longitudinally and rotationally with the superstructures. However, concrete for diaphragms is cast between forms that are usually fastened to and supported by rigid abutments. In addition, rigid abutment bridge seats would be covered by fillers and bearings that will serve as stationary bottom forms for diaphragm concrete. So, if the ambient temperature at a bridge site changes during and shortly after diaphragm concrete placement, superstructure stringers will be either elongating or shortening in response to these changes resulting in differential movement between the stringer ends and the rigidly formed and supported fresh diaphragm concrete. If this movement is appreciable, and occasionally it can be, damage to the freshly setting diaphragm concrete will result. This problem is more acute for the more thermally responsive steel stringer bridges. It is magnified in longer bridges, and it can be compounded in geographical locations where rapid and significant temperature changes could occur during diaphragm concrete placement and setting. Generally, it is not practicable to restrict concrete placement to those days of the year with the smallest ambient temperature ranges and consequently to those periods with the smallest potential for large temperature changes and superstructure movements. But for more moderate-length semi-integral bridges, it is practicable to limit concrete placement to those days when large and rapid temperature changes are not expected and to periods during the day when superstructure movement is smallest, generally shortly after the ambient temperature approaches, reaches, and departs from the day’s peak temperature. A plan note to provide such control and protection for freshly placed diaphragm concrete could be phrased somewhat as follows:

136

Integral and Semi-integral Bridges

CONCRETE for end diaphragms shall be placed during days when sudden temperature changes are unlikely. Placement shall be completed at least 4 hours before the concreteplacement-day’s peak ambient temperature.

For longer structures where such placement controls may not be sufficient to protect fresh concrete, end diaphragms can be constructed with two separate concrete placements. The first placement, up to but slightly below the superstructure stringers, can be placed without concern for superstructure movement. Then, after an appropriate cure time, the stringers and diaphragm forms can be attached to and be supported by this first placement (see Figure 6.1, Chapter 6). Subsequently, the first placement, the diaphragm side forms, and the stringers will all move in unison so that the remainder of the diaphragm concrete can be placed at any convenient time without regard for ambient temperature changes. Finally, another procedure can be used to help minimize the potential differential movement between the superstructure stringers and freshly placed end-diaphragm concrete. Before diaphragm concrete placement, deck slab concrete can be placed and completed longitudinally up to within a foot or two (0.30–0.61 m) of the diaphragms. Such placement will unify the deck slab and stringers, thereby reducing the thermal responsiveness of the composite superstructure to ambient temperature changes. Such a procedure will also be beneficial because placing deck slab concrete first will help to minimize the amount of superstructure end rotation that would otherwise occur during continuous placement of the superstructure concrete.

Approach slabs Approach-slab construction procedures should also take into account the adverse effects of thermally induced superstructure movements. These movements could take place while approach-slab concrete is being placed on rigid subbase material. Significantly, these movements could also take place while fresh approach-slab concrete is beginning to set. As approach slabs of semi-integral bridges must be attached to potentially moving superstructures, the fresh connection between these two structures could be damaged if the amount of superstructure movement is substantial. Consequently, construction procedures should be used to control the placement of approach-slab concrete. For moderate length semi-integral bridges, a plan note somewhat as follows could be used: APPROACH-SLAB CONCRETE shall be placed towards the superstructure during days when sudden temperature changes are unlikely. Concrete placement shall be completed at least 4 hours before the concrete placement day’s peak ambient temperature.

To protect the connections between approach slabs and a superstructure, attempts should be made to reduce the force necessary to move approach slabs. This can be accomplished by providing a smooth surface (i.e., a polyethylene sheet) to serve as a bottom form for the slabs. For longer semi-integral bridges, it may be necessary to place an approach slab in two separate placements. The first placement can extend from the far end of the slab up to a construction joint located about 3 ft. (0.91 m) from superstructure.

Chapter 8 Semi-integral Bridges: Movements and Forces

137

Then, after this first segment has been placed and cured, it can be connected to the superstructure with several longitudinal tie bars with mechanical connectors. The remaining portion of the slab can then be placed using a note similar to the one given above to protect fresh approach-slab concrete. If there is superstructure movement during final placement of approach-slab concrete, the mechanically connected tie bars should be sufficient to pull the approach slab without stressing the fresh approach-slab concrete. This will relieve the second-stage concrete placement from movement-induced stresses. Backfill As superstructures of semi-integral bridges are restrained in place longitudinally by backfill at both abutments, and to some extent by the shearing resistance of elastomeric bearings, placement of this backfill needs to be controlled to avoid unbalancing backfill pressures, otherwise unbalanced pressures could be exerted on the fixed piers of multiple-span structures or they could permanently shift the superstructure longitudinally with respect to the abutments of single-span structures. Therefore, a backfill placement procedure is necessary to ensure that backfill is placed simultaneously at both abutments. In addition, as noted above, it might be advantageous to place and consolidate backfill during low temperature periods to improve confinement of the superstructure. In this respect, during hot weather, placement of backfill at night also should be considered.

Summary The first semi-integral bridge built in the State of Ohio was SR 555, Muskingum River Bridge, Zanesville, Ohio. It was constructed in 1979. It has characteristics similar to those that have been described in this chapter (see also photograph at start of Chapter 10 and Figure 6.1). It is a 540 ft. (164.6 m) long, unskewed, threespan, steel girder structure. At the rear abutment, turn-back wingwalls were employed to engage or embrace the backfill. Since then a great number of similar shorter semi-integral bridges have been constructed. The concept has been used most often to retrofit existing single- and multiple-span continuous bridges with deck joints at the superstructure/abutment interface. Currently, many other semiintegral bridges are being planned for both new and retrofit applications, some with significant skews. With respect to the semi-integral bridges built by other transportation departments, see Chapter 9. The response of local maintenance engineers to these bridges has been good. It was primarily through their continual urging that many of these bridges were built. Except for Washington State and Pennsylvania, where the semi-integral concept is favored for most of their relatively short bridges, the construction of integral bridges is presently the main emphasis for most transportation departments. However, for those applications where rigid abutments are unavoidable, the semi-integral bridge concept is now being adapted and used with increasing regularity. The actual performance of these initial bridges will determine the continued viability of the semiintegral bridge concept for future applications.

138

Integral and Semi-integral Bridges

References 1. Department of the Navy. “Foundations and Earth Structures,” Design Manual 7.2, Department of the Navy, Naval Facilities Engineering Command, Alexandria, Virginia, 1982. 2. Wu, T. H., Soil Mechanics, T. H. Wu, Worthington, Ohio, 1982, pp. 276–280. 3. Burke, M. P. Jr., “Bridge Approach Pavements, Integral Bridges and Cycle-Control Joints,” Transportation Research Record No. 1113, Transportation Research Board of the National Academies, Washington, D.C., 1987. 4. Elgaay, M., Sandford, T. C., Colby, C. B., Monitoring of the Forks Bridge to June 6, 1990 (plus Report Supplement No. 1), University of Maine, August, 1992. 5. Sandford, T. C., Elgaay, M., Skew Effects on Backfill Pressure at Frame Bridge Abutments, Transportation Research Record No. 1415, Transportation Research Board of the National Academies, Washington, D.C., 1993.

Chapter 9

Emergence of Semi-integral Bridges

Most new things are not good, and die an early death; but those that push themselves forward and by slow degrees force themselves on the attention of mankind are the unconscious productions of human wisdom, and must have honest consideration, and must not be made the subject of unreasoning prejudice.

Thomas B. Reed

Introduction This chapter describes the semi-integral bridge concept as it was envisioned and developed by the author for the Ohio Department of Transportation (DOT) more than 30 years ago. Although the Ohio DOT first originated the integral bridge concept with the construction of the Teens Run Bridge in 1938 (see Chapter 5), the concept could not be used at site locations where established integral bridge limitations (structures not longer than 300 ft. [91 m], skews not greater than 30 °, and flexible abutment piling not shorter than about 10–15 ft. [3–4.6 m]) would not be violated. At such locations, single- and multiple-span, continuous end-jointed bridges continued to be used. Incidentally, the original integral bridge limitations were established by the Ohio DOT primarily to control secondary stresses, passive pressures, superstructure rotation, and pile flexure stresses to the extent that these 139

140

Integral and Semi-integral Bridges

effects could be ignored during ordinary analysis and design. The excellent behavior of hundreds of these integral bridges throughout the past half-century attests to the effectiveness of these original integral bridge applications. However, as end-jointed bridges created such a significant problem for the maintenance department, the semi-integral concept was developed primarily to extend the application range of bridges without movable deck joints to beyond the 300 ft. (91 m) range and/or to site locations where flexible piling could not be used (i.e., where bedrock was at or near the ground surface). At the time when the semiintegral concept was first developed, it too had some recommended limitations (skews not greater than about 10–15 °, and bridge length not longer than the capacity of the pavement cycle-control joint). After a short introductory period, the use of semi-integral bridges was increasing at a relatively fast rate. This bridge type was then not only being used for new bridge applications, but also found especially well suited for the retrofitting of existing single- and multiple-span continuous bridges with movable deck joints at the superstructure/abutment interface. However, the semi-integral bridge concept was being abused by those not familiar with its behavior during thermal expansion cycles, by being used for large skew applications where substantial eccentric forces were being neglected or ignored. To combat such abuse, the author prepared the paper “Semi-integral Bridges: Movements and Forces” (see Chapter 8), a paper that was intended to clarify the peculiar characteristics of this unusual concept with enough specificity so that future applications could be more suitably designed. Such structures that were more carefully designed would yield not only economical and durable semi-integral bridges for a wide application range, but also more functional ones. It was hoped that such improved short-term bridge performance would: (a) help ensure the success of the semi-integral bridge concept to the extent necessary to guarantee its continued acceptance and use; (b) facilitate development of improved and more functional structure movement systems (see Chapter 10); and (c) encourage the gradual proliferation of this structure type to other state transportation departments and other transportation agencies. However, the author was unprepared for the revelations that were to occur subsequent to the semi-integral bridge paper presentation. These revelations are the subject of this chapter.

The Ohio experience In describing the rotational problems peculiar to semi-integral bridges at a Transportation Research Board’s (TRB) Annual Meeting, in Washington, D.C., superstructure/abutment details similar to those used by the Ohio DOT were presented (Figure 9.1). As can be observed by reference to Figure 9.1, in place of the usual structure movement system characteristics of bridges with movable deck joints at abutments (movable deck joints, deck joint seals, and some type of movable bearings), semi-integral bridges have a more complex structure movement system. This movement system is composed of the following:

•

A jointless superstructure that moves longitudinally essentially independent of rigidly supported abutments

Chapter 9 Emergence of Semi-integral Bridges

141

Figure 9.1 A flawed semi-integral bridge concept similar to Ohio’s for singleand multiple-span continuous deck-type bridges on rigid abutment foundations. (Note that a more functional design requires the removal of bridge seat joint-forming fillers after placement of end-diaphragm concrete; see Figure 8.1.)

• • • •

Reinforced concrete end diaphragms that function in unison (or compositely) with backfill to provide both longitudinal and lateral support for superstructures; they also bear vertically on elastomeric bearings and, where necessary, laterally against guide bearings Attached approach slabs that bear vertically on abutment backfill and superstructure end diaphragms Sealed bridge seat joints intended to prevent compressible backfill from penetrating the joints and fouling bridge movements Cycle-control joints between approach slabs and approach pavements; these compressible joints facilitate the cyclic movement of the bridge and its attached approached slabs.

Such structure movement systems not only avoid the use of troublesome deck joints, but also, by virtue of the longitudinal separation of the superstructure and abutments, superstructures can, while interacting compositely with backfill, move longitudinally, essentially independent of the abutments. This basic characteristic of the semi-integral bridge concept made it possible to use semi-integral bridges for most applications where abutments had to be supported by rigid foundations, applications where integral bridges should not be used. After the development of the semi-integral concept, the application range of bridges without movable deck joints was expanded to the extent illustrated in Figure 9.2. With respect to Figure 9.2, notice that integral and semi-integral bridges are now favored by the Ohio DOT

142

Integral and Semi-integral Bridges

Figure 9.2 Bridge Transportation.

application

limitations

of

Ohio

Department

of

for most bridges with lengths of up to 400 ft. (122 m) and with skews of up to 30 °. As mentioned above, the semi-integral concept should not be used for the large skews unless they are provided with some form of lateral support at abutments (see Chapter 8).

Washington State revelation After the semi-integral bridge presentation at the TRB Annual Meeting, Charles Rhoeder of the University of Washington was moved to comment about the movement system illustrated in Figure 9.1. He stated that the expanded polystyrene (and similarly, preformed expansion joint filler [PEJF] identified in Figure 9.1) used to form bridge seat joints, if left in place by the contractor, would compromise the function of the bearings by offering a substantial amount of resistance to joint movement both vertically and horizontally, movements that the elastomeric bearings were intended to facilitate. He said that this observation was made by one of his colleagues at the University of Washington, Marc O. Eberhard, who had just completed a seismic research project [1] for Washington State DOT. As the design of an efficient and fully functional structure movement system for semi-integral bridges is a constant concern of those who are responsible for the design and function of such structures, a copy of the report by Eberhard et al. was obtained and examined. Upon examining the report, it was surprising to discover that the seismic research was performed on a semi-integral, three-span, continuous, prestressed concrete bridge, a bridge that was constructed in 1966, fully a decade before the construction of the Ohio DOT’s first semi-integral bridge (see the photograph at the start of this chapter). Also, on viewing a photograph of the bridge in the report, it appeared evident that it was one of a type and not merely a unique early example of the semi-integral concept. In other words, the report appeared to indicate that the semi-integral concept may have originated and been developed as a standard bridge type by the Washington State DOT in the early 1960s or earlier. Subsequent correspondence and conversations with members of the Washington State DOT bridge staff confirmed that the semi-integral concept was in fact devel-

Chapter 9 Emergence of Semi-integral Bridges

143

oped and adopted by the state for its typical bridge designs. In addition, their success with this concept was attested to by the fact that more than 80 percent of the structures being built by that state were based on the semi-integral concept. Further inquiries determined that the Washington State DOT’s semi-integral bridge concept was originated primarily by Willis B. Horn, former chief bridge design engineer for that state. Excerpts from Horn’s response to questions about the origination of this concept are as follows: I was motivated by a desire to eliminate expansion joint maintenance and abutment costs. In addition, I wanted to eliminate the unsightly staining and the occasional damage to concrete that expansion joint leakage can cause at piers. I also believe[d] that aesthetic quality [was] improved by eliminating joints at intermediate piers.

With respect to a question about the use of some type of lateral support at abutments for skewed structures, Horn responded: I am not aware that we put out any designs that did not have transverse girder stops. I believe that we had transverse stops [at abutments] going back to the “Standard Prestressed Girder Plans” that were developed in the 1950’s. Regardless, they are needed to resist transverse loads.

With respect to soil/structure interaction and the composition of abutment backfill, Horn responded: Extensive research into soil-structure interaction occurring with bridge length changes was not accomplished. It is my recollection that we limited bridge lengths to about 500 ft. (152 m) with an abutment at each end. This type of abutment has withstood “actual full-scale testing,” which speaks well of its performance for a period of about 30 years.

As implied by Horn’s recollections, the semi-integral concept was introduced by the Washington State DOT as a part of the change to the use of prestressed concrete girders made continuous at piers for live loads. It was also interesting to observe that part of the motivation for change was not only to reduce construction and maintenance costs but also to improve bridge aesthetics. This dual concern was rather unusual for the austere construction budget days of the early 1960s and shows reawakening to the importance of aesthetics in Washington State earlier than it occurred in most other states.

California connection A 1989 National Cooperative Highway Research Program Report, NCHRP Synthesis of Highway Practice Report 141: Bridge Deck Joints, contains one chapter devoted exclusively to integral bridges [2]. To develop sufficient background for that synthesis, an inquiry was sent to all state transportation departments requesting specific information about state practices and experience with bridges with deck joints and integral bridges. In response to that survey, many states that were building integral bridges provided typical details for their designs [2]. Only the California DOT

144

Integral and Semi-integral Bridges

responded with details of a semi-integral concept [2, p. 23]. Thus, it seemed that at the time the synthesis was prepared (1986 through 1987), only the states of Ohio and California had early designs based on the semi-integral bridge concept. Consequently, with the emergence of yet another design by Washington State, it began to appear unbeknownst, to the transportation profession at large, that other states may also have independently originated their own version of this unusual structure concept.

Semi-integral bridge experience In the conception of any new or unusual design, be it a bridge bearing, deck joint, material application, or structure type, there are always a number of design issues that must be resolved by bridge engineers. In general, there are usually a number of different suitable solutions for each particular issue. In some instances issues are interrelated to such an extent that effective solutions to some issues have a detrimental effect on others. The semi-integral bridge is a structure type in which a number of interrelated design issues must be considered and suitably resolved. For example, of particular concern is the development of an effective structure movement system (see Chapter 10) for design of semi-integral structures. Such a system must facilitate its own construction, provide sufficient movement capacity, and be protected by joint seals so that backfill is prevented from infiltrating joints and fouling bridge movements. In addition, joint seals must be suitable for skewed structure applications and be of extremely durable construction. This is so because they will be buried within the backfill and be inaccessible for inspection, maintenance, and/or replacement. Other design issues of particular concern are the development of suitable guide bearings or other devices to prevent the horizontal rotation of superstructures of skewed bridges [3], adoption of standard design details for effective approach slabs and suitable approach-slab joints, adoption of suitable abutment foundation types that will minimize settlement and provide effective vertical and horizontal support for bearings and joint seals, selection of abutment backfill composition suitable for composite interaction with and providing longitudinal support for superstructures, and vertical support for approach slabs, etc. With the emergence of the Washington State DOT experience, coupled with the knowledge that the states of both Ohio and California had also developed their own versions of the semi-integral concept, as mentioned above it was thought that other state transportation departments may also have originated their own versions of this concept and gained some experience with the design, construction, and maintenance of this type of structure. If so, a summary of such experience would form an valuable resource of time-tested structural details and materials for bridge engineers charged with the responsibility of developing suitable design concepts for their own department’s use. With these thoughts in mind, it was decided to make a countrywide inquiry to determine the number of transportation departments that had built semi-integral bridges, and to determine the age and prevalence of such structures.

Chapter 9 Emergence of Semi-integral Bridges

145

Semi-integral bridge inquiry Following the Washington State revelation, a postcard inquiry was made by the author of all state and province transportation departments. This initial survey was then followed by letters and telephone calls to bridge staff engineers of those departments that were found to have been building some version of the semi-integralbridge concept for more than 20 years. The inquiry was narrowed in this way because these departments would probably be the ones with the most valuable experience to share with their colleagues who would be considering the use of semiintegral bridges for the first time. Summaries of the original inquiry responses are provided in Figures 9.3–9.5. Figure 9.3 indicates that nine state transportation departments have been constructing “many” semi-integral bridges for 20 years or more. These states are California, Minnesota, Nebraska, Nevada, Oregon, Pennsylvania, Tennessee, Vermont, and Washington State. Surprisingly, two states (California and Nebraska) have been constructing semi-integral bridges for four decades.

Figure 9.3 As of 1997, the number of state transportation departments that have built semi-integral bridges for 20 years or more (based on responses from 50 states).

Figure 9.4 Trends in the number of state transportation departments building semi-integral bridges (based on responses from 50 states).

146

Integral and Semi-integral Bridges

Figure 9.5 As of 1997, the contiguous states of the United States that have built semiintegral bridges for more than 20 years.

The “many” of Figure 9.3 takes on significance when the experience of two states is considered. For the states of Nebraska and Washington, “many” represents about 80–90 percent of the bridges built by these states. These statistics from Nebraska and Washington are a clear indication of the success that these states have had with the effectiveness and durability of their particular version of the semi-integral bridge concept. Other states, such as California and Ohio (Figure 9.2), favor a mix of structure types including integral, semi-integral, and single- and multiplespan continuous bridges with movable deck joints at abutments. They can then choose the structure type that best (cost, function, durability, etc.) suits particular applications. However, complete satisfaction with the semi-integral concept was not unanimous. When commenting upon his state’s experience with semi-integral types of construction, Edward P. Wasserman, Engineering Director, Structures Division, Tennessee DOT, stated, “We used to do it that way. Now we think that we do it better!” Doing it better refers to Tennessee DOT’s current preference for integral bridges with fewer of the geometric limitations imposed by other states for integral bridge construction. This inquiry uncovered another surprising statistic. As illustrated in Figure 9.4, as of 1997, the number of departments that built many, few, and some semi-integral bridges totaled 26. Thus more than half of all US State transportation departments expressed sufficient interest in the semi-integral concept to have provoked them to construct a number of such structures. Also, most departments that constructed a few or some of these structures constructed them within the 20-year period between 1977 and 1997. One of these later departments was the Ohio DOT the experience of which initiated this inquiry. Semi-integral bridge construction in Ohio did not begin until 1975. Since then, the Ohio DOT has built many of these structures, with the numbers increasing with the passing years.

Chapter 9 Emergence of Semi-integral Bridges

147

The upper summary line of Figure 9.4 illustrates the trend in the number of state transportation departments building semi-integral bridges. The slope of this line suggests the increasing interest being focused on such bridges by other transportation authorities as well. As illustrated in Figure 9.5, the conception of semi-integral bridges appear to have originated spontaneously in two separate regions of the country (the contiguous states in the western seismic zone, and states scattered throughout the northern regions of the country where bridge roadways are regularly subjected to winter de-icing chemical applications). On the basis of the details submitted in response to the semi-integral inquiry, it appears certain that cross-fertilization of semiintegral bridge details occurred in the western region because all states in that region (California, Washington State, Oregon, and Nevada) have evolved similar basic details. Seismic activity in that region appears to be partly responsibly for this similarity. The suitability of semi-integral bridges for that region is attested to by the results of the Northridge Earthquake in California, during which many of these structures provided adequate performance, whereas portions of bridges with movable deck joints collapsed [4].

Representative semi-integral bridge details Figures 9.6 and 9.7 illustrate representative details for semi-integral bridges from some of the states with more than 20 years of experience with such structures. When examining these examples, it should be immediately apparent that a considerable amount of originality, intuition, and creativity has been used by state bridge engineers to meet the challenge of constructing bridges without troublesome movable deck joints. When one considers the length of time that has elapsed since the original versions of these bridge details were conceived, it is apparent that early semi-integral bridge designers achieved a level of success that is truly enviable. Most of the illustrated designs, such as those of California, Minnesota, Nebraska, Nevada, Oregon, Pennsylvania, Tennessee, Vermont, and Washington State, are designs intended for typical bridges. Some of these same states (Oregon, Tennessee, and Washington State) use modified versions of these details for long-span bridges or for bridges with unusually long lengths. Some states with the most experience with integral bridges, such as California, Nevada, and Washington State, have published design directives to govern their application. These directives (see “Policy statements and reports” below) contain useful comments about lateral force control, approach slab joints and anchorage preferences, wing-wall types and orientations, bearing types and placement, etc. In place of design directives, the state of Pennsylvania has produced a standard drawing that provides complete details for typical short-span semi-integral bridges. Other states have achieved semi-integral bridge details but have not perfected structure movement system details (joints, bearings, joint seals, backfill, approach slabs, cycle-control joints) to provide and ensure fully functional and durable structures. Some states do not have standard details for providing lateral force control for skewed applications. Others use structural expanded polystyrene or preformed expansion joint filler together with elastomeric bearings, similar to the early

148

Integral and Semi-integral Bridges

(a)

(c)

(b)

(d)

Figure 9.6 Representative semi-integral bridge details used by some states that have construction such structures for more than 20 years: (a) California, (b) Nebraska, (c) Pennsylvania, (d) Nevada. See also Figure 9.7.

Washington design that was examined and unfavorably critiqued by Eberhard et al. [1], and similar to the example in Figure 9.1 that provoked Rhoeder’s comment at the TRB semi-integral bridge presentation, the comment that motivated the semiintegral-bridge inquiry. Still others specify movement system devices incapable of preventing compressed backfill from penetrating bridge seat joints, impairing

Chapter 9 Emergence of Semi-integral Bridges

149

(b)

(a)

(c)

(d)

Figure 9.7 Representative semi-integral bridge details used by some states: (a) Oregon, (b) Tennessee, (c, d) Washington State. See also Figure 9.6.

bearing function, and fouling bridge movements, while others specify materials by generic or trade names only (without performance requirements or prequalification tests). It is the author’s opinion that the semi-integral design as it has evolved in Washington State (Figure 9.7d) appears to be the most effective of all the designs illustrated in this chapter.

150

Integral and Semi-integral Bridges

Policy statements and reports Design, construction, and maintenance When it was learned that many West Coast state transportation departments had long-term familiarity with the design, construction, and maintenance of semiintegral bridges, their published policy statements relative to these aspects of these structures took on particular significance. Therefore, some of these statements were obtained and have been excerpted and quoted below.

California The lateral earth pressure on a skewed abutment produces a resultant horizontal force that is eccentric to the center of resistance of the substructure element. This horizontal force must be resisted at the abutment or taken into account in the design of the bents. Furthermore, with respect to bridge length effects, California limits its abutment movements to 0.5 in. (13 mm) or 1 in. (25 mm) when unattached or attached approach slabs, respectively, are proposed.

Nevada Integral and Dozer [semi-integral] abutments may be used on bridges with flares, skewed supports, or curvature only where the effect on these items is limited. Determination of acceptable limits is based more on experience than theory. Flares and curvature are limited since the soil pressures must be reasonably balanced such that the unbalanced force can be readily accommodated. Structure skew also results in unbalanced soil pressures since the line of actions of the soil pressure on the two abutments do not coincide.

Furthermore, Nevada limits the skew of its two semi-integral bridge types to 35 ° and 45 ° and to bridge lengths not exceeding 250 and 400 ft. (76 and 122 m) for steel and concrete structures, respectively.

Washington For skewed structures with earth pressure against the end diaphragm … the need for girder stop bearings shall be investigated. When required, these bearings are placed vertically against the girder stop to transfer the skew component of the earth pressure to the abutment without restricting the movement of the superstructure in the direction parallel to the centerline.

Furthermore, the state’s design manual contains charts for determining the lateral forces that must be used in the design of the girder stop bearings of skewed bridges. The manual also states that “some form of girder stop is required for all abutments.” With respect to bridge length effects, Washington State limits the length of its various semi-integral bridge types on the basis of the composition of its primary members, as follows: steel, 300 ft. (91 m); cast-in-place concrete, 400 ft. (122 m); post-tensioned concrete, 350 ft. (107 m); prestressed concrete, 450 ft. (137 m).

Chapter 9 Emergence of Semi-integral Bridges

151

Design of approach slabs, approach slab anchors, and cycle-control joints The design of approach slabs, approach slab anchors, and cycle-control joints appears to have caused the most problems and elicited the most criticism of semiintegral bridges. Consider the following policy statements. California The major benefit of the diaphragm abutment [integral or semi-integral] is its initial lower first cost to construct; however, maintenance problems generally preclude its use except for short structures. … The two main disadvantages of the diaphragm abutment are (a) continuous maintenance of the approach embankments due to the settlement of the pavement notch and along the wingwalls, and (b) damage to the approach embankment and pavement caused by water intrusion between the abutment diaphragm and approach roadway. With water intrusion being the main problem for diaphragm abutments, the current structure approach [slab] has been designed to prevent water from entering behind the abutment and along the wingwalls. The structure approach [slab] is connected directly to the abutment and extends over the wingwalls. An underlying drainage system provides insurance against erosion.

Minnesota In a letter response to questions about integral bridge performance, Minnesota replied that: … most reported problems appear related to leakage from the expansion joint which we locate between the abutment and approach panel. Review of our plan file indicates the semi-integral fixed and expansion designs have been used in Minnesota at least as early as 1961. However, because some bridges with earlier versions of the semi-integral abutment continue to develop maintenance problems, their reputation has suffered and use of the semi-integral abutment has declined over the past 10 years, until today; they are used only on local roads. Recently, based on success in other states, we have begun to design a few integral abutments with the expansion joint located at the end of the approach panel and away from the abutment. This abutment design appears to eliminate many problems that had been associated with our previous semi-integral design.

Oregon In telephone conversations, Oregon reports trouble with its longest bridges (deep structures), with which they have experienced problems with the backfill that is continually being compressed and released by the thermal movement of superstructures. Apparently, backfill adjacent to abutments settles considerably. For bridges without approach slabs, maintenance forces must continually repair the junction between abutments and approach pavements. General observation Leakage of bridge deck drainage between turn-back wingwalls and approach slabs has caused staff engineers to recommend the use of curbs or barriers on

152

Integral and Semi-integral Bridges

approach slabs and the placement of wingwalls outside or under the edge of approach slabs. A number of western states provide movable joints between abutments and approach slabs, joints with special anchorage devices that limit maximum joint movements (Figure 9.7c,d). The main design consideration for such joints is that they must be provided with seals that will be exposed to traffic and continual maintenance. Also, approach slabs not rigidly anchored to superstructures will not provide immediate resistance (weight and friction) to longitudinal and lateral forces. Without rigidly attached approach slabs, semi-integral bridges without fixed piers are almost wholly dependent on composite backfill/end diaphragm interaction for their longitudinal support.

