Adjacent Precast Concrete Box Beam Bridges: Connection Details (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

TRANSPORTATION RESEARCH BOARD WASHINGTON, D.C. 2009 www.TRB.org NAT IONAL COOPERAT IVE H IGHWAY RESEARCH PROGRAM NCHRP SYNTHESIS 393 Research Sponsored by the American Association of State Highway and Transportation Officials in Cooperation with the Federal Highway Administration SUBJECT AREAS Bridges, Other Structures, Hydraulics and Hydrology, and Maintenance Adjacent Precast Concrete Box Beam Bridges: Connection Details A Synthesis of Highway Practice CONSULTANT HENRY G. RUSSELL Henry G. Russell, Inc. Glenview, Illinois

NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM Systematic, well-designed research provides the most effective approach to the solution of many problems facing highway administra- tors and engineers. Often, highway problems are of local interest and can best be studied by highway departments individually or in coop- eration with their state universities and others. However, the accelerat- ing growth of highway transportation develops increasingly complex problems of wide interest to highway authorities. These problems are best studied through a coordinated program of cooperative research. In recognition of these needs, the highway administrators of the American Association of State Highway and Transportation Officials initiated in 1962 an objective national highway research program employing modern scientific techniques. This program is supported on a continuing basis by funds from participating member states of the Association and it receives the full cooperation and support of the Federal Highway Administration, United States Department of Trans- portation. The Transportation Research Board of the National Research Coun- cil was requested by the Association to administer the research pro- gram because of the Board’s recognized objectivity and understanding of modern research practices. The Board is uniquely suited for this purpose as it maintains an extensive committee structure from which authorities on any highway transportation subject may be drawn; it possesses avenues of communication and cooperation with federal, state, and local governmental agencies, universities, and industry; its relationship to the National Research Council is an insurance of objec- tivity; it maintains a full-time research correlation staff of specialists in highway transportation matters to bring the findings of research directly to those who are in a position to use them. The program is developed on the basis of research needs identified by chief administrators of the highway and transportation departments and by committees of AASHTO. Each year, specific areas of research needs to be included in the program are proposed to the National Research Council and the Board by the American Association of State Highway and Transportation Officials. Research projects to fulfill these needs are defined by the Board, and qualified research agencies are selected from those that have submitted proposals. Administration and surveillance of research contracts are the responsibilities of the National Research Council and the Transportation Research Board. The needs for highway research are many, and the National Coop- erative Highway Research Program can make significant contributions to the solution of highway transportation problems of mutual concern to many responsible groups. The program, however, is intended to complement rather than to substitute for or duplicate other highway research programs. NCHRP SYNTHESIS 393 Project 20-5 (Topic 39-10) ISSN 0547-5570 ISBN 978-0-309-098304 Library of Congress Control No. 2009900728 © 2009 Transportation Research Board COPYRIGHT PERMISSION Authors herein are responsible for the authenticity of their manuscripts and for obtaining written permissions from publishers or persons who own the copyright to any previously published or copyrighted material used herein. Cooperative Research Programs (CRP) grants permission to repro- duce material in this publication for classroom and not-for-profit pur- poses. Permission is given with the understanding that non of the mate- rial will be used to imply TRB, AASHTO, FAA, FHWA, FMSCA, FTA, or Transit development Corporation endorsement of a particular product, method, or practice. It is expected that those reproducing the material in this document for educational and not-for-profit uses will give appropriate acknowledgment of the source of any development or reproduced material. For other uses of the material, request permission from CRP. NOTICE The project that is the subject of this report was a part of the National Co-operative Highway Research Program conducted by the Transpor- tation Research Board with the approval of the Governing Board of the National Research Council. Such approval reflects the Governing Board’s judgment that the program concerned is of national impor- tance and appropriate with respect to both the purposes and resources of the National Research Council. The members of the technical committee selected to monitor this project and to review this report were chosen for recognized scholarly competence and with due consideration for the balance of disciplines appropriate to the project. The opinions and conclusions expressed or implied are those of the research agency that performed the research, and, while they have been accepted as appropriate by the technical com- mittee, they are not necessarily those of the Transportation Research Board, the National Research Council, the American Association of State Highway and Transportation Officials, or the Federal Highway Administration of the U.S. Department of Transportation. Each report is reviewed and accepted for publication by the tech- nical committee according to procedures established and monitored by the Transportation Research Board Executive Committee and the Governing Board of the National Research Council. Published reports of the NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM are available from: Transportation Research Board Business Office 500 Fifth Street, NW Washington, DC 20001 and can be ordered through the Internet at: http://www.national-academies.org/trb/bookstore Printed in the United States of America NOTE: The Transportation Research Board of the National Academies, the National Research Council, the Federal Highway Administration, the American Association of State Highway and Transportation Officials, and the individual states participating in the National Cooperative Highway Research Program do not endorse products or manufacturers. Trade or man- ufacturers’ names appear herein solely because they are considered essential to the object of this report.

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

nchrp coMMiTTee For proJecT 20-5 chair GARY D. TAYLOR, CTE Engineers MeMBers KATHLEEN S. AMES, Illinois DOT STUART D. ANDERSON, Texas A&M University CYNTHIA J. BURBANK, PB Americas, Inc. LISA FREESE, Scoot County (MN) Public Works Division MALCOLM T. KERLEY, Virginia DOT RICHARD D. LAND, California DOT JAMES W. MARCH, Federal Highway Administration MARK A. MAREK, Texas DOT JOHN M. MASON, JR., Auburn University ANANTH PRASAD, HNTB Corporation ROBERT L. SACK, New York State DOT FRANCINE SHAW-WHITSON, Federal Highway Administration LARRY VELASQUEZ, New Mexico DOT FhWa Liaison WILLIAM ZACCAGNINO TrB Liaison STEPHEN F. MAHER acKnoWLedGMenTs I would like to thank the Virginia Department of Transportation for providing the drawings. Cover figure: Washington Street Bridge, Dayton, Ohio, precast concrete box beam bridge. cooperaTiVe research proGraMs sTaFF CHRISTOPHER W. JENKS, Director, Cooperative Research Programs CRAWFORD F. JENCKS, Deputy Director, Cooperative Research Programs NANDA SRINIVASAN, Senior Program Officer EILEEN DELANEY, Director of Publications nchrp sYnThesis sTaFF STEPHEN R. GODWIN, Director for Studies and Special Programs JON M. WILLIAMS, Program Director, IDEA and Synthesis Studies GAIL STABA, Senior Program Officer DONNA L. VLASAK, Senior Program Officer DON TIPPMAN, Editor CHERYL KEITH, Senior Program Assistant Topic paneL DONALD W. HERBERT, Pennsylvania Department of Transportation AHMED IBRAHIM, California Department of Transportation STEPHEN F. MAHER, Transportation Research Board BARNEY MARTIN, Modjeski & Masters, Poughkeepsie, NY BASILE RABBAT, Portland Cement Association ANDREA SCHOKKER, Pennsylvania State University MICHAEL B. TWISS, New York State Department of Transportation JULIUS F.J. VOLGYI, JR., Virginia Department of Transportation BENJAMIN GRAYBEAL, Federal Highway Administration (Liaison) GARY JAKOVICH, Federal Highway Administration (Liaison)

Highway administrators, engineers, and researchers often face problems for which infor- mation already exists, either in documented form or as undocumented experience and practice. This information may be fragmented, scattered, and unevaluated. As a conse- quence, full knowledge of what has been learned about a problem may not be brought to bear on its solution. Costly research findings may go unused, valuable experience may be overlooked, and due consideration may not be given to recommended practices for solving or alleviating the problem. There is information on nearly every subject of concern to highway administrators and engineers. Much of it derives from research or from the work of practitioners faced with problems in their day-to-day work. To provide a systematic means for assembling and evaluating such useful information and to make it available to the entire highway commu- nity, the American Association of State Highway and Transportation Officials—through the mechanism of the National Cooperative Highway Research Program—authorized the Transportation Research Board to undertake a continuing study. This study, NCHRP Proj- ect 20-5, “Synthesis of Information Related to Highway Problems,” searches out and syn- thesizes useful knowledge from all available sources and prepares concise, documented reports on specific topics. Reports from this endeavor constitute an NCHRP report series, Synthesis of Highway Practice. This synthesis series reports on current knowledge and practice, in a compact format, without the detailed directions usually found in handbooks or design manuals. Each report in the series provides a compendium of the best knowledge available on those measures found to be the most successful in resolving specific problems. Bridges built with adjacent precast, prestressed concrete box beams are a popular and economical solution in many states because they can be constructed rapidly and most deck forming is eliminated. Bridges constructed with box beams have been in service for many years and have generally performed well. A recurring problem, however, is cracking in the longitudinal grouted joints between adjacent beams, resulting in reflective cracks forming in the wearing surface. This in turn may lead to leakage, corrosion, and, in severe cases, complete cracking of joints and loss of load transfer. This study discusses current design and construction practices that are reported to reduce the likelihood of longitudinal crack- ing in box beam bridges. Information for the study was gathered through a literature review. In addition, state and Canadian provincial transportation agencies were surveyed, and the survey was aug- mented with selected individual interviews. Henry G. Russell, Henry G. Russell, Inc., Glenview, Illinois, collected and synthesized the information and wrote the report. The members of the topic panel are acknowledged on the preceding page. This synthesis is an immediately useful document that records the practices that were acceptable within the limitations of the knowledge available at the time of its preparation. As progress in research and practice continues, new knowledge will be added to that now at hand. ForeWord preFace By Jon M. Williams Program Director Transportation Research Board

conTenTs 1 SUMMARY 3 CHAPTER ONE INTRODUCTION Background, 3 History, 4 Scope, 4 6 CHAPTER TWO STRUCTURAL DESIGN AND DETAILS Span Lengths, 6 Skew Angles, 6 Beam Cross Sections, 6 Composite Versus Noncomposite Designs, 6 Keyway Configurations, 7 Transverse Tie Details, 8 Design Criteria for Connections, 10 Exterior Beam Details, 12 13 CHAPTER THREE SPECIFICATIONS AND CONSTRUCTION PRACTICES AASHTO Standard Specifications, 13 AASHTO LRFD Specifications, 13 Bearing Types, 14 Construction Sequence, 14 Differential Camber, 16 Keyway Preparation, 16 Grouting Materials and Practices, 16 17 CHAPTER FOUR LONG-TERM PERFORMANCE, MAINTENANCE, AND REPAIRS Types of Distress, 17 States Reporting Little or No Observed Distress, 17 Other Observations, 18 Maintenance Procedures, 20 Repair Procedures, 20 Factors Affecting Long-Term Performance, 20 23 CHAPTER FIVE INSPECTION PRACTICES Visual Inspection, 23 Identification of Corrosion, 23 Other Practices, 23 25 CHAPTER SIX RESEARCH Materials Research, 25 Structural Research, 25 27 CHAPTER SEVEN CONCLUSIONS AND SUGGESTIONS FOR FUTURE RESEARCH Conclusions, 27 Suggestions for Future Research, 28 29 REFERENCES

