Integral and Semi-integral Bridges Martin P. Burke Jr. PE
A John Wiley & Sons, Ltd., Publication
Integral and Semi-i...
762 downloads
3387 Views
9MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
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 wa