Attributes Semi-integral bridges as described and illustrated herein have a number of primary attributes that appear to be responsible for their adoption, development, and increasing popularity. These include the following:

• • • • •

Deck joints: The absence of movable deck joints has all but eliminated belowdeck deterioration and corrosion caused primarily by roadway de-icing chemicals. It is the one attribute probably most responsible for the development and use of semi-integral bridges in place of their more vulnerable bridge counterparts with movable deck joints at the superstructure/abutment interface. Stress levels: Unlike flexible abutment piling of long integral bridges, all parts of semi-integral bridges can be configured and designed well within conventional stress levels. Longitudinal forces: Unlike bridges with movable deck joints at abutments, semi-integral bridges are highly resistant to the huge longitudinal forces characteristic of jointed-rigid pavements [4, pp. 51–55] and earthquakes. Span ratios: Due to the presence and weight of end diaphragms, approach slabs, and cantilevered wingwalls, multiple-span continuous semi-integral bridges generally can be constructed with span ratios as low as 0.5 without special details. Jointed bridge retrofit: As the semi-integral-bridge concept is suitable for bridges with rigid abutment foundations, the concept can be used to eliminate movable deck joints from most existing moderate-length bridges.

Limitations Prior experiences and this study have revealed that semi-integral bridges have a number of limitations that should be recognized and accounted for during structure-type studies and detailed design. These include the following:

•

Settlement: As settlement correction of semi-integral bridges cannot be accomplished without extensive and time-consuming reconstruction, such structures

Chapter 9 Emergence of Semi-integral Bridges

• • • • • •

•

153

should be used only where settlements are minimized to tolerable levels by the use of secure foundations (piles, pedestals, drilled shafts, etc.). Otherwise, a more settlement-tolerant structure with movable deck joints should be employed. Buoyancy: Superstructures of semi-integral bridges may be buoyant. Such buoyancy should be avoided, minimized, or prevented [3, p. 12]. Bridge length: As embankment and pavement distress are reported for semiintegral bridges near and beyond the 400 ft. (122 m) range, the lengths of such bridges should be limited to less than this range unless special provisions have been developed and incorporated in the design that will accommodate longer lengths. Continuity: To balance the longitudinal forces that may be induced due to the restrained growth of jointed rigid approach pavements, and those forces due to temperature and moisture-related elongation of superstructures, multiple-span semi-integral bridges should be of continuous construction. Alignment: To improve resistance to potentially huge longitudinal compressive forces (pavement growth, earthquakes, etc.), main superstructure members should be straight. Lateral forces: For superstructures that may be subjected to large lateral forces (collisions, earthquakes, stream debris, bridge skews, etc.), guide bearings should be provided at abutments to help resist such forces. Approach slabs: The most dissatisfaction with the semi-integral bridge concept appears to be related to those bridges without approach slabs or with approach slabs not effectively attached to superstructures. Consequently, approach slabs should be provided and they should be effectively attached to the superstructures. For bridges where roadway drainage is confined by curbs or barriers of some sort, approach slabs should be provided with 6 in. (150 mm) high curbs or barriers to conduct deck drainage to approach inlets or to side slope gutters. Cycle-control joints: The success of the semi-integral bridge concept probably will depend to a great extent on the development of effective and durable cyclecontrol joints [5, pp. 61–64].

Summary The semi-integral bridge inquiry made by the author has revealed that these bridges are not as unique as was originally believed. Apparently, many bridge engineers working independently have conceived of bridges with superstructures that move longitudinally, essentially independent of rigid abutments, but that are restrained by composite interaction between backfill and superstructure end diaphragms. Although before 1997, transportation departments of only nine states used semiintegral bridges for more than 20 years. As of 1997, at least 26 states had built one or more of these structures. The relative success of these structures is attested to by the fact that fully 80–90 percent of the small- and medium-size bridges being built by two states were based on the semi-integral concept.

154

Integral and Semi-integral Bridges

Where poor performance was reported, generally it was related to the effects of uncontrolled roadway drainage on bridge approaches. Consequently, approach slabs should be provided and they should be attached to superstructure end diaphragms. When bridge or approach roadway drainage is confined by curbs, barriers, piled snow, shoulder rutting, excessive plant growth on shoulders, etc., bridge approach slabs should be provided with reasonably high curbs or barriers to protect abutment backfill and as an aid in the conduction of surface drainage to approach inlets or open side-slope flumes. With respect to the quality of the various semi-integral abutment details that have been developed thus far, it appears that the superstructure-abutment configuration evolved by Washington State (see Figure 9.7c,d) are the most functional and durable. In addition, superstructure-mounted turn-back wingwalls (a Washington State DOT practice) avoids the use of movable joints between superstructures and wingwalls. The problem with abutment-mounted wingwalls is that the movable joints between superstructures and abutments must be effectively sealed (an extremely difficult problem) not only to achieve fully functional sealed joints but durable ones as well. As described in Chapter 10 (Figure 10.7) and Chapter 11 (bridge seat and bearings of the I-90 B. N. Railroad Bridge), many of the superstructure-abutment designs illustrated in Figures 9.1, 9.6b, and 9.7b are flawed. They should be modified to eliminate the probably inappropriate interaction of joint fillers and elastomeric bearings. Possibly, in the near future, it may be possible for some elastomeric specialist to compound an economical sponge-type elastomer that could function not only as a joint filler and sealer but also as a movable elastomeric bearing. Such a triple duty device would considerably simplify the design and construction of semiintegral bridges. The superstructures of skewed bridges need to be laterally restrained at abutments by guide bearings. Effective cycle-control joints need to be developed and perfected to facilitate differential longitudinal movement of superstructures and attached approach slabs with respect to approach pavements. Although semi-integral bridges have been constructed by some states for more than four decades, the critical composite interaction of backfill with superstructure end diaphragms is still based on supposition. Consequently, it is hoped that the experience summarized above will help to motivate soil/structure interaction research. Such research is necessary if the semi-integral concept is to have the necessary experimental background to ensure the continued success of these remarkable structures.

Acknowledgments The number of responses received from state and province transportation departments to the initial inquiry that helped to form the background for this chapter was extraordinary. Every one of the 50 states and most of the Canadian provinces responded. As a result of such outstanding cooperation and assistance, the author wants to take this opportunity to express his sincere appreciation to all transportation departments, and most especially to those staff engineers who graciously

Chapter 9 Emergence of Semi-integral Bridges

155

responded to subsequent questions and further requests for copies of published policy statements and sample bridge plans. Their cooperation and assistance are gratefully acknowledged. The text of the paper that formed the basis of this chapter was co-authored with Charles Gloyd, formally of the Washington State DOT. The author is grateful to Charles Gloyd for his permission to use parts of that paper for this chapter.

References 1. Eberhard, M. O., et al., Lateral Load Response of a Reinforced Concrete Bridge, Washington State Transportation Center, Seattle, Washington, 1993. 2. Burke, M. P., Jr., National Cooperative Highway Research Program Report 141: Bridge Deck Joints, Transportation Research Board of the National Academies, Washington, D.C., 1989. 3. Burke, M. P., Jr., “Semi-integral Bridges: Moments and Forces,” Transportation Research Record No. 1460, Transportation Research Board of the National Academies, Washington, D.C., 1994. 4. Concrete Reinforcing Steel Institute, “Performance of Reinforced Concrete Bridges in the Northridge Earthquake,” Special Report, Concrete Reinforcing Steel Institute, Schaumburg, Illinois, 1994, pp. 6, 11. 5. Burke, M. P., Jr., “Bridge Approach Pavements, Integral Bridges, and Cycle-Control Joints,” Transportation Research Record No. 1113, Transportation Research Board of the National Academies, Washington, D.C., 1987.

Chapter 10

Elementalistic and Holistic Views for the Evaluation and Design of Structure Movement Systems

In dealing with ourselves and the world around us, we must take into account the structural fact that everything in this world is strictly interrelated with everything else.

Alfred Korzybski

Introduction This chapter had its beginning during a recent Annual Meeting of the Transportation Research Board in Washington, D.C. Before that meeting, the General Structures Committee had a Bridge Bearings Subcommittee and a Bridge Joints Subcommittee. At the time it was argued that the mere existence of these two separate and independent subcommittees gave professional legitimacy to the relatively common and mistaken practice of elementalistically treating movable deck joints and movable bearings as discrete and structurally independent components of a bridge. Instead, such joints and bearings should more properly be treated holistically as coordinated components of an integrated and fully functional structure movement system. As this argument suggested, these two separate subcommittees should be merged into a single subcommittee that would be concerned with the design, construction, and maintenance of structure movement systems (and subsystems) 157

158

Integral and Semi-integral Bridges

Figure 10.1 Composite structure (consisting of a bridge and its supports, backfill at abutments, approach embankments, approach pavement, in situ subsoils, etc.).

of composite structures (Figure 10.1 and see Appendix 2). These systems are composed of responding bridges, moving (rocking, rolling, sliding, compressing, etc.) bridge bearings, moving deck joints, flexing supports (i.e., piers, columns, piles, etc.), compressing backfill, moving approach slabs, growing approach pavements, consolidating embankments, consolidating surcharged subsoils, etc. In effect, all components of composite structures form a structure movement system gestalt, an integrated (intentional or unintentional) structure movement system. As the scope of the new subcommittee implies, the General Structures Committee’s new subcommittee would henceforth be known as the Structure Movement Systems Subcommittee. Designing movable deck joints and bearings as independent parts of a bridge can be considered comparable to designing pistons and camshafts as independent parts of a motor. Mechanical engineers would be quick to recognize the ludicrous nature of such a practice. Not so with many bridge engineers. Design examples abound in which bridge designers appear not to have concerned themselves with the alignment and dimensional coordination of two or more components of a structure movement system, or with how the deformations of one component of a movement system would affect the deformations and performance of other components. In fact, it is not now at all unusual for components of some movement systems (e.g., movable deck joints and movable bridge bearings) to be designed by different commercial firms, different transportation departments, or different individuals. Admittedly, motor pistons and camshafts move at significantly higher speeds than bridge joints and bearings. Nevertheless, when the alignment and fit of two or more components of a movement system are uncoordinated, or when deformations under loading of these components are different, one can expect quick wear and early replacement of motor parts in the first case, and similarly, early binding and distress of joints, bearings, connections, and supports in the second. A couple of actual examples of defective structure movement systems should make this clear. Surprisingly, these

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

159

examples were commonplace before the advent of elastomeric joint seals for movable bridge deck joints.

Movable deck joints, movable bearings, and grade One of the most blatant examples of a defective structure movement system was brought to the author’s attention by a bridge design colleague, Charles Dorian. It is illustrated in Figure 10.2. Surprisingly, the plan detail shown in Figure 10.2 was not the mistake of a single individual designer. Instead, the detail was exhibited for many years on a state standard design drawing, a drawing that was supposed to be used as a guide for the preparation of plans for countless bridges. It was intended to show how movable deck joint components and abutment backwalls were to be positioned and oriented for bridges that were to be constructed on steep roadway grades. Although the bridge seat surface was made a part of this design detail, the appearance of its horizontal surface apparently had little effect on the thought processes of the countless designers who made use of this detail in the preparation of individual bridge plans. Apparently, their attention must have been elementalistically focused on the fit of the joint components and their orientation with respect to the sloping roadway surface, and not holistically on the function of the structure movement system (the movable joints and bearings). If the structure movement system shown in Figure 10.3 (i.e., the bridge joints and bearings) were oriented and positioned as shown in Figure 10.2 to accommodate a bridge that was to be constructed on a steep roadway grade, expansion of the bridge superstructure would tend to force bridge deck joint components at the high end of the bridge to move parallel to the sloping roadway surface. As the bridge seat surface for the movable bearings is horizontal, such forced movement of the superstructure parallel to the sloping surfaces of the roadway would tend to raise the end of the superstructure up and off the bearings (At the low end of the bridge,

Figure 10.2 One state’s elementalistic view illustrating the required orientation of abutment backwall faces and bridge joint components for bridges on sloping grades.

160

Integral and Semi-integral Bridges

Figure 10.3 A structure movement system composed of sliding plate, movable deck joint, and rocker bearings.

expansion of the superstructure would cause the superstructure to move horizontally parallel to the bridge seat surface, resulting in a separation of the sliding surfaces of the deck joint.) Depending upon the rigidity of the fixed pier (the pier with the fixed bearings) to resist longitudinal forces, the sliding fit and the flexural capacity of the upper mating joint components at the high end of the bridge, the grade of the bridge roadway, and the length of the bridge superstructure, superstructure reactions at the high end of the bridge would in some cases be supported by abutment backwalls and not by the movable bearings. One could argue that the above example of a defective structure movement system is a hypothetical example that has little relevance for actual bridge designs because the sloping backwall surface illustrated in Figure 10.2 is probably distorted for clarity, the fit of mating joint components would actually be loose, and the flexural capacity of the bridging component illustrated in Figure 10.3 would be negligible relative to the magnitude of the usual superstructure reaction. However, such was not the case. In fact, the practice of the state bridge staff responsible for the Figure 10.3 detail was to have the two separate mating components of the joint bolted tightly together during fabrication. Also, they were to remain bolted together during construction until after the backwall concrete was placed and before the ambient temperatures started lowering and the superstructure began withdrawing. This tight fitting was done to make sure that the mating components of joints were as water tight as practicable. In addition, bridging components of the joints were usually specified to be up to 1 in. (25 mm) thick for primarily roads with their greater vehicular loadings and also to make them more resistant to the impact of snow plow blades. Finally, this detail was used for structures that were to be constructed on grades, some of which were as high as 11 percent. Consequently, for some structures provided with structure movement systems similar to the one illustrated in Figures 10.2 and 10.3, binding of mating joint components during superstructure expansion

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

161

would result in structure distress and substantial rehabilitation costs. Usually, for bridges with rigid piers, binding of joints would result in the raising of superstructures and transfer of superstructure reactions to the top of abutment backwalls. Such transfers would result in either fracturing of backwalls immediately below joint bridging components, or progressive and incremental closure of joint openings between superstructures and backwalls. This latter effect would be caused by binding of mating joint components, subsequent contraction of superstructures along with the frictionally attached abutments, reconsolidation of abutment backfill, subsequent re-expansion of superstructures and further binding, etc. Eventually such cyclic movement of superstructures and abutments would result in the complete closure of the movable deck joints and the subsequent fracture of abutment backwalls near bridge seats or between the wingwalls. Fortunately, the development and use of elastomeric joint seals have generally eliminated rigid bridging components and consequently have eliminated this binding of joint components and subsequent bridge distress.

Movable deck joints and hangers An incompatibility of movable deck joints and movable bearings similar to that described above is found when hangers are used as bearings to support steel stringers adjacent to intermediate movable deck joints. This appears to be another case of designers elementalistically focusing their attention on the design of deck joints, or on the design of bridge bearings, and not holistically on the design of a structure movement system composed of both joints and bearings. Figure 10.4 illustrates the details that were used almost universally by countless bridge design staffs to accommodate the movements of especially long, steel stringer

Figure 10.4 A structure movement system composed of a sliding plate movable deck joint, and hanger bearings.

162

Integral and Semi-integral Bridges

bridges. As indicated, differential longitudinal thermal movement, in either direction, of the lower pins of suspended stringers with respect to the upper pins of supporting stringers would be accompanied by slight vertical movement of the lower pins, Δv, as the lower pins moved circumferentially with respect to the upper pins. When such hangers were used in conjunction with sliding plate or fingerplate joint components, oriented as illustrated in Figure 10.4, longitudinal thermal movements of suspended spans would tend to lift the suspended spans and pry up the cantilevered bridging plates. Flexural resistance to vertical movement of these plates would induce compression on sliding surfaces, tension in plate connections, bending of plates, tension in hangers, and shear in pins. Joint compression and sliding surface friction would usually be sufficient to bind the joints and thereby resist longitudinal movement of suspended spans. In effect such joints would resist the structure movements that they were intended to facilitate. Joint binding would result in the generation of longitudinal forces against fixed bearings and supporting piers, unless the fit of mating plates was loose enough or the plates flexible enough to partially compensate for the vertical movement of suspended spans at the joints. As a result of a structure’s actual dimensions, this problem would be compounded by shorter spans with their shallower stringers and shorter hangers, and by longer structures that experience greater longitudinal thermal movements. Similar binding of movable deck joint plates, as described above, could occur if the plates were installed and rigidly attached to stringer flanges before the placement of concrete deck slabs. Depending upon the orientation, fit, and flexibility of mating components of a joint, the end rotations of the suspended spans about the lower pins, θ, due to the placement of the concrete deck-slab dead load, would usually be sufficient to induce stresses similar to those described above for longitudinal thermal movements. Both situations tend to lift one side of the joints with respect to the other. Where the orientation of joint components would not permit such movements, distress would be induced and damage to components would likely result. To minimize such problems, some designers have specified that joint components should not to be placed until after deck slabs have been placed. The author had the opportunity of examining a bridge with a structural movement system similar to that illustrated in Figure 10.4. Placement of deck-slab concrete on the suspended span resulted in fracture of welds between the components of one of its joints. Upon examination of the structure, it was determined that the two major components of one joint had to be removed and reversed to eliminate the binding that occurred during placement of the deck-slab concrete. As the components of the joint at the other end of the suspended span had the same joint orientation with respect to the roadway grade as the failed joint, it was not affected by placement of deck concrete and did not need to be modified. Without holistic views of structure movement systems composed of movable deck joints and hangers, it would not be possible to predict the deformations and distress likely to develop from such an inappropriate combination of structural components as illustrated in Figure 10.4. On the other hand, holistic views of structure movement systems during design will reveal those characteristics of the system that need to be adjusted so that the system can be configured to facilitate all anticipated structure movements.

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

163

It appears evident that many designers of structure details such as those illustrated in Figure 10.4 have elementalistically focused their attention on the design of the hangers and separately on the design of the components of the joints. Actually, it appears to be the practice of many bridge design organizations to assign the design of girders and its hangers to one set of designers (usually the more experienced design engineers) and the design of joints to another set of designers (usually novice or less experienced engineers or, in some instances, technicians). Apparently, the design review engineers, who should know better, also appeared to have elementalistically focused their attention in the same way. Consequently, without using holistic views of the design of the structure movement systems, designers and reviewers have produced and will continue to produce defective movement systems, defective structural designs, and eventually large structure-repair costs.

Elementalism and holism Before proceeding further, a clarification of terms seems to be warranted, especially the terms “elementalism” and “holism.” Elementalism is the name given to the concept that reality is composed of discrete and independent objects that have been noticed and named. Holism, on the other hand, is the name given to the concept that reality is composed of a synergistically functioning whole. In bridge engineering, elementalism is manifest in the practice of mentally and verbally segmenting a structure for design purposes (superstructure, piers, abutments, joints, bearings, etc.) so that each component receives the concentrated attention during design necessary to ensure its adequacy with respect to agency standards and bridge design codes. Holism is manifest in the practice of visualizing an assembly of components and their unified response to applied loads and environmental changes to ensure that the composite structure functions in an efficient and effective manner. Elementalistic views help to make the components. Holistic views help to ensure their fit and function. An example of an holistic view in design taken to its zenith has been described by the electrical genius, Nikola Tesla: By that faculty of visualizing, I have evolved what is, I believe, a new method of materializing inventive ideas and conceptions. … Before I put a sketch on paper, the whole idea is worked out mentally. In my mind I change the construction, make improvements, and even operate the device. … I can give the measurements on all of parts to the workmen, and when completed all these parts will fit, just as certainly as though I had made the actual drawings. It is immaterial to me whether I run my machine in my mind or test it in my shop. [1, p. 257]

In a fashion similar to the method described by Tesla, effective bridge design consists of the conscious interplay between elementalistic and holistic views within the design process. Elementalistic views are mental images or visualizations of a part or component of a composite structure (see Figure 10.1), or line simulations of a component of the composite structure responding to a portion of the total static and dynamic loads. These views may or may not include the effects of environmental changes (temperature, moisture, etc.) and material changes (elastic

164

Integral and Semi-integral Bridges

deformations, creep strain, soil consolidations, etc.). They may be given substance by means of stationary displays on computer screens or by manual sketches. Holistic views, on the other hand, are comprehensive mental images or visualizations of a composite structure, selected components of a composite structure, or line simulations of a composite structure responding to static and dynamic loads, and environmental and material changes. These views may be given substance by means of moving displays on computer screens or by a series of manual sketches. Although primarily elementalistic views are generally adequate for the design of some simple nonredundant structures, holistic views gain in importance as the complexity and redundancy of structures increase. Without holistic views to guide and verify designs, a structure may later be found to be inadequate. For moderatesize jointed bridges, such inadequacies are usually related to a faulty structure movement system. An ineffective interplay between elementalistic and holistic aspects of attention has been noticed in many professions. For example, a holistic aspect is either deficient or absent in medicine when disease is considered apart from the patient, in the behavioral sciences when the mind is considered apart from the body, in business when the organization is considered apart from its members, in education when the teaching method is considered apart from the subject matter being taught, in music when the score is considered apart from the perceptions of the audience, and in engineering when structure components are considered apart from the unified structure movement system. Manifestations of an essentially elementalistic approach to design in bridge engineering are prevalent. Consider the following few examples:

•

•

•

National bridge design specifications: These national specifications provide designers of deck-type highway bridges with live-load distribution factors for proportioning vehicular live loads for stringer design. The convenience of this elementalistic design procedure encourages designers to visualize the loading and deformation of a member of a complex structural system (i.e., individual stringers) functioning independently of the system. As a result of habitually using such elementalistic views, some designers are later surprised by the fatigue cracking of secondary braces and stringer connections, and by the peculiar response of mechanical bearings at skewed supports (see below). Manufacturers’ catalogues: These catalogues and some state transportation standard drawings for bridge bearings and deck joints rarely, if ever, holistically qualify the use of these devices, one with respect to the other. What these documents fail to explicitly state is that “joints” and “bearings” are just names for two components of structure movement mechanisms or movement systems. As such, these catalogues do not specify that the structural, mechanical, and material characteristics of the one (the bearings) should be holistically coordinated with the structural, mechanical, and material characteristics of the other (the movable deck joints) to ensure that the composite structure will be provided with a synergistic and functionally efficient structure movement system. Transportation departments: Some State transportation departments have elementalistically devised standard details and construction procedures for

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

•

•

165

eliminating deck joints of simply supported, multiple-span bridges. They do this by retrofitting the discontinuous superstructures at piers with continuously reinforced concrete slabs, surface sealants, and bituminous overlays (see Figure 1.7b in Chapter 1). Yet missing from these details and procedures is any holistic consideration for how this change in superstructure continuity would change the rotational movements and forces on existing side-by-side bolster bearings, the longitudinal forces on remote fixed bearings and their supporting piers, or the superstructure-lengthening effect of such conversions on the performance of the movement systems at abutments. Bridge design supervisors: Some design supervisors elementalistically assign the various components of composite structures to different engineers and technicians for design and detailing without requiring them to share their design decisions with each other to ensure that they achieve holistically compatible components with synergistic movements. As a result, many structures are built with inappropriate structure movement systems (typically defective movable deck joints and/or bearings) that require either early modifications or, in some cases, early replacement. Author’s example: The author remembers his experience with the design of a relatively small composite structure consisting of a deck truss with short suspended end spans and small stub-type abutments supported near the top of high approach embankments. After the first year of service, he was informed that the end floor beams of the truss were seriously deformed. Upon examination of the structure, it was discovered that the approach embankment supporting the abutments had consolidated over 1 ft. (0.30 m). This magnitude of embankment consolidation and abutment settlement, the short suspended end spans, and inappropriate stringer supports on the truss end floor beams resulted in the prying up of the truss deck and the excessive deformation of the floor beams. Later he was shocked to discover that the foundation engineers were aware of the possibility of such large embankment consolidations but routinely did not include embankment settlement estimates in their foundation reports.

Unfortunately, the author and the foundation engineers were using elementalistic views of their individual responsibilities and not holistic views of the composite structure. As a result of his elementalistic view of bridge design, he provided a structure movement system that was extremely sensitive to abutment settlement. With a more holistic view of the composite structure, he would have anticipated the possibility of embankment consolidation and abutment settlement, and consequently would have configured the truss and end-span support connections to provide a composite structure with a more appropriate structure movement system.

Skewed superstructures Many structures designed with the aid of elementalistic views provide suitable structures with reasonable service lives, especially short-span highway structures and those with modest skews. However, for some structures, especially those with long spans and large skews, the practice of habitually using only elementalistic

166

Integral and Semi-integral Bridges

views has resulted in defective structures requiring structural modifications and occasionally large modification costs. As mentioned earlier, elementalistic views appear to have dominated the thought processes of some designers to such an extent that holistic views of the deformations of integrated superstructure systems (longitudinal stringers, transverse distribution members, lateral bracing, deck slabs, curbs or parapets, etc.) responding to vertical loads have been ignored or neglected. These holistic views of an integrated superstructure system are especially important when one is considering the behavior of framing connections and stringer bearings for structural systems with large skews. For example, consider the behavior of the USR I-76, East Market Street Bridge of Akron, Ohio. This is a simply supported 140 ft. (42.67 m) long steel structure. It is supported by riveted girders, mechanical bearings, and heavily skewed wall-type abutments (Figure 10.5). After about 20 years of service, maintenance engineers became concerned about the stability of the structure when they observed that the movable deck joints at abutments appeared to be moving laterally instead of longitudinally in response to the passage of heavy vehicular traffic. An examination of the structure revealed that, due to the movement of vehicular traffic, the ends of the superstructure were, as expected, rotating about axes that were located at the level of the bearings. However, unlike the axes of square bridges, the rotational axis of the superstructure at each abutment of this structure was essentially parallel to the abutment and not normal to each and every girder web, as an elementalistic view of girder deflections and rotations would suggest. Consequently, as the joints moved circumferentially about the superstructure rotational axes, they were in this

Figure 10.5 USR I-76, East Market Street Bridge, Akron, Ohio, 1958. This long, single-span, deck-type structure is supported by riveted steel girders and heavily skewed wall-type abutments.

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

167

case moving essentially normal to the abutments and not parallel to the span of the superstructure. It was this abnormal movement or rotation of the ends of the skewed superstructure that gave the impression that the superstructure was moving laterally in response to the movement of traffic. It was also apparent that the large circumferential surface movements of the joints were due to the unusual flexibility of this long-span steel structure, the concomitant large superstructure rotations at the bearings, and the magnification of the rotationally related surface movements by the large depth of the superstructure. With respect to the behavior of movement systems at skewed supports, holistic views of an integrated and skewed superstructure system responding to the movement of vehicular traffic will reveal that these systems typically rotate about axes that are essentially parallel to superstructure supports (piers and abutments) and not about axes that are normal to each and every stringer web. Consequently, for structures with the usual types of mechanical bearings (rockers and bolsters with their bearing axes placed normal to the supported stringer webs), superstructure rotation at supports will result in bearing edge loading, local stringer distortions, and excessively high bridge seat pressures. For bearings that have been set on sheet lead pads, such periodic bearing edge loading, which accompanies each and every superstructure rotation, will eventually compress, flatten, and extrude sheet lead somewhat like well-rolled pastry dough (Figure 10.6).

Figure 10.6 A typical bolster bearing of the bridge shown in Figure 10.5. Notice that the upper rotational axis of the bearing has been placed perpendicular to the centerline of the girder, a manner of placement that was typical for the era and the type of bearing being furnished. Also notice that the 20-year-old sheet lead-bearing pad has been and continues to be gradually extruded by the high pressures on the front edge of the bearing due to the abnormal rotations of the superstructure.

168

Integral and Semi-integral Bridges

Designers who visualize the behavior of stringers elementalistically as independent components of a structure deflecting vertically and rotating about axes that are normal to stringer webs will be dumbfounded by inoperative bearings and extruded sheet lead. However, engineers who routinely contemplate holistic views of superstructure behavior will recognize in advance the adverse effect of skew on superstructure rotations and will avoid the use of typical mechanical bearings (or steel hangers and pins) in favor of elastomeric or compound bearings that can accommodate the abnormal rotations of superstructures at skewed supports [2].