31 APPENDIX A SURVEY QUESTIONNAIRE 43 APPENDIX B SUMMARY OF RESPONSES TO SURVEY QUESTIONNAIRE 67 APPENDIX C BEAM AND CONNECTION DETAILS 74 APPENDIX D RESEARCH PROBLEM STATEMENT

Bridges built with adjacent precast, prestressed concrete box beams were first introduced in the 1950s and are a common and economical solution in many states because they can be constructed rapidly and most deck forming is eliminated. Today, box beam bridges are used in about two-thirds of the states. Based on the survey conducted for this synthesis, the current practice for box beam bridges is as follows: Approximately half of the states with box beam bridges use AASHTO/PCI cross-• sectional shapes. Span lengths range from less than 20 ft to more than 80 ft.• The most common maximum skew angle between the abutment and the perpendicu-• lar to the bridge centerline is 30 degrees. Most states use simple spans with a cast-in-place concrete deck.• Where a composite deck is used for continuous spans, the bridges are generally • designed to be continuous for live load. Most longitudinal keyways between adjacent box beams are partial depth.• The most common transverse tie consists of unbonded post-tensioned strands • or bars. Approximately half the states grout the keyway before post-tensioning and approxi-• mately half after post-tensioning. There is no consensus about the number of transverse ties and the magnitude of post-• tensioning force. Exterior and interior beams generally use the same design.• Most bridges have either full-width support or two-point supports on each end.• More states use plain elastomeric bearings than laminated elastomeric bearings.• In single stage construction, all beams are generally connected transversely at • one time. In two-stage construction, a variety of sequences is used.• Approximately half the states require sandblasting of the keyway before erection. • The sandblasting is always done before shipment. The most common grout used for the keyways is a nonshrink grout.• Approximately half the states require the use of wet curing or curing compounds for • the grout. Respondents to the survey included 35 state departments of transportation, five Cana- dian provincial transportation agencies, three railroads, and 13 U.S. counties. Bridges constructed using box beams have been in service for many years and have generally performed well. However, a recurring problem is cracking in the longitudinal grouted joints between adjacent beams, resulting in reflective cracks that form in the wear- ing surface, if present. The cracking appears to be either initiated by stresses associated with temperature gradients and then propagates as a result of live load, or is caused by a combination of stresses from temperature gradients and live load. In bridges with partial- depth keyways, the cracking may be initiated by tensile stresses caused by the post-ten- sioning. In most cases, the cracking leads to leakage, which allows chloride-laden water sUMMarY adJacenT precasT concreTe BoX BeaM BridGes: connecTion deTaiLs

2 to saturate the sides and bottoms of the beams. This eventually can cause corrosion of the non-prestressed reinforcement, prestressing strand, and transverse tie. In severe cases, the joints crack completely and load transfer is lost. Unless deterioration or leakage at the joint is evident from the underside of the bridge, there is no way to know easily the extent of dete- rioration at internal joints. Consequently, it is better to design and build box beam bridges so that cracking does not occur. The following practices can reduce the likelihood of longitudinal cracking in box beam bridges: Design Practices• Requiring full-depth shear keys that can be grouted easily – Providing transverse post-tensioning so that tensile stresses do not occur across the – joint Requiring a cast-in-place, reinforced concrete, composite deck with a specified con- – crete compressive strength of 4,000 psi and a minimum thickness of 5 in., to limit the potential for longitudinal deck cracking Construction Practices• Using stay-in-place expanded polystyrene to form the voids – Sandblasting the longitudinal keyway surfaces of the box beams immediately before – shipping to provide a better bonding surface for the grout Cleaning the keyway surfaces with compressed air or water before erection of the – beams to provide a better bonding surface for the grout Grouting the keyways before transversely post-tensioning to ensure compression – in the grout Using a grout that provides a high bond strength to the box beam keyway surfaces – to limit cracking Providing proper curing for the grout to reduce shrinkage stresses and ensure proper – strength development Providing wet curing of the concrete deck for at least 7 days to reduce the potential – for shrinkage cracking and to provide a durable surface It is suggested that the following practices be avoided: Design Practices• Using nontensioned transverse ties, because they do not prevent cracking – Construction Practices• Using an asphalt wearing surface unless a waterproofing membrane is used, because – water accumulates below the asphalt Using nonprepackaged products for grout in the keyways – This synthesis has identified the need for additional research related to the design, dura- bility, and repair of adjacent box beam bridges. A research problem statement addressing research on the design issues is included.

3 There is a new thrust to use these bridges for rapid con- struction under the FHWA Highways for LIFE program. The purpose of this program is to advance Longer-lasting high- way infrastructure using Innovations to accomplish the Fast construction of Efficient and safe highways and bridges. An all-precast concrete bridge that used adjacent box beams for the superstructure and was constructed in 30 days is shown in Figure 4. According to recent National Bridge Inventory data, adjacent concrete box beam bridges constitute about one-sixth of the bridges built annually on public roads. BACKGROUND Bridges built with adjacent precast, prestressed concrete box beams are a common and economical solution in many states, because they can be constructed rapidly and deck forming is eliminated. The bridges may be single span as shown in Figure 1 or multiple spans as shown in Figure 2. They have proved to be economical for major river crossings, such as shown in Figure 3. Although this bridge resembles an arch, the superstructure consists of adjacent precast, prestressed concrete box beams. CHAPTER ONE INTRODUCTION FIGURE 1 Single-span box beam bridge over railroad (Source: New York State DOT). FIGURE 3 Adjacent box beam bridge to replicate an arch bridge (Source: Henry G. Russell). FIGURE 4 Adjacent box beams used for the superstructure of the Davis Narrows Bridge, Maine (Source: Maine DOT). FIGURE 2 Three span box beam bridge over a ravine (Source: Illinois DOT).

4 between adjacent beams was 0.25 in. For spans up to 40 ft, a transverse tie rod was used at the center of the span. For span lengths greater than 40 ft, the tie rods were used at third points along the span. The ties were located at mid-depth of the box beam. The specified thickness of the asphalt wearing course was 3 in. In subsequent years, Michigan made numerous design changes including deeper shear keys, wider beam spacings, use of closed instead of open stirrups, use of post-tensioning tendons for the transverse ties, replacement of cardboard void forms with polystyrene, additional transverse ties, dif- ferent cross-sectional dimensions for the box beams, and the use of a 6-in.-thick composite reinforced concrete deck with a single layer of reinforcement. Since the early use of box beams, many changes in pre- stressed concrete design and construction have contributed to changes in box beam details (Macioce et al. 2007). These changes include the following: Improved design criteria and design methods• Higher concrete compressive strengths• Lower permeability concretes• Larger diameter strands• Low-relaxation strands• Epoxy-coated non-prestressed reinforcement• Expanded polystyrene to form the voids• Curing practices• Thicker concrete cover• scope This synthesis documents the different types of grout key configurations, grouts, and transverse tie systems that cur- rently are being used in the United States and Canada, and how each type has performed. The synthesis includes the following: Practices and details that have proven to enhance the • performance of box beam bridges Practices and details to avoid • Specific areas of interest, including the impact of the • following: Span range – Bridge skew – Bearing types – Topped and nontopped beams – Transverse tie details – Phased construction – Waterproofing membranes – Exterior beam details, including connections to the – barrier and parapet Grout specifications• The box beams are generally connected by grout placed in a keyway between each of the units, and usually with trans- verse ties. Partial-depth or full-depth keyways are typically used, incorporating grouts using various mixes. Transverse ties, grouted or ungrouted, vary from a limited number of nontensioned threaded rods to several high-strength tendons post-tensioned in multiple stages. In some cases, no topping is applied to the structure, whereas in other cases, a noncom- posite topping or a composite structural slab is added. Bridges constructed using box beams have been in service for many years and generally have performed well. How- ever, one recurring problem is cracking in the grouted joints between adjacent units, which results in reflective cracks forming in the wearing surface. In most cases, the cracking leads to leakage, which allows chloride-laden water to satu- rate the sides and bottom of the beams, eventually causing corrosion of the non-prestressed reinforcement, prestressing strand, and transverse ties. In severe cases, the joints crack completely and load transfer is lost. There is no design method for shear keys in the AASHTO Standard Specifications for Highway Bridges or the AASHTO Load and Resistance Factor Design (LRFD) Bridge Design Specifications. Most shear-key details in use are regional “standard details” of uncertain origin, and there is no infor- mation on the magnitude of forces induced in the shear keys or on the ability of a given detail to resist these forces. hisTorY According to Miller et al. (1999), the use of prestressed con- crete adjacent box beams started in about 1950 for bridges with span lengths of 30 to 100 ft, and these box beams are widely used today for these span lengths. The beam design evolved from an open channel design. Shear keys in the top flange were used to transfer the load between adjacent beams. When the load is transferred this way, torsion occurs in the section and, hence, a bottom flange is needed to convert the open section into a torsionally stiff closed section. Macioce et al. (2007) reported that adjacent box beam bridges constructed of noncomposite prestressed concrete with an asphalt wearing surface were developed during the interstate construction period to provide a shallow super- structure, rapid uncomplicated construction, and low initial costs. In many circumstances, this bridge type was used on low-volume roads. The history of adjacent box beam bridges in Michigan was described by Attanayake and Aktan (2008). Box beam bridges were first introduced in 1954 and consisted of either single-cell or double-cell units. The stirrups were open and did not extend to the bottom flange. The specified spacing

5 Information gathered in this synthesis provides a basis for understanding the behavior of adjacent concrete box beam bridges and will help establish the most practical and effi- cient details to reduce maintenance costs and extend bridge service life. The remaining text of this synthesis is organized as follows: Chapter two identifies and discusses the items that are • generally considered during the design stage. These include span lengths, skew angles, and beam cross sec- tions; composite versus noncomposite design; keyway configurations; transverse tie details; design criteria for connections; and exterior beam details. Chapter three reviews information that is available in • the AASHTO Specifications. Construction practices that affect keyway performance, such as bearing types, construction sequence, differential camber, keyway preparation, grouting materials, and grouting prac- tices, are discussed. Chapter four identifies the types of observed distress, • maintenance procedures, repair procedures, and fac- tors affecting long-term performance. Chapter five identifies what inspection techniques, • other than visual inspection, have been used. Chapter six summarizes relevant recent and ongo-• ing research in materials technology and structural design. Chapter seven summarizes the best practices and • details that have proven to enhance the performance of box beam bridges and reviews practices and details to avoid. Design and construction issues requiring fur- ther research are listed. Appendixes provide the survey questionnaire (Appendix A), a summary of the responses to the questionnaire (Appen- dix B), beam and connection details (Appendix C), and the research problem statement (Appendix D). Inspection practices• Bridge maintenance, including rehabilitation and ret-• rofitting techniques Sources of ongoing or completed analytical or experi-• mental research pertaining to the design or construc- tion of this type of bridge Design and construction issues that require further • research and evaluation This information was gathered from literature reviews and surveys of state highway agencies through the AASHTO Highway Subcommittee on Bridges and Structures; Cana- dian Provinces through the Transportation Association of Canada, Class 1 railroads; U.S. counties through the National Association of County Engineers; and industry through the Precast/Prestressed Concrete Institute (PCI). Some infor- mation on Japanese practice is included. Fifty-eight complete responses were received, including 21 from owners who do not use adjacent box beam bridges. A follow-up was made with those states that did not respond to determine whether they used box beams. The usage by state highway agencies is illustrated in Figure 5. In subse- quent chapters, the information from the survey is summa- rized as state responses and total survey responses. In some cases, the percentages total more than 100 because more than one answer was possible, and some states reported the use of multiple practices. FIGURE 5 Usage of box beam bridges by state.