Composite structures and structure movement systems In the above discussions the terms “composite structures” and “structure movement systems” have been used by the author in his attempts to describe the structural conceptions that are routinely used by more experienced engineers in their structural design considerations. Although names have been given to these conceptions, so far he has neglected to describe them more explicitly and in more detail. For those who may be interested in a more comprehensive description of these conceptions, the following two sections of this chapter have been prepared for that purpose. For novice engineers, it is recommended that the following descriptions be reviewed before considering the remainder of the chapter. Composite structures Composite structures are generally thought of as structures composed of two or more different types of structural components or materials integrated into a single functional whole. The most familiar example of a composite structure is that obtained by mechanically joining a cast-in-place concrete deck slab to prefabricated or precast stringers with shear connectors or shear reinforcement. Although not generally designated or thought of as such, there are numerous other examples of composite structures in bridge engineering. Consider, for example, common reinforced concrete-retaining walls supported by spread footings on in situ subsoils. The footings of such structures are integrated with suitable foundation subsoils, usually with shear keys, to engage deeper soil masses. This composite interaction of footing and subsoils enables them (the composite structure) to resist greater lateral earth-pressure forces. In critical cases, further integration or composite behavior of retaining walls and subsoils can be accomplished by providing dead-men or soil anchors. Such earth-retaining systems can be considered composite structures. At the turn of the century, most small- and medium-size highway bridges consisted of superstructures with movable deck joints supported by substructures with spread footings on in situ subsoils. For these structures, purposeful composite construction was generally limited to retaining walls and wall-type abutments. Abutment construction was designed and configured not only to provide vertical support for superstructures but also to isolate superstructures from backfill earth pressures and roadway pavement growth. These abutments, which were integrated with

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

169

subsoil masses (by the use of footing keys or tiebacks), can be considered composite structures. The successful adoption of continuous superstructures in the early 1930s depended upon a more conservative design of substructure foundations to minimize differential settlement and settlement-induced, secondary superstructure stresses. Where spread footings on in situ subsoils were considered inadequate for providing suitable vertical support for the continuous superstructures, foundations were integrated with deeper and more extensive soil masses by the use of friction piles. By such integration, substructure loading was spread over a greater subsurface area, thereby diminishing soil pressures, differential substructure settlements, and concomitant superstructure secondary stresses. Such footing foundations with their integrated subsoil masses can be considered composite structures. The 1950s saw the introduction and extensive use of rigid frame bridges. For this composite-type structure, the bridge frame, backfill, and embankment act in unison to provide vertical support for the bridge frame and vehicular traffic. Although initiated in the early 1930s by the Ohio Department of Transportation (DOT) for the construction of continuously reinforced concrete slab bridges, and used extensively for these and similar bridges ever since, it was not until the mid-1960s that integral bridges (bridges without movable deck joints) received nationwide attention and widespread applications. These bridges are single- or multiple-span, continuous, deck-type, jointless bridges with stub-type abutments supported on earthen embankments and single rows of flexible piles. Such bridges are supported not only vertically by the interaction of the piling with deep soil masses, but also both longitudinally and laterally by composite interaction of the abutments and abutment piling with embankments and abutment backfill. As their name and construction implies, these can also be considered composite structures. The 1960s saw the introduction of semi-integral bridges by the Washington State and California DOTs (see Chapter 9). These are single- or multiple-span, continuous, deck-type bridges with jointless superstructures supported by rigid, nonintegral abutments and movable bearings in movable longitudinal joints at the superstructure/abutment interface. However, at abutments, the superstructures of these bridges are also supported longitudinally by composite interaction of the superstructure and backfill and by approach-slab/subbase friction. These too can be considered composite structures. The last three decades saw the introduction of two other examples of composite structures that are explicitly recognized as such. One is the long-span arch culvert in which composite interaction of thin, structural, steel-arch plates with selected backfill are used to support deep overburden and vehicular traffic. The other is the retaining wall in which composite interaction of precast concrete wall panels and mechanically reinforced or stabilized embankment are used to retain backfill and support superimposed vehicular traffic. Structure movement systems Because a system is defined as a complex unity formed of many diverse segments or components subject to a common plan and serving a common purpose, and because composite structures consist of an assembly of components (bridges, in situ

170

Integral and Semi-integral Bridges

subsoils, embankments, backfill, approach slabs, etc.) that must be configured and designed to appropriately move (compress, shear, roll, slide, consolidate, etc.) in response to applied loads and material and environmental changes, a composite structure can be conceived of as the primary structure movement system consisting of a coordinated assembly of structure movement systems or subsystems. With respect to the name “structure movement system,” the term “structure” was chosen to induce designers to consider an aggregate or composite structure (consisting of a bridge, backfill, approach pavement, approach embankments, in situ subsoils, etc.) to be a complex unity composed of both natural and constructed segments or components that, when completed and unified, would function as a composite structural system in response to applied loads and to material and environmental changes (see Figure 10.1). Second, the term “movement” was chosen to induce designers to recognize and consider: the characteristics of structure movements, such as a structure’s dynamic response to vehicular loading, wind pressure, earthquakes, etc.; its gradual responses as a result of dead-load application, stream flow pressures, thermal and moisture changes, etc.; and its long-term responses and changes due to creep of stressed concrete members, consolidation of embankments and in situ subsoils, scour of embankments and streambeds, growth of jointed concrete pavement, corrosion immobilization of metal bearings, age hardening of elastomeric bearings, etc. Third, the holistic term “system” was chosen instead of such elementalistic terms as movable deck joints, bearings, flexible piers, compressible backfill, consolidating embankments and subsoils, flexible piling, etc., to induce designers to consider a complex unity formed of many diverse movable and/or deformable segments or components subject to a common plan and serving a common purpose. Furthermore, when contemplating a movement system approach to the design of composite structures, it has been found advantageous to conceive of such systems as having several levels or subsystems depending upon the complexity of the structure being designed. For example, with respect to the composite structure shown if Figure 10.1, a structure movement system could be conceived of as follows:

• • •

Primary movement system: The composite structure constitutes the primary movement system. Its main components are the bridge, its foundation supports, abutment backfill, approach embankments, jointed approach pavements, in situ soils, bedrock, etc. Secondary movement system: Each component of a primary movement system functions as a movement subsystem. The bridge, for example, would be the most complex component or movement subsystem of the composite structure. Its components include the superstructure, substructures, bearings, movable deck joints, approach slabs, etc. Tertiary movement system: Each component of a secondary movement system functions as a movement subsystem. One of the more complex movement subsystems of bridges would be the bearings, or bearing components, in the case of pot bearings, include pistons, pots, sliding surfaces, anchor bolts, etc. Other components of bridge subsystems would be deck joints. As modular deck joints of the largest bridges are actually complex mini-bridges, they consist of many components that also function as movement subsystems.

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

171

For an effective structure design, not only must each component (each movement subsystem) be designed as an efficient and functional system in its own right, but each subsystem must also be designed as a suitable component of the larger secondary system (the bridge) and of the primary system (the composite structure). In other words, all components of structure movement systems and subsystems must be considered with respect to how they would facilitate the proper and effective functioning of the composite structure. An examination of the various movement system components of a relatively simple composite structure type will help to illustrate the importance of evaluating the suitability of such components for ensuring the proper and effective response of the movement system to applied loads and to material and environmental changes.

Deceptive simplicity As another example of a defective structure movement system, consider the bridge details illustrated in Figure 10.7. Surprisingly, the basic components of this detail were exhibited on a recent state transportation department drawing that was made to illustrate standard details of a new bridge concept. Undoubtedly, this concept was devised primarily to eliminate movable deck joints between bridge decks and approach slabs, and thereby prevent bridge roadway drainage from penetrating below the deck slab and adversely affecting the durability of primary bridge members, bearings, and bridge seats. As illustrated, the bridge deck slab and approach slabs were to be joined together at the construction joints between them with lower longitudinal reinforcing steel. Sliding surfaces have been provided between the deck slab and top of the abutment backwalls and, apparently, there will be some type of movable joints between approach slabs and approach pavements to accommodate longitudinal superstructure expansion and contraction.

Figure 10.7 One state’s design concept for a bridge with an unusual movable joint at the superstructure/abutment interface, a concept with uncoordinated structure movement system components.

172

Integral and Semi-integral Bridges

Preventing the movable joints from projecting up through the deck slab to the roadway surface appears to have been the limit of the designer’s interest and attention. Notice, in this version of the jointless bridge concept, that the main members of the superstructure (the stringers) are supposedly supported by elastomeric bearings. Bridge engineers who use holistic views of structure movement systems and who are aware of the deformational characteristics of elastomeric bearings will readily recognize the primary flaw in this jointless bridge concept. This concept is just one simple example of where designers, checkers, and reviewers of a bridge design staff did not recognize that, although the names “joints” and “bearings” are used by the profession to identify and classify two dissimilar mechanical devices, these devices are actually intended to function as critical components of structure movement systems. As such, the characteristics of the one component (the movable joints between the deck slab and top of the abutment backwalls) must be coordinated with the characteristics of the other (the bearings), to ensure that all will function in unison and facilitate the response of the structure to applied loads and to material and environmental changes. As such coordination of characteristics was obviously not done, structures based on this concept will not perform as expected. Ultimately, many of the larger bridges based on this concept will require substantial repair or modification to keep them in serviceable condition. More specifically, when vehicular live loads are applied to spans adjacent to an abutment with a structure movement system similar to the one illustrated in Figure 10.7, the superstructure end reactions would tend to compress the elastomeric bearings, and compression of the bearings would tend to lower the superstructure. However, since the extended deck slab bears upon and is supported by the abutment backwall, the relative rigidity or stiffness of the cantilever deck slab would resist superstructure lowering and compression of the elastomeric bearings. Thus, vertical reactions would be transmitted through the cantilevered deck slab and down the abutment backwall, and not down through the stringers and bearings, as was obviously contemplated for this design. The above analysis is based on the questionable assumption that there would be absolute contact between the surfaces of the components of the movement system illustrated in Figure 10.7. However, as the application of live load on the span adjacent to an abutment would also be accompanied by vertical deflections and end rotations of the superstructure stringers (and deck slab), the exact distribution of the end reactions between the bearings and backwall would depend upon the rigidity of the cantilevered deck slab, the compressibility of the bearings, the magnitude of the superstructure end reactions, and the original fit of the sliding surface between the deck slab and backwall. For the longest bridges with thicker, more compressible bearings, it is probable that most superstructure live load reactions at abutments would be, with the movement system illustrated, supported by the relatively rigid deck slab and abutment backwalls and not by compressible elastomeric bearings. As a direct result of this type of load distribution, stringers would hang from the deck slab instead of being supported from below by bearings. Depending upon the strength of the deck-slab/stringer connections, chances would be good that these connections would be overstressed. Also, stringer webs would be subjected to stresses that were the reverse of those contemplated during design, stresses that could cause web damage and rapid stringer deterioration.

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

173

Similarly, the distribution of the superstructure dead load should also be of concern. Depending upon the deck-slab placement direction, placement speed, and concrete set time, it is probable that elastomeric bearings would provide initial support for all or most superstructure dead-load reactions. Portions of the superstructure dead load initially supported by abutment backwalls would probably be negligible. Over time, however, elastomeric bearings under a sustained load would continue to compress at a gradually decreasing rate because of a forced realignment of the elastomer’s molecular structure. It has been estimated that such sustained load compression (called creep strain) can be as much as 25–35 percent of the initial compression due to load application on typical bearing elastomers. However, where creep is prevented, molecular realignment of loaded elastomer will still take place, resulting in a relaxation of the elastomer and a shifting of the sustained dead load from the elastomer to those more rigid portions of the movement system preventing creep. Consequently, when deck slabs are rigidly supported by abutment backwalls, portions of superstructure dead-load reactions at the abutments will be shifted from compressible bearings to rigid backwalls. This type of load redistribution would be more significant for the longer bridges with thicker bearings, those bearings that would experience the greatest total creep effects. In summary, it appears that the use of elastomeric bearings as components of the movement systems of the bridge concept illustrated in Figure 10.7 is inappropriate because these bearings are deformationally incompatible with the rigid deck slab that extends over and is supported by rigid abutment backwalls. It also appears clear that the originators of this bridge concept were focusing their attention elementalistically on two supposedly disparate components of a design instead of holistically on the suitability of the movable joints and bearings as synergistic components of a structure movement system. Had the probable movements and deformations of this structure’s movement system components been more appropriately visualized, designers would have recognized the flaws in this concept, and been motivated to choose more compatible components to facilitate the effective response of this bridge type to applied loading and environmental changes.

Deceptive simplicity, continued The above evaluation of the jointless bridge concept of Figure 10.7 focused somewhat elementalistically on the defective structure movement system of the idealized concept as it is presented in the design sketch. It neglected to holistically evaluate the structure movement system of an actual composite structure based on this concept as the structure might likely be found several years after its construction. The purpose of the brief evaluation that follows is to compensate for that neglect. The jointless bridge concept, as it was presented in the state’s exhibit drawing, did not have any qualifier with respect to the type of foundation that had to be provided to support the abutments of a structure based on this concept. Consequently, it is a distinct possibility that the concept could be used with abutments supported by spread footings located at the apex of newly constructed embankments. In such a situation, the embankments become primary components of the composite structure and thus would function as primary components of its

174

Integral and Semi-integral Bridges

Figure 10.8 The adverse effect of differential settlement on the bridge concept illustrated in Figure 10.7.

structure movement system. As a result, an holistic evaluation of such a composite structure must consider how the embankment components would likely affect the performance of its movement system. Figure 10.8 was prepared to visually magnify the effect that differential settlement of one abutment (due to after-construction consolidation of an embankment) with respect to the other would have on the orientation and position of the superstructure with respect to the supporting abutment. Such settlement is rather commonplace, even for projects that have mandatory waiting periods between embankment completion and abutment construction. Notice that embankment consolidation and concomitant settlement of an abutment would cause the bearings to become separated from the bridge seat. Geometrically, it is estimated that dimension Z would be about equal to ⅛ in. (3 mm) for a bridge span of 50 ft. (15 m) and an abutment settlement of 3 in. (76 mm). Z would be proportionally larger or smaller depending on the magnitude of the actual settlement and dimension X. As illustrated, as abutment settlement would raise the bearings off of the bridge seat, such settlement would transfer the total live-load and dead-load reactions from the bearings to the backwall, with adverse consequences similar to those associated with elastomeric bearings described above. Notice also that this adverse condition caused by settlement would occur regardless of the type of bearings used. Although not illustrated, such settlement of one abutment with respect to the other would also have adverse consequences at the other abutment, although at the other abutment the superstructure would be supported by and rotate about the bearings. Such rotation would separate the deck slab and connected approach slab from the top of the abutment backwall, leaving the approach supported solely by

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

175

the end of the deck slab, or more properly by the lower reinforcement at the end of the deck slab. Due to lack of shear resistance at the construction joint, it is probable that the live-load and dead-load end reactions of the approach slab would result in the fracture of the lower concrete encasement of the deck-slab continuity reinforcement and the gradual degradation of both slab ends at this location. As a result of such potentially adverse responses, the bridge concept illustrated in Figure 10.7 should not be used at locations where appreciable settlement is probable. From a systems perspective, this evaluation can be further generalized by stating that newly constructed abutment embankments should not be used as components of a structure movement system unless the magnitude of post-construction consolidation of embankments and surcharged in situ subsoils has been estimated with reasonable accuracy and that other components of the structure movement system have been articulated in such a way that they would not be adversely affected if such consolidations were realized.

Holistic view boundaries Bridge engineers’ first visualizations of structure deformations and movements were typically attained during their college courses on the analysis and design of continuous members and rigid frames. The method of choice for such problems was Hardy Cross’s “Analysis of Continuous Frames by Distributing Fixed End Moments” [3]. Cross’s method begins with an elementalistic view of each member and generally concludes with holistic views of the structure. With the continued use of this method, these engineers gradually become able to visualize the movements and deformations that take place throughout a structure as a result of the application of a load at any particular point on the structure. However, even for some structures for which holistic views of structure movement systems were obviously used during analysis and design, engineer’s elementalistic tendencies would occasionally be manifest on design drawings of proposed bridges and in the structural details of completed bridges. These tendencies are most evident where limited or bounded holistic views of structure movement systems were apparently assumed during design. Although examples of these tendencies are numerous, two are described below to substantiate this observation. Pier piling A glimpse of such a tendency is given by the column footing connection illustrated in Figure 10.9. This connection is intended for an exterior column of an unusually long cap-and-column pier. Presumably, the engineers (designer, checker, and reviewer) responsible for the design of this pier proposed the hinged column joint shown in Figure 10.9 to minimize frame moments induced by shrinkage and contraction of its long pier cap. For practical purposes, holistic views must have boundaries. Such boundaries are necessary so that attention can be limited or focused on those local aspects of a composite structure (see Figure 10.1) that have more than a negligible effect on the performance of a structure. Except for seismic effects, bedrock is a suitable

176

Integral and Semi-integral Bridges

Figure 10.9 A superfluous hinged joint provided primarily because of an unnecessarily limited holistic view during frame design of the structure’s movement system.

boundary. Pile inflection points or pile tips are other valid boundaries. Usually 200 ft. (61 m) of approach embankment and 1,000 ft. (305 m) of jointed approach pavement can also be considered suitable boundaries for holistic views. Depending upon the particular structure being considered, boundaries of holistic views can be narrowed as appropriate. As suggested by the details illustrated in Figure 10.9, however, it appears that the design engineers’ holistic views of the rigid frame pier were limited by assuming a fixed joint for the pile-supported footing (an absolutely rigid holistic view boundary at the level of the bottom of the footing). Consequently, to compensate for this unrealistic and limited holistic view boundary, the engineers were forced to provide hinged column/footing joints for the exterior columns to minimize frame moments and yield more reasonable section properties for the members of the pier frame. On the other hand, a less restricted holistic view would have recognized that the piles supporting the column footings of this pier frame were integral with, and probably flexible components of, the frame’s movement system. As such, they should have included all or part of them in holistic views of the structure movement system. As such, frame analysis based on broadened holistic views would have yielded more realistic frame moments and forces (eliminating the need for column/footing joints with hinged reinforcement) and a simpler structure. Incidentally, with respect to very long cap-and-column piers, holistic views of the structure movement system of the composite structures could be made more realistic (but more complex) by including differential axial deformation of piling and differential lateral translation of footings responding to pier cap shrinkage, creep, and temperature changes.

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

Figure 10.10

177

USR I-480, Cuyahoga River Bridge, Cleveland, Ohio, 1975.

Pier footings The I-480 Bridge over the Cuyahoga River Valley in Cleveland, Ohio (Figure 10.10), is another example of where an apparently limited or unnecessarily bounded holistic view of the structure’s movement system was used in design. The designers of this outstanding high-level bridge provided its superstructure with movable deck joints at abutments and at several span inflection points located throughout the length of the bridge. These intermediate joints resulted in several somewhat independent superstructure segments that are each supported by either two or three hammer-head-type piers. Fixed bolster bearings were provided at one pier of each unit and roller bearings were provided at the others (Figure 10.11). This use of roller bearings on these tall flexible piers leads one to suspect that the designers of the bridge limited or bounded their holistic views of the structure’s movement system at the tops of the piers. Apparently ignored was the significant effect of pier flexibility on the structure’s movement system. Bridge inspector’s photographs of the sprocketed roller bearings seem to support this observation. Notice, in Figure 10.11a, that the amount of anticipated longitudinal superstructure movement at this pier is suggested by the difference in the width between the near scupper and the open-headed downspout. For such a small amount of anticipated differential movement, pier flexibility should have allowed this pier and the superstructure to be connected together with the less complicated and less expensive bolster bearings. In addition, use of bolster bearings would have allowed tight, waterproof connections to be made between superstructure scuppers and pier downspouts. Such connections would have minimized drainage problems and later maintenance efforts.

178

Integral and Semi-integral Bridges

(a)

(b)

Figure 10.11 Pier inspection photographs of a roller bearing (a) and a bolster bearing (b) of the bridge in Figure 10.10.

Also, as evident in Figure 10.12, vertical keil marks on the sprocketed rollers and guide bars made during a previous inspection indicate that the bearing had not moved longitudinal. At other piers, uncracked bearing paint indicated this same lack of roller bearing movement. Consequently, the condition of the roller bearings

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

179

Figure 10.12 A close-up photograph of another roller bearing of the bridge in Figure 10.10. The inspector’s keil marks on the sprocketed rollers and guide bars indicate that this bearing had not moved since the last biannual inspection of the structure.

indicates that all longitudinal movements of the superstructure at the piers of this structure were being accommodated by pier flexure alone. In other words, it appears that holistic views of the primary movement system of this bridge should not have been bounded at the pier tops. Instead, they should have been bounded at the tops of the pier footings so that the tall flexible pier stems could have been included with the deck joints and bearings as major components of the primary movement system of the bridge. Such views would have resulted in not only a fully functioning and efficient structure movement system but also the exclusive use of less complicated and less expensive bolster bearings at the piers, a closed and more efficient deck drainage system at the piers, less long-term maintenance, and, ultimately, a more durable bridge.

Multidimensional views Holistic views as the name indicates need not, and for some structures should not, be limited to visualizing the performance of structure movement systems in only one plane or dimension. Typically, views in one dimension are adequate (i.e., stringers in square deck-type bridges moving longitudinally and rotating about the bearings in a vertical plane. Occasionally, views in two dimensions are found useful (i.e., tall piers deflecting both longitudinally and laterally, or embankment surcharged subsoils being compressed vertically and bulging or translating both laterally and

180

Integral and Semi-integral Bridges

longitudinally). Sometimes, however, holistic views of the performance of movement systems in several dimensions are necessary (i.e., the bearings for skewed or curved superstructures that are required to move longitudinally and rotate in both longitudinal and lateral planes, or superstructures of skewed semi-integral bridges that tend to rotate in a horizontal plane, to move longitudinally in an essentially horizontal plane, and rotate about the bearings in a plane essentially perpendicular to the centerline of the piers and abutments). An example of the multidimensional movement system of a single-span, skewed, semi-integral bridge is partially described below and in Chapter 8. Figure 9.6a (see Chapter 9) illustrates an example of the primary details of one state’s version of the semi-integral bridge concept. Numerous other versions of this concept are illustrated in Chapter 9. As evidenced by Figure 9.6a, the composite structure and its movement system components consist of the reinforced concrete end diaphragms, superstructure backfill, approach slabs, and movable bearings located at the superstructure/abutment-footing interface. For lateral and longitudinal stability, this bridge type depends primarily upon composite interaction of the end diaphragms and backfill at both abutments. As suggested by Figure 9.6a, longitudinal expansion or contraction of the superstructure will be accommodated by backfill compression or expansion, respectively. Active backfill pressure provides some restraint for the superstructure during contraction, whereas passive pressure provides restraint during superstructure expansion. As approach slabs are attached to the end diaphragms, their weight and the friction between the slabs and their subbases will act in unison to restrain longitudinal movements of the superstructure in either direction. Also functioning to restrain longitudinal superstructure movements would be the elastomeric bearings, fillers between the approach slabs and the approach pavement, and. ultimately, the approach pavements themselves. It should be particularly instructive to consider the effects of backfill pressures on the superstructure of a skewed semi-integral bridge during expansion of the superstructure. As noted above and in Chapter 8, expansion of the superstructure results in compression to the backfill at both abutments. The resultant forces (due to the summation of these pressures), Pp, at both abutments, together with end-diaphragm/ backfill-interface friction, act to restrain the superstructure both longitudinally and laterally. However for a skewed superstructure, these resultant forces against the end diaphragms do not coincide (see Figure 8.4). This lack of coincidence causes unbalanced moments in a horizontal plane and possibly an unstable structure. Such stability depends primarily upon the structural characteristics of the backfill and the composite interaction of the backfill at both abutments with the superstructure (the effect of the approach slab and approach pavement neglected), the three primary components of the structure movement system. Figure 8.4 illustrates the results of the interaction between the superstructure and the backfill of a skewed semi-integral bridge during superstructure expansion. As a result of the skew, the two resultant passive pressure forces acting perpendicular to the end diaphragms would tend to induce clockwise rotation (Pp L sin θ) while two resultant shearing forces acting parallel to the end diaphragms would tend to resist clockwise rotation (Pp tan θ L cos θ). The shear force components (Pp tan θ) represent either friction of the end-diaphragm surface moving relative to the backfill (Pp tan δ)

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

181

or the shearing of the backfill (Pp tan Ø), whichever is the smallest. Rotation would not take place if: PP L sin θ ≤ Pp tan δL cos θ or PP tan ∅L cos θ. However, if the former moment (Pp L sin θ) is greater than either of the latter moments, then the superstructure would be unstable and would, for the skew illustrated in Figure 8.4, tend to move rotationally in a clockwise direction during each expansion of the superstructure. If guide bearings were not provided to resist this rotation, the bridge superstructure (although initially restrained by the shearing and frictional resistance of the backfill against the end diaphragms, the weight and frictional resistance of attached approach slab, and the slight shearing resistance of elastomeric bearings) would incrementally and progressively rotate towards the acute corner of the bridge. As can be ascertained from the movement systems of the composite structures described above, designing an effective and durable structure movement system for a single span bridge with semi-integral abutments is difficult. That difficulty is immensely complicated when the semi-integral bridge abutments are skewed. For such a system, movable bearings, movable longitudinal joints, superstructure backfill, and guide bearings must be designed and provided that will support superimposed vertical, lateral, and longitudinal loads and accommodate (facilitate or prevent) various types of superstructure movements, including: (1) support vertical dead and live loads, and superimposed lateral and longitudinal loads (2) facilitate cyclic longitudinal elongation and shortening of the superstructure (3) facilitate cyclic abnormal vertical rotation of the superstructure at abutments and piers (4) prevent cyclic incremental and progressive horizontal rotation of the superstructure at abutments (5) prevent or minimize vertical settlement of the superstructure at piers and abutments. Fortunately, elastomeric bearings can be designed and manufactured to support the loads enumerated in (1) and to facilitate the movements described in (2) and (3). Vertically oriented elastomeric bearings can be designed and manufactured to prevent the movements described in (4) while still facilitating the movements described in (2) and (3). Friction or end-bearing piles, pedestals, or drilled shafts can be designed to support the loads enumerated in (1) while minimizing the movements described in (5). Finally, it is now possible to design and manufacture elastomeric joint seals that will accommodate the movements described in (2) and (3). However, when thought is given to the complexity involved in designing semiintegral bridge abutments with movement systems that support the loads of (1) and facilitate the movements described above in (2) and (3) while preventing the movements described in (4) and (5), it becomes increasingly clear why bridges with integral abutments, such as the one illustrated in Figure 7.7 (Chapter 7), without any joints whatsoever, have become the structure of choice by many transportation departments where such structures can be supported by flexible piles. This structure

182

Integral and Semi-integral Bridges

type has been made possible only because bridge design engineers have taken advantage of the fact that the single row of vertically driven flexible abutment piles of integral bridges provide support not only for the loads of (1), but also for all of the movements described in (2)–(5). Either friction or end bearing, these flexible piles minimize settlement described in (5); they resist the horizontal rotational movements described in (4); and they facilitate the longitudinal and abnormal rotational movements described in (2) and (3). In effect, the flexible piles of integral abutments constitute the primary components of the composite structure’s movement system. They eliminate the need for movable bearings and movable joints, and also the need for elastomeric joint seals. To accomplish these beneficial effects from such a simple structure movement system, however, design engineers must be willing to subject the piling of integral abutments to a complex of loads and bending moments that are easy to control but difficult to estimate. So until computer programs are developed to accurately estimate integral abutment piling stress levels, engineers designing composite structures with integral bridges must be willing to use their intuition to limit the dimensional and geometrical application range of such structures to ensure reasonable stress magnitudes and stress ranges for the critical piling that make the simplified movement system of integral bridges possible.

Summary Today, small- and medium-size highway bridges are being built that fill the spectrum of bridge types from the single- and multiple-span simply supported superstructures with movable deck joints at the superstructure/abutment interface, to composite structures consisting of jointless superstructures, flexible substructures, compressible backfill, well-consolidating embankments, slideable approach slabs, and movable pavement joints. The one constant in this design and construction of highway bridges is the necessity to compose and configure the various components of composite structures into well-coordinated structure movement systems. Such coordination will help to ensure that all components of movement systems will respond appropriately to applied loads and environmental changes. Consequently, to be more effective in designing today’s small- and medium-size highway bridges, bridge design engineers must become somewhat familiar with and make allowance for the behavior of movement system components (surcharged in situ subsoils, new embankments, abutment backfill, movable pavement joints, etc.) that historically have been the primary responsibility of others. In this respect, modern bridge design and construction have become multidisciplinary efforts. Consequently, bridge design engineers should work cooperatively with and receive guidance from many engineering specialists (foundation, hydraulic, materials, soil, geotechnical, pavement, and maintenance specialists) to achieve fully functional structure movement systems. As the few examples presented and examined above suggest, it appears that some (possibly many) bridge designers have not as yet allied themselves with specialist colleagues to achieve such results. Instead, there appears to be a tendency for bridge designers to design structure components elementalistically without regard for how

Chapter 10 Elementalistic and Holistic Views for Evaluation and Design

183

the structure components, especially components of composite structures, will interact with each other to facilitate the responses of structures to applied loads and to material and environmental changes. It is hoped that this chapter and the new TRB Structure Movement Systems Subcommittee will serve to encourage cooperation between transportation specialists so that future construction will be viewed holistically. Such views will help to ensure that all components of composite structure movement systems will be properly selected, configured, composed, and designed to function synergistically, and yield more effective structure movement systems and more functional and durable bridges.

References 1. O’Neil, J. J., Prodigal Genius: The Life of Nikola Tesla, Ives Washburn Inc., New York, 1944, p. 257. 2. Burke, M. P. Jr., “The Abnormal Rotation of Skewed and Curved Bridges,” Transportation Research Record No. 903, Transportation Research Board of the National Academies, Washington, D.C., 1983. 3. Cross, H., “Analysis of Continuous Frames by Distributing Fixed End Moments,” ASCE Proceedings, American Society of Civil Engineers, New York, 1930, pp. 919–928.

Chapter 11

Awareness of Reality in Bridge Design

The range of what we think and do is limited by what we fail to notice. And because we fail to notice that we fail to notice, there is little that we can do to change, until we notice how failing to notice shapes our thoughts and deeds.

R. D. Laing

Introduction While doing the extensive library research that was necessary for the development of the Bridge Aesthetics Bibliography [1], a brief thought-provoking article by Bruce Johnson was found in the January 1965 issue of Civil Engineering. Titled “An Awareness of Reality,” the article mentioned Hardy Cross’s concern for his students because of their general lack of awareness. Johnson [2] stated: Hardy Cross affirmed that an important duty of teachers was to force students repeatedly back into the field of reality and, even more, to teach them to force themselves back to reality. He continually jerked his students out of the textbook, out of the theory, back to a consideration of the real structure. … He went on to say that the sense of reality seems to decrease with elaborateness of equipment and may finally disappear completely from laboratory work. [2] 185

186

Integral and Semi-integral Bridges

This article struck a responsive chord since it appeared to explain the predilection of many bridge design engineers, particularly novice engineers, toward particular types of design errors. This predilection appeared to be true not only for novices and taciturn individuals with a limited aptitude for design, but also for engineers with decades of experience – experience that appeared to be limited to new bridges designed with the aid of standard drawings, typical construction techniques, and the usual agency design guidelines. Consequently, upon reading this article, this author was motivated to make a collection of blatant examples of bridge design errors, errors that appeared to owe their existence to a deficient awareness of reality. Finally, this collection of design errors provoked a study that resulted in the observations, speculations, and suggestions given below. Although a more general paper on this subject was published several years ago, the examples given below and the observations and speculations given herein are generally related to the design and construction of integral and semi-integral bridges. Although these examples are based mostly on the design and construction of Ohio structures, presumably, similar examples could be described for the bridges of other states as well.

Reality Reality is commonly defined as “that which exist, as contrasted with what is fictitious, or merely conceived of.” Another definition that accounts for the process nature of reality probably would be “what is going on.” Unfortunately, due to years of conditioning, many willingly use secondary or tertiary sources for information about “what is going on.” They have not learned the value of the more vivid and thought-provoking impressions that are obtained from first-hand personal observations. Mort Walker, the creator of the Beetle Bailey comic strip, made an observation similar to those of Bruce Johnson and Hardy Cross. In one of his strips, Sargent Snorkle is sitting on his cot putting on a pair of socks. He turns and asks Zoro, one of the men of his company, “What’s the weather like outside, Zoro?” Zoro eagerly responds, “I’ll turn on the T.V. and find out.” In reaction to that response, Sargent Snorkle, with an exasperated scowl on his face, observes, “This generation has forgotten to look out the window” [3]. For the purpose of this chapter, the type of errors selected as typical examples have been limited to those that appear blatantly related to a deficient awareness of reality, or to the design engineer’s failure to “look out the window.” More specifically, these errors appear to be caused by a deficient awareness of “things,” a deficient awareness of “change,” a deficient awareness of “differences,” and a deficient awareness of “similarities.”

Awareness of things A thing or object is commonly defined as “that which exists as a separate entity; an inanimate object,” or “that which is designated, as contrasted with the word or symbol used to denote it.” A more scientific definition would probably be something

Chapter 11 Awareness of Reality in Bridge Design

187

like “An event consisting of almost imperceptible changes.” Although it seems improbable that well-educated engineers would not have a heightened awareness of things or objects, the following examples show that this assumption is not always true. Depending upon circumstances, certain engineers appear to be oblivious to important aspects of reality, or they have failed to consider what effect the presence of these aspects would have on the constructability of their designs. The elusive stream The design consultant for a bridge replacement over a tributary to the Black River in Medina County, Ohio, hired a surveyor to prepare a contoured site plan for the proposed bridge site. After obtaining cross-sections, the surveyor prepared the site plan that the consultant used to establish the structure type, spans and skew, and other structure and embankment characteristics. Apparently, neither the surveyor nor the design consultant bothered to evaluate the site plan by comparing it with the actual surface conformations at the bridge site. During review of the consultant’s proposed bridge type, spans, and skew, it was noticed, by means of new site photographs, that the site plan contours did not appear to properly reflect the ground conformation or stream channel configuration. To aid site plan review, a search was made and an old aerial survey of the area found. This aerial survey revealed the reason for the peculiarity of the consultant’s site plan. A small tributary stream was converging with the main channel just upstream of the proposed structure. The submitted site plan indicated that neither the surveyor nor the bridge design consultant was aware of this tributary stream. Upon rejection of his initial proposal, the design consultant had a second survey made and a new site plan prepared. He then changed the characteristics of the proposed structure to accommodate the altered stream channels and ground contours. When comparing the originally proposed structure with the final re-contoured site plan, it showed that the tributary stream would have flowed behind the left wingwall of the forward abutment. If the originally proposed structure had been approved and the project sold, such a mistake would easily have been noticed by project supervisors. Consequently, this error would have resulted in a quick redesign of the structure or, more likely, a termination of the bridge construction contract. Unobtrusive dam Another engineer who was making a study of stream hydraulics for a bridge site made a similar but less serious error. A review of his work revealed that he was completely unaware of an upstream dam that could be observed from the existing bridge roadway. Inaccessible dowels As this author was preparing to describe this example of a designer’s lack of awareness of the contractor’s reality for this chapter – the reality that the contractor would be experiencing as he attempted to accomplish what was specified by the

188

Integral and Semi-integral Bridges

designer on the bridge rehabilitation plans – he became aware that he too had also been guilty of a similar lack of awareness in his review of the final revised project plans. Presumably, he became so interested in and preoccupied with the obvious errors in the design proposal that he neglected another aspect of the design – an aspect that had also escaped the attention of the designers and primary reviewer of the project. Hopefully, the construction project superintendent became aware of the construction problem left unresolved by the designer, checker, reviewers, and the author (described below) before permanent damage was done to the structure. Surprisingly, after this example of lack of awareness was described in a previously published paper, not a single engineer contacted the author to question him about a problem that may have been left unresolved on the final contract plans. As the reader follows the author’s presentation below, it would be of interest to know if he or she becomes aware of this unresolved problem before it is described at the end of this discussion. A four-span steel beam bridge over USR I-76 in Portage County, Ohio, was to be modified and rehabilitated. As part of the project, its 30-year-old concrete deck was to be replaced and the superstructure raised to improve the vertical clearance under the structure. To support the superstructure in its new higher position, the designer engineer proposed to provide relatively thick reinforced concrete slabs on the caps of the cap-and-column piers. The slabs were to be attached to the pier caps with vertical dowel bars placed and grouted into field-drilled holes. A pier partial plan and elevation view (Figure 11.1, left) shows the consultant’s intention to

Figure 11.1 Placement of dowel bars in the caps of reinforced concrete capand-column piers.

Chapter 11 Awareness of Reality in Bridge Design

189

provide equally spaced dowel bars to anchor the new reinforced concrete slabs to the original bridge seats. Notice in the plan view of Figure 11.1 (left) the position of beam centerlines. What the details in Figure 11.1 (top left) failed to illustrate and what the design consultant failed to remember and visualized is that, although the existing beams were to be supported in a raised position, they would remain in their original lateral location. Notice, in the part plan view of Figure 11.1 (top right), where the beams are more realistically illustrated, that some of the dowel locations are under a beam flange, a location that makes it impossible for the contractor to drill vertical dowel holes in the pier caps. This situation would occur at both ends of each pier and at all three bridge piers. As can be surmised from this error, the designer’s original plan view, with its centerline of beams, was not a fair representation of the actual structure at the time when dowel holes were to be drilled and dowels grouted into place. Also, it is clear that when the plan view of Figure 11.1 (left) was being prepared and dowel bars spaced, the design engineer had completely forgotten the presence of the bridge’s entire superstructure. Instead, the designer appears to have been concentrating so exclusively on the representation of the pier cap that he or she was solving a paper problem, a commonplace type of design error. As it is difficult to retain clear images of portions of an existing structure by thought processes alone, the bridge design profession has found it helpful and customary to supplement such images with photographs of various structure details, sketches, and more realistic plan details. Obviously, to make these plan details more realistic, they need to contain representations of all important elements of a structure that are present when intended field work is to be performed. Otherwise, these elements are often forgotten, resulting in proposed work that cannot be accomplished. After being informed of his or her design error, the design engineer changed proposed pier bridge seat details by respacing the dowel bars to clear existing beam flanges (Figure 11.1, center right). However, in revising the design, the designer again failed to visualize the complete steel framing of the existing superstructure. For most Ohio steel beam bridges, a series of equally and closely spaced transverse cross-frames are provided with some frames located over piers (Figure 11.1, bottom right). Notice, in Figure 11.1 (bottom right), that the revised dowel bar spacing places some dowel bars at inaccessible locations directly under existing cross-frames. The plans were again returned to the designer so that a dowel bar spacing could be provided that would be clear of all of superstructure elements. The review of the bridge plans for this structure provided one clear example of where a design engineer seemed to be oblivious of important elements of an existing bridge, even after his or her original oversight was brought to his or her attention. Although this is only one example of where a design engineer appeared to be unaware of the presence of a bridge’s entire steel superstructure, similar inaccessible dowel bar locations have been mistakenly specified on the bridge plans by many other engineers. It was a common type of mistake. Such bridge rehabilitation plan errors suggest the importance of accurate and realistic representations of existing structure details. Otherwise, for those who have not been given the opportunity to become thoroughly familiar with existing structures (site visits, recent detail photographs, inventory records, etc.), those who fail

190

Integral and Semi-integral Bridges

to visualize important components of existing structures, and those who generally produce incomplete and/or small-scale detailed drawings, adverse construction situations are usually not realized or anticipated. These unrealized situations usually result in plan errors and omissions, necessary revisions of plans and contract requirements, project delays, and extra work charges.

Inaccessible dowels epilogue It appears that the designer’s, checker’s, and reviewer’s (including this author’s) preoccupation with appropriate dowel bar spacings for these pier-cap additions resulted in no recognition being given to the probability that the final approved location for these dowel bars could have an adverse effect on the upper reinforcement of the pier caps being modified. Cap-and-column pier frames usually contain generous longitudinal reinforcement adjacent to the upper horizontal surface of the pier caps. As a result, there is always the probability that drilling dowel bar holes into these caps would result in cutting some of this reinforcement. Although efforts are usually made in the field to avoid cutting existing reinforcement when placing dowel bars, it is usually prudent for bridge designers to anticipate such problems by choosing dowel bar locations that are least likely to coincide with existing reinforcement. Nevertheless, with the dowel bar locations specified for this and other similar projects, there was the possibility that these locations would result in damage to some of the primary reinforcement of these pier frames and, as a result of the reviewer’s and this author’s preoccupation with dowel bar spacings, some of this type of damage was likely. For the pier-cap modifications like the one described above, where composite behavior was not considered, it was not really necessary to have a uniform dowel bar spacing. Actually, it appears that the use of dowel bars is sometimes more of a habit than a necessity. In this particular application, it appears that the limited benefit achieved by the use of dowel bars did not justify the probable damage to upper pier-cap reinforcing steel that was likely to occur during the drilling of the dowel holes. In applications such as this one, dowel bars are used only to ensure that the horizontal construction joint between the pier caps and the new bridge seat slabs were joined together to provide adequate horizontal shear resistance, and to help prevent the ingress and freezing of water in the joints – not significant problems for horizontal construction joints subjected to sustained compression. Consequently, if some dowel bars are considered absolutely necessary for situations like that described above, only a nominal number of dowel bars should be provided. They should be spaced not only to avoid superstructure members (to facilitate dowel hole drilling) but also to clear upper pier-cap reinforcement, or more specifically to clear upper pier-cap reinforcement in negative moment regions (to safeguard the structural integrity of the caps and thus minimize potential problems for the bridge maintenance engineers). The situation described above, although it was for a relatively small and relatively simple structure, is a clear example of how important it is for bridge engineers, especially design, checking, and review engineers, to have a vivid awareness of reality.

Chapter 11 Awareness of Reality in Bridge Design

191

Bar splices Ohio Bridge No. HIG-28-0280 (Figure 11.2), a single-span, prestressed, concrete box-beam, integral bridge, was to be constructed in stages while one lane of two-way traffic was to be maintained. Abutment and box-beam end-reinforcement for this bridge concept are shown in Figure 11.3.

Figure 11.2 Ohio Bridge No. HIG-28-0280, 1992: Two-stage construction of a prestressed concrete, box-beam, integral bridge.

Figure 11.3 Abutment and superstructure continuity reinforcement details for prestressed concrete, box-beam, integral bridges.

192

Integral and Semi-integral Bridges

Figure 11.4 An abutment view of the Figure 11.12 bridge just before placement of the first beam of the second construction stage.

Stage construction proceeded without delay until the first beam of the second stage was to be placed (Figure 11.4). On attempting to move this beam into position adjacent to the first-stage superstructure, it was discovered that the protruding firststage deck-slab reinforcement prevented placement of the beam vertically so that the reinforcement protruding horizontally from the beam and vertically from the abutment seat could be merged. Obviously, the bridge designer, checker, and reviewer all failed to visualize the second-stage erection situation (the contractor’s reality) with respect to the reinforcement protruding from the abutment, box-beam, and first-stage deck slab. This problem could have been avoided if the first-stage construction joint in the deck slab had been provided with nonprotruding-type mechanical connectors for splicing the first- and second-stage transverse deck-slab reinforcement. The lesson to be taken from the construction of this bridge is that all of those involved in the preparation of bridge plans need to visualize all stages of construction (to become familiar with the contractor’s changing reality) to ensure that bridge plans do not create situations for the contractor that prevent him from expeditiously completing the project.

Awareness of change Change is defined as “a passing from one phase, form, place, etc., to another.” It is the term used to recognize the process nature of reality and that everything con-

Chapter 11 Awareness of Reality in Bridge Design

193

tinuously changes from instant to instant. As Heraclitus observed, “One cannot step twice into the same river.” Not only does the water continuously change, but so do individuals change from instant to instant. Consequently, each step is by a different individual into a different river. Obviously, the changes considered in Heraclitus’s aphorism are microscopic imperceptible changes occurring within inconsequential time intervals. Other changes are barely recognizable, and still others are significant and unmistakable (earthquakes, for example). For the purpose of this chapter, only gradual changes are considered, changes with manifestations that are nevertheless unmistakable for knowledgeable and observant engineers who have learned to look for an awareness of change. The Golden Gate Bridge is a familiar example of such changes. A bridge in its setting is typically a unique and relatively slow event. The Golden Gate Bridge is such a structure, while in bridge terms the first Tocoma Narrows Bridge, which collapsed shortly after its completion, can be considered (in bridge terms) a relatively fast event. Although it is considered a slow event, the Golden Gate Bridge is such a large structure that changes in its characteristics are apparent from day to day. As a result of its oceanfront setting and prevailing winds, the steel of this structure is exposed to salt-contaminated water vapor. To combat the deleterious effect of this vapor, the paint system on this bridge is continually being renewed. To accomplish this renewal, a permanent staff of painters paints this bridge in one continuous cycle after another. The bridge is now old enough that some individuals have devoted their entire working lives to painting just this one bridge. There have been other more significant changes to this structure. For example, all its suspenders and the complete bridge roadway have been replaced. The bridge also has recently been retrofitted to improve its response to earthquakes. Actually, it would not be inappropriate to say that, due to changes that have been made to this structure during the last 70 years, the Golden Gate Bridge now standing at the entrance to San Francisco Bay is not the Golden Gate Bridge that was built at the same location 70 years ago. A fairer name for the structure would be the Golden Gate Bridges because such a name would make individuals aware of the changes that have taken place in the structure from instant to instant throughout the many years that “they” have served the City of San Francisco and neighboring communities. With such changes in mind, engineers assigned the responsibility of designing bridge modifications should avail themselves of the opportunity to inspect assigned bridges and become thoroughly familiar with them and the changes that they undergone since they were first constructed. Based on such familiarity, they will then be in a better position to provide the modifications that will restore the bridges to serviceable condition. Unfortunately, many engineers seem oblivious to manifestations of change as the following few examples will attest. Phantom deck The superstructure of a 50-year-old three-span steel beam bridge was to be replaced by a wider superstructure while one lane of two-way roadway traffic was to be maintained. The design engineer for this bridge modification contract submitted a superstructure cross-section for review, to show his proposed lane width for

194

Integral and Semi-integral Bridges

maintaining traffic on a portion of the existing superstructure while the other portion of the superstructure was removed and replaced. That superstructure cross-section appeared to have been based on details from original construction plans for the structure. A check of the state’s inventory records and the original construction plans for the structure indicated that the superstructure details (roadway width and railing design) agreed with those that were illustrated in the deck cross-section submitted for review and approval. However, examination of the actual structure revealed a significantly different superstructure cross-section. Apparently, after being modified many years ago, the changes to the superstructure of this bridge had not been documented in the inventory records. Consequently, the design engineer for this project, who depended on data from the original plans and inventory records, was spending his or her time working on a paper problem. The design engineer concluded the preliminary design effort by proposing the alteration of a bridge superstructure that no longer existed. Forlorn railing The superstructure of a three-span, continuous, reinforced concrete, integral slab bridge was to be replaced while one lane of two-way roadway traffic was to be maintained. This is another case where the design engineer did not appear to be the least bit curious about the changes that must have taken place in the 38-year-old deck concrete of this structure to have provoked the state to schedule its removal and replacement. It appeared certain that this design engineer had never seen this structure and that he or she based the proposed stage construction preliminary design proposal on the bridge’s original plans. This was evident from an examination of the proposed stage construction typical sections of the existing deck slab. That section view indicated the designer’s intention to maintain first-stage roadway traffic close to the bridge’s right railing while the left half of the deck slab was removed and replaced. However, examination of the actual structure revealed that the deck slab adjacent to the right railing (Figure 11.5) had so severely deteriorated that the right railing was only tenuously attached to the structure and incapable of resisting vehicular impacts. Actually the deck-slab and railing connections were in such poor condition that it was surprising that the state had not installed temporary railing until the deck of this structure could be replaced. As the designer had proposed to move one lane of two-way temporary traffic immediately adjacent to this right railing, his or her proposal would have seriously imperiled vehicular traffic. Had the design engineer familiarized him- or herself with the actual characteristics of this structure, the proposal to place roadway traffic adjacent to the forlorn railing would not have been made. The designer’s use of secondary sources (original plans and inventory records) to guide the development of stage construction proposals did not account for changes that the superstructure of this bridge had undergone throughout the years that it was exposed to environmental deterioration. Fortunately, the deck of this structure was wide enough that a portable concrete barrier could be installed and anchored to the deck slab inside the right railing to safeguard vehicular traffic while the left half of the deck slab was being removed and replaced.

Chapter 11 Awareness of Reality in Bridge Design

195

Figure 11.5 Ohio Bridge No. WAY-585-0859, 1952, Rev. 1991: A view of the bolted connections of its forlorn railing.

196

Integral and Semi-integral Bridges

Awareness of differences Difference is defined as “the state, quality, or degree of being unlike another.” It is the term used to recognize the process nature of reality and that no two things (events) are the same, and no one thing is the same from instant to instant. That some bridge design engineers are not fully aware of differences is illustrated by the next example. Integral or semi-integral bridges In 2001, the manual, Precast/Prestressed Integral Bridges, was published by the Precast/Prestressed Concrete Institute (PCI) as a service to the transportation profession and to help promote construction of these remarkable little structures [4]. The manual provides bridge design engineers and bridge administrators with much helpful commentary, illustrations, and design examples for the design and construction of precast/prestressed concrete integral bridges. In general, the text and illustrations are well done. However, the primer portion of the manual mistakenly does not recognize the significant structural differences between integral and semiintegral bridges. This mistake is made plain by an examination of the primer [4] illustrations. For example, “Figure 4.1: Integral Abutment Types” contains illustrations for five different abutment concepts, two of which are for semi-integral abutments; “Figure 4.3: Typical Semi-Integral Abutment” contains an illustration for an integral abutment concept for prestressed box-beam bridges; and “Figure 4.7: Integral Bridge Abutment” contains an illustration for a semi-integral abutment concept for girder bridges. (Figure 4.7 also presents a flawed structure movement system concept similar to those of Figures 9.6b,c and 9.7c in Chapter 9. See Figure 9.1 for another flawed example and Chapter 10 for a discussion that should help to clarify the problems associated with these flawed structure movement system concepts.) The structural differences between these two different abutment concepts result in significant behavioral differences, especially for semi-integral bridges constructed with large skews or curves. This behavioral difference of semi-integral bridges, if left uncontrolled, can result in superstructure rotational movements in a horizontal plain, progressive superstructure, pier and abutment distress, and eventually structural damage. The primary structural differences between integral and semi-integral bridges is that the abutments of integral bridges, as the name implies, are fully integrated with jointless superstructures, while the abutments of semi-integral bridges are not. In addition, the abutments of integral bridges are usually supported by flexible foundations (typically, single rows of slender piles) while the abutments of semi-integral bridges are supported on rigid foundations (rock, pedestals to rock, large-diameter drilled shafts, multiple rows of both vertical and battered piles, etc.). As a result of these structural differences, the superstructures of integral bridges are restrained longitudinally and rigidly restrained laterally at abutments. But the superstructures of semi-integral bridges are restrained both longitudinally and laterally at the abutments. Full definitions of these two structure types are given in Appendix 2. Illustra-

Chapter 11 Awareness of Reality in Bridge Design

197

tions of typical integral bridges are given in Figures 1.4 and 1.5 (Chapter 1) and of typical semi-integral bridges in Figures 9.6 and 9.7 (Chapter 9). As a result of structural differences between integral and semi-integral bridges, the maximum reasonable bridge length that can be achieved with these two structure types is quite different. A suitable length for integral bridges is generally limited by the horizontal flexural capacity of abutment foundations (flexible piles). However, as superstructures of semi-integral bridges are longitudinally and horizontally restrained at abutments, the length of semi-integral bridges is generally limited only by the movement capacity of compressible backfill and/or the cycle-control joints located at the approach slab/approach pavement interface. Due to the resistance of abutment backfill to compression, superstructures of skewed integral and semi-integral bridges tend to rotate in a horizontal plane toward their acute corners during superstructure expansion. For bridges with integral abutments, this rotational tendency of superstructures at abutments is resisted not only by friction at the abutment/backfill interface, but also by the lateral resistance of abutment piling. For integral bridges with large skews, one or two piles at acute corners of integral abutments could be battered laterally to help resist such rotation. However, as semi-integral bridges are not restrained by the abutments, such bridges constructed with large skews need to be provided with guide bearings of some sort at abutments to help resist horizontal superstructure rotation. Refer to Chapter 8 for a more thorough discussion of this tendency of skewed integral and semi-integral bridge superstructures to rotate horizontally toward their acute corners. In addition, examine Figure 6.6 (Chapter 6) for an example of abutment damage of a semi-integral bridge that was not provided with guide bearings, and see the photograph at the start of Chapter 8 for an example of a heavily skewed semi-integral bridge and Figure 8.7 for design details of the accessible and replaceable guide bearings that were provided for this structure. Obviously, relatively simple and less expensive accessible guide bearings can be provided for smaller structures and structures with lesser skews. For those who use the PCI manual and who will be depending upon its directions, they are hereby warned that ignoring the differences between integral and semiintegral bridges will eventually result in the construction of some defective semiintegral bridges, especially skewed semi-integral bridges (see Chapter 8). Those defects will be extremely difficult or economically not feasible to correct after structures are completed.

Awareness of similarities Described below are examples of three different structure types where design engineers or, more properly, design engineering staffs, did not recognize or were not fully aware of the similarity between the structure movement system behavior of their structures, or structure segments, and the compressive behavior of elastic composites subjected to axial loading. The first example is the URS 422 McFarland Creek Bridge of Geauga County, Ohio. The second is the I-90 B. N. Railroad Bridge of Washington State. Surprisingly, the third example is a design example from the PCI manual, Precast/Prestressed Integral Bridges [4].

198

Integral and Semi-integral Bridges

Figure 11.6 USR 422 McFarland Creek Bridge, Geauga County, Ohio, 1986: A three-span, prestressed concrete, box-beam, integral bridge supported by new embankments and by cap-pile piers and abutments.

McFarland Creek Bridge This composite structure (see Chapter 10) is composed of a three-span, prestressed concrete, box-beam integral bridge on a slightly curved alignment with nominal span lengths of 50 ft. center-to-center of bearings, new abutment embankments on loose and compressible subsoils, and capped pile piers and abutments (Figure 11.6). To achieve suitable composite interaction between the 18 in. diameter reinforced concrete friction piles and the subsoil strata, the piles were specified to be driven at least 70 ft. to achieve the necessary bearing capacity. The behavior of the composite structure movement system was partially accounted for by the design in that the plans for this structure contained an embankment procedure note requiring a 60-day waiting period after approach embankments were placed and compacted before abutment piles could be driven. This requirement accounted for the fact that embankment-loaded subsoil strata would continue to compress vertically at a gradually decreasing rate with time, and that the piles were not permitted to be placed until most of this vertical consolidation of the embankments and subsoils had taken place. This procedure was a reasonable attempt to minimize, as much as practicable, the amount of dead load on the piles due to friction at the pile/embankment interface (draw-down forces), continual consolidation of subsoil strata, and the concomitant differential settlement of embankments with respect to the piles. There was a second type of behavior of the structure movement system of this composite structure, or a secondary component to the behavior described above, which appears to have been neglected. That behavior is similar to what would be expected of any partially confined elastic material being subjected to vertical loading. It is generally understood that when a three-dimensional homogeneous elastic material is subjected to a concentric axial load, it will be compressed vertically and extended or translated laterally (or circumferentially) with respect to the load. The denser, more laterally confined materials will be compressed the least by the same load. This type of behavior is utilized in the design of elastomeric bearings and

Chapter 11 Awareness of Reality in Bridge Design

199

compressible earthen embankments. Elastomeric bearings are reinforced with steel laminates and earthen embankments reinforced with synthetic laminates; these laminates reinforce the elastic material, thus minimizing lateral translation and, consequently, vertical compression (i.e., bearings) and vertical settlement (i.e., embankments). However, where an elastic material (subsoil strata) is subjected to an eccentric vertical load (bridge approach embankments) and the material is not uniformly and laterally confined (due to lack of the overburden at the toe of embankments), the elastic material (the upper subsoil stratum and, to a lesser and lesser extent, the lower soil strata) will tend to translate towards the lesser confined areas (laterally and longitudinally toward the toe of embankments). This was the structure movement system behavior that appears to have been neglected and that was evident at this bridge site. It should have been anticipated during the preparation of the plans for this structure. Unfortunately it was not, and the contractor chose to drive the pier piles and complete the piers before the abutment embankments were placed. Not surprisingly, after the embankments were placed, the newly surcharged subsoil stratum adjacent to the original ground surface translated laterally and longitudinally toward the toe of the embankment slopes. The pressure related to this translating subsoil stratum against the newly driven and unsupported pier piles deflected the rear-capped pile piers toward the center of the bridge. The top of these 25 ft. piers were deflected laterally about 6–9 in. as a result of this problem, the construction of the bridge had to be delayed until the earth pressure against the piles could be relieved, and the pier straightened to the extent that the bearings and the precast and prestressed beams could be placed. The present Design Regulations of the Ohio Department of Transportation (DOT) contain the following instructions for bridges similar to the one described above: 605.2 ABUTMENTS AND PIERS ON PILES IN NEW EMBANKMENTS: For abutments on piles and capped pile piers placed in new embankments the following note should be used: [23] PILE DRIVING CONSTRAINTS: Prior to driving piles, the spill through slopes and the bridge approach embankment behind abutments shall be constructed up to the level of the subgrade elevation for a minimum distance of ____ behind each abutment. The excavation for the abutment footings and the installation of the abutment and pier piles, for pier (s)____ , shall not begin until after the above required embankment has been constructed. [5]

Depending upon the density of subsoil strata and the height of embankments, a waiting period is also sometimes specified to ensure that most soil consolidation has been completed before driving of piles. Unfortunately, the above plan note from the Design Regulations, and a similar embankment procedure note on the plans for the McFarland Greek Bridge, do not and did not have any limitations for the placement of pier piles, presumably because the pier piles were not “in” the embankments. As a result of this omission, it appears clear that the Ohio Design Regulations and the designers of the McFarland Bridge were concerned only with minimizing pile “draw-down forces.” Neglected by the regulations and the designers of the McFarland Bridge was any recognition of the

200

Integral and Semi-integral Bridges

similarity between the behavior of an elastic material subjected to vertical compression, and a compressible soil stratum (and strata) subjected to the weight of new embankments. Both types of movement system behaviors should have been recognized if “draw-down forces” were to be minimized, and if subsurface lateral pressures on newly driven pier piles were to be prevented or minimized. Consequently, for those transportation agencies just beginning to consider the construction of integral bridges supported by new embankments on compressible subsoils and capped pile substructures, the pile-driving precautions used by the Ohio DOT should be considered to minimize draw-down forces on abutment and pier piles. In addition, similar precautions should also be taken to ensure that driving of pier piles in or near the toe of embankment slopes will be appropriately delayed.