6 essary to modify some of the responses based on each state’s definition of skew angle. The responses are summarized in Figure 8 for the skew angle defined in Figure 7a. Some states indicated that they do allow exceptions to their normal maxi- mum values. BEAM CROSS SECTIONS In the survey conducted for this synthesis, 50% of the state respondents and 54% of the total respondents reported that they use AASHTO/PCI-shaped box beams. Approximately 30% use state standards and the remainder use other cross sections. The other cross sections used by respondents were reported as PCI Northeast and Canadian PCI standards. Drawings of the AASHTO/PCI cross sections are included in Appendix C. COMPOSITE VERSUS NONCOMPOSITE DESIGNS In the survey conducted for this synthesis, respondents iden- tified the types of box beam superstructures that they build. The responses, shown in Figure 9, indicate that the type used by most respondents consists of simple spans with a cast-in-place concrete wearing surface. For all bridges with cast-in-place concrete wearing surfaces, the specified mini- mum thickness ranged from 4.5 to 6 in. for state agencies SPAN LENGTHS In the survey conducted for this synthesis, respondents reported on the span lengths for which adjacent box beams were used. The results, shown in Figure 6 individually for the states and the total survey, indicate that adjacent box beams are used for span lengths ranging from less than 20 ft to more than 80 ft. The PCI Bridge Design Manual (PCI 1997/2004) includes preliminary design charts for AASHTO box beams with span lengths ranging from 40 to 140 ft. SKEW ANGLES The survey asked about the maximum skew angle used for box beam bridges. The skew angle of a bridge can be defined in two ways, as shown in Figure 7. Consequently, it was nec- CHAPTER TWO STRUCTURAL DESIGN AND DETAILS FIGURE 6 Survey results for span lengths. FIGURE 8 Survey responses for skew angles. FIGURE 7 Denition of skew angle. a. b.

7 In New York State before 1992, adjacent box beams were connected through a grouted shear key designed to transfer shear force between individual beam units (Lall et al. 1997, 1998). The keyway extended to a depth of about 12 in. from the top of the beam. A cast-in-place deck, at least 6 in. thick and reinforced with welded wire reinforcement, was made composite with stirrups projecting from the beams. Lon- gitudinal cracks began appearing in the concrete overlays immediately after construction. Over time, cracks devel- oped over nearly all shear keys. A survey indicated that 54% of the box beam bridges built between 1985 and 1990 had developed longitudinal cracks over the shear keys (Lall et al. 1997, 1998). In 1992, a design change was made to increase the depth of the shear key to almost the full depth of the precast unit. A change to the transverse tie requirements, as discussed in the next section, was also made. A 1996 survey of 91 bridges built from 1992 through early 1996 found that 23% of the bridges had shear key–related longitudinal cracking compared with 54% for the previous inspection. Analysis of the data by age of structure at time of inspection showed that the new full-depth shear-key system reduced the percentage of decks with cracks by about 50% (Lall et al. 1997, 1998). and 3 to 9 in. for the total survey. For owners that build con- tinuous spans with a cast-in-place concrete deck, approxi- mately 90% design the bridges for live-load continuity. The “other” superstructure type, reported by Massachusetts, New Hampshire, and three Canadian provinces, was a cast- in-place concrete topping with waterproofing membrane and asphalt wearing surface. El-Remaily et al. (1996) reported that composite topping is not a structurally efficient solution for the transfer of forces at the longitudinal joint, because it does not control differ- ential rotation of the box and it is not an economical solution because a composite concrete topping costs about four times as much as a thin layer of bituminous concrete. KEYWAY CONFIGURATIONS Keyway configurations are generally defined as partial depth or full depth. The depth refers to that of the grout and not the depth of the box beam. Therefore, a full-depth keyway does not extend to the bottom of the beam because a gasket must be placed near the bottom of the beam to prevent the grout from falling out. Typical keyway configurations and a new Illinois Department of Transportation (DOT) detail are shown in Figure 10. In the survey conducted for this synthesis, 82% of the state respondents and 73% of the total respondents reported that they use a partial-depth keyway. Most of the others reported using a full-depth keyway. FIGURE 10 Examples of keyway congurations. FIGURE 9 Survey responses for superstructure types. a. Simple spans with no cast-in-place concrete or bituminous wearing surface b. Simple spans with cast-in-place concrete wearing surface c. Simple spans with bituminous wearing surface only d. Simple spans with waterproofing membrane and bituminous wearing surface e. Continuous spans with composite cast-in-place concrete wearing surface f. Integral abutments with no cast-in-place concrete or bituminous wearing surface g. Integral abutments with composite cast-in-place concrete wearing surface h. Other Typical keyway configurations Illinois DOT new details The left and center drawings are typical keyway configurations. The right drawing is the Illinois DOT new details. a. Partial Depth b. Full Depth

8 the three joint configurations shown in Figure 12. Three dif- ferent grout strengths of 3.5, 7.0, and 10.5 ksi were assumed. The tensile strength of the grout was assumed to be ksi. Comparisons based on the maximum principal stresses in the grout indicated stresses in Joint A below the assumed tensile strength. Stresses in Joints B and C exceeded the ten- sile strength for all grout strengths. Based on their research, Miller et al. (1999) suggested that a full-depth shear key may stop the joint from acting like a hinge and prevent the joint from opening. With a partial- depth grouted keyway, the area below the keyway is open and free to move. If the area is grouted, the movement at the joint may be reduced. Nottingham (1995) reported on the use of a wider full- depth joint between precast units as used in Alaska. The minimum joint width was 2 in., with the sides of the keyway sandblasted and washed. The wider configuration accom- modates panel tolerances more readily and helps ensure full grout-to-beam contact. With the wider joint, a form is needed below the joint instead of the joint packing or backer rod often used for narrower joints. TRANSVERSE TIE DETAILS In the survey conducted for this synthesis, respondents identified the types of transverse ties used between the box beams. The results are shown in Figure 13. Some respon- dents use more than one type of tie. It may be concluded that the most common types of trans- verse ties are unbonded post-tensioned strands or bars with Keyway configurations in Japan have been described by Yamane et al. (1994) and El-Remaily et al. (1996). Cross-sec- tional shapes are similar to those in the United States except for the size and shape of the longitudinal joint between beams, as shown in Figure 11. In Japan, cast-in-place con- crete is placed in relatively wide and deep longitudinal joints between beams as opposed to the narrow mortar-grouted joints used in the United States. About 6 in. of clear spacing in the longitudinal joint is used in Japan to accommodate differential camber between adjacent beams. All highway bridge decks are covered with 2 to 3 in. of concrete or an asphaltic concrete wearing surface. El- Remaily et al. (1996) reported that longitudinal cracking is seldom reported for Japanese adjacent box beam bridges of this type. The performance of three different grouted joint con- figurations used between the flanges of adjacent decked prestressed concrete bridges was compared analytically by Dong et al. (2007). The forces on the joint were determined from an analysis of typical decked prestressed concrete bridge superstructures subjected to various live-load config- urations. Various combinations of normal horizontal force, transverse bending moment, vertical shear, and horizontal longitudinal force were applied to finite element models of FIGURE 11 Japanese keyway. FIGURE 12 Keyway congurations analyzed by Dong et al. (2007). FIGURE 13 Survey results of types of transverse ties used. a. Transverse unbonded post-tensioning strands b. Transverse unbonded post-tensioning bars c. Non-prestressed unbonded reinforcement d. Transverse bonded post-tensioning strands e. Transverse bonded post-tensioning bars f. Non-prestressed bonded reinforcement

9 In Japan, four to seven equally spaced diaphragms, including the end diaphragms, are commonly provided for box beam bridges (El-Remaily et al. 1996). The reported Japanese design philosophy is that concrete diaphragms with transverse post-tensioning produce a more durable system and more efficient load distribution between beams (Yamane et al. 1994). The diaphragms and cast-in-place concrete between adjacent beams are integrated by post-ten- sioning through the diaphragms. The amount and location of the post-tensioning are determined by the flexural design. A nontensioned reinforced concrete connection between adjacent box beams has been proposed by Hanna et al. (2007, some owners using both types. Eighty-two percent of the respondents used post-tensioning strands or bars that were either bonded or unbonded. The survey indicated a range of ways in which the post- tensioning force is defined, including force per bar or strand, force per duct, and torque on a threaded bar. The number of transverse tie locations varied from one to five per span, depending on span length. The Illinois DOT calculates the number of ties (N) according to the following equation (Anderson 2007): and rounded up to the nearest integer. Ties were located at the ends, midspan, quarter points, and third points, depending on the number of ties. The dif- ferent arrangements are shown schematically in Figure 14. Approximately 70% of the respondents reported that the ties were placed at mid-depth. If two strands or bars were used at one longitudinal location, they were placed at the third points in the depth. Other responses included specific loca- tion depths. In New York State before 1992, transverse ties were not used in adjacent box beam bridges with spans up to 50 ft (Lall et al. 1997, 1998). For spans from 50 to 75 ft, one trans- verse tendon was used at the center. For spans longer than 75 ft, tendons were used at the outer quarter points. The tendons were stressed to a force of 30 kips. In 1992, the number of transverse tendons was increased to three for spans of less than 50 ft and five for spans equal to or greater than 50 ft. This change may have contributed to the reduction in crack- ing discussed in the previous section. For many years, Pennsylvania DOT (PennDOT) used 1.25-in.-diameter steel rods or strands to tie beams together (Macioce et al. 2007). In practice, the beams were only min- imally pulled together by tightening nuts on the rods. For some older bridges, deterioration of the grout in the shear key led to severe corrosion of the tie rod and to eventual failure. Today, the strand is continuous from fascia to fascia and post-tensioned to about 30 kips per location resulting in an average force of about 0.6 kips/ft. The distribution of transverse stress caused by applying 40-kip post-tensioning forces at 12-ft centers to a three-beam assembly was investigated analytically by Huckelbridge and El-Esnawi (1997). They determined that the force was effec- tive only over a distance of about 2.5 ft. They observed that for the transverse force to be fully effective would require such close spacing that much of the economic attractiveness of the box beam system would be sacrificed. Research by Hawkins and Fuentes (2003) showed that, if the tie rods remain snug, they contribute significantly to load distribution among the beams. FIGURE 14 Schematic of transverse tie spacings. a. One Tie b. Two Ties c. Three Ties d. Four Ties e. Five Ties