USR I-90 B. N. Railroad Bridge This semi-integral bridge was located in Grant County, Washington State, 13 miles east of the city of Moses Lake, Washington (eee photograph at start of chapter). It was a three-span, prestressed concrete, girder bridge with a 40 ft. wide bridge deck supported on cap-and-column piers and reinforced-concrete abutments, founded on spread footings on new embankments. (It was constructed in 1966, one of the many early semi-integral bridges built by the Washington State Department of Transportation – WSDOT.) As the bridge spanned an abandoned railroad line, it was no longer needed to maintain traffic. Consequently, it was an ideal structure for seismic testing. Before abandoning the bridge, WSDOT, apparently concerned about the lateral earthquake resistance of this structure and others like it, funded a research project to assess the structure’s transverse stiffness and to estimate each support’s contribution to that stiffness. Presumably, WSDOT was particularly concerned about such stiffness because the bridge was a semi-integral structure without lateral support for the superstructure at abutments, apart from the resistance provided by the superstructure ends that were embedded in the earthen embankments and the lateral resistance provided by abutment bearings. Of particular interest in this bridge is the design and performance of the abutment bridge seat and bearings. The configuration of the semi-integral bridge’s superstructure at the abutments is shown in Figure 11.7 (left). This area appeared somewhat similar to a similar Ohio bridge (Figure 11.7, right) except the Ohio bridge was designed for rolled-steel beams instead of prestressed concrete I-beams. Although not the subject of this particular critique of the structure’s bearing design, the summary of the research report should be of interest to all concerned about the lateral integrity of semi-integral bridges. In part, the summary of the report states: The bridge was extremely stiff and strong. In two cycles to a load equal to 45 percent of the bridge’s weight, the maximum bridge displacement was 0.15 inch. During these cycles damage was minimal. At a load equal to 65 percent of the bridge’s weight, the pier displacement was 0.30 inch. After the bridge was excavated [Figure 11.8], the stiffness decreased to 15 percent of its initial stiffness. The stiffness further decreased to 8 percent of the initial stiffness after the superstructure had been isolated from the abutments.

Chapter 11 Awareness of Reality in Bridge Design

2'' 3''

201

e5–7 SPA e 9'–5'–3'' 2'–10''

RAILING CURB LEVEL

4'–0''

e4 – 4 SPA e 11'

e4

2''

PSC GIRDER EMBEDDED 2'' INTO ABUTMENT DIAPHRAGM e4 BEARING PADS AND EXPANDED POLYSTYRENE PEDESTAL

e4–5 SPA e1'–6'' 5'–6''

2'–6''

Figure 11.7 I-90, B. N. Railroad Bridge, Grant County, Washington State, 1966: Left: plan details for its semi-integral abutments. Right: abutment photograph of a similar Ohio bridge.

The University of Washington (UW), California Department of Transportation (CALTRANS) and WSDOT models underestimated the stiffness of the bridge in its initial state. The UW model probably overestimated the resistance of the Polystyrene at the abutments and underestimated the stiffness of the soil at the wingwalls [these were turn-back wingwalls embedded in the embankment]. The CALTRANS model was too flexible because it neglected the resistance of the bearings and Polystyrene. The WSDOT model was too flexible because it neglected the resistance of the bearing pads and Polystyrene, and underestimated the soil stiffness. [6]

With respect to the lateral stiffness of this bridge, no mention was made in the report about the presence or absence of attached approach slabs. Presumably, there were none. Had this been the case, the resistance of this structure to lateral forces would have been dramatically increased. In this respect, for new semi-integral bridges that potentially would be exposed to substantial lateral forces, it would be prudent to attach such slabs to the bridge’s superstructure with generous amounts of reinforcing steel. Concerned about the participation of polystyrene with the bearings in supporting the superstructure’s reaction at abutments, and consequently participating with the bearings in resisting lateral movement of the superstructure at the abutments, the WU researchers subjected both the existing bearing elastomer and new polystyrene to load-deflection tests. The data from those tests were used by the author to produce the stress–strain chart of Figure 11.9. By the use of this chart, the actual participation of polystyrene in supporting the abutment reaction can be determined. Based on a 40 ft. width of superstructure, a bridge seat width of 1′ 6″, a skew of 12.8 °, and a total design reaction of 180 kips, the distribution of the abutment reaction to elastomeric bearings and polystyrene form boards can be determined as follows: Area, bridge seat = (40 ft.)(1.5 ft.) (sec 12.8°) = 61.5 ft.2 = ( 61.5 ft.2 )(144 in.2 ft.2 ) = 8, 860 in.2

202

Integral and Semi-integral Bridges

Figure 11.8 Part of the superstructure of the I-90 B. N. Railroad Bridge after the embankment above the abutment was removed in preparation for lateral loading of the superstructure.

Chapter 11 Awareness of Reality in Bridge Design

203

Figure 11.9 Compressive stress–strain relationship for ⅞ in. thick, 27-year-old, elastomeric bearing pads, and new, ⅞ in. thick, expanded polystyrene form boards. Data for this relationship extracted from Eberhard [6] by the author.

Area elastomer, Ae = (10 5 8 in. )(10 5 8 in. )( 6 ) = 677.3 in.2 Area polystyrene, AP = 8860 in.2 − 677.3 in.2 = 8183 in.2 Total design reaction = 180 kips. As the deflection or strain of both the bearing elastomer and polystyrene due to a reaction of the 180 kips would be the same, using the stress–strain chart of Figure 11.9, the strain associated with the 180 kips reaction can be determined as follows: assume a strain; determine the associated pressures from the stress–strain chart of Figure 11.9; multiply these pressures by the associated areas computed above to determine the component loads; and then total the component loads and compare the total with the known 180 kip reaction. By a cut-and-try process, a strain of about 0.028 can be determined. Based on this strain, the load participation between the two bridge seat materials will have been determined. For example: Elastomer: Strain = 0.028 ( Figure 11.9 )

Stress = 68 psi Load = ( 68 psi ) ( 677.3 in.2 ) 1000 = 47.4 kips. Polystyrene: Strain = 0.028 ( Figure 11.9 )

Stress = 16 psi

204

Integral and Semi-integral Bridges

Load = (16 psi ) ( 8183 in.2 ) 1, 000 = 130.9 kips. Total Reaction: 47.4 kips + 130.9 kips = 178.3, say 180 kips. The original bearing design reaction was 180 kips/6 bearings = 30 kips per bearing. However, based on the participation of polystyrene, the beam-bearing reaction was 47.4 kips/6 bearings = 7.9 kips. Consequently even though polystyrene was considerably softer than the elastomeric bearings, the significantly greater area of polystyrene caused the polystyrene to reduce the bearing reactions from 30 kips to just 7.9 kips, a reduction of almost 74 percent. With respect to bridge-bearing design, the neglect of polystyrene represents a significant error in design assumptions. These computations illustrate the significant effect that expanded polystyrene can have on the bearing reaction, and consequently on the performance of the bearings in providing for longitudinal movement of the superstructure. Considering the amount of friction developed at the polystyrene/concrete interface, polystyrene would have a detrimental effect on the longitudinal movement of the superstructure. However, due to the skew of the structure, this friction would provide lateral resistance and thus help to prevent the lateral rotation of the superstructure (see Chapter 8). Polystyrene was beneficial in another respect: it prevented abutment backfill from penetrating the bridge seat and thus avoided the need to provide some other form of joint seal to serve this purpose. WSDOT eventually solved this problem in another way, however. They moved the superstructure end diaphragm to behind the abutment thereby eliminating the need for polystyrene form boards. They also projected the end diaphragm down and below the bridge seat, thereby providing a barrier between the backfill and the bridge seat. This design improvement is illustrated in Figure 9.7c,d. Of particular interest for this chapter is the fact that polystyrene form boards in the abutment bridge seat were presumably specified not only to serve as bottom form boards for the superstructure’s cast-in-place concrete end diaphragms, but also to function as joint seals to prevent backfill from penetrating the bridge seat and fouling bridge movement at the bearings. They were neglected in the design of the bearings. Apparently, the designers failed to recognize the compression deformational similarity of bearing elastomers and polystyrene form boards. This is not surprising when one considers that the structure was constructed at a time when bearing elastomers were first being introduced to the bridge engineering profession, and design engineers were understandably not familiar with the behavior of elastomeric type materials. Their unfamiliarity with expanded polystyrene was even greater. The author can make these positive statements about the general unfamiliarity with elastomers and polystyrene because he was personally involved with this elastomeric bearing introductory period. As a member of the Ohio Highway Department’s Bureau of Bridges, he worked closely with both sales and research engineers of the General Tire and Rubber Company, not only to introduce elastomeric bearings into Ohio bridge design standards, regulations and construction specifications, but also to assist these engineers in the development of the first elastomeric bearing design specifications for consideration by the bridge committee of the American Association of State Highway and Transportation Officials (AASHTO). Conse-

Chapter 11 Awareness of Reality in Bridge Design

205

quently, although he was deeply immersed in the subject of elastomeric behavior, performance and durability, he was also aware of his own ignorance and uncertainty and those of his colleagues concerning the different types of elastomers and how they responded to compression, shear, sustained loads, low temperatures, etc. Consequently, it was not at all unusual at the time for other early designers of semiintegral bridges to also be unfamiliar with the materials that they were specifying on the plans for these structures. However, the successful performance and longterm durability of these semi-integral bridges attest to the actual practical success of these early design engineers.

2001 PCI precast/prestressed integral bridges With respect to the integral abutment concept illustrated in Figure 4.3 on page 13 of the manual (Figure 11.10), the following statement is given on page 12 of the manual: … In the semi-integral concept, the transfer of rotational displacement to the piles is minimized. Rotation is generally accomplished by using a flexible bearing surface at a selected horizontal interface in the abutment backwall. Allowing rotation at the pile top generally reduces pile loads. Figure 4.3 shows a typical semi-integral abutment. [4] (emphasis added by the author)

Figure 11.10 Abutment concept for prestressed concrete, box-beam, integral bridges. This figure is a close copy of Figure 4.3 that appears on page 13 of the 2005 PCI manual Precast/Prestressed Integral Bridges.

206

Integral and Semi-integral Bridges

First, the designation of the Figure 4.3 abutment concept (Figure 11.10) as a semi-integral abutment is a mistake in that the illustrated abutment is an integral abutment (see Appendix 2 and Chapters 1 and 9). The Figure 4.3 abutment is fully integrated with the superstructure and the single row of flexible piles. Although there is an elastomeric erection strip providing temporary support for the precast beams, it becomes merely a nonfunctioning construction artifact after the remainder of the abutment concrete has been placed and allowed to solidify. Incidentally, the integral abutment concept illustrated in the PCI manual’s Figure 4.3 is for prestressed, side-by-side, box-beam bridges. The abutment design for prestressed I-beam bridges is similar except that the face of that abutment and its reinforcing steel would extend straight up to the deck slab. For this latter abutment, the ends of the beams and the elastomeric bearings would be totally encased within the abutment and, for this abutment concept, the elastomeric bearings would also eventually become merely nonfunctioning construction artifacts as described below. In composing the above manual statement about the purpose of elastomeric erection strips (or elastomeric erection bearings), the manual author was apparently rationalizing about their purpose rather than intuitively recognizing the similarity between the behavior of integral abutments with elastomeric erection devices and that of those without them. The behavior in both cases would essentially be the same. The validity of this statement is substantiated by the discussion and brief computation presented below. To estimate the effect of an elastomeric erection strip, or of an elastomeric bearing, on the behavior of a steel-reinforced concrete member, the elastic moduli of the three different materials is needed. For the more commonly used materials, the elastic modulus of steel (Es) and concrete (Ec) may conservatively be assumed to be equal to 30 × 106 psi and 3.6 × 106 psi, respectively. However, as the elastic moduli of elastomers (Ee) vary with the amount of stress to which they are subjected, a typical stress must first be computed for an elastomer under the application condition being considered. For this purpose, the bearing pressure applied to an elastomeric erection strip by the weight of the prestressed box-beams of an assumed 50 ft. bridge span will first be computed. Using 48 in. × 27 in. box-beams, the abutment reaction conservatively equals about 20 kips/beam, or about 5 kips/ft. on 4 in. erection strips. Compressive stress: = 5.0 kips 48 in.2 = 0.104 kips in.2 = 104 psi Shape factor: = (12 in.)(4 in.) (24 in. + 8 in.)(1 in.) = 48 in. 32 in. = 1.5. Using a pressure of 104 psi, a shape factor of 1.5, and a typical hardness of 60, the load deflection chart of Figure 14.6.5.3.3-1 of the AASHTO Bridge Design Specifications [7] indicates a deflection of 4 percent, or a strain of 0.04 for a 1 in. thick strip. Since E = stress strain, Ee =104 psi 0.04 = 2.6 × 103 psi. As a check of this computation, reference can be made to the stress/strain chart of Figure 11.9. With an assumed strain of 0.04, the stress given by the chart for a

Chapter 11 Awareness of Reality in Bridge Design

207

64 hardness elastomer is about 103 psi. This stress agrees very closely with the stress (i.e., 104 psi) used above to determine the 2.6 × 103 psi modulus computed using the AASHO shape factor chart. A typical cross-section and a transformed cross-section for a 1 ft. width of the PCI abutment of Figure 4.3 in the manual is shown in Figure 11.11. A brief comparison of the transformed areas of the various components of the transformed cross-section will show the insignificance of the elastomeric device with respect to the rest of the cross-section. Steel: Es = 30 × 106 psi Concrete: Ec = 3.6 × 106 psi Elastomer: Ee = 2.6 × 103 psi Transformed areas: Concrete: Ac′ = Ac = (26 in.)(12 in.) = 312 in.2 Steel: As′ = As [ Es Ec − 1] = 0.66 in.2 [8.33 − 1] = 4.84 in.2 Elastomer: Ae′ = Ae [ Ee Ec ] = 48 in.2 [2.6 × 103 psi 3.6 × 106 psi ] = 0.035 in.2 Total transformed areas: At′ = Ac′ + As′ + Ae′ At′ = 312 in.2 + 2 (4.84 in.2 ) + 0.035 in.2 = 321.7 in.2 Ratio of Ae′ to At′ : Ae′ At′ = 0.035 in.2 321.7 in.2 = 0.0001, or Ae′ = 0.01% of At′ or 1 100% of At′. Based on these computations, it is clear that the presence of an elastomeric device in an abutment cross-section subjected to axial compression is structurally inconsequential. An analysis of the same cross-section subjected to flexure reveals the same insignificance. With respect to the transformed section of Figure 11.11, the inclusion of an elastomeric device changes the location of the neutral axis by only 0.0016 in. These computations do not take into account the relaxation or realignment of the molecular structure of an elastomer subjected to a sustained deflection. As illustrated, for all practical structural purposes, the effect of temporary elastomeric strips or elastomeric bearings in a solidified concrete abutment cross-section is negligible. Consequently, the suggestion in the PCI manual that these elastomeric devices change an integral abutment to a semi-integral abutment is baseless. Hopefully, before many of these manuals are circulated, these erroneous remarks will be removed, and the misleading illustrations regarding semi-integral bridges appropriately revised. The author also suggests that designers should resist using the pier-stem stiffness equations presented on pages 27 and 28 of the PCI manual until they are satisfied

208

Integral and Semi-integral Bridges

Figure 11.11 Horizontal cross-sections at the bridge seat construction joint of the PCI Figure 4.3 integral abutment concept. Left: a typical cross-section; right: a transformed cross-section.

with the assumptions upon which they are based and the use to which they are to be put.

Summary It is as well to recognize that the lack of awareness exhibited by the design engineers whose design efforts have been selected for discussion in this chapter were generally not working alone. Typically, these engineers work together with their colleagues, especially those who checked their work, and their engineer mentors who had the responsibility of reviewing their work. Consequently, the lack of awareness described above was shared by members of design staffs and not merely by lone novice engineers working with no professional guidance. These observations with respect to novice design engineers suggest that there is a vital component missing in their intellectual ability, a component that is needed if these engineers are to grow and eventually assume the responsibility for the safety and durability of the nation’s bridges. That component is the habit of being acutely aware of the characteristics of real structures, and of becoming acutely aware of various aspects of real structures during the various phases of construction that will be encountered by construction contractors. With respect to the more experienced engineers who were responsible for the review of the work of these novices, these observations suggest that they did not, as Hardy Cross suggested, jerk their novice’s attention away from the plans, away from the computations, and force them instead to become aware of and to contemplate real structures. Most of the design examples selected and described in this chapter (they represent only a very small portion of the type and number of design mistakes that have come or been brought to the author’s attention over the last two decades) were selected for presentation here because they seemed to best represent some of the most blatant examples of the kind of design mistakes that apparently are not untypical of certain bridge engineering staffs. They suggest that the engineers responsible for these mistakes had a deficient awareness of reality or in the words of Sargent Snorkel, they generally failed to “look out the window.”

Chapter 11 Awareness of Reality in Bridge Design

209

There appears to be many contributing reasons for this lack of awareness including economics, preoccupation, habit, language, age, etc. For some bridge design projects, economics appears to play a leading role.

Economics With respect to bridge rehabilitation design assignments or bridge rehabilitation design contracts, where it becomes evident that design engineers had not visited and examined the bridges that they were assigned to rehabilitate, it appears probable that they or their supervisors were less concerned with the actual condition of the bridges that they were assigned to rehabilitate, and more concerned with spending as few person-hours as possible on their design assignments either to impress their supervisors or to maximize contract profits. Such engineers have lost sight of their professional responsibilities, their moral obligations, and their personal reputations.

Preoccupation Preoccupation with other obligations and responsibilities, and with unresolved structural problems, has a significant effect on awareness, especially on the awareness of novice engineers. In this respect, it is appropriate to recognize that most of the errors that have been described above were probably made by recent engineering graduates with little or no experience with typical analytical procedures, agency design standards, electronic analysis programs, shop fabrication and casting practices, or field construction practices and procedures. They were just beginning to familiarize themselves with detailed drawings, computer drafting, and bridge plan preparation procedures. Consequently, most were concerned with learning the mechanics of design and plan preparation. This concentrated attention on office practices, procedures, and methods left little time or attention to consider real structures, structures that were so remote from the design office that they sometimes completely passed from memory. As described above, bridge design at this stage of the young engineer’s life is essentially a paper preparation process, and many of the errors that were made appear to owe their existence to the fact that the novice engineer’s attention was otherwise preoccupied.

Habit Although most engineers have learned and intellectually accepted the concept of reality as process, many seem to relate or limit such knowledge to the submicroscopic levels of reality, those levels where they have the least contact and experience. At the macroscopic level where they do their living and where structural changes are so gradual as to be almost imperceptible, these same engineers appear to habitually ignore processes. Differences go unnoticed or are neglected. They seem oblivious to change. They observe noticeable events (things, i.e., structures) as permanent immovable fixtures of the environment. So, although they have accepted the concept

210

Integral and Semi-integral Bridges

of reality as process, they have not altered the basic structure of their understanding to use this orientation in their perceptions and evaluations.

Language The anthropologist and linguist Edward Sapir has written: Human beings do not live in the objective world alone, nor in the world of social activity as ordinarily understood, but very much at the mercy of the particular language which has become the medium of expression for their society. It is quite an illusion to imagine that one adjusts to reality essentially without the use of language and that language is merely an incidental means of solving specific problems of communication or reflection. The fact of the matter is that the “real world” is to a large extent unconsciously built up on the language habits of the group. … We see and hear and otherwise experience very largely as we do because the language habits of our community predisposes certain choices of interpretation. [8]

For example, in all of grade school and on into high school, children are taught and reminded that the principal parts of English grammar are the verbs (words that denote action, i.e., run, jump, move, grow, etc.) and, by contrast, the nouns (words that do not denote action, i.e., places, things, etc.). As they continue to learn and think about the world of words as either verbs or nouns, and what they denote, they gradually become aware of the ethereal-type nouns, nouns that denote “things” that have no substance and no action (i.e., vacuum, void, space, shadow, shade, night, etc.), of apparition-type nouns, nouns that denote “things” that appear to be all action but no substance (i.e., time, gravity, day, present, my, mine, ghosts, etc.), and of process-type nouns, nouns that denote “things” that are substance in readily perceivable action (i.e., lightening, fire, heat, rain, steam, etc.). With these exceptions, they come to recognize the great multitude of static-type nouns, nouns that denote “things” that appear to be all substance and no action (i.e., buildings, bridges, mountains, lakes, etc.). This remaining category of nouns must be the “true” nouns, the nouns that grammar intended to be classified as such. However, as students age and gradually become more aware, they learn that even the “things” denoted by the nouns of this latter category are substances that are gradually changing, a type of action that may not be instantly perceivable, but action nevertheless. At this stage of learning, some fortunately become aware that all worldly “things” are in the process of change, including their own bodies, thoughts, memories, and emotions. As a result of such continuous changes, they finally recognize that the words that denote these “things” should all more correctly be classified as verbs and not nouns. Grammar has deceived them. It is only by casual circumstances that some become aware of these verbal and mental errors. In becoming aware that all is process and change, the true nature of the reality in which they do their living, they also learn that the language labels that they use for these nouns also tend to obscure differences and similarities and obfuscate change. As another example of language’s thought-numbing power, consider the singular noun that is used to designate “Pittsburgh,” the city that is located at the junction of Allegheny, Monongahela, and Ohio Rivers in the southwestern corner of

Chapter 11 Awareness of Reality in Bridge Design

211

Pennsylvania. Innumerable changes have taken place in this city since the time, in 1754, when Indians standing on the high bluff across the Monongahela River could observe Captain William Trent of the Virginia Militia and his frontiersmen cohorts building a timber palisade fort at the junction of the three rivers (Figure 11.12a).

(a)

(b)

(c)

(d)

(e)

Figure 11.12 Pittsburgh, Pennsylvania: (a) 1754; (b) 1890; (c) 1952; (d) 1966; (e) 2006. The city continues to change.

212

Integral and Semi-integral Bridges

From that same vantage point in the late 1930s and early 1940s, the author can remember the pungent odor of the atmosphere, the night-time glow from blast furnaces reflected on the cloud cover overhead, and the sight and sound of steamboats as they maneuvered strings of barges up and down the rivers (Figure 11.12b). Now, at the turn of the Millennium, not only are most of the blast furnaces and all of the steamboats gone, but also many of the major bridges and almost all of the buildings in the point area between the rivers have since been removed and replaced (Figure 11.12c, d, and e). When examining these five images of the city of “Pittsburgh,” it should be remembered that every generation of Pittsburghers actually experience a significantly different city. Yet, each and every one of them calls it by the same name. Can any other example better illustrate the obfuscating effect that common language labels have on the awareness and consciousness of those destined to live their lives in language’s pervasive shadow? Perhaps, instead of the name “Pittsburgh,” the plural form “Pittsburghs,” or the verbal form “Pittsburghing,” or the indexing form “Pittsburgh1929,” were used, all would be continually reminded of the ever-changing dynamic nature of this city, and be prompted to become aware of the continual changes that are taking place all around them all of the time. With language conventions like this and the fact that structures change so gradually, is it any wonder that most individuals habitually ignore the process nature of reality? In professional work, when time is actually taken to examine an existing structure, many engineers are insensitive to differences, and thus are unaware of changes that have taken place in the structure since it was first constructed. Rarer still are the individuals who are sensitive to the changes that are taking place as they view the structure (i.e., crack propagation, material corrosion and deterioration, etc.). For example, during the recabling of the General US Grant Suspension Bridge of Portsmouth, Ohio, daily examinations of the existing and partially exposed corroded cables were made. This examination revealed that as many as four to six of the lower cable wires were breaking each week. Previous examinations of the cables of this structure documented the fact that this wire breakage, which was probably occurring at a lesser rate, had been going on undetected for several years. Rarely is awareness so critical. Nevertheless, many of the major bridge tragedies that most are familiar with have occurred because the engineers in responsible charge were not sufficiently aware of the changes that were occurring, or did not appreciate the significance of those changes. In addition to economics, habit, preoccupation with other responsibilities and interests, and an uncritical use of language, there are probably many other reasons for a general lack of awareness. These include a person’s youth, early indoctrination to the imaginary (and its persistent emphasis in our society), with little or no indoctrination to the real, lack of visualizing skills, physical limitations, etc. To fully understand the physical and intellectual limitations that novice or apprentice engineers bring into the profession, one should have the benefit of well-conceived research. Research The errors described in this chapter suggest that the education and professional training of bridge design engineers and possibly all civil engineers should be

Chapter 11 Awareness of Reality in Bridge Design

213

improved to heighten their awareness of reality. To accomplish this effectively, students and educational methods should be subjected to formal research to establish the best methods to accomplish this goal. As transportation departments routinely dedicate portions of their funding to support both basic and applied research, what could be more important to the future of transportation facilities than to more properly educate and train its planers and designers?

Epilogue The Silver Bridge of Gallipolis, Ohio, the Mianus River Bridge of Greenwich, Connecticut, the Schoharie Creek Bridge of New York State, and the USR I-35 W Mississippi River Bridge of Minneapolis, Minnesota, collapsed after years of gradual changes, changes in most cases due primarily to the corrosion of critical structural components (the Schoharie Creek bridge collapsed due to scour of the creek bed at one of its piers). Not reported in the news media were the many hundreds of other bridges where structural changes were noticed and corrected before they progressed far enough to become a danger to the traveling public. Structure changes fill the spectrum from the slow and imperceptible to the rapid and quite visible. The observant engineers are those who have become sensitive to these changes and who have prepared themselves to notice similarities and differences. However, the truly effective engineers are not only observant and notice similarities and differences, but also the ones who are capable of properly evaluating the significance of structural changes, and have the self-confidence to appropriately respond to adverse structural changes to protect not only the integrity of structures, but also, and most importantly, the safety of the traveling public. When all is said and done, being a good looker may be the most important quality a scientist, – or a doctor – can possess. K. C. Cole The author would like to add to K. C. Cole’s observation: or a bridge design engineer.

References 1. Burke, Martin P. Jr., “Bridge Aesthetics Bibliography,” Bridge Aesthetics Around the World, Transportation Research Board of the National Academies, Washington, D.C., 1991, pp. 241–263. 2. Johnson, B. G., “An Awareness of Reality,” Civil Engineering, American Society of Civil Engineers, Reston, Virginia, 1965, pp. 58, 59. 3. Walker, Mort, “Beatle Bailey,” The Columbus Dispatch, King Features Syndicate, Inc., May 12, 1990. 4. Precast/Prestressed Concrete Institute, Precast/Prestressed Integral Bridges, PCI, Chicago, Illinois, 2005. 5. Ohio Department of Transportation, Bridge Design Regulations, Ohio DOT, Columbus, Ohio, 2007, p. 6–14.

214

Integral and Semi-integral Bridges

6. Eberhard, M. O., et al., Lateral Load Response of a Reinforced Concrete Bridge, Washington State Transportation Center, University of Washington, Seattle, Washington, 1993. 7. American Associations of State Highway and Transportation Officials, Standard Specifications for Highway Bridges, 17th edn, American Associations of State Highway and Transportation Officials, Washington. D.C., 2007, p. 348.6. 8. Sapir, E., “The Status of Linguistic as a Science,” Culture, Language and Personality, University of California Press, Berkeley, California, 1958, p. 69. 9. Cole, K. C., Mind Over Matter: Conversation with the Cosmos, Harcourt, New York, 2003, p. 36.

Appendix 1

The Pavement Growth/Pressure Phenomenon: The Neglected Aspect of Jointed Pavement Behavior

Acceptance is not a state of passivity or inaction. … It is, in fact, the first step to successful action. If you don’t fully accept a situation precisely the way it is, you will have difficulty changing it. Moreover, if you don’t accept the situation, you will never really know if the situation should be changed. Peter McWilliams I cannot say whether things will get better if they change, what I am saying is they must change if they are to get better. Georg Christoph Lichtenberg

Introduction The original purpose behind the development of the documentation, illustrations, and analyses of this appendix was to provide this author with suitable background 215

216

Appendix 1

to discuss and critique the jointed concrete pavement maintenance recommendations emanating from a Midwest State Department of Transportation, recommendations that appeared to suggest that their proponents were oblivious of the pavement growth/pressure (G/P) phenomenon and its destructive potential with respect to both pavements and bridges [1, 2]. However, after reviewing other contemporary reports and papers on jointed pavement growth, pressure, distress, blowups, etc., it became apparent at that time that many other pavement maintenance engineers and research specialists were also oblivious, or underestimated the significance, of the pavement G/P phenomenon. This is the phenomenon responsible for most of the serious pavement and highway bridge damage that had taken place throughout most of the last century. As a result, it was this author’s opinion that the advice and recommendations given by these authorities would not only not ameliorate the stressed condition of the nation’s pavements and bridges, but their recommendations actually had the potential to make bad situations even worse. Specifically, their recommendations would probably result in continued widespread pavement pressures, higher pressures, pavement distress, bridge damage, restricted traffic flow during emergency repairs, and vehicular accidents and personal injuries that usually accompanied such restrictions. This bleak future could be changed, but it would require many individuals to re-examine some of their basic assumptions with respect to pavement pressure generation, pressure magnitudes, and the detrimental effect that these potentially huge pressures could have on the integrity of pavements and abutting bridges. Consequently, the original focus of that study was then changed to provide a brief discussion and explanation of the pavement G/P phenomenon to help motivate this re-examination to the extent that future pavement research, subsequent design details, and pavement maintenance practices would be changed to control this critical aspect of jointed pavement behavior.

Background To clarify the evaluations and comments that follow, it is first necessary to present and briefly explain the G/P phenomenon, the perspective from which these comments originated. Background for this explanation comes from an extended examination and personal study of pavements and bridges located throughout Ohio, which at that time were adversely affected by the response of jointed rigid pavements and abutting bridges to cyclic environmental variables [3]. This subject was approached a second time where the effects of the G/P phenomenon on movable deck joints, approach slabs, and integral bridges were discussed and illustrated [4]. In most cases, pavements abutted multiple-span continuous bridges with movable deck joints at the superstructures/abutment interface. In many cases, they abutted multiple-span integral bridges. Response of such pavements to environmental variables resulted in the generation of longitudinal pavement stresses and pressures great enough to fracture and shatter bridge approach pavements and approach slabs, and to seriously impair the integrity of abutting bridges. The potentially huge magnitude of these stresses and pressures has been partially verified by pavement strain measurements using rock mechanics techniques [5]. However, the explanations

Appendix 1

217

given below should still be considered somewhat speculative because the strain measurements noted above, to the author’s knowledge, have not been replicated elsewhere. In addition, accelerated aging tests with pavement simulations have not been done, and this author is not aware of any researcher who has attempted to monitor pavement stress generation in aged pavements. As the manifestations of huge pavement stresses have been observed by countless individuals for over a century, this author finds it almost incredible that pavement stress generation research has not been done. In addition, apparently such research is probably not even being contemplated by federal, state, or local transportation research organizations. The fact that comprehensive controlled research has not been done is probably the primary reason why the G/P phenomenon is either ignored entirely or misunderstood by many pavement practitioners. However, there are other practitioners who are aware that this phenomenon is one of the most destructive phenomena known to transportation professionals. They are the individuals who are most concerned that current difficulties (the limited funds available to achieve optimum pavement maintenance) are going to be made worse by the abandonment of current, empirically evolved, joint maintenance practices. It is also worth noting here that the tremendous forces associated with this phenomenon, probably second only to the effects of earthquakes, are now not even recognized by the bridge engineering profession, to the extent that pressures transmitted through bridge approach pavements and approach slabs are not mentioned in American Association of State Highway and Transportation Officials’ (AASHTO’s) Standard Specifications for Highway Bridges [6]. Consequently, it almost appears that there is a tacit agreement among transportation professionals that the emperor is well dressed, while a few observant and pragmatic practitioners can see that he is striding about stark naked. Structures For most structures, applied forces (F) and internal unit stresses (f ) are usually sufficient to describe the response of structures to applied loads. However, for pavements, the terms “stress” (f ) and “pressure” (p) are both in common use, and are used somewhat interchangeably. So in keeping with this practice, both terms are used in this appendix to describe the response of pavements (segments and joints) to total restraint forces, with the terms “stress” or “pressure” related to the effect of restraint forces on segments or joints, respectively. Assuming straight pavements, uniform cross-section areas (A) and uniform distribution of temperature, moisture, and joint debris: f = p = F/A. Analogy is a useful method for forming a conception of pavement pressures and pressure distribution. For example, in Figure A1.1 three different structures are illustrated – a, b, and c – with similar force, stress, and pressure characteristics. Figure A1.1a is a structure composed of a series of identical rigid blocks bound together by an elastic tie (i.e., wood blocks with a rubber band). Figure A1.1b is a structure composed of a series of identical rigid blocks bound together by stretched internal post-tensioning bars or strands (i.e., concrete segments of a post-tensioned bridge superstructure). Figure A1.1c is a structure composed of a series of identical

218

Appendix 1

Figure A1.1 Three different structures with similar stress or pressure characteristics.

masonry blocks or slabs confined between immovable end restraints (i.e., a short length of about 2,000 ft. [610 m] of jointed rigid pavement). In Figure A1.1a and b, the elastic band, bars, or strands binding or pulling the blocks of the structures together with a compressive force (F) create compressive unit stresses throughout the structures’ cross-sections and throughout the structures. Similarly, in Figure A1.1c, immovable end restraints prevent the blocks or slabs of the structure from expanding to lengths commensurate with high temperature and moisture levels. In effect, they keep the structure’s length constant during increasing temperature and moisture levels, thereby inducing proportionally increasing compressive unit stresses throughout the structure’s cross-section and throughout the structure. They also create compressive unit pressures (p) on pavement segments as shown in detail A, Figure A1.1. Although the three structures of Figure A1.1 have similar force, stress, and pressure characteristics, there are differences in the way that they respond to rising temperature and moisture levels. For example, increased moisture levels would cause only the blocks of structure Figure A1.1a to swell, thereby lengthening the structure, stretching the band, and increasing the confining forces and internal stresses. With respect to increases in temperature, as the thermal coefficient of elastomer is greater than that of wood, the band would become (with higher temperatures) longer than the blocks, thereby loosening the band and diminishing the restraint forces and internal stresses. For the structure illustrated in Figure A1.1b, an increase in moisture levels would cause only the segments to swell, thereby lengthening the structure, stretching the bars or strands, increasing their tension (and total restraint forces), and conse-

Appendix 1

219

quently increasing segment internal compressive stresses. With respect to temperature increases, restraint forces and internal compressive stresses would be for practical purposes unchanged because both segments and bars or strands would have essentially the same elongation response to temperature increases. With respect to the pavement type structure illustrated in Figure A1.1c, increases in both temperature and moisture would tend to increase the structure’s length. By being confined between immovable end restraints, however, the potentially longer structure is kept compressed, resulting in proportionally higher compressive forces, segment stresses, and joint pressures with increasing temperature and moisture levels.

Pavement G/P phenomenon Pavement pressure If design had to contend only with the response of structures (pavement and bridges) to ambient temperature ranges, the achievement of an efficient and functional transportation system would be relatively easy. However, concrete shrinkage and the less than ideal behavior of pavement contraction joints responding to environmental variables, vehicular traffic, and customary traffic maintenance practices have compounded the problems of pavement and bridge maintenance engineers. With respect to an extensive length (≥2000 ft. [610 m]) of jointed concrete pavement similar to the hypothetical structure illustrated in Figure A1.1c and a typical contraction joint as illustrated in Figure A1.2, the following brief and simplified explanation should help to illustrate how the G/P phenomenon generates such high compression stresses throughout a pavement’s cross-section and throughout extensive lengths of such pavements. These brief computations should clarify and help to substantiate the huge problems that pavement and bridge maintenance engineers must contend with when their transportation departments do not make it a practice to monitor pavement stress levels and implement pressure-relief practices to protect pavements, bridges, and vehicular traffic from the consequences of uncontrolled pavement pressure generation. For example, consider an extensive length of jointed concrete pavement similar to the structure illustrated in Figure A1.1c. Pavement segments between the saw-cut contraction joints contract and shrink in response to lowering temperature and moisture levels, thereby causing contraction joints to crack below the saw-cut surface and open. Conversely, these joints and cracks become narrower in response to raising temperature and moisture levels. However, after joints open and remain open at low temperatures and moisture levels, compression-resistant fine roadway debris infiltrates the cracks below the saw-cut surfaces. This infiltration prevents complete closure of the crack to an as-constructed condition. Joint opening, infiltration of debris, and partial closing continue in sequence with daily and yearly temperature and moisture cycles. Of course, debris infiltration is facilitated where de-icing chemicals are used to ensure dry pavements (and, consequently, open contraction joints) during low winter temperatures. Complete closure is prevented due to compressive-resistant debris that infiltrates joints while they are open.

220

Appendix 1

Figure A1.2

Cyclic movement at pavement contraction joints.

Continuous cyclic infiltration and compression of debris and the subsequent restrained elongation of pavement segments generate longitudinal compressive stresses and pressures that eventually culminate in either progressive minor fracturing of pavement and bridge abutments or violent rupturing of either pavement or bridge members. Such temperature- and moisture-driven cyclic infiltration and compaction of joint debris induces longitudinally oriented pavement stresses and pressures well in excess of 1,000 psi (7 MPa). Such stress and pressure levels have been documented by analytical means, field measurements, and as manifested by progressive pavement fracturing, ruptured pavements (blow-ups), closed movable deck joints, and fractured bridges [3, pp. 54–60]. A brief elaboration should help to clarify the pavement G/P phenomenon. For example, Figure A1.2 illustrates the cyclic movement and debris infiltration that occur at a typical pavement contraction joint. As noted above, this movement is caused by the response of the pavement to changes in the pavement’s moisture content and temperature. In drying, following a wet curing period, concrete shrinks a maximum of about 0.0005 of its length. This is the well-established average value of the total free shrinkage, namely from a saturated to a dry condition. Most of the shrinkage can be recovered by a thorough re-wetting. As concrete pavement in contact with a subgrade probably retains a substantial amount of moisture, a coefficient of 0.0003 may be used to represent the initial shrinkage of the average pavement. After being cast, concrete pavement responding to a loss of moisture, cooling after the heat of hydration, and a lowering of the ambient temperature tends to

Appendix 1

221

shorten. This shortening is resisted by the tensile strength of the concrete. Ultimately, the tensile strength is exceeded and the pavement cracks at pre-sawed contraction joints. Ignoring the effects of concrete hydration, pavement reinforcement, subgrade friction, etc., the initial cracking due to shrinkage, as illustrated in Figure A1.2a, may be assumed to be equal to about 0.0003 Ls, where Ls equals the length of a pavement section between contraction joints. Responding to ambient temperature changes, the initial shrinkage crack opens wider at temperatures lower than normal (Figure A1.2b) and closes at temperatures higher than normal (Figure A1.2c). With daily fluctuations in temperature, and magnification of these fluctuations due to seasonal temperature extremes, this movement at the contraction joint continues to cycle, as illustrated in Figure A1.2b,c. Infiltration of these joints by compressive-resistant debris begins almost as soon as the joints become cracked. Unsealed joints are infiltrated from the top, sides, and bottoms. For “sealed” joints (actually a misnomer because only the upper surfaces of such joints are sealed), initial infiltration begins at the open ends and bottoms. This infiltration is facilitated by the movement of water, which penetrates pavement and shoulder joints from above, and groundwater, which seeps through shoulders and migrates along the subbase below. As joint seals begin to fail because of a combination of age degradation, low temperature stiffening, traffic abrasion, neglect, etc., debris infiltration accelerates from both above and below. Owing to the presence of debris, cyclic movement at contraction joints is restrained by compression of debris and by the restrained expansion (compression) of the pavement, Δcp (compare Figure A1.2d with A1.2c. As stress is proportional to strain ( fc = Ec ε), the stress induced by this restrained expansion (compression) can be estimated by assuming a value for the strain associated with the condition illustrated in Figure A1.2d. By assuming Δcp of Figure A1.2d to be about equal to the original pavement shrinkage (Δs of Figure A1.2a equal to 0.0003 Ls), the unit strain, ε = 0.0003. With the weight of concrete, Wc, equal to 145 lb/ft.3 (23.20 kg/m3), the 28-day cylinder strength of concrete, fc, equal to 4,000 psi (27.57 MPa), and the unit strain, ε, equal to 0.0003, the concrete compression stress, fc, equals about 1,000 psi (7.00 MPa). English units:

Metric units:

fc = Ecε

fc = Ecε

Ec = W (33) ( f c′)

Ec = Wc1.5 (33) ( f c′)1 2

Ec = 1451.5(33) (4,000)1/2

Ec = (23201.5)(0.040) (27.57)1/2

Ec = 3.64 × 106

Ec = 23,470

fc = (3.64 × 106) (0.0003)

fc = (23,470) (0.0003)

fc ≈ 1,000 psi

fc ≈ 7.0 MPa

1.5 c

12

This is the stress associated with a pavement compression Δcp about equal to the original shrinkage crack width, Δs. Obviously, other assumptions will yield other stresses, but any reasonable assumptions will yield stresses of similar magnitudes.

222

Appendix 1

Figure A1.3 Hypothesized stress or pressure generation curves for jointed rigid pavement (brick, stone block, or jointed concrete).

Using rock mechanics techniques, stresses of these magnitudes were documented in Ohio pavements almost 30 years ago [5]. The procedure used at that time for measuring pavement strain consisted essentially of drilling a 1½ in. (38 mm) diameter hole in pavement suspected of being compressed and bonding strain gauges to concave surfaces of the hole. At the same location, the pavement was then overcored with the hole located at the center of the core. After the core was removed, changes in the strain gauges mounted within the core indicated that the magnitude of stresses that were compressing the core before its removal. Of 13 locations sampled in various Ohio counties, 3 cores indicated stresses in excess of 900 psi (6.2 MPa). Two of these three cores were removed from pavement on bridge approaches. Other cores removed indicated a complete spectrum of stresses from about 70 psi (0.48 MPa) up to and including a stress of 1,064 psi (7.34 MPa). The generation of such stresses may be visualized as suggested in Figure A1.3, which shows an “idealized” chart (curve a) of the yearly maximum longitudinally oriented compressive stress, fc, in an extensive length of restrained pavement without movable joints (i.e., no pavement expansion or relief joints or movable bridge deck joints). Initially, the stress or pressure is insignificant because joints are relatively clean and joint seals are intact and functioning. However, as the years pass and joints begin to fill with debris, the yearly maximum pressure increases at a growing rate. As joints continue to fill, compression-resistant debris functions to minimize infiltration of additional material, slowing the rate of joint infiltration and pressure generation. Somewhere along this hypothesized pressure generation curve, the pavement will fracture adjacent to a joint, relieving some of the pressure (Figure A1.4, foreground); if a pressure relief joint is not installed, the pavement will eventually blow up, relieving all the pressure at the location of the blow-up (Figure A1.4, background). If a rigid full-depth repair were made after a blow-up, pavement pressure generation would start again and continue at a relatively fast rate (Figure A1.3, curve c) to the extent that pavement distress, fractures, or blow-ups should be expected

Appendix 1

223

Figure A1.4 Blow-up of the M11 pavement at Harlow, Essex, UK, 1993: As evidenced by the presence of the large asphalt patch adjacent to the buckled pavement, this pavement experienced prior blow-up fractures requiring the earlier asphalt repair (Figure A1.3, curves a and c).

within subsequent years unless significant changes were made in joint maintenance practices. Figure A1.3 is an idealized, yearly, maximum pressure generation curve (a) for one particular pavement. Owing to the many factors that affect joint infiltration and compression, innumerable similar stress/time curves could be illustrated. This suggests that fracturing could occur at an earlier or later time depending on the number of factors that combine to affect stress generation. From observations of various projects throughout Ohio and elsewhere, it appears that the major factors contributing to joint infiltration, and consequently to pavement pressure generation, are as enumerated below. When considering these factors, it should be obvious that the sequence in which they are given does not suggest their significance with respect to the generation of pavement pressure.

G/P generation factors 1. 2. 3. 4. 5. 6. 7.

Pavement joint design and spacing Sealant quality and durability Subgrade composition and porosity Subgrade drainage Concrete strength and reinforcement Temperature range and duration Rainfall and pavement moisture content

224

Appendix 1

8. 9. 10. 11.

Traffic volume Deicing applications (including grit) Sealant maintenance. Pavement age

It is apparent from examining this list of factors that care should be exercised in the original design and construction to ensure the best functioning of pavement contraction joints by careful and thoughtful attention to (1) through (5). However, after the project has been constructed and exposed to environmental and traffic variables (6 through 8), the most efficient functioning of contraction joints can be influenced only by periodic attention and maintenance to (9) and (10). Where modification and repair can be justified, it may be possible to improve the functioning of (2) and (4). As Figure A1.3 and the list of factors would suggest, however, a pavement that has experienced a broad temperature range (requiring de-icing applications), good subgrade drainage, high-quality sealants, modest traffic volumes, and a good joint maintenance program should be expected to survive 25–30 years before pressure generation reaches a level where pavement distress becomes evident. On the other hand, a similar pavement that has experienced the same temperature range and de-icing applications, but with poor subgrade drainage, high traffic volumes, and minimal or no joint maintenance, should be expected to exhibit pavement distress and substantial bridge damage within 10 years. Blow-ups Pavement damage related to the pavement G/P phenomenon can take many forms including transverse cracking, longitudinal fracturing, progressive spalling at contraction joints, or instantaneous fracturing and/or buckling (blow-up) at contraction joints. Pavement maintenance engineers generally understand the term “blow-up” to mean an instantaneous, explosive disintegration and/or buckling of pavement, generally occurring at a contraction joint. A blow-up may be triggered by the movement of vehicular traffic across pavement joints that are simultaneously being subjected to high longitudinal compressive forces. Sustained longitudinal compression forces in a pavement, and the dynamically changing vertical shear forces due to vehicular axle loads moving across highly compressed joints, combine to rupture (or buckle) the weakest or most defective joint in a local stretch of jointed pavement. The blow-up thus relieves or releases the sustained compressive forces in the pavement. Blow-ups are unmistakable indications of high pavement pressures. As they impede the movement of vehicular traffic, their occurrence is usually reported. When they occur with considerable regularity, the numbers of them that occur within a certain period are counted. These reports serve as an indication that high pavement pressures are not local peculiarities. Instead, they indicate that high pavement pressures are, or have been, before the advent of pressure relief joint use, rather commonplace. The term “blow-up” is generally understood to mean an instantaneous fracture or buckling of pavement or both. Blow-ups may be triggered by the movement of vehicular traffic, but they are caused primarily by high longitudinal compression

Appendix 1

225

stresses within the pavement. Compressive stresses are relieved or released by blowups (Figure A1.3, curves a and c). The size of blow-ups has not been quantified. They can consist of minor localized joint fractures and slight buckling, major fractures with little or no buckling, and occasionally minor fracturing with significant buckling. Based on a count of blow-ups in 11 districts in Ohio, it was estimated … that there were in excess of 500 blowups in Ohio in 1970. During the late 1960’s and early 1970’s, Michigan reported 1,000 or more blowups a year. A bulletin of the Associated Press in Detroit contained a report stating that in 1971, Michigan experienced 1,387 blowups in the month of June alone. New York is reported to have experienced 1,590 blowups in 1 year with most of them occurring on the same day, July 3, 1966. [4, p. 56]

A 1970 article in the Milwaukee Journal contained an interview with Mr. Charles R. Ryan, the Chief District Maintenance Engineer for the eight counties in southeastern Wisconsin. Mr. Ryan is quoted as saying “During the last two weeks more than 200 sections of road … have buckled and many more are under stress and are ready to blow” [7]. Blow-ups are not just recent manifestations of pavement compressive stresses. C. D. Buck had a clear conception of the problem when he reported his experiences with blow-ups in Delaware in 1925 [8]. Following the construction of jointed concrete pavement in Delaware at the start of the twentieth century, this type of construction proliferated throughout the United States during the remainder of the century. Such pavement construction was followed by the generation of pavement pressures and ultimately by pavement blow-ups (see Figure 2.8). The author had a personal experience with the blow-up of a stone block street in the early 1940s. On his way home from a part-time job with the Pittsburgh Post Gazette newspaper, he was intrigued by the sight of a four door sedan sitting on its rump in the middle of Carson Street with its rear wheels and axle behind it on the other side of a large pile of stone blocks. Apparently, the stone block pavement blew up as the vehicle was passing over it. The rear wheels and axle were torn loose as the rest of the vehicle skidded to a stop on the other side of the blow-up. The massiveness of the blow-up was probably due partly to the fact that the pavement was reinforced by embedded streetcar tracks which made the stone block street significantly more resistant to generated pavement pressures. Ultimately, though, even this reinforced pavement was unable to withstand the pressures generated by sustained high temperatures that apparently preceded the blow-up. It would not be surprising to this author to learn that the Romans had some familiarity with these high-pressure manifestations in some of their more precisely fitted stone block ways.

Major blow-up records For a magazine article, “Reducing Bridge Damage Caused by Pavement Forces” [9], an international survey was made to record some of the major blow-ups that occurred throughout the world. Twenty of these major events are identified, described, and tabulated in Table A1.1, which also includes some data concerning

Nottinghamshire, UK

Akron, OH

Al Tuxford bypass

USR 21

I-77

USR 21

Akron, OH

Cambridge, OH

Elsinore Fwy

USR 25

Denmark

Wood County, OH

Carson St.

?

Delaware

Pittsburgh, PA

Identity

1975 (06/24) 1976 (06/25)

1971 (06/29)

1971 (06/28)

1963 (07/28) Before 1969

1941+

1925

Date

C C

No

C

C

C

C

SB

C

Yes

Yes

Yes

Yes

Yes

No

Yes

Mid 60s

1956

1964

1956

1954-56

1951

–

1911+

80 (24)

40 (12)

40 (12)

40 (12)

20 (6)

40 (12)

0.5 (1.6)

?

Contraction joint space (ft. [m])

Pavement data

Photo Type Construction date (year)

Blow-up

Major pavement blow-ups

Location

Table A1.1

Yes

Yes

Yes

Yes

?

Yes

No

No

Joint seala

85, 85, 87, 88, 87 ?

19b 12

80, 80, 86, 87, 99

87, 89, 79, 93, 97

89, 90, 91, 91, 89 –

–

–

Peak daily temperature before blow-ups (°F)

7

15

14

12

–

10–15

Age at blowup (years)

?

Trace

None

1 + (25)

–

Humid

–

–

Rainfall (in. [mm])

Environment

Between Brønsholm and Lundtofte Bridge (SUM21-0693) approach slab 1 Mi N & S of I-77/I-70 interchange South of USR 162 Article in NCE 7/1/76

Blow-ups commonplace Just south of Smithfield St.

Comments

226 Appendix 1

B, brick; C, concrete; SB, stone block. a May or may not have been maintained. b Evidence of earlier blow-ups. c For rehabbed pavement. °C = [°F − 32] × 0.556.

I-94 and I-494

M4, Heston

London, UK

St. Paul, MN

M11, Harlow

London, UK

SB I-395 (Old SR 350)

Highway 51

Wausau, WI

Arlington, VA

Allen Road

Toronto, Canada

I 505 Fairfield

A2 Motorway

Austria (Lower)

Sacramento, CA

9th St. S.W. Rt. 27, E.B.

Identity

Canton, OH Hauppauqe, NY

Location

C

C

C

Yes

No

No

1999 (07/08)

2000 (07/28)

C

C

C

Yes

No

C

C

Yes Yes

B C

No Yes

12 (3.7)

16 (4.9)

80 (24)

80 (24)

17 (5)

26(8)

0.33 (0.1) 60 (18)

1969

27 (8.2)

Original pavement Late 40s 50 (15)

1974

1965

1975

1964 & 1973

1970

1956

1930s 1958

Contraction joint space (ft. [m])

Pavement data

Photo Type Construction date (year)

No

1993 (06/7-8) 1993 (06/7-8) 1998 (07/19)

1983+ 1983 (06/24) 1986 (07/01) 1986 (07/06) 1987 (06/14)

Date

Blow-up

Yes

Yes

No

Yes

Yes

Yes

Yes

No

No Yes

Joint seala

– 77, 77, 83, 88, 88 –

50± 25

?77–86? ?77–86? 100, 102, 101, 104, 105 99, 102, 103, 94, 93 80, 83, 78, 73, 76

14 or 23 18b 28b 27b

31b

29c

78, 68, 76, 87, 88 66, 76, 83, 94, 97

16

21b

Peak daily temperature before blow-ups (°F)

Age at blowup (years)

0.5 + (13)

None

None

Yes

Yes

0.37 (9)

Trace

–

– 0.25 (6)

Rainfall (in. [mm])

Environment

Other blow-ups several years earlier Rehabbed and overlayed in 1970 Star Tribune newspaper

Toronto Star article Marathon County at 8–10 locations Five blow-ups mentioned In 6/17/93 NCE

By Duber Ave. N.E. of Phyllis Dr.

Comments

Appendix 1 227

228

Appendix 1

characteristics of the damaged pavement and the environmental conditions that preceded the events. The author located these data in various commercial, public, and professional publications, in various public records, and, with the help of others, in transportation department records. Tabulation of these data was done to help determine some of the design aspects that affect the pavement G/P phenomenon: progressive disintegration of pavement and movable bridge deck joints; progressive fracturing of abutment walls and fixed pier columns and bridge seats; and, ultimately, instantaneous (blow-up) failure of pavements. With respect to the pavement blow-up data provided in Table A1.1, some pertinent observations are possible as listed below. Historical events Undoubtedly, pavement blow-ups have been occurring ever since the first jointed, rigid pavements were constructed. They appear to be inevitable behavioral characteristics of stone block streets, and of brick (see Figure 2.8a) and jointed concrete pavements. Blow-ups of stone block and brick streets were probably rather commonplace before 1900 but rarely of a significance that merited notice of them in the newspapers of the time. Although the author witnessed the results of a major blow-up of a stone block street in Pittsburgh in 1941, or slightly thereafter, he was unsuccessful in locating a published record of that event in the local newspapers of the time. Nor was he able to locate records of more recent stone block pavement blow-up events occurring in either the United States or London. Presumably, the blow-ups (or minor buckling) of the few stone block streets or pavements that remain in service today have not been large enough, or do not sufficiently disrupt the movement of traffic that their occurrence is reported in the media or mentioned in pavement maintenance records. Jointed concrete pavements began replacing stone block streets at the turn of the last century. Shortly thereafter, the major blowups of these pavements began to be experienced and reported periodically in the print media (Figures A1.4 and A1.5, and see Figure 2.8 and Table A1.1). Contraction joint spacing It was expected that there would be a clear relationship between the longitudinal spacing of contraction joints and the age of pavements when blow-ups occurred. There does appear to be a (barely discernible) trend toward longer pavement lives with shorter contraction joint spacings (40 ft. [12 m] or less). But no absolutely clear trend could be determined from the tabulated data, probably because of the small number of events tabulated and the numerous pavement structure characteristics and maintenance practices (including prior repairs and pressure relief joints) that were unknown. Environmental variables It is clear that the presence of some moisture facilitates blow-ups, but, typically, not much moisture was recorded before most of the tabulated events. Except for the

Appendix 1

229

Figure A1.5 Allen Roadway, Toronto, Canada: Officer Dan Sutton examined the blow-up that occurred on July 6, 1986, when this pavement was just 16 years old.

Sacramento blow-up (where there was no recent rain with exceptionally high temperatures), it appears that several days of sustained high temperatures, with or without rain, are harbingers of exceptionally high pavement pressures and possible blow-ups.