10 without composite topping. The methodology involves the use of rigid post-tensioned transverse diaphragms as the primary wheel load transfer mechanism between adjacent boxes. The diaphragms are provided at the ends, quarter points, and midspan. Two post-tensioning tendons through the diaphragms are stressed after the longitudinal joints are grouted. One tendon is placed near the top and the other near the bottom of the diaphragm. A grid analysis was used to determine member forces for various combinations of box depths, bridge widths, and span lengths. The required transverse force was almost linearly proportional to the span length and increased sig- nificantly with bridge width. Shallower sections required more post-tensioning than deeper sections. A design chart was provided for preliminary determination of the post- tensioning required for standard beam depths and common bridge widths. Values of prestressing force for the midspan diaphragm ranged from 4 to 14 kips/ft. The design chart was subsequently updated by Hanna et al. (2007) to include changes in the AASHTO LRFD specification for live load and dynamic load allowance. Their analysis indicated that the impact of the skew angle on the required post-tensioning force is minimal for deep beams and increases slightly with skew angle for shallow beams. To avoid the necessity of grid analysis for every bridge, Hanna et al. (2007) developed the following equation as a best fit for the data points of all cases analyzed: 2008). In this connection, the joint has a width of 8 in. with loop bars protruding from the sides of the boxes and overlap- ping loop bars from the adjacent boxes. Longitudinal bars are placed through the loops and cast- in-place concrete is used to fill the joint. This detail is similar to the deck joint detail used by the Japanese and the French Poutre Dalle system (Ralls et al. 2005). DESIGN CRITERIA FOR CONNECTIONS Eighty-one percent of states and 89% of the respondents to the survey stated that they did not make any design calcu- lations to determine the number of transverse ties between box beams. As part of the survey for this synthesis, some respondents provided information about the post-tensioning force used for each transverse tie and the spacing of ties. Based on this information, the average transverse force per unit length for various numbers of ties was calculated. The results are shown in Figure 15 for 11 states. Where a single value is shown, it is based on the specified maximum spac- ing between ties. If the ties are closer than the minimum, the force will be higher than shown in the figure. For some states, a range of forces is presented because these states used a fixed number of ties for a range of span lengths. El-Remaily et al. (1996) proposed a methodology for the transverse design of precast concrete box beam bridges FIGURE 15 Range of average transverse post-tensioning forces. Note: PCI BDM = PCI Bridge Design Manual (1997/2004).

11 are included in Figure 15. The forces determined from the Hanna et al. equations have similar values to those used by Michigan and New York, but these forces are higher than those used by the other states. Badwan and Liang (2007) values are very high, because they reported their values as required transverse post-tensioning compressive stresses. For Figure 15, it was assumed that these stresses acted over the full depth of the section used in their analyses. Figure 15 also includes the AASHTO values based on a compressive stress of 0.25 ksi applied over a keyway depth of 7 in., which is discussed in chapter three. A design chart to determine the required effective trans- verse post-tensioning force is provided in the PCI Bridge Design Manual (PCI 1997/2004). This chart is based on the work of El-Remaily et al. (1996) and was described earlier. Values from the PCI manual are included in Figure 15. An analysis of 16 Michigan bridges selected at random showed that the transverse post-tensioning was generally greater than required by the PCI procedure, depending on span length and beam depth. Field inspection data revealed that the use of a high level of post-tensioning force did not prevent reflective cracking (Attanayake and Aktan 2008). In an example of a Japanese box beam bridge, Yamane et al. (1994) reported an average transverse post-tensioning force of 11 kips/ft. When compared with the data in Figure 15, this value is higher than used by many states. In addition to consideration of the magnitude of the trans- verse post-tensioning force, the vertical location of the force relative to the depth of the keyway is equally important. Consider a partial-depth keyway that is grouted before post- tensioning with the post-tensioning located at mid-depth of the box. A typical partial-depth grouted keyway has a depth of 12 in. The minimum size box beam has a depth of 27 in. The center of the post-tensioning force is 13.5 in. below the top of the box or 1.5 in. below the bottom of the keyway. When the post-tensioning force is applied, tensile stresses are induced in the grout at the top of the box and the box sec- tion will rotate until the bottom flanges come into contact. At that time, the post-tensioning force becomes approximately concentric on the joint. Box beams may not be perfectly straight, however, and are difficult to place in intimate con- tact along their complete length. Consequently, with a par- tial-depth grouted keyway and the post-tensioning applied at mid-depth of the box, it is likely that tensile stresses and cracking can occur when the post-tensioning is applied. The situation becomes more critical with deeper boxes because the grout depth remains constant. Next consider a full-depth keyway that is grouted before post-tensioning and the post-tensioning force is located at mid-depth of the box. The post-tensioning force is now almost concentric with the grouted keyway, and compres- where P = transverse force/ft D = box depth, ft W = bridge width, ft L = bridge span, ft θ = skew angle, degree KL = correction factor for span-to-depth ratio = KS = correction factor for skew angle = 1.0 + 0.002 θ Badwan and Liang (2007) presented an analysis using grillage analogy to calculate the required transverse post- tensioning stress for a deck built with precast multibeams. Variables included in the analyses were span length, deck width, beam depth, and skew angle. A solid section beam was assumed with depths ranging from 15 to 20 in., which are less than the smallest box beam cross sections. A full- depth grout joint was assumed. The skew angle of the deck was found to be the most criti- cal factor affecting the required transverse post-tensioning stress. This appears to contradict the conclusion by El- Remaily et al. (1996) that the effect of skew angle was mini- mal. The Badwan and Liang (2007) analysis showed that the required stress in skewed decks was generally less than that required in a tangent (no-skew) deck. For no-skew bridges, the required stress was independent of deck width but decreased as deck width increased for skew angles greater than 15 degrees. The required stress increased as span length increased, which agrees with the results of El-Remaily et al. (1996). The required post-tensioning stress parallel to the skew varied from about 0.15 to 0.27 ksi and in most situa- tions was less than the 0.25 ksi specified in the LRFD Speci- fications (AASHTO 2007/2008). A stress of 0.15 ksi applied uniformly to a beam depth of 15 in. is 27 kips/ft and a stress of 0.27 ksi on a depth of 20 in. is 65 kips/ft. However, the stress does not necessarily need to be applied to the full depth of the section (R.Y. Liang, personal communication, Aug. 2008). For comparison purposes, the range of values based on the work of Hanna et al. (2007) and Badwan and Liang (2007)

12 According to Macioce et al. (2007), barrier weight is 50% to 100% of the self-weight of a box beam. A survey of five states by PennDOT indicated that the following assumptions are used by different states for distribution of barrier dead load: 100% when analyzing the fascia beam and 50% when • analyzing the first interior beam 50% to fascia beam and first interior beam• 33% to fascia beam, first interior beam, and other inte-• rior beam Equally distributed to all beams• Equally distributed to all beams unless evidence shows • beams acting independently, then 100% to the fascia beams The Illinois DOT’s practice is to avoid the use of a con- crete barrier on this type of structure whenever possible based on the belief that the barrier stiffens the fascia girder and live-load differential deflection will accelerate deterio- ration of the keyway between the fascia girder and first inte- rior girder (Macioce et al. 2007). Research conducted at the University of Pittsburgh showed that the eccentric load effect caused by the barrier load on the edge of the beam resulted in a minimal reduction in the flexural capacity of box beams (Harries 2006). sive stresses will be applied to both top and bottom of the keyway. These stresses must be overcome by the external forces before cracking along the keyway occurs. Therefore, if the post-tensioning is applied at about mid-depth of the box, it is better to have a full-depth keyway. When the transverse post-tensioning is applied at mid- depth of the box, it is necessary to have a diaphragm to prevent the anchorage from punching through the web. However, the diaphragm functions as a stiff lateral member and, therefore, most of the post-tensioning force goes into the diaphragm and is not distributed longitudinally to pro- vide a uniform compression across the keyway. At the end diaphragms, some of the post-tensioning force is transferred into the bearings, which causes them to deform laterally. The amount of force absorbed by the bearings depends on their stiffness. If the bearing is sufficiently flexible, the force will not be great. If the bearings are stiff, they could absorb a large amount of the post-tensioning force. eXTerior BeaM deTaiLs According to the results from the survey for this synthesis, approximately two-thirds of the respondents use an exte- rior beam design that is the same as the interior beams. The main reasons given for using different designs were the dead load of the parapet, curb, railing, and sidewalk and live-load distribution.

13 connection is enhanced by either transverse post-tensioning with intensity of 0.25 ksi or by a reinforced structural over- lay or both. The commentary in the article cautions that the use of transverse mild steel rods secured by nuts should not be considered sufficient to achieve full transverse flexural continuity unless demonstrated by testing or experience. Generally, post-tensioning is thought to be more effective than a structural overlay if the intensity of 0.25 ksi as speci- fied above is achieved. Article 5.14.4.3 and its related commentary—beginning with C5.14.4.3.2—contain the following provisions for pre- cast deck bridges made using solid, voided, tee, and double- tee cross sections: 5.14.4.3.2 Shear Transfer Joints Precast longitudinal components may be joined together by a shear key not less than 7.0 in. in depth. For the purpose of analysis, the longitudinal shear transfer joints shall be modeled as hinges. The joint shall be filled with nonshrinking grout with a minimum compressive strength of 5.0 ksi at 24 hours. C5.14.4.3.2 Many bridges have indications of joint distress where load transfer among the components relies entirely on shear keys because the grout is subject to extensive cracking. Long-term performance of the key joint should be investigated for cracking and separation. 5.14.4.3.3 Shear-Flexure Transfer Joints 5.14.4.3.3a General Precast longitudinal components may be joined together by transverse post-tensioning, cast-in-place closure joints, a structural overlay, or a combination thereof. C5.14.4.3.3a These joints are intended to provide full continuity and monolithic behavior of the deck. 5.14.4.3.3c Post-Tensioning Transverse post-tensioning shall be uniformly distributed in the longitudinal direction. Block-outs may be used to facilitate splicing of the post-tensioning ducts. The compressed depth of the joint shall not be less than 7.0 in., and the prestress after all losses shall not be less than 0.25 ksi therein. C5.14.4.3.3c aashTo sTandard speciFicaTions The majority of adjacent box beam bridges in existence today were probably designed in accordance with the AASHTO Standard Specifications for Highway Bridges. Article 3.2.3.4 of the 17th edition (AASHTO 2002) addresses load distribu- tion in multibeam bridges constructed with prestressed con- crete beams that are placed side by side on supports: The interaction between the beams is developed by continuous longitudinal shear keys used in combination with transverse tie assemblies which may, or may not, be prestressed, such as bolts, rods, or prestressing strands, or other mechanical means. Full-depth rigid end diaphragms are needed to ensure proper load distribution for channel, single- and multi-stemmed tee beams. A procedure is then provided to calculate the distribution of wheel loads to each beam. Section 9 of the specifications addresses prestressed con- crete analysis and design. Article 9.10.3.2 states that, for pre- cast box multibeam bridges, diaphragms are required only if necessary for slab-end support or to contain or resist trans- verse tension ties. Other than the articles cited previously, the AASHTO Standard Specifications do not provide any guidance for the design or construction of the connection between adjacent box beams. It is, therefore, not surprising that different prac- tices have developed. aashTo LrFd speciFicaTions Article 4.6.2.2 of the AASHTO LRFD Bridge Design Spec- ifications (2007/2008) addresses approximate methods of analysis of beam-slab bridges. The commentary of Article 4.6.2.2.1 states that the transverse post-tensioning shown for some cross sections is intended to make the units act together. A minimum 0.25 ksi prestress is recommended. The depth over which the 0.25 ksi is applied is not clearly defined in this article. However, an illustration in Table 4.6.2.2.1-1 shows the force applied at the level of the top flange in adjacent box beam bridges without a concrete overlay. Precast con- crete bridges with longitudinal joints are considered to act as a monolithic unit if sufficiently interconnected. The inter- CHAPTER THREE speciFicaTions and consTrUcTion pracTices