Pavement age The most revealing data of Table A1.1 are the ages of pavements when they first blew up. As tabulated, most pavements failed before they reached the age of 20 years, some earlier than 15. It is quite clear from the ages of pavements when they failed that the generation of longitudinal pressures began almost immediately after the pavements were constructed. Pressures reached high levels before the age of 10 years (evidently high enough that they probably resulted in excessive forces against bridge abutments, compression closure of movable deck joints, and some visible distress of bridge approach pavements). After the bridge joints were compressed and closed, the pavement G/P phenomenon continued to generate longitudinal pressures and induced progressive distress of bridge members similar to that described in chapter 2 for the three bridges with intermediate movable deck joints. At comparable ages, similar serious distress, as described below, would be expected for the more common and more numerous continuous bridges with movable deck joints at the superstructure/abutment interface. However, less serious bridge distress would be expected for

230

Appendix 1

the more compressive-resistant integral and semi-integral bridges. Finally, however, generated pavement pressures would reach such high levels that either progressive minor fractures would take place or instantaneous pavement blow-ups would be expected. Pavement structure As mentioned previously, for long stretches of jointed concrete pavement constructed without movable joints (effective expansion or pressure relief joints), the pavement G/P phenomenon will generally cause pavements to fail in one of two ways. Where longitudinal pressure on pavement joints are not uniformly distributed throughout the pavement’s cross-section, and where pavement joints are not uniformly supported from below by a relatively rigid subbase, localized and progressive fracturing of pavement joints will take place. Otherwise, where longitudinal pressures are more uniformly distributed throughout a pavement’s cross-section (such as pavements with straight, relatively flat, or slightly depressed alignments), and where pavement joints are adequately supported from below, such pavements are able to resist great longitudinal pressures for long periods of time (>20 years). However, such large sustained pressures will ultimately cause pavements to fail instantaneously as illustrated in Figures A1.5 and 2.8. Pavement growth An examination of the behavior of pavement pressure relief joints and adjacent pavements helps to explain the behavior of a pavement structure such as that illustrated in Figure A1.3 (curve a), after it has experienced a catastrophic blowup. The behavior of such a pavement is similar to what would be expected after the installation of a pressure-relief joint, or immediately after the completion of a bridge with movable deck joints at the superstructure/abutment interface. At the Stanley Avenue Bridge in Dayton, Ohio, concrete approach pavements were cut transversely so that 3 ft. (0.91 m) wide bituminous-filled pressure-relief joints could be installed (Figure A1.6). The need for these relief joints became necessary when the movable deck joints at the superstructure/abutment interface were found closed and evidences of substantial longitudinal pressure were evident. Periodic observations of these relief joints were made over a 5-year period. The cutting of pavements (and the release of pressure) was followed by a gradual and progressive closure of the relief joints. At one of the two joints, for instance, the movement of the approach pavement into the joint amounted to 7½ in. (190 mm). As this movement occurred over a 5-year period, an average pavement growth rate of 1½ in. (38 mm) per year was experienced. The approaches to this bridge consist of two pairs of two 12 ft. (3.66 m) wide pavements with a separately cast 4 ft. (1.22 m) wide raised median between them. When the pavements and separate median were constructed, the pre-sawed contraction joints were placed to coincide with the vertical joints in median curbs. The longitudinal movement of pavement was manifest not only by compression of the relief joints, but also by a differential movement of pavement joints in relation to median curb joints. Joints closest to the bridge showed the greatest differential

Appendix 1

231

Figure A1.6 The rear approach roadway of the Stanley Avenue, B&O Railroad Bridge, Dayton, Ohio, 1960.

movement – 7½ in. (190 mm) – while joints further removed from the bridge showed progressively less movement, with joints located approximately 1,000 ft. (300 m) from the bridge showing no appreciable movement. Consequently, based on behavior of pavement approaches to the Stanley Avenue Bridge, and to similar pavements of many other bridges, it appears that up to 1,000 ft. (300 m) or more of pavement can contribute to movement of pavement at pressure-relief joints. As these pavement movements are both progressive and accumulative, and as a substantial length of pavement contributes to this movement, it has come to be called “growth” to distinguish it from “expansion,” the term usually used to refer to the minor component of temperature-related cyclic movement. Pavements with pressure relief joints experiencing growth are also subjected to substantial pressures. However, instead of relatively constant pressures being distributed throughout an extensive length of jointed pavement, as is typical of the restrained pavement illustrated in Figure A1.1c, pressures vary linearly along the length of pavement experiencing growth. Pressures are minimal at pressure-relief joints. (They equal the resistance of relief joints to compression, the pressures indicated by the relatively flat portion of curve b in Figure A1.3.) They approach a maximum at a pavement pressure plateau located at some distance (±1,000 ft. [300 m]) from relief joints. Pavement between these two boundaries will continue to grow toward relief joints, with most growth occurring at relief joints and progressively less growth at distant locations (± 1,000 ft. [300 m]) where maximum residual pressures have been maintained. Between these two limits, pavement will continue to grow toward relief joints until the relief joints are closed. Thereafter, with pavement being restrained from further growth, pressure generation increases (Figure

232

Appendix 1

A1.3, curve b) within the lengths of formally growing pavement until increasing pressures are relieved by progressive fracturing of pavement joints, blow-ups, or the replacement or modification of formerly closed relief joints. Movable bridge deck joints The behavior of pavements adjacent to continuous bridges provided with movable deck joints at the superstructure/abutment interface is similar to that of pavements adjacent to pressure relief joints (Figure A1.3, curve b). As bridge abutments are not designed to resist the huge pressures characteristic of restrained pavements, approach pavements will grow toward jointed bridges, jamming abutments toward superstructures and progressively closing the movable deck joints. Subsequently, after the deck joints are closed, pavement pressure generation continues to increase until it reaches high enough levels to begin fracturing pavements, abutments (Figures A1.7 and A1.8), or rigid bridge piers. In this context, pressures generated by the restrained growth of approach pavements are compounded by pressures generated by the restrained expansion of bridge superstructures. Supplementing each other, they cause both pavement and bridge distress and fracturing earlier than would have been expected from the restrained growth of approach pavements alone. Generation of pavement pressure and the generation of pavement growth appear to be two sides of the same coin, or two major aspects of the same phenomenon. Debris infiltration of contraction joints will result in pressure generation where

Figure A1.7 USR I-90, 140th Street Bridge, Cleveland, Ohio, 1976: This fourspan, continuous deck-type, steel-beam bridge with movable deck joints at the superstructure/abutment interface, along with similar companion bridges on the same route, was seriously damaged by the pavement G/P phenomenon (see Figure A1.8).

Appendix 1

233

Figure A1.8 This view of an abutment bridge seat of the bridge shown in Figure A1.7 illustrates how the pavement G/P phenomenon jammed the stubtype abutments (founded on two rows of piles) at least 3 in. (75 mm) into the superstructure. Subsequently, the generated pavement pressures and bridge superstructure expansion jammed the abutments under the movable deck joint armor, thus raising the superstructure up and off of the rocker bearings, and the superstructure’s dead- and live-load reactions on the abutment backwalls were great enough to shatter the backwalls.

pavements are restrained (no pressure relief joints or movable deck joints (Figure A1.3, curve a), or growth generation where pavements are not restrained (with relief joints or movable deck joints). In many instances, growth will take place until all available space has been consumed. All available space refers to space provided in pavement expansion joints, space available in pressure-relief joints by compression of the filler, and space provided in movable bridge deck joints. Then, as pavements are restrained from further growth, pressure generation commences along the second portion of the pressure-generation curve (Figure A1.3, curve b). As both pressure and growth generation appear to be directly related to debris infiltration of contraction joints, it goes without saying that the factors that have a significant effect on pressure generation have a similar effect on growth generation. Ideally, the solution to this problem is simple. All that is needed is a pavement contraction joint design that would completely seal such joints against the intrusion of all foreign materials, even at the lowest site temperatures. Designs that are somewhat less than ideal would be suitable because a reasonable service life could be attained. However, it should be clear that current technology is not sufficiently developed to provide a cost-effective solution to the significant problem of debris infiltration of contraction joints.

234

Appendix 1

Many of these words were written over 30 years ago. At that time, pavement and bridge maintenance engineers adopted programs to install pavement pressure-relief joints to control stress levels and protect both structures and vehicular traffic from the adverse consequences of pavement pressure generation. However, as described and illustrated below, a few current pavement research and maintenance engineers are proposing an entirely new direction for pavement maintenance, a direction that appears to ignore the major characteristics of the pavement G/P phenomenon, and the inevitable consequences of allowing generating pressures to reach destructive levels. Unfortunately, if this new direction is accepted by pavement maintenance agencies, this author predicts dire consequences in the not too distant future for the transportation profession in both human and monetary terms.

New direction commentary Wisconsin research In a series of recent papers (1996–1997), the engineers of Wisconsin Department of Transportation (DOT) have preached the use of unsealed contraction joints. In the most recent paper of this series [2], it continually stressed that the primary objective of all joint sealant research should be total pavement performance and not merely sealant performance. However, when examining this latest report, it was found that the evangelist is guilty of the sin that is condemned. Like most prior joint sealing research that is disparaged as being too narrowly focused, the same can be said about current Wisconsin research because it completely ignores the effect of unsealed joints on the generation of pavement pressures. Wisconsin’s latest report was for five 1,000 ft. (300 m) long pavement test sections with ages of 8, 13, and 22 years. The report concludes with several statements to the effect that joint sealing has no significant effect upon pavement distress or life, upon material integrity, or upon ride quality. It is also stated that blow-ups are a function of joint spacing and not joint sealing. Although many pavements reach the age of 10–15 years without visible signs of pressure-related joint distress, it would be unusual, but not unexpected, that Wisconsin’s 8- and 13-year-old unsealed test sections could reach these ages without visual signs of compression joint damage. But it would be remarkable for a 22-yearold pavement to reach such an advanced age unscathed. Consequently, it was not surprising to learn that these older test sections had in fact been modified with a number of full-depth repairs. John W. Bugler, a former pavement maintenance engineer for the New York State DOT, examined Wisconsin’s test pavements a number of times over the last two decades. During one of these examinations he found that five full-depth lane-width repairs were made to the “unsealed” 20 ft. (6.1 m) test sections of USR 51, the oldest test section reported upon in the paper. Pavement markings indicated that these repairs were made in June, July, and August (the hottest months) of 1993, at the time when pavements had aged 19 years. Although the research principal claimed that this section of pavement experienced premature joint spalling due to an irregularity in the location of reinforcement, two pavement joints (identified as joints

Appendix 1

235

37 and 41 on the test pavements) appeared to have been in very good condition in 1987, just 6 years before their full-depth repair. Consequently, although misplaced reinforcement may have contributed to distress at some joints, it appears that high compressive stresses were probably responsible for most distress that necessitated the full-depth repairs to this oldest unsealed test section. It is instructive to note that, in 1993, there were no full-depth repairs made to a companion sealed test section (N4F), even though it presumably was constructed at the same time, by the same contractor. Also, apart from the sealed joints, it had a similar design and was exposed to similar weather extremes, traffic volumes, and truck loadings. Therefore, it appears that the reported results of Wisconsin’s research on unsealed pavements were inconclusive because there was no attempt made to monitor and report about the relative magnitude of generated pavement pressures for both the sealed and the unsealed pavement test sections. As a result, joint sealing recommendations coming from this research are of questionable validity. As there are 50 state transportation departments that have adopted their own sealing practices with respect to the maintenance of jointed concrete pavements, why should this author (and his colleague, John Bugler) be so concerned about the pavement maintenance recommendations originating from a single state like Wisconsin? The reason for the concern was the fact that Wisconsin was recommending the adoption of unsealed pavement contraction joints, a recommendation that was contrary to the generally accepted sealing practices of most States and to the practices that were recommended by most authoritative pavement maintenance specialists, and, as described above, because that recommendation was based on the complete neglect of the pavement G/P phenomenon and its destructive potential with respect to both pavements and bridges. Nevertheless, Wisconsin’s pavement maintenance recommendations were given particular significance by pavement engineers of the Federal Highway Administration (FHWA) who were, without any other research justification, promoting the adoption of Wisconsin’s unsealed pavement practices by other state transportation departments, including those of Ohio and New York. That is the reason why the unfortunate Wisconsin episode has demanded recognition and a brief elaboration in this appendix on the pavement G/P phenomenon. Pressure-relief joints Some pavement researchers appear to have an ambivalent attitude with respect to the efficacy of pressure relief joints. They define a pressure relief joint as “a transverse joint installed to relieve compressive stress for the purpose of reducing deterioration of existing joints, pavement blowups, and protecting abutments” [10, p. 3]. In this definition, there appears to be a vivid recognition that pavement compressive stresses and pressures are primarily responsible for both pavement and bridge damage, and that pressure-relief joints, as their name implies, are intended to prevent such damage by moderating pavement pressures to tolerable levels. On the other hand, they also state that “Because they provide no load transfer, deflections at the relief joints tend to be high. Significant pumping, faulting, corner brakes, and slab deterioration can thus occur in the vicinity of a pressure relief joint.” They unfortunately conclude these statements by recommending that: “Pressure relief

236

Appendix 1

Figure A1.9 Performance of a foam-filled pavement pressure relief joint [9, p. 32], with a clarifying detail added by the author.

joints should be used only on pavements which have experienced blowups or are pushing bridges” [10, p. 25]. Even though they recognize the relationship between high pavement pressure and pavement and bridge damage, they caution against the early use of pressure-relief joints. In other words they seem to be suggesting that patients with high pressures should not be treated until after the patient has experienced a life-threatening rupture. Such a prescription is questionable to say the least. That same report contains a number of informative charts of relief joint measurements which illustrate the response of such joints (polymer-foam-filled joints 4 in. [100 mm] wide) to the pressure-generated growth of pavement. Figure A1.9 is one such chart and can be considered somewhat idealized but nevertheless typical for applications of foam-filled pressure-relief joints that were initially 4 in. (100 mm) wide. Notice that the joint width is progressively compressed during the hot summer months and attains almost complete closure in about 3 years’ time. Subsequently, as the joint is fully closed, it will function to restrain pavement growth and thus permit the subsequent generation of higher pavement pressures as indicated by curve b in Figure A1.3. However, astoundingly, the report concludes with the recommendation that “most new pressure relief joint widths should be limited to 1 to 2 in. (25 to 50 mm) (maximum) to reduce the possibility and severity of over-relieving the pavement” [10, p. 124]. As indicated with respect to Figure A1.9, such a recommended joint width would have an effective service of only a year or two, after which it would have to be replaced periodically to prevent pavement pressures from reaching destructive levels. As most individuals would recognize the adverse economical and political consequences of closing pavement lanes to traffic every couple of years

Appendix 1

237

merely to replace such short-lived relief joints, could such a recommendation suggest that these researchers incorrectly assume that placing a relief joint terminates the pressure generation process?

Contraction joint spacing In the introduction to the final report of the research project “Pressure Relief and Other Joint Rehabilitation Techniques,” the following statement is given: Most of these problems [sealant deterioration, intrusion of incompressibles, concrete deterioration, etc.] are associated with long-jointed reinforced concrete pavements, and not with short-jointed plain concrete pavements. [10, p. 1]

On page 13 of the same report [10], the following statement is given: Pavement growth due to intrusion of incompressibles produces far more severe problems in pavements with long slab lengths. States such as California which have thousands of miles of short-jointed pavement (i.e., slab length less than 20 ft. [6.1 m]) rarely experience blowups. States such as Illinois, Michigan, and Virginia which have 40 to 100 ft. [12.2 to 30.5 m] slabs frequently experience blowups.

Current reports from Wisconsin reflect this view: Incidentally, blowups were a major problem in Wisconsin for pavements with 80 and 100 foot (24 and 30 meter) joint spacing. The use of closer spacing (15 to 20 feet [5 to 6 meters]) has virtually eliminated blowups. [1, p. 6]

The problem with these statements is that they imply that pressure generation. and consequently pressure-related pavement distress such as spalling, fracturing, and blow-ups. can be avoided by reducing just the joint spacing from about 80– 100 ft. (24–30 m) down to 15–20 ft. (5–6 m). Such recommendations completely ignore the numerous other factors that significantly influence joint infiltration (and pressure generation) such as ambient temperature range and duration, sealant quality and maintenance, rainfall, porosity of subgrade and subsoils, etc. They also conveniently ignore the potential long-term consequences of using pavements with a short contraction joint spacing. Assuming that all aspects of pavements are the same except for joint spacing, this author thinks that it is improbable that pavements with the “long” spacing will blow up whereas the pavement with the “short” spacing will not. Structures do not operate on the basis of good or bad; they operate in a continuum. Consequently, it very well may be that as joint spacing diminishes, joint shrinkage cracks will become smaller and, depending upon particle sizes of compression-resistant debris, debris infiltration and consequently pressure generation may be slowed. However, ultimately, pressure generation will reach such levels that blow-ups will occur in both long-jointed and short-jointed pavements, one later than the other. For example, how else can one account for the growth and blow-ups of brick and stone block streets that have joint spacing of only 6 in. (150 mm) or less?

238

Appendix 1

Bridge damage Pavement and bridge engineers have different experiences and consequently different attitudes about the effect of pavements on abutting bridges. A glimpse of these attitudes may be gained by examining the terminology that they use when they talk or write about this subject (emphasis added): Another problem caused by pavement growth is bridge pushing … the pavement will push against the bridge approach slabs. Incidents of cracked abutments and bridge decks being pushed nearly off the abutments have been documented … . [10, p. 7] Pressure relief joints should be used only on pavements which experienced blowups or are pushing bridges. [10, p. 25]

In addition to “pushing,” other pavement researchers use the terms “shoving” or “thrusting.” In contrast, consider how they feel when they are speaking about other pavement devices. Many other types of secondary structures can also be damaged by pavement growth. These include manholes and other drainage and access structures in the pavement surface. They can be crushed, collapsed, or rendered nonfunctional as they are moved by the pavement. Curbs and traffic islands are also subject to shattering, breakup, upheaval, and failure as the surrounding pavement expands. [10, p. 7]

Notice, with respect to bridges, that they say “push,” “shove,” or “thrust.” But with pavement items they say “crushed,” “collapsed,” “shattered,” “breakup,” etc. Obviously, these researchers have not had the opportunity to examine bridges that have been compressed by high pavement pressures. In bridge engineers’ terms, abutments and fixed piers are “fractured,” backwalls are “crushed,” girders are “buckled,” bearings “displaced,” etc. In Wisconsin, bridges were not even an issue. In evaluating of the relative efficacy of sealed or unsealed pavement joints, pavement maintenance engineers qualified their research as follows: Wisconsin’s research relates to PCC highway pavement slabs on grade; it does not consider airfields, interior slabs, buildings, bridges, etc. [2, p. 15]

Notice the sequencing of the items in this list. It appears that these researchers think of bridges as isolated and completely independent structures like “airfields,” “buildings,” “interior slabs,” etc., and not as integral parts of a highway facility that could affect and be affected by pavement behavior. Incidentally, the offending statement quoted above was subsequently modified in the published report after criticism of the original report by this author. Although Wisconsin engineers did not consider bridges in their evaluation of pavement performance, they were well aware that pavement compressive stresses could reach high levels. With respect to such stresses, they state that pavement stresses “can only amount to 7,000 to 14,000 kPa (1,000 to 2,000 psi) maximum, well below the compressive strength of concrete” [2, p. 30]. These are significant stresses, even for concrete pavement. However, with respect to bridges, such pressures from

Appendix 1

239

a two-lane pavement could amount to from 25 to 50 times the force that stub abutments are designed to resist. In bridge terms, such forces are irresistible. Consequently, where pavement engineers ignore the potential for high pavement pressure against bridges, serious bridge damage could be the result. As bridges with movable deck joints have functioned as huge and expensive pavement expansion joints [3], they have been progressively closed by the pressureinduced growth of jointed pavements. Subsequent to such closure, bridges have experienced severe damage due to the combined effect of the response of restrained pavements and abutting bridge decks to high temperature levels. Pragmatic bridge engineers have become aware of the apparent futility of constructing bridges with movable deck joints adjacent to rigid pavements. Consequently, they are now electing in increasing numbers to construct integral bridges (jointless bridges), which are significantly more resistant to pavement pressures. Such a response to jointed bridge damage can compound the pavement engineer’s problems if such bridges are constructed without suitable joints in bridge approaches. Otherwise, in the future, blow-ups will become more commonplace on bridge approaches. Pavement expansion joints Two quotes from recent pavement research literature help to illustrate that some pavement research engineers fail to understand the performance of pavement expansion joints when such joints are located in pavements that are under pressure. Construction of expansion joints is currently recommended only when major blowups or other pressure-related damage has occurred. [10, p. 25] Bridge approach expansion joints will usually provide sufficient pressure relief within 500 feet (150 meters) and additional relief is not needed. [10, p. 25]

First, as typical pavement expansion joints are designed to provide for relatively small cyclic thermal movements, their use as pressure-relief devices in pavements being subjected to progressively higher and higher pressure is inappropriate. Generating pavement pressure will close such devices within a year or two, after which they will become merely rigid pavement artifacts incapable of further extension or compression. Second, with respect to both statements, how can a class of objects be safely recommended for use based upon its generic name only? Does the joint width and filler type not have some significance in this respect? For example, a 1 in. (25 mm) wide joint filled with asphalt-impregnated cane will, when subjected to compressive stresses, close to such an extent in 1 year’s time that it would be completely ineffective in relieving subsequent pavement pressure. Foam-filled pressure-relief joints in widths of at least 4 in. (100 mm), or asphalt-filled relief joints 12 in. (300 mm) or wider, are intended specifically for this purpose. Expansion joints in compressed pavements are beneficial in only one respect. Their somewhat elastic type of filler helps to minimize localized stress concentrations and thereby protects pavement joints from localized spalling and fracturing. But they will not, as apparently some pavement engineers believe, prevent the generation of pavement pressures.

240

Appendix 1

Pavement research Most pavement professionals recognize that, in addition to reactive aggregates, it is the contamination of contraction joints by compressive-resistant debris that is responsible for the development of pavement pressures and associated joint distress such as spalling, fracturing, and blow-ups. They also recognize that the G/P phenomenon takes up to 8 or more years for pressures to reach levels where distress becomes manifested. Consequently, it was distressing to learn that a new pavement research project was initiated by the Ohio DOT with the encouragement of the FHWA to, in part, evaluate the relative efficacy of sealed or unsealed contraction joints throughout a pavement service life of only 5 years. In addition, it was also learned that the monitoring of pavement stresses and pressures was not to be a part of this project [11]. As one of the primary purposes of sealing contraction joints is to retard the generation of destructive pavement pressures, it appears ludicrous to this author to try to compare the relative efficacy of sealed and unsealed contraction joints without monitoring their effect on the pavement pressures that they are intended to control. Also, as this project was programmed for only 5 years, the period illustrated by the initial flat portion of curve a in Figure A1.3, how could the researchers accurately determine differences between the pressure-generation curves for the pavements with sealed contraction joints or the pavements with the unsealed contraction joints, or predict the ultimate effects of these differences? Since such research requires clairvoyance and speculations to achieve useful results, it must be considered bad science at best and pseudoscience at worst.

Summary As discussed in this chapter, the focus of recent pavement research on pavement joint design and maintenance is startling to engineers who recognize that high pavement stresses and pressures are responsible for most of the progressive damage suffered by jointed pavements and abutting bridges. For example, one of the most comprehensive recent examinations of current pavement maintenance practices [10] uses the term “pressure” over 450 times. That is correct, more than 450 times as counted by this author. Also, the report for this project contains an extensive bibliography of 149 references to other papers of pertinent pavement research. Yet, remarkably, this bibliography did not contain a single reference to research that focused on the generation of pavement pressure. Another recent report about extensive research on sealed and unsealed pavement joints [2] did not mention the term “pressure” once, although it is widely recognized that joint performance is one of the primary factors that affect the generation of destructive pavement pressures. To put it bluntly, it appears that some current pavement researchers are not making any attempt to envision events occurring beyond the realm of ordinary sense perceptions. With respect to the effect of the G/P phenomenon on bridges that abut jointed pavement, some researchers ignore bridges entirely or assume that some device named “expansion joint” will somehow magically protect a bridge from the

Appendix 1

241

inexorable pavement growth, and subsequently from the crushing pressures that are characteristic of such pavements. Others minimize pavement/bridge interaction by using such terms as bridge “pushing,” “shoving,” or “thrusting” to help avoid the recognition that jointed pavements and abutting bridges are self-destructing. With respect to the relationship between jointed pavement design and maintenance, and the generation of pavement pressure, many pavement maintenance engineers intuitively recognize that there are at least a dozen factors that influence pavement pressure generation and high pressure levels. Yet the relative significance of these factors has not been demonstrated by research that controlled these factors and measured pavement strains and/or pressures. Without controlled monitoring of the G/P phenomenon, pavement design details and maintenance practices will continue to depend upon the advice of presumed “authorities.” Unfortunately, the transportation profession will then continue to base its decisions on personal philosophies and flawed assumptions, and not on scientifically demonstrated and replicated proof. Contrary to the recommendations of some current pavement research authorities, generating pavement pressures should not be allowed to reach destructive levels. Otherwise, the transportation profession will continue to be familiar with pavement fracturing (see Figures A1.4, A1.5, and 2.8), bridge fracturing (see Figures A1.8, 2.2, 2.4, 2.6, 2.10, and 2.12), restricted traffic flow at bridge repair sites, vehicular accidents, and personal injury. There is a better way. Simple methods should be devised to monitor the generation of longitudinal pavement stress and pressure levels, the relative significance of the factors that affect contraction joint behavior and the generation of pressures should be established, and design details and maintenance practices should be devised that will retard pressure generation and limit or control generated pressures to tolerable levels. Then pavement design and maintenance will be given a different direction, a direction that should produce a significantly more durable and safer highway environment for those who have entrusted the transportation profession with their welfare.

Acknowledgment The brief critique given above about the pavement maintenance research conducted by pavement maintenance engineers of the Wisconsin Department of Transportation would not have been possible without the personal interests and project site inspections made by John W. Bugler, formerly a pavement maintenance engineer with of the New York State Department of Transportation. His interest and concern about that project over many years, his many examinations of the test pavements, and his reports of those examinations were invaluable in making an independent evaluation of the Wisconsin pavement maintenance research possible. His professional efforts in this respect were outstanding. Consequentially, the author of this book is pleased to have this opportunity to acknowledge John Bugler’s contribution to this work. That contribution will be of benefit not only to the members of the transportation profession in general but also and particularly to the highway travelers who will eventually be provided with a safer highway environment primarily because John Bugler took his pavement maintenance responsibilities so seriously.

242

Appendix 1

References 1. Shober, S. F., “The Effect of PCC Joint Sealing on Total Pavement Performance,” Fourth World Congress on Joint Sealants and Bearing Systems for Concrete Structures, American Concrete Institute, Sacramento, California, 1996 (preprint). 2. Shober, S. F., “The Great Unsealing: A Perspective on PCC Joint Sealing,” Transportation Research Record No. 1597, Transportation Research Board of the National Academies, Washington, D.C., 1997, pp. 22–33. 3. Burke, M. P. Jr., “The World’s Most Expensive Pavement Expansion Joints,” Ohio Transportation Engineering Conference Proceedings, The Ohio State University, Columbus, Ohio, 1972. 4. Burke, M. P. Jr., “Bridge Approach Pavements, Integral Bridges, and Cycle-Control Joints,” Transportation Research Record No. 1113, Transportation Research Board of the National Academies, Washington, D.C., 1987, pp. 54–65. 5. Richards, A. M., Causes, Measurements and Prevention of Pavement Forces Leading to Blow-Ups, The University of Akron, Akron, Ohio, 1976. 6. AASHTO. Standard Specifications for Highway Bridges, 17th edn, American Association of State Highway and Transportation Officials, Washington, D.C., 2002. 7. Milwaukee Journal Newspaper, July 7, 1970. 8. Buck, C. D., “Repair of Concrete Road Blow-Ups in Delaware,” Engineering News Record, Vol. 95, No. 11, 1925, pp. 432 and 433. 9. Burke, M. P. Jr., “Reducing Bridge Damage Caused by Pavement Forces,” Concrete International, American Concrete Institute, Farmington Hills, Michigan, January and February, 2004. 10. Smith, K. D., et al., “Pressure Relief and Other Joint Rehabilitation Techniques,” Report No. FHWA/RD-86/XXX, Federal Highway Administration, McLean, Virginia, 1987. 11. State Project No. 14668, Ohio Route 50, Joint Sealing Experiments (November 1996– 2001).