14 the respondents reported experiences with uneven seating. This was more prevalent in bridges that used a full-width support at each end. consTrUcTion seQUence Results from the survey conducted for this synthesis indi- cate that, in single-stage construction, three-quarters of the respondents erect all beams and then connect them together at one time as illustrated in Figure 16a. About 40% of the respondents erect and connect the first two beams, erect the third beam and connect it to the second beam, and so on as shown in Figure 16b. Some respondents permit both methods. In two-stage construction, one of the following sequences is used: Continue the sequence of erecting and connecting • one beam at a time. The first beam of the second stage is erected and connected to the last beam of the first stage. The second beam of the second stage is then erected and connected to the first beam of the second stage and so on. All beams in the second stage are erected and then con-• nected at one time with the second-stage transverse ties spliced to the ties of the first stage. All beams in the second stage are erected and then con-• nected at one time with the second-stage transverse ties passing through the first-stage beams. This requires two sets of holes in the first-stage beams. Another variation is to connect all second-stage beams to the last beam of the first stage. The construction sequence is also dependent on the skew of the bridge and the use of skewed or right-angle interme- diate diaphragms. With a skewed bridge and perpendicular intermediate diaphragms, the beam-to-beam connection system is easier. With a skewed bridge and skewed dia- phragms, either a beam-to-beam approach or all beams con- nected at one time is possible. Approximately one-half of the respondents to the survey for this synthesis reported that the keyways were grouted before transverse post-tensioning and one-half after post-tensioning. Post-tensioning before grout- ing places a higher transverse stress in the beams where they are in contact because the bearing area is less. The grout then functions as a filler and may transfer some shear force, but it will transfer only the compressive stress of any trans- verse bending moments. Post-tensioning after grouting puts a compressive stress in the grout and across the interface between the grout and the beams, and it provides a higher moment capacity before the precompression is overcome. The decision to grout before or after post-tensioning appears to be related to the construction sequence. When When tensioning narrow decks, losses due to anchorage setting should be kept to a minimum. Ducts should preferably be straight and grouted. The post-tensioning force is known to spread at an angle of 45 degrees or larger and to attain a uniform distribution within a short distance from the cable anchorage. The economy of prestressing is also known to increase with the spacing of ducts. For these reasons, the spacing of the ducts need not be smaller than about 4.0 ft. or the width of the component housing the anchorages, whichever is larger. 5.14.4.3.3d Longitudinal Construction Joints Longitudinal construction joints between precast concrete flexural components shall consist of a key filled with a nonshrinkage mortar attaining a compressive strength of 5.0 ksi within 24 hours. The depth of the key should not be less than 5.0 in. If the components are post-tensioned together transversely, the top flanges may be assumed to act as a monolithic slab. However, the empirical slab design specified in Article 9.7.2 is not applicable. The amount of transverse prestress may be determined by either the strip method or two-dimensional analysis. The transverse prestress, after all losses, shall not be less than 0.25 ksi through the key. In the last 3.0 ft. at a free end, the required transverse prestress shall be doubled. C5.14.4.3.3d This Article relates to deck systems composed entirely of precast beams of box, T- and double-T sections, laid side- by-side, and, preferably, joined together by transverse post- tensioning. The transverse post-tensioning tendons should be located at the centerline of the key. Articles 5.14.4.3.3c and 5.14.4.3.3d clearly require a trans- verse prestress of at least 0.25 ksi on a compressed depth of at least 7 in. This amounts to a transverse force of 21 kips/ ft. This requires 0.5-in.-diameter, 270-ksi low-relaxation strands stressed to 189 ksi after losses at 16.5-in. centers. If based on a 5-in. depth as stated in Commentary C5.14.4.3.3d, the force would be 15 kips/ft. Commentary C5.14.4.3.3d states that the post-tensioning tendons should be located at the centerline of the key, whereas 68% of the respondents to the survey for this synthesis reported that the ties were placed at mid-depth of the section. BearinG TYpes Bearing types are either plain elastomeric or laminated elas- tomeric. In the survey conducted for this synthesis, approxi- mately three-quarters of the respondents reported the use of plain elastomeric bearings. Forty-two percent of the states and 56% of the total respondents use one full-width support on each end, whereas two-point supports at each end are used by 42% of the states and 38% of the total respondents. The remainder, with one exception, use two-point supports at one end and one-point supports at the other. The exception reported the use of partial-width bearings with preformed asphalt joint filler under the remaining area. With two sup- ports at one end, adjacent beams may be supported by the same bearing pad that extends under adjacent beams. Half

15 bridge. In Phase II, the second half of the bridge, consist- ing of five box beams, was constructed. The beams were installed sequentially along with the installation and tight- ening of the threaded rods. However, only the shear keys between the five beams were grouted. The shear key at the construction joint between the two phases was not grouted because of movement in the first-phase beams caused by traffic. Without the shear key grouted, movement of the last beam in Phase I caused spalling on the bottom flange of the first beam in Phase II. To arrest the spalling, the tension in the threaded rods across the construction joint was relieved. Subsequently, the traffic was rerouted during the night, the threaded rods were retightened, and the shear keys were grouted with a fast-setting magnesium phosphate grout. all beams are post-tensioned at one time, the option exists to grout all keyways before or after post-tensioning without delaying construction. When beams are connected in pairs, it becomes necessary to allow the grout in the first keyway to gain strength before the first pair of beams can be post- tensioned and the next beam placed. This extends the con- struction time. Greuel et al. (2000) reported on the erection of an Ohio bridge in two phases. In Phase I, one-half of the old bridge was removed and replaced with seven adjacent box beams. After transverse threaded rods were tightened using a torque wrench to pull the beams together, the shear keys were grouted and the longitudinal joints were sealed. Traf- fic was then rerouted onto the completed half of the new FIGURE 16 Transverse tie and diaphragm arrangements on skew bridges. Skew diaphragms Perpendicular diaphragms a. Single transverse tie Small skew angle Any skew angle b. Perpendicular diaphragms with staggered transverse ties connecting adjacent beams

16 high-bond, high early strength grout with user-friendly char- acteristics and low-temperature curing ability was needed. If a prepackaged grout mix was available to construction work- ers and an exact prescribed amount of water could be added, high-quality joints would be obtained more consistently. The material most closely meeting these requirements was a prepackaged magnesium-ammonium-phosphate grout often extended with pea gravel. According to El-Remaily et al. (1996), West Virginia DOT investigated several high-volume, heavily loaded bridges that had joint failures and topping cracking. The investiga- tors concluded that vertical shear failure in the keys was most likely the result of inadequate grout installation and transverse tie force. The ties used for the failed joints were 1-in. diameter, ASTM A36 rods spaced at the third points along the span and tightened with an approximate torque of 400 ft-lb. As a result of the investigation, the West Virginia DOT changed its practices to include the following: A pourable epoxy instead of a nonshrink grout in the • shear key Sandblasting of surfaces in contact with the grout• Post-tensioned ties to be used• In the Andover Dam Bridge in Upton, Maine, the shear keys between boxes were made wider than the Maine DOT’s standard width and were filled using a self-consolidating concrete modified with the addition of a shrinkage-compen- sating admixture. This allowed the shear keys to be grouted rapidly. On the Davis Narrows Bridge in Brooksville, Maine, a pea stone concrete mix was used in the shear keys instead of the conventional sand grout. This reduced the possibility of discharging material into the river. As an additional mea- sure, the foam backer rods in the shear keys were bonded to the beams before erection and then were compressed into place during the erection of adjacent beams (Iqbal 2006). To improve the quality of grout used in the joints, Illinois DOT now requires that a mechanical mixer be used to mix nonshrink grout. The grout could be worked into place with a pencil vibrator, and the surface needs to be troweled smooth and immediately covered with cotton mats for a minimum of 7 days. The curing period may be reduced if the contractor determines that the grout cube strength exceeds the specified strength. In no case shall the curing time be less than 3 days (Illinois DOT 2008). diFFerenTiaL caMBer In the survey conducted for this synthesis, approximately one-third of the respondents reported that they had limita- tions on the differential camber between adjacent beams. One-half of those with limitations indicated that the maxi- mum was 0.5 in. Other variations included 0.25 in. in 10 ft, 0.75 in. maximum, and 1 in. between the high and low beam in the same span. Methods to remove excessive cam- ber included (1) loading the high beam before grouting and post-tensioning, (2) placing the barrier on the high beam, (3) adjusting bearing seat elevations, (4) accommodating the differential in the concrete or asphalt overlay, and (5) preas- sembling the span before shipment to obtain best fit. KeYWaY preparaTion Forty-five percent of the states responding to the survey stated that the keyways are sandblasted before the beams are installed. Of all the other respondents, only one—a county— reported using sandblasting. When sandblasting was used, it was always done before shipment. Approximately one-third of the respondents reported that additional preparation other than sandblasting was performed for the interior faces of beams. This generally involved cleaning with compressed air or water. One state reported applying a sealer to limit absorption from the grout. For the exterior face, only one- sixth of the respondents reported additional preparation. A forensic investigation performed during the demoli- tion of an adjacent box beam bridge in Michigan showed that shear-key mortar adherence to the beams was poor (Attanayake and Aktan 2008). GroUTinG MaTeriaLs and pracTices Results from the survey conducted for this synthesis showed that about 40% of the respondents use a nonshrink grout, about 25% use a mortar, and others use epoxy grout, epoxy resin, or concrete topping. The predominant method used to place the grout is by hand. Approximately 40% of the respondents provide no curing, 5% use curing compounds, and 45% wet cure. The remainder follows manufacturers’ recommendations. Based on experiences in Alaska, Nottingham (1995) stated that a high-quality, low-shrinkage, impermeable,

17 STATES REPORTING LITTLE OR NO OBSERVED DISTRESS In the survey, several states reported little or no observed types of distress. Their practices are described here. Massachusetts’ current standard is to use either simple or continuous spans with a 5-in.-thick cast-in-place concrete topping, waterproofing membrane, and a 3.5-in.-thick bitu- minous wearing surface. The transverse ties are unbonded post-tensioning strands tensioned to 44 kips. For spans less than 50 ft, the transverse ties are located at the ends and midspan. For spans greater than 50 ft, the ties are at the ends, TYPES OF DISTRESS In the survey for this synthesis, respondents identified the types of distress that they have observed at the joints between adjacent box beams. The results are summarized in Figure 17. The most common types of observed distress are longi- tudinal cracking along the grout-to-box beam interface and water and salt leakage through the joint. Reflective cracks are often visible in the riding surface, as shown in Figures 18 and 19. CHAPTER FOUR LONG-TERM PERFORMANCE, MAINTENANCE, AND REPAIRS FIGURE 17 Observed types of distress along the joints between adjacent box beams. a. None b. Longitudinal cracking along the grout and box beam interface c. Cracking within the grout d. Spalling of the grout e. Spalling of the corners of the boxes f. Differential vertical movement between adjacent beams g. Corrosion of the transverse ties h. Corrosion of the longitudinal prestressing strands i. Freeze-thaw damage to the grout j. Freeze-thaw damage to the concrete adjacent to the joint k. Water and salt leakage through the joint l. Other

18 Although Oregon listed six types of distress, its survey response indicated that the types were not widespread. Ore- gon’s practice is to build simple spans with a waterproofing membrane and bituminous wearing surface or continuous spans with a composite concrete deck. Keyways have a par- tial depth of 12 in. and are filled with a nonshrink grout by hand after post-tensioning. Adjacent box beams are con- nected in pairs with unbonded transverse post-tensioning located at mid-depth of the box. Transverse ties have a maxi- mum spacing of 24 ft and are tensioned to 39 kips. This is equivalent to a force of at least 1.63 kips/ft. Wyoming reported that the state has only one adjacent box beam bridge, which was built in 2004. It is a simple span structure with a cast-in-place concrete wearing surface and a span length of 120 ft. Partial-depth shear keys are used. Details of the transverse ties were not provided. OTHER OBSERVATIONS Shenoy and Frantz (1991) and Rao and Frantz (1996) reported on structural tests of box beams removed from a 27-year old bridge in Connecticut. The 54-ft simply supported bridge was made of 13 prestressed concrete box beams. Lateral post-tensioning was provided at each end and midspan. The practice at the time of construction was to have no water- proofing membrane between the top of the beams and the 2-in.-thick asphalt wearing surface. As a result, water con- taining deicing salts was able to seep down into the beams, penetrate the concrete, and cause severe deterioration in some of the beams. Many of the shear keys leaked, as shown by efflorescent stains on the sides and bottoms of the beams. The concrete on the sides and bottoms of some beams was cracked and spalled, exposing prestressing strands and lead- ing to rupture of one strand. quarter points, and midspan. For a 50-ft-long span, the aver- age transverse force is 2.64 kips/ft. For a 100-ft-long span, the average force is 2.2 kips/ft. The full-depth keyways are grouted before post-tensioning with a two-component poly- mer-modified cementitious, fast-setting mortar. All beams are connected at one time. Michigan reported that its only problem was spalling of the grout. Michigan’s practice is to build simple and continu- ous spans with a composite concrete deck. Keyways have a partial depth and are filled by hand with a mortar after trans- versely post-tensioning. As shown in Figure 15, Michigan uses the second-highest amount of transverse post-tension- ing of the 11 states that provided sufficient information to determine the transverse post-tensioning forces. The trans- verse ties are bonded post-tensioning bars tensioned to 104.5 kips for HS25 loading and 82.5 kips for HS20 loading. All beams are connected at one time. Missouri reported that it recently started using box beam bridges and found no major leakage through the joints. Mis- souri’s practice is to build simple spans with a cast-in-place deck or seal coat and asphalt wearing surface and continuous spans with a composite concrete deck. Keyways are partial depth and are filled with a nonshrink grout by hand. Adja- cent box beams are connected in pairs using nontensioned unbonded reinforcement located at mid-depth of the box. New Mexico’s practice is to use simple spans with a 5-in.- thick composite cast-in-place concrete deck. Transverse ties consist of two bonded post-tensioning bars stressed to 50 kips with five ties per span spaced no more than 25 ft apart. This is equivalent to a force of at least 4 kips/ft. The par- tial-depth keyways are grouted after post-tensioning using a mortar. Adjacent box beams are connected in pairs. The New Mexico survey response indicated that the state does not build many of these types of bridges. FIGURE 18 Longitudinal cracking in asphalt riding surface (Source: Henry G. Russell). FIGURE 19 Longitudinal cracking in composite concrete deck (Source: New York State DOT).

19 depth keyways instead of partial depth, and elimination of stiffened cardboard as the internal void former. According to Ahlborn et al. (2005), longitudinal cracking was present in both pre-1974 and post-1985 bridge decks. All the pre-1974 bridges showed signs of prolonged exposure to moisture along beam edges and bottom flanges. Some also had calcium carbonate deposits on the underside. Similar deposits were also visible on the outside face of the fascia beam as a result of leakage through the concrete barrier to bridge deck interface. Rust stains were visible around drain holes, indicating some form of active corrosion. Delamina- tion, spalling, and breakage of some tendons were concen- trated along beam edges. Longitudinal cracking in box beam bottom flanges appeared to be caused by corrosion of the strand. For bridges built after 1985, longitudinal cracking in the bottom flange and concrete spalling were not observed. Attanayake and Aktan (2008) reported on the monitoring of a bridge constructed in Portage, Michigan, in 2007. The two-span straight bridge has span lengths of 79 ft and a width of 93 ft 5 in. The cross section consists of twenty-two 48 in. by 33 in. box beams. The total transverse force between boxes is about 16 kips/ft applied at six transverse locations using pairs of strands. Cracks along the shear key and beam interface were observed before and after post-tensioning was applied. The bridge deck was cast about 24 days after the transverse post-tensioning was applied and moist cured for 7 days. Fifteen days after casting the deck, the research team observed through thickness cracks that stemmed from the top surface of the deck above the abutments. Miller et al. (1995) reported the removal of a sidewalk support beam from a bridge over the Maumee River in Defi- ance, Ohio, for testing. The bridge superstructure consisted of sixteen 33-in.-deep, 36-in.-wide, 76-ft-6-in.-long box beams with a cast-in-place composite deck. The two outside beams on each side of the bridge supported the sidewalk and were slightly raised and separated from the main bridge beams by a gap. This gap provided for drainage from the bridge. The exposed side of the sidewalk beams adjacent to the roadway was intended to have a waterproof coating, but there was no indication that the coating was ever applied. This was later confirmed by tests. Damage to the beam was caused by chloride-laden water penetrating the unprotected side of the beam and corroding the prestressing strands on the roadway side of the beam. The beam was cast in 1980 and contained eighteen 0.5-in.- diameter strands. At the time of the tests, three strands in the corner of the beam adjacent to the roadway had corroded. The strand closest to the corner was missing along the entire length of the beam. The second strand had broken individual wires at various places. The third strand showed less corro- Whiting and Stejskal (1994) reported on a field survey of the condition of prestressed concrete bridge elements in adverse potentially corrosive environments. Their survey included two box beam bridges; one located in a temperate marine environment in Oregon and one in a deicing environ- ment in Michigan. On the Oregon bridge, some amount of efflorescence between the box beams was noted, indicating migration of water from the deck surface down the sides of the beams. The box beam bridge in Michigan was in signifi- cantly worse condition than the one in Oregon. Longitudi- nal cracks had formed in the asphaltic wearing course and allowed salt-laden water to run down the exterior sides of the box beams. The chlorides had penetrated the concrete and caused corrosion of the prestressing steel. Heavy deposits of efflorescence and corroded strands could be seen at various locations on the bridge soffit. At most locations of missing concrete, exposed strands exhibited 100% loss of section. Needham and Juntunen (1997) studied different types of distress in prestressed concrete box beams in Michigan and discussed the causes and effects of each (Ahlborn et al. 2005). During this study, chloride samples were taken from seven box beam structures and five I-beam structures to determine whether the chloride concentration in a repre- sentative number of typical prestressed concrete beams was high enough to initiate corrosion in the beams. The aver- age measured chloride content of the box beam bridges was 0.42 kg/m3 (0.71 lb/yd3) on county highways and 0.98 kg/ m3 (1.65 lb/yd3) on state highways. The average measured chloride content was 0.63 kg/m3 (0.94 lb/yd3) on the I-beam bridges. For both structures, the chloride content was higher at the ends of the beams. Based on the investigation of chlo- ride content in box beam bridges, Needham and Juntunen (1997) concluded that the condition of the box beam bridges on county roads is better than that on state highways. The reason is likely the result of decreased traffic loads and fewer chemical deicers applied to the county roads. The investi- gation revealed that chloride contamination is primarily the result of leaky joints and filtration of water through the deck. The ends of the box beams exhibited greater deterioration than at other locations. A survey conducted in conjunction with an investiga- tion by PennDOT identified that seven states (Colorado, Florida, Illinois, Indiana, Ohio, Pennsylvania, and Virginia) have reported failures of box beam bridges (Macioce et al. 2007). Ahlborn et al. (2005) reported on an inspection of eight bridges built before 1974 and seven bridges built after 1985 in Michigan. Attanayake and Aktan (2008) reported on the same 15 bridges plus two more. In Michigan between 1974 and 1985, significant changes were made in the construction of adjacent box beams. These included a change from an asphalt wearing surface to a 6-in.-thick reinforced concrete deck because of water leakage to the shear key, use of full-