Appendix 2

Glossary

AASHTO: American Association of State Highway and Transportation Officials. Composite structure: An assembly of two or more different structural components integrated into a single functional whole. Cycle control joint: A transverse movable joint provided between bridge approach slabs and approach pavements to facilitate the longitudinal cyclic movement of bridge superstructures and attached approach slabs. DOT: Department of Transportation. Elementalism: The name given to the concept that reality is composed of discrete and independent objects that have been noticed and named. End diaphragm: A transverse reinforced concrete member used to integrate superstructure stringers and reinforced concrete deck slabs at abutments of semiintegral bridges. Expansion joint: Refer below to Movable joint and movable deck joint. FHWA: The Federal Highway Administration. It is an administrative agency of the US Department of Transportation. It is empowered by the US Congress to administer federal funds allocated for the design and construction of state and local highway projects. It encourages and provides funding for and supervision of transportation-related research projects intended to improve not only highway safety but also the materials and methods used in highway construction. G/P phenomenon: A temperature- and moisture-driven phenomenon whereby jointed rigid pavement periodically and progressively generates longitudinal pavement growth and/or pressure. 243

244

Appendix 2

Guide bearing: A movable bearing provided at abutments of skewed semi-integral bridges to facilitate differential longitudinal movement of superstructures with respect to abutments while providing lateral support for superstructures. Holism: The name given to the concept that reality is composed of a synergistically functioning whole. Integral abutment: An abutment that is constructed integrally with a bridge superstructure. Integral bridge: A single- or multiple-span continuous deck-type bridge without movable deck joints at the superstructure/abutment interface. It is generally supported by embankments and stub-type abutments on single rows of vertically oriented flexible piles, by flexible piers constructed compositely with the superstructure, or by semi-rigid piers with fixed and/or movable bearings. Movable deck joint: See Movable transverse deck joint. Movable joint: As bridge joints should be designed to accommodate anticipated multi-directional movements of bridge members due to various phenomena such as thermal changes, moisture changes, creep, elastic shortening, superimposed loadings, etc., it will be used in this book instead of the archaic misnomer “expansion joint.” Movable transverse deck joint: A transversely oriented movable joint in bridge superstructures intended to facilitate differential movement of superstructures or superstructure segments with respect to piers, abutments, or other superstructure segments. Pressure relief joint: A transversely oriented movable pavement joint (usually 1–4 ft. [0.3–1.2 m] wide and usually filled with compressible asphalt concrete) provided between jointed rigid pavement and bridge approach slabs to protect both pavements and bridges from the longitudinal pressures generated by the pavement G/P phenomenon. Semi-integral bridge: A single- or multiple-span, continuous, deck-type bridge without movable deck joints in its superstructure but with movable longitudinal joints between its superstructure and abutments. Abutments of such bridges should be rigidly supported. Piers can be flexible and constructed integrally with the superstructure, or rigid with fixed and/or movable bearings. Structure movement system: A complex unity of diverse components configured and designed to facilitate structure movements (rock, roll, slide, shear, flex, compress, consolidate, etc.) in response to applied loads and to material and environmental changes. TRB: The Transportation Research Board is one of six major divisions of the National Research Council. Its 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.

Appendix 3

Captions for Photographs

These photographs appear on the first page of each chapter of this book. Introduction: Twisp River Bridge, Twisp, Okanogan County, Washington State, 2001. This innovated semi-integral bridge is supported by field-spliced, prestressed, post-tensioned, high-performance concrete girders. With a spectacular single span of 197 ft. (60 m), it is probably the longest single span semi-integral bridge in the world. Chapter 1: SR 50, Happy Hollow Creek Bridge, Hickman County, Tennessee, 1996. This is probably the world’s longest integral bridge. Its overall length is 1,175 ft. (358 m). Chapter 2: SR 21, West Fork of Black River Bridge, Reynold County, Missouri. This nine-span, 765 ft. (233 m), prestressed, concrete integral bridge has a centerline of roadway radius 2865 ft. (873 m). Chapter 3: Rainbow Bridge National Monument. It spans a tributary of the Colorado River near the Utah–Arizona border. It has a span of 275 ft. (84 m) and a rise of 290 ft. (88 m). It is the world’s largest natural integral bridge. Chapter 4: Naibekoshinai River Bridge, Hokkaido, Japan, 1996. This is one of the first of two early integral bridges constructed in Japan. It has three continuous spans with an overall length of about 560 ft. (110 m). Chapter 5: SR 7, Teens Run Bridge, near Eureka, Ohio, 1938. This five-span, continuous, reinforced-concrete structure, with each of its integral stub-type 245

246

Appendix 3

abutments supported by a single row of flexible piles, is thought to be the first integral bridge constructed in Ohio. It is probably the first such bridge constructed in the United States and possibly the world. It has an overall length of about 144 ft. (44 m). Chapter 6: DooDong Bridge, Ham Yang Co., South Kyung Province, South Korea, 2001. The original project plans for this bridge provided it with movable deck joints at piers and abutments. Those plans were revised (except for pier details) before construction to provide Korea with its first integral bridge. Chapter 7: Route 401, Prospect Avenue Bridge, Toronto, Ontario, Canada, 1995. This relatively new, steel box-beam integral bridge is about 490 ft. (150 m) long. Chapter 8: USR 23, SR 32 Bridge, Pike County, Ohio, 1996. This twin, heavily skewed, semi-integral bridge was provided with two accessible guide bearings at each abutment of both structures (see Figures 8.6 and 8.7). Chapter 9: USR I-90 B. N. Railroad bridge, Grant County, Washington State, 1966. This early semi-integral bridge performed well in simulated seismic tests in 1993 just before its removal. Chapter 10: SR 555, Muskingum River Bridge, Zanesville, Ohio, 1979. This is the first semi-integral bridge constructed in Ohio. It has three continuous spans with an overall length of 540 ft. (165 m). Abutment details are illustrated in Figure 1.8. Chapter 11: Price-Hillards, Scioto Darby Creek Road Bridge, Franklin County, Ohio, 1986. This is the first integral bridge with drilled-shaft piers constructed in Ohio. Appendix 1: A27 – Brockhampton Road Bridge, Hampshire, United Kingdom, 2001. This 215 ft. (65.5 m), three-span, continuous, steel-girder integral bridge is supported by both integral abutments and piers and cast-in-place concrete piles. Appendix 2: Route 401, Franklyn Boulevard Underpass, Ontario, Canada, 1990. The original, continuous precast, prestressed, concrete box-girder integral bridge was just recently widened using the same type of primary members. Appendix 3: SR 725, Sycamore Creek Bridge, Miami Township, Montgomery County Ohio, 1963. This was the first integral steel-beam bridge constructed in Ohio. The steel portions of this structure were cleaned and painted, and the original concrete fascias of the deck were sealed shortly before this 1999 photograph was taken.

Index

Bridges Ashtabula River Valley Viaduct, Ashtabula, Ohio, 71–74, 78, 106, 107 A 27 Brockhampton Road Bridge, Hampshire, United Kingdom, 215, 246 Broad Road Viaduct, Bedford, Ohio, 76 Brookpark Viaduct, Cleveland, Ohio, 74 Cuyahoga River Valley Bridge, Brecksville, Ohio, 76, 106–107, 115 Doodong Bridge, Ham Yang Co., South Kyung Sang Prov., South Korea, 81, 246 General US Grant Suspension Bridge, Portsmouth, Ohio, 212 Golden Gate Bridge, San Francisco, California, 193 Happy Hollow Creek Bridge, Hickman County, Tennessee, 1, 8, 43, 57, 118, 245 John F. Kennedy Memorial Bridge, Louisville, Kentucky, 23, 25–27, 29 Long Island Bridge, Kingsport, Tennessee, 6 Mianus River Bridge, Greenwich, Connecticut, 213 Naibekoshinai River Bridge, Hokkaido, Japan, 59, 71, 78, 245 Ohio Bridge No. GEA – 422 – 0057, Geauga County, Ohio, 197

Ohio Bridge No. HIG – 28 – 0280, Highland County, Ohio, 191, 192 Ohio Bridge No. WAY – 585 – 0859, Wayne County, Ohio, 194, 195 Old North Hill Viaduct, Akron, Ohio, 103–106, 115, 118 Old Third Street Viaduct, Cincinnati, Ohio, 23–25, 29 Patterson-Riverside Great Miami River Bridge, Dayton, Ohio, 4, 5 Pecos River Bridge, Carlsbad, New Mexico, 23, 27–29 Price-Hillards Scioto Darby Creek Road Bridge, Franklin County, Ohio, 246 Rainbow Bridge National Monument, Utah/Arizona Border, 41, 245 Route 401, Franklyn Boulevard Underpass, Ontario, Canada, 243, 246 Route 401, Prospect Avenue Bridge, Toronto, Canada, 99, 246 Salginatobel Bridge, Schiers, Switzerland, 109 Schoharie Creek Bridge, Montgomery County, New York State, 213 Silver Bridge, Gallipolis, Ohio, 213 SR 21 Barberton Reservoir Inlet Bridge, Akron, Ohio, 35, 36 SR 21 West Fork Black River Bridge, Reynold County, Missouri, 21, 245 SR 180 – USR 33 Bridge, Hocking County, Ohio, 91 SR 250 – 148th Avenue N. E. Bridge, King County, Washington State, 91, 92 SR 555 Muskingum River Bridge, Zanesville, Ohio, 137, 157, 246 247

248

Index

SR 725 Sycamore Creek Bridge, Miami Twp., Montgomery County, Ohio, 245, 246 SR 771 Big Branch Creek Bridge, Highland County, Ohio, 89–90 Stanley Avenue – B&O Railroad Bridge, Dayton, Ohio, 230, 231 Teens Run Bridge, Gallia County, Ohio, xii, 71–72, 75, 76–77, 78, 139, 245 Tocoma Narrows Bridge, Tocoma, Washington State, 193 TR 45 – CSXT Railroad Bridge, Pickaway County, Ohio, 44, 95 Twisp River Bridge, Twisp, Okanogan County, Washington State, 11, 245 USR 23 – SR 32 Bridge, Pike County, Ohio, 121, 129–131, 246 USR 52 Little Scioto River Bridge, Portsmouth, Ohio, 33, 34 USR 62 Yankee Creek Bridge, Trumbull County, Ohio, 82, 83 USR 422 McFarland Creek Bridge, Geauga County, Ohio, 197–199 USR I-35W Mississippi River Bridge, Minneapolis, Minnesota, 213 USR I-76 – East Market Street Bridge, Akron, Ohio, 166, 167 USR I-77 – SR 18 Bridge, Summit County, Ohio, 97 USR I-90 – B. N. Railroad Bridge, Grant County, Washington State, 139, 197, 202, 246 USR I-90 – 140th Street Bridge, Cleveland, Ohio, 232, 233 USR I-271 – Wilson Mills Road Bridge, Cleveland, Ohio, 37 USR I-480 Cuyahoga River Valley Bridge, Cleveland, Ohio, 177, 179

General AASHTO (American Association of State Highway and Transportation Officials), 22, 217, 243 AASHTO Standard Design Specifications for Highway Bridges, 37, 39, 112–113, 204, 217

abutment backfill. See backfill capped pile, 1, 2, 51, 198 continuity connection, 16, 48, 54, 62, 65 embankments. See embankments flexible, 19, 72, 75, 76 flexible supports, 114 integral. See bridges, integral non-integral, 7, 9, 169 pile cap, 97 pile foundations. See piles semi-integral. See bridges, semi-integral settlement, 2, 18, 60 stub-type, 2, 7, 42, 47, 52, 59, 69, 76, 82, 116, 169 wall-type, 7, 82, 166 wingwalls, 161 reinforcement, 9, 69, 117 turn-back, 12, 137, 151 American Iron and Steel Institute, 118 anchor bars, holes, 49–50 angle of internal friction, 62, 127 approach pavement, 170 asphalt (bituminous) concrete, 93 concrete, 28 growth. See G/P phenomenon jointed, 2, 21, 22, 23, 26, 55, 170 rigid, 6, 39, 55 approach slabs, 180 anchors, 2, 151 curbs, 69 cycle-control joints, 95, 133, 151, 244 compression seals, 96 curb inlets, 97 drainage troughs, 55, 97 fingerplates, 97, 162 pressure-relief joints, 45, 55, 56, 57, 117, 245 strip seals, 96 diagonal tie bars, 93, 116 full-width, 55, 94, 97, 102, 134 mechanical connectors, 54, 137 polyethylene sheets, 55, 88, 136 seats, 93 assumptions realistic, 99 simplifying, 65–67, 100, 115 awareness, lack of, contributing reasons economics, xii, 209, 212 habit, 209–210, 212,

Index

language, 210, 212 preoccupation, 207, 212 awareness of change, 186, 192–195, 212 of differences, 186, 196–197, 212 of reality, 185–186, 190, 208, 212, 213 of similarities, 197 of things, 186–192 backfill compressible, 170, 197 compression, 56, 62, 125 consolidation, 55, 63 erosion, 12, 55, 134 expansion, 56 frictional resistance, 126 granular, 62, 69 placement procedures, 137 shearing resistance, 137 well-drained, select granular, 12, 69, 117 beams, uplift. See buoyancy countermeasures bearings abutment, 113 anchor bars, 49–50 bolster, 23, 167, 177 compound, 168 elastomeric, 48, 94, 123, 133, 168, 170, 172, 173, 180, 181, 198–199, 204, 206 fixed, 25, 47, 177 guide, 93, 124, 125, 126, 129, 130–132, 144, 154, 181, 197, 244 accessible, 131, 197 replaceable, 197 movable, 2, 14, 22, 41, 42, 43, 44, 47, 49, 59, 65, 158, 159, 161, 169, 170, 173, 180, 181 rockers, 167 roller, 177, 178 blow-up Allen Road, Toronto Canada, 229 approach slabs, 216 pavement, 31–32, 35, 215, 223 records, 225–228 stone-block streets, 31, 228 bridge, cast-in-place concrete, 103, 150, 168 characteristics aesthetics, 109 durability, 109 economy, 109, 116

249

function, 109 safety, 116 collapsed, 46, 88 construction, foundations integral, 52, 60 procedures, 10, 53, 57, 68 stage, 46, 192 continuous, concrete slab, 71, 72, 74 end-jointed, 14, 22, 32, 35 integral, 7–8, 35, 48 multiple-span, 3, 60, 72, 75, 137, 216 prestressed box-beam, 5, 16, 71, 72, 191, 192, 198, 205 steel stringer-type, 74 superstructures, 19 bridge, composite concrete conversion techniques, 17 cost-effective, 2 deck-type, xii, 89, 94, 103, 112 deflections, 64, 87 durability, 183 functional, 183 foundation restraint, 10 foundations, old, 48 grade separation, 50, 107, 113 Inspector’s Training Manual, 37, 38 bridge, integral abutment, 2, 4, 11, 12, 19, 86, 118, 169, 206, 207, 244 aesthetics, 116 attributes broad span ratios, 50, 51 compressive resistant, xii, 230 cost-effective, 2 dry construction, 45 durable, 2, 43, 100, 116 load distribution, 113 pressure resistant, 22, 42 rapid construction. See Bridge, integral, construction safety, 116 simple design, 43–44 simple replacement, 51 simple widening, 51 concept, 93, 244 concrete, 7, 59 construction, 43, 53, 59 broad tolerances, 46, 47 embankments, 46 few parts, 27 no cofferdams, 46

250

Index

small excavations, 46 simple beam seats, 47 simple forms, 47 vertical piles, 46 cost-effective, 2, 42 integrity, 2, 43, 100 limitations alignment, xi application range, 52, 68 approach guard rail connections, 53 approach slabs required, 12, 55, 56, 88 buoyant. See buoyancy, countermeasures continuity required, 117 curvature, 57, 112, 116, 118 cycle-control joints. See approach slabs flexible abutments, 9, 53, 59, 182 length, 7–8, 11, 43, 52, 57, 69, 116 overburden depth, 5, 57 pile length, 13 pile stresses, 10, 13, 51–52, 139 settlement control, 64–65 skew, 13, 43, 57, 69, 76, 112, 116, 118 uplift. See buoyancy load capacity, 59 multiple-span continuous, 3, 43, 47, 59, 69, 216 pier. See piers precast prestressed, 1 prestressed concrete, 10, 16 reinforced concrete slab, 3, 16, 113, 116, 194 replacement, 89 restraint, longitudinal active earth pressure, 68 approach slab/subbase friction, 169 backfill compression, 56 bearing shear, 124 passive pressure, 11, 55, 59, 112 wingwall/backfill friction, 7 rolled steel beams, 3 bridge, semi-integral attributes aesthetics, 142 broad application range, 123 broad skew range, 91 compression resistant, xii, 124 durable, 140 earthquake resistant, 50, 124

jointless deck, 122, 132, 147, 152, 169, 239 length, 153 rigid foundations, 114, 121, 152, 196 concept, 121, 134, 140–142, 180, 205, 244 experience, 144 limitations alignment, 153 application range, 140 approach slabs, 136, 141, 153 buoyant, 153 continuity required, 153 cycle-control joints, 141, 153, 154, 197 lateral force control, 153 length, 153 restricted settlement, 152 rigid foundations, 132, 141, 153 representative details, 147 seismic research, 200 use, 142 bridge, settlement, 2 Bugler, John W., 234, 241 buoyancy, countermeasures counterweights, 67 drain holes, 132 floodwater clearance, 53 integral abutments, 86, 118, 197, 206, 207, 244 integral piers, 132 mechanical hold-down connections, 53, 88 vent holes, 53 cofferdams, 46 columns flexible, 73, 196 slender, 196 composite structure. See structure movement systems concrete closure placements, 86, 88, 97 crack sealers, 85, 87 curing blankets, 61, 84 diaphragms, 16 end diaphragms, 84, 85, 87, 89, 91, 95, 136, 180 finishing, machines, 84, 85 forms, 61 high performance, 85 high strength, 85

Index

placement, 16, 47 days, 54, 87, 88, 91 night placement, 85–86, 87 procedures, 84 rapid, 85 sequences, 84, 85 set-retarding admixtures, 85 water curing, continuous, 87 construction accelerated, 82 all-weather, 82 cast-in-place, 10 continuous, 69, 116 jointless, 44–45 procedures, 53 specifications. See AASHTO Standard Design Specifications for Highway Bridges stage, 89, 192, 194 continuity, connections abutments, 54, 61, 62, 88 cast-in-place, 10, 50, 86, 116 moments, 3, 61, 65 reinforcement, 16, 65, 69, 191 superstructure/abutment, 44, 48, 67, 86, superstructure/pier, 16, 44, 68 continuous, construction deck slab, 73, 75, 112 frames, 3, 9 highway bridges, xi, xii, xiii, 3, 74 multiple span bridges, 69, 216 contraction, 9, 10, 103, 107–108, 175 conversions, integral, 15, 17, 82 corrosion, 7 counterweights. See buoyancy cracking diagonal, bridge deck, 82, 97 early age deck slab, 82, 83–85, 86, 87, 97 flexural, deck slab, 9, 85 sealant, 23 transverse, bridge deck, 83, 84, 85, 113 creep. See stresses, secondary Cross, Hardy, 3, 99–100 cycle-control joints, 2, 55, 57, 95, 151, 243 debris, infiltration, roadway, compression resistant, 30, 93, 102, 221, 237, 240 deck, drainage curb inlets, 102 downspouts, 102

251

hinges, 17 horizontal conductors, 102 joints. See joints scuppers, 102 deck slab closure placements, 89, 90, 91 concrete placement procedures, 84, 86, 162 sequences, 84, 86, 112 speed, 112 machine finishing, 84, 85, 87 transverse cracking, 83, 84, 86, 87, 224 de-icing chemical deterioration, xii, 7, 8, 113, 122, 152 design intuition, 182 prudent judgment, 19 simplified assumptions, 62 design, specifications. See AASHTO Standard Design Specifications for Highway Bridges deterioration, environmental, 194 diaphragm concrete end, 141, 204 concrete placement, 135–136 transverse, end, 83 drainage, deck, 7, 102, 151 earthquake North Ridge, 147 resistance, 200 elastomeric compression seals, 7, 96, 133 erection devices bearings, 48, 50, 51, 122, 123, 124, 133, 145, 180, 181, 198–199, 203, 206–207 strips, 48–49, 206–207 joint seals, 7, 47, 96, 97, 124, 133, 181 membrane. See deck slab elementalism, 157, 159, 163–165, 166, 243 elementalistic, approach, 157, 159, 175 embankment benches, 63, 69, 117 consolidating, approach, 93–94, 116, 158 consolidation, 82, 94, 117, 129, 165, 170, 175, 182, 198 construction, 44, 53–54 erosion, 82, 97 mechanical stabilized, 169 placement procedure, 84, 137

252

Index

scour, 170 settlement, 165 side-slope drainage flumes, 97, 102, 134, 154 spill-around slopes, 63 stable, 169 waiting period, 69 end diaphragms concrete, 85, 180, 243 concrete placement, 87, 89, 135–136, 180 factors distribution, 65, 164 stiffness, 43, 65 fatigue design categories, 114 procedures, 114 specifications, 114 Federal Highway Administration, 14, 235 forces lateral, 112 longitudinal, 13, 31, 39, 64, 118, 124, 152, 162, 165 foundations, abutment flexibility, 72 flexible, end-bearing flexible piles, 74, 121, 181 flexible capped piles, 68, 75, 82, 197 rigid battered piles, 82, 196 bedrock, 75 drilled shafts, 121, 181, 196 pedestals on bedrock, 75, 82, 121, 122, 181, 196 semi-rigid, 12, 19 G/P generation factors concrete strength, 223 de-icing chemicals applications, 7, 16, 30, 77, 103, 219, 224 joint design and spacing, 223 joint spacing, 228, 237 pavement age, 16–17, 25, 28, 229–230 rainfall, 223 sealant maintenance, 224 sealant quality, 223 subgrade composition, 223 subgrade drainage, 223 temperature range, 54, 223 traffic volume, 31, 224 G/P phenomenon, xi, xii, 11, 21–39, 43, 45, 112, 123, 133, 215–242, 243

General Structures Committee, 157 guard rail connections, 53 Hindman, William S., 76–77 hinges, plastic. See piles holes drain. See buoyancy vent. See buoyancy holism, 157, 244 holistic, approach, 157 evaluation, 174 views, 159, 162, 163–165, 166, 167, 168, 180 view boundaries, 175, 176, 177 integral construction. See bridges, integral integrated superstructure system, 18–19, 166 interface abutment backfill, 170, 197 approach-slab/aggregate-base, 93 approach-slab/approach-pavement, 2, 53, 132, 133, 197 approach-slab/approach-sidewalk, 96 approach-slab/superstructure, 93 bridge/embankment, 50, 96 end diaphragm/backfill, 126, 129, 135, 151, 180 polystyrene/concrete, 20 superstructure/abutment, 2, 7, 17, 19, 23, 25, 32, 34, 35, 42, 43, 59, 82, 85, 152, 171, 180, 182, 215, 232 joint closed, 45 contraction sealed or unsealed, 240 spacing, 237 cycle-control. See approach slabs, 2 debris infiltration, 30–31, 233 elastomeric, seal, 7, 47 expansion, 8, 9, 73, 74, 106 fillers compressible, 133 polyethylene, 25 movable, deck, 2, 14, 22, 41, 42, 45, 59, 67, 71, 73, 82, 86, 94, 96, 103, 106, 107, 114, 122, 158, 159, 161, 162, 168, 171–172, 173, 177, 182, 215, 233, 239, 244 longitudinal, 2, 23, 82, 91, 132, 181 transverse, 1, 95 open, 109

Index

pavement expansion, 233, 239 pressure relief, 25, 26, 27, 29, 34, 35, 38, 39, 45, 56, 60, 96, 117, 134, 233, 235, 236, 239 saw-cut, 219 sealed, 7, 45, 221, 235 sliding plate, 7, 96, 162 transverse deck, 1 unsealed, 30, 221, 234–235 Land of No Special Computations, xiii, 12, 108, 109, 111, 115, 116, 118 loads dead, 60 horizontal, 68 lateral, 118 live distribution, 113 surcharge, 55 longitudinal, 118 ultimate capacity, 60 mechanical connectors, 88 modulus, elastic, 60, 94 moments continuity connection, 16, 61, 62 negative, 15, 61, 62 positive, 16, 50, 61, 62, 65 movement systems. See structure movement systems movements abnormal rotations, 167 differential, 86 horizontal, 93 longitudinal, 87 rotational, 87 National Bridge Inventory, 118 Naval facilities command, 70, 138 nouns apparition type, 210 ethereal type, 210 process type, 210 singular, 210–211 static type, 210 Odd Albert’s Method, 115 pavement approach blow-up, 31–32, 35, 215, 225, 235, 236

253

compressive stresses, 31, 225 contraction joints, transverse, 30, 219 expansion joints, 38, 96, 106, 133, 240 forces. See pressure growth. See G/P phenomenon restrained, 21, 37, 64 jointed concrete, 2, 22, 23, 31, 32, 35, 45, 60, 170 pressure, 6, 34, 45 pressure relief joints. See pressure relief joints pumping, 96 PCI manual, 197 PCI Precast/Prestressed Integral Bridges, 16, 197, 205 piers cap-and-column, 175, 188, 200 capped pile, 1, 51, 198 continuity connections, 43, 44, 47 fixed, 44, 68, 160 flexible capped pile, 68, 75 flexible integral, 2, 42, 43, 44 49 self-supporting, 2, 43, 49, 50, 59, 65, 68 semi-rigid, 2, 42, 65, 68 pile battered, 46, 68, 196 cap, connection reinforcement, 9, 44 capped piles. See piers cast-in-place, 10, 52 driving constraints, 199 elastic range, 52 flexible, 42, 59, 65, 69, 82, 116, 169, 197, 206 flexural resistance, 52, 63 flexural stresses, 9, 52 plastic hinges, 13, 52 prebored holes with granular material, 13, 52, 53, 63, 117 precast concrete, 52 prestressed reinforced concrete, 16, 52 single row, 8, 52 steel H, 8, 9, 13, 34, 52, 116 test research, 52 vertically driven, 2, 116, 196 weak axis, 116 polyethylene, sheets, 88, 136, 203 polystyrene, expanded, 94, 142, 145, 201, 204 pressure active, 128 distribution, 63

254

Index

generation curve, 31, 222, 233, 240 generation of, 6, 31, 237 pressure relief joints asphalt (bituminous) concrete, 56, 93, 239 polymer foam filled, 236 sleeper slabs, 56, 96 subbase drains, 56 problems elimination, 100, 101–102 ignore nonproblems, 100, 101–102 recognizing, 100–101 redefining, 100, 102–103 simplifying, 100, 103–110 solving, 100 reinforcement, negative moment, 62 research creep studies, 56, 62 half-scale model, 61 integral-bridge, 8 passive pressure, 56, 62 pile test, 52 seismic, 94, 200 shrinkage, xii, 16, 56, 62 soil/structure interaction, 154 restraint, foundation lateral, 123, 124 longitudinal, 123–124 rotational, 125 roads, interstate primary system, 43, 179 secondary system, 43 roadway shoulders curb inlets, 102 drainage, 95 erosion, 95 side-slope flumes, 154 subsidence, 94 underdrains, 97, 134 rock mechanics, techniques, 215, 222 rotation abnormal, 167 horizontal, 93, 129 superstructure, 197 seals deck joint, 67, 82 elastomeric compression, 7 elastomeric joint, 97, 159 elastomeric sheet, 133 joint, 67, 82

secondary effects. See stresses settlement differential, 174 post-construction, 175 vertical, 181 shrinkage, xii, 9, 30, 44, 51, 60–61, 87, 103, 107–108, 175, 176 skew, xii, 43, 76, 91, 112, 118, 125, 130, 144, 165, 196 skew limitations, 5, 11, 43, 150 snow plows, 67 span continuous, 15 simply supported, 14, 89 width ratio, 113 stage reconstruction, 89 strength fatigue, 90 ultimate, 90 stresses primary buoyancy, xii dead load, 51, 60, 66, 67, 68, 85 earthquakes, xii, 50, 67, 124, 199, 217 flexural, 9 live load, 51, 55, 60, 66, 67, 68, 75, 112 pavement pressure, 7, 16–17, 31, 34, 38, 43, 112, 218–224 secondary creep, 44, 51, 60, 61–62, 65, 66, 67, 68, 176 passive pressure, xii, 11, 44, 52, 55, 57, 60, 61, 65, 66, 68, 139, 180 settlement, xii, 60 shrinkage, xii, 10, 16, 44, 51, 60–61, 66, 67, 68, 75, 87, 107–108, 175 stream flow, 68, 113 temperature, 87 thermal gradients, xii, 52, 60, 61, 65, 67, 68, 75, 84, 176 wind, 68, 113 stringer support bolts, 88 structure durability, 106 integrity, 106 movement subsystems movement systems, xii, 1–2, 93, 97, 157, 159, 161, 162, 165, 168–171, 172, 173, 175, 176, 179, 180, 182, 183, 197, 198, 200, 244 movement systems, 140, 144

Index

primary, 170 secondary, 170 tertiary, 170 structures continuous, 168 grade separation, 50 Study Tour of North America, 8 subsoil consolidation and translation, 69, 158 stability, 53, 57 stable, 50 surcharged, 69, 116–117, 158, 175, 179 substructure, 170 capped pile, 200 flexibility, 2, 51 superstructure, 197 superstructure restraint, longitudinal, 123–124 active earth pressure, 68, 124 approach-slab/subbase friction, 124, 127, 129, 169 backfill compression, 56, 62, 125 bearing shear, 124 passive pressure, 11, 55, 60, 124, 125–126, 127, 128, 180 wingwall/backfill friction, 7, 124 temperature ambient, 54, 93, 160, 219 changes, 29 levels, 65

255

movement coefficient, thermal. See stresses range, 54 thermal gradients, xii, 65, 176 traffic maintained, 89, 193, 194 vehicular, 31, 35, 50, 55, 166, 167, 224 Transportation Research Board (TRB), 157, 244 TRB Structure Movement Systems Subcommittee, 158 uplift. See integral and semi-integral bridge limitations buoyancy. See buoyancy deck placement, 54 mechanical hold-down connections, 53, 88 vehicular traffic. See traffic views elementalistic. See elementalism holistic. See holism multidimensional, 179 welder pre-qualification tests, 3 welding, butt beams, 3 field, 3 fillet, moment plates, 3 splices, 3