20 Use waterproofing membrane over the entire surface • and reseal the deck When installing a new concrete deck, it is important to thoroughly clean the top surface of the beams and, if nec- essary, add reinforcement dowels into the webs to provide composite action between the deck and the beams. Illinois believes that replacing asphalt wearing surfaces with a thicker reinforced concrete wearing surface is effective in prolonging the life of beams if they are in good condition and have not experienced salt exposure from leaking keyways. The current practice in Illinois is to use reinforced concrete overlays on new box beam bridges (Macioce et al. 2007). FACTORS AFFECTING LONG-TERM PERFORMANCE In the survey conducted for this synthesis, respondents reported the methods of construction that they have found to be most effective in preventing deterioration along the joints. The two items that were identified more than others were sufficient transverse post-tensioning and use of a concrete topping slab. Items that were identified as being noneffective included asphalt wearing surface with or without a water- proofing membrane, phased construction, and inadequate concrete overlay. When asked to identify the factors that affect the long- term performance of adjacent box beam bridges, the survey responses were varied (see Table 1). TABLE 1 FACTORS INFLUENCING LONG-TERM PERFORMANCE Factor States (%) Total Survey (%) Yes No Yes No Span Length 25 75 20 80 Simple Spans vs. Continuous Spans 27 73 24 76 Skew 57 43 48 52 Bearing Types 8 92 10 90 Topped vs. Untopped 80 20 72 28 Integral Abutments 27 73 24 76 Phased Construction 31 69 21 79 Waterproof Membrane 45 55 47 53 Exterior Beam Details 0 100 7 93 Maintenance 36 64 48 52 The survey also asked what problems have been observed with joints between adjacent units. The two major problems identified were longitudinal cracking along the grout-to-box beam interface and water and salt leakage through the joint. When a concrete topping was used, 65% of the responding sion. The corroded strands caused a loss of concrete in the lower corner region of the beam adjacent to the roadway. Although the concrete barrier is not considered to act compositely with the fascia beam, it behaves compositely because it is rigidly attached. When open deflection joints are provided in the barrier, a change in stiffness occurs at the joint. This results in a concentrated rotation in the box beam below the joint. Extensive cracking can occur at this location. The joint also provides a path for salt-contaminated deck drainage to attack the exposed fascia girders. For these reasons, Macioce et al. (2007) recommend that barriers be made continuous. According to a draft PCI report (PCI 2009), the pre- dominant distress observed in adjacent box beam bridges is reflective cracking of the deck along the shear keys between beams and the associated degradation below the cracks. The cracking allows water and deicing chemicals to penetrate through the deck and may cause freeze-thaw damage to the concrete of the box beam or corrosion of the transverse tie. MAINTENANCE PROCEDURES Suggestions to maintain adjacent box beam bridges from survey respondents include the following: Seal the deck• Remove the asphalt topping• Seal the cracks• Wash the decks annually• Macioce et al. (2007) report that limited maintenance activities are associated with this type of bridge. During removal and replacement of the asphalt wearing surface, a waterproofing membrane can be installed before the new wearing surface to improve performance. Open deflection joints in barriers that result in a concentrated rotation in the beam below the joint should be closed. This can be accom- plished by removing the concrete in the barrier on both sides of the joint to expose the reinforcement, lap splicing a rein- forcement across the joint with the exposed reinforcement, and recasting the concrete (Macioce et al. 2007). REPAIR PROCEDURES When asked what methods have been used to rehabilitate or retrofit adjacent box beam bridges, survey respondents mentioned the following: Add a reinforced concrete deck• Add supplemental tie rods• Replace the asphalt wearing surface with a concrete • deck

21 tive cracks in the wearing surface on the bridges built in the late 1980s and early 1990s. These agencies have emphasized the importance of eliminating these cracks. According to the survey (Hanna et al. 2007), the states and provinces have recommended the following preventive actions based on les- sons learned in the last two decades: Use cast-in-place concrete deck on top of the adjacent • boxes to prevent water leakage and to uniformly dis- tribute the loads on adjacent boxes Use nonshrink grout or appropriate sealant instead of • the conventional sand/cement mortar in the shear keys, in addition to blast cleaning of keyway surfaces before grouting Use full-depth shear keys owing to their superior per-• formance over the traditional top flange keys (recom- mended by a few states) Use transverse post-tensioning to improve load distri-• bution and minimize differential deflections between adjacent box beams; adequate post-tensioning force should be applied after grouting the shear keys to mini- mize the tensile stresses that cause longitudinal crack- ing at these joints Use end diaphragms to ensure proper seating of adja-• cent boxes and intermediate diaphragms to provide the necessary stiffness in the transverse direction Use wide bearing pads under the middle of the box to • eliminate the rocking of the box while grouting the shear keys (the use of sloped bearing seats that match the surface cross-slope is also recommended) Use adequate concrete cover and corrosion inhibitor • admixtures in the concrete mix to resist the chloride- induced corrosion of reinforcing steel Eliminate the use of welded connections between adja-• cent boxes and avoid dimensional tolerances that result in inadequate sealing of the shear keys Based on a PCI survey of 45 states and three Canadian provinces, the following design, fabrication, and construc- tion practices have been shown to improve the performance of adjacent box beam bridges (PCI 2009): Design• Use high-performance or high-strength, low-per- – meability concrete in the beams and deck slab Provide shear-key geometries that allow deck con- – crete to fill the key or use full-depth shear keys Provide a minimum of 1.5 in. of cover to all rein- – forcement; use 2 in. of cover where practical Use strand patterns that omit longitudinal prestress- – ing strands in the exterior corners Design for composite action with a reinforced con- – crete deck slab that has a minimum thickness of 5 in. Minimize skews where practical – Provide lateral restraint at piers and abutments – states and 55% of the total respondents reported reflective cracking in the topping. Following the collapse of a fascia box beam of the Lakev- iew Drive Bridge over I-70 in Washington County, Pennsyl- vania, the bridge was closed to traffic and selected beams removed for a detailed investigation by Lehigh University (Naito et al. 2006) and the University of Pittsburgh (Har- ries 2006). The four-span structure, built in 1960, had span lengths of 54, 89, 89, and 42 ft and a 51-degree skew. The cross section consisted of eight 4-ft-wide adjacent box beams. The fascia beam contained sixty 0.375-in.-diameter, 250-ksi strands. The wearing surface consisted of a 2-in.- thick asphalt layer applied directly to the top flange of the beams. No waterproofing membrane was provided. The transverse ties between adjacent beams were 1-in.- diameter steel rods threaded at both ends and passing through 2.25-in.-diameter holes in the diaphragms. To accommodate the skew, the ties were staggered along the beam’s length. These ties were reported to be heavily corroded such that their strength was seriously reduced. The report indicated that the corrosion may have occurred because of poor con- solidation and poor construction of the longitudinal joint between adjacent beams (Naito et al. 2006). High chloride contents were measured on the interior and exterior faces of the webs and bottom flanges of the interior beams. Harries (2006) reported that few of the beams located at the test site had evidence of intact shear keys. In cases in which shear keys were present, the grout was poorly consoli- dated. It was common to find asphalt material on the lower ledge of the shear key, indicating that the shear-key grout had not been present when the bridge was last paved. Using the results of their bridge inspection (Lall et al. 1997), New York State DOT conducted studies to determine whether the incidence of cracking was related to such factors as span length, skew, average annual daily traffic (AADT), or bearing type. They concluded the following: Frequency of shear-key cracking did not increase with • either span length or total bridge length Bridge skew angle was not directly related to frequency • of shear-key cracking Some evidence indicates that bridges with higher • AADT crack more often Bridges with fixed bearings crack somewhat more • often than those with expansion bearings, but it was difficult to say whether the difference was significant According to Hanna et al. (2007), the PCI subcommittee on adjacent member bridges conducted a survey on the cur- rent practices in the design and construction of adjacent box beam bridges in the United States and Canada. Most of these transportation agencies have experienced premature reflec-

22 joints and minimize differential deflections between boxes Sandblast shear keys before grouting or concreting – When using small shear keys, use epoxy grout in – keyways (some agencies report success with non- metallic, nonshrink grout) Post-tension transverse ties before grouting shear – keys on skewed bridges and post-tension after grouting on square bridges Grind concrete pier and abutment surfaces, if neces- – sary, to achieve a uniform bearing surface In staged construction, provide a minimum gap of – 1 ft between the last beam of the first stage and the first beam of the second stage to provide a closure pour When differential camber occurs, use force to – remove the differential, when practical, or use the joint grout material to provide smooth transition In their study of New York state bridges, Lall et al. 1997, 1998 concluded that the frequency of longitudinal cracking was unrelated to maximum span length, total bridge length, and bridge skew. Use corrosion inhibitor in the concrete mix design – for the beams Provide waterproofing between the top of the struc- – tural member and the overlay if a noncomposite overlay is to be used Fabrication• Use polystyrene material to form the voids – Provide consistent casting conditions to minimize – differential camber in beams Properly anchor void forms to prevent floating of – forms during casting Provide vent and drainage holes in boxes – When extending stirrups for shear connection to – slab, consider the bent shape of bar so that it does not interfere with placement of void forms When extending reinforcing steel at ends of beams, – provide straight bars and bend after fabrication Construction• Consider a three-point bearing system to minimize – rocking of beams Provide transverse post-tensioning to compress the –

23 IDENTIFICATION OF CORROSION The two visible means to identify corrosion are as follows: (1) rust stains that appear on the surface or (2) spalled con- crete exposing corroded reinforcement. By the time either of these are visible, active corrosion has been ongoing for some time. Procedures to identify corrosive environments and active corrosion in concrete have been reported by ACI Committee 222 (2001). Electrical methods such as the half-cell potential and linear-polarization methods may be used to evaluate corro- sion activity in concrete (ACI Committee 228 1998). The half-cell potential method can be used to identify regions in which there is a high probability that corrosion is occur- ring at the time of the measurement. The linear-polarization method determines the instantaneous corrosion rate of the reinforcement located below the test point. Both methods require a connection to the embedded reinforcement, and the reinforcement must be electrically connected. The tech- niques are not applicable to epoxy-coated reinforcement and require experienced personnel to test and interpret the measurements. OTHER PRACTICES Forty-five percent of the respondents to the survey for this synthesis reported that they inspected drain holes for debris. Seventeen percent did not inspect drain holes and 38% reported that inspection of drain holes was not applicable. VISUAL INSPECTION Visual inspection is the current state-of-the-practice used to document the condition of beams (Macioce et al. 2007). This was verified by the survey for this synthesis, which had 100% of the states and 90% of total respondents using only visual inspection for box beam bridges. The two other meth- ods mentioned were chain dragging and full deck survey. Visual inspections are unable to detect the corrosion of unexposed prestressing strands. Research by Hawkins and Fuentes (2003) found that the high tensile stress in the strands causes them to corrode at a faster rate than that of conventional reinforcement and welded wire reinforcement. Cardboard forms were often used in past construction to form internal voids. These forms were susceptible to damage from water entering the voids through seepage along the tie rod or through steam vent holes in the top flange of the box. Drain holes are needed in the bottom flange of the box, but these holes can become clogged and need to be unclogged on a regular basis. Harries (2006) reported that there does not appear to be a practical manner to assess the condition of the shear keys between adjacent beams. However, water dripping from the joints between beams during a period of rain (or icicles as shown in Figure 20) or longitudinal staining caused by water “wicking” along the beam soffit (see Figure 21) represents observable evidence that the shear key is degraded. CHAPTER FIVE INSPECTION PRACTICES FIGURE 21 Eforescence on the underside, indicating leakage through the joints (Source: New York State DOT). FIGURE 20 Leakage through longitudinal joints (Source: Pennsylvania DOT).

24 Examine wearing course for longitudinal cracks• Most clogged drain holes were cleaned by rodding out the debris in various ways. Illinois DOT now requires that all loose and delami- nated concrete be removed from the underside of precast, prestressed concrete deck beam bridges during inspection. This requirement enhances the accuracy of bridge inspec- tions and reduces the rate at which the prestressing strands corrode by reducing the presence of trapped moisture adja- cent to the strands (Modeer and Anderson 2005). Macioce et al. (2007) have summarized key inspection factors as follows: Document exposed strands• Document cracking patterns• Document areas of exposed concrete• Identify areas of delaminated concrete by sounding• Document visible rust stains• Define strand corrosion• Evaluate barrier and barrier connections• Clear clogged drain holes• Check for evidence of tie rod failure•

25 be the best material for transverse joints in terms of strength, bond, and mode of failure. However, Issa et al. (2003) recom- mended the use of proprietary products containing magne- sium-ammonium-phosphate mortars because of their ease of use and satisfactory performance. In cases in which the joint is subject to excessive stresses or a quick resumption of traf- fic is critical, polymer concrete was recommended. sTrUcTUraL research Shahawy (1990) conducted punching shear tests on a half- scale model of a prestressed concrete double-tee bridge sys- tem. The model consisted of three adjacent double-tee beams that were post-tensioned together transversely through their top flange. Initially, the double tees were post-tensioned to a value of 75 psi at the middle portion of the span and to 150 psi at the 3-ft-long end regions. After two initial punching tests were performed, it was concluded that the amount of transverse post-tensioning was insufficient to achieve mono- lithic behavior of the bridge slab. Therefore, the value was increased to 150 psi at the middle portion and 300 psi at the ends. Based on seven tests, Shahawy (1990) concluded that an average effective post-tensioning of 150 psi across the lon- gitudinal joint resulted in monolithic behavior and produced punching shear resistance similar to that of cast-in-place concrete slabs in multibeam bridges. He also concluded that the 150 psi effective prestress seemed to ensure adequate fatigue life of the longitudinal joints. Huckelbridge et al. (1995) and Huckelbridge and El- Esnawi (1997) investigated the performance of grouted shear keys located at the longitudinal joints between adja- cent beams of multibeam prestressed concrete box beams. The test specimens consisted of a 12-in. longitudinal slice of a three-beam-wide assembly with a loaded center beam. Transverse ties were not included. All the test specimens exhibited relative displacements across some of the joints, which indicated a fractured shear key. The study proposed to locate the 6-in.-deep shear key at the neutral axis level of the cross section rather than starting 6 in. below the top of the beam. The static shear load capacity of the new shear-key MaTeriaLs research Gulyas et al. (1995) conducted laboratory tests to compare component material tests and composite grouted keyway specimen tests using nonshrink grouts and magnesium- ammonium-phosphate mortars. Comparative composite specimens were tested in vertical shear, longitudinal shear, and direct tension. Results indicated significant differences in performance between materials. Composite testing of grouted keyway assemblies rather than component materi- als testing was shown to be a more accurate way to evaluate the performance of a grouting material, because the effects of grouting materials, keyway shapes, curing, substrate exposure, and texture can be evaluated. Composite assem- blies made with magnesium-ammonium-phosphate mortars provided much higher vertical shear, horizontal shear, and direct tensile tests than assemblies made with nonshrink grouts. The use of a keyway surface that was sandblasted before grouting the keyway with magnesium-ammonium- phosphate mortars provided higher strengths than a keyway surface that was allowed to carbonate. Based on test results, Gulyas et al. (1995) recommended that consideration be given to grouting materials that have inherent bond strength to the keyway surface. Grout mate- rials should have low shrinkage as measured using ASTM C157, all keyway surfaces should be provided with an aggres- sive grit-blasted surface at the plant immediately before shipment, and the use of nonshrink grouts for keyway appli- cations should be discouraged unless the proposed material meets specific criteria for bond strength, drying shrinkage, chloride absorption, and shear strength. (Note, however, that the authors of the paper are employees of the company that sells the magnesium-ammonia-phosphate mortars used in the tests.) Issa et al. (2003) reported an evaluation of the perfor- mance of four different grout materials for use in precast concrete deck systems. Thirty-six full-scale specimens were tested for vertical shear, direct tension, or flexural capacity. The precast slab joint surfaces were sandblasted until the coarse aggregate was slightly exposed followed by air and high-pressure water washing. Polymer concrete was found to CHAPTER SIX research

26 they would leak. Thus, the authors concluded that the main problem associated with shear-key cracking appears to be leakage rather than structural load transfer. Grace and Jenson (2008) performed an experimental and analytical study to examine the influence of the level of transverse post-tensioning and the number of transverse diaphragms on the performance of a bridge in the transverse direction. The experimental program included the con- struction, instrumentation, and testing of a one-half scale, 30-degree skew adjacent box beam bridge model with an effective span of 31 ft. The model consisted of four adjacent box beams with full-depth shear keys, reinforced composite deck slab, and unbonded carbon fiber reinforced polymer (CFRP) transverse post-tensioning. Testing of the model consisted of load distribution tests with an uncracked, cracked, and repaired deck slab. The distribution of the transverse strain that developed at the top surface of the deck slab and the deflection across the width of the bridge were examined for different numbers of transverse diaphragms and different levels of transverse post-tensioning force. The results indicated that increasing the transverse post-tensioning force improved the transverse load distribution and that five diaphragms were better than three in terms of load distribution. The analytical study included finite element analyses to simulate a wide range of bridges with different span lengths and widths. Different loading cases were evaluated to estab- lish an adequate number of diaphragms and appropriate transverse post-tensioning forces to prevent longitudinal cracks. The Grace and Jenson study (2008) showed that live load alone is not the major cause of the longitudinal cracks. Combining temperature gradient with live load can lead to the development of longitudinal cracks between adjacent beams. The required number of diaphragms was found to be a function of the span length, whereas the required trans- verse post-tensioning force was a function of the bridge width. For 48-in.-wide box beams and span lengths up to 50 ft, the authors recommended a minimum of five diaphragms. Beyond 50 ft, the minimum number of diaphragms recom- mended increased to nine at a span length of 110 ft. For 36-in.-wide box beams, the recommended minimum num- ber of diaphragms increased from five for span lengths up to 50 ft to eight for span lengths of 80 ft. The appropriate post-tensioning force ranged from 120 kips/diaphragm for a 24-ft-wide bridge to 160 kips/diaphragm for a 78-ft-wide bridge that was defined as having a recently constructed slab. These forces are equivalent to 10 to 16 kips/ft. When the recommendation for more diaphragms as the bridge span increases is combined with a fixed force per diaphragm, the total transverse post-tensioning force increases as the span length increases. design was 2.4 times that of the previous shear-key design with nonshrink grout in the keyways. The fatigue life of the new shear-key design was extended to more than 8,000,000 cycles from about 100 cycles with the previous key design. Hlavacs et al. (1997) used a full-scale portion of an adja- cent box beam bridge to test the performance of grouted shear keys under environmental and cyclic loads. Two sepa- rate tests were conducted. In the first, shear keys that were grouted in late autumn cracked soon after casting, before any load had been applied. Data from instruments embedded in the beams and shear keys showed large discontinuities in strains caused by freezing temperatures. These strains were much larger than strains that occurred under loads cor- responding to the weight of an HS20-44 truck. The beams were subjected to 41,000 cycles of loading that simulated HS20-44 wheel loads. No new cracking occurred from the loading, but cracks caused by temperature propagated under these loads. In the second test, the keys were grouted in the summer. Higher temperatures caused by the sun’s heat on the top of the beams again caused large thermal strains, which cracked the shear keys. These keys were subjected to 1,000,000 cycles of load corresponding to an HS20-44 wheel load. As in the first test, the load itself did not cause new cracks, but the existing thermal cracks propagated under the load. Miller et al. (1999) conducted outdoor full-scale testing of shear keys using four-beam-wide full-scale assemblies, 33 in. deep and 75 ft long. Three different shear-key configura- tions were studied: Configuration 1: A current detail in which the shear • key is approximately 10 in. deep from the top of the beam and grouted with nonshrink grout Configuration 2: The same keyway used in the first • configuration but grouted with epoxy Configuration 3: A proposed mid-depth keyway • grouted with nonshrink grout Five transverse tie rods were provided as required by the Ohio DOT specifications. The tie rods were tightened with a torque wrench before casting the shear keys to pull the beams together. The test results showed that the currently used shear-key design cracks because of thermal stresses that are generated as the beams deflect up and down during daily heating and cooling cycles. The mid-depth keyway was less suscep- tible to these stresses and was more resistant to cracking. The epoxied shear key did not crack. Miller et al. (1999), however, expressed concerns about thermal compatibility between the epoxy and the concrete. Load tests on the assemblies showed that cracked shear keys still transfer load, but dye penetration tests showed that

27 Approximately half the states require the use of wet • curing or curing compounds for the grout. Although box beam bridges have generally performed well, design and construction practices have changed over the years as experience with the use of box beams has grown. Nevertheless, a recurring problem with this type of construction is cracking in the grouted longitudinal joints between adjacent beams. This cracking appears to be initi- ated by stresses associated with temperature gradients, and then propagates as a result of live load, or it is caused by a combination of stresses from temperature gradients and live load. In bridges with partial-depth keyways, the crack- ing may be initiated by tensile stresses caused by the post- tensioning. Once the cracking has occurred, chloride-laden water can penetrate the cracks and saturate the sides of the beams. Eventually, this can lead to corrosion of the non-pre- stressed reinforcement, prestressing strands, and transverse ties. From a structural aspect, cracking at the keyway can lead to a reduction in the bending moment or vertical shear trans- ferred across the joint. Therefore, the live-load distribution may not occur as assumed in design, and individual beams could be overloaded. This overloading can cause one beam to deflect more than the adjacent beams, which can lead to further deterioration of the joint. Useful practices Based on information received from the survey and litera- ture review, the following practices can eliminate or reduce the likelihood of longitudinal cracking and joint deteriora- tion and, therefore, enhance the performance of adjacent box beam bridges: Design Practices • Using full-depth shear keys that can be grouted – easily Using transverse post-tensioning so that tensile – stresses do not occur across the joint Using a cast-in-place, reinforced concrete, com- – posite deck with a specified concrete compressive strength of 4,000 psi and minimum thickness of 5 in., to limit the potential for longitudinal deck cracking concLUsions Bridges built with adjacent precast, prestressed concrete box beams are used in about two-thirds of the states. The bridges are economical, can be constructed quickly, and eliminate most deck forming. Based on the survey conducted for this synthesis, the current practice for box beam bridges is as follows: Approximately half the states with box beam bridges • use AASHTO/PCI cross-sectional shapes. Span lengths range from fewer than 20 ft to more than • 80 ft. The most common maximum skew angle between the • abutment and the perpendicular to the bridge center- line is 30 degrees. Most states use simple spans with a cast-in-place con-• crete deck. Where a composite deck is used for continuous spans, • the bridges are generally designed to be continuous for live load. Most longitudinal keyways between adjacent box • beams are partial depth. The most common transverse tie consists of unbonded • post-tensioned strands or bars. Approximately half the states grout the keyway before • post-tensioning and approximately half after post- tensioning. There is no consensus about the number of transverse • ties and the magnitude of post-tensioning force. Exterior and interior beams generally use the same • design. Most bridges have either full-width support or two-• point supports on each end. More states use plain elastomeric bearings rather than • laminated elastomeric bearings. In single-stage construction, all beams are generally • connected transversely at one time. In two-stage construction, a variety of sequences is • used. Approximately half the states require sandblasting of • the keyway before erection. The sandblasting is always done before shipment. The most common grout used for the keyways is a non-• shrink grout. CHAPTER SEVEN concLUsions and sUGGesTions For FUTUre research

28 proofing membrane is used because water accumu- lates below the asphalt Using nonprepackaged products for grout in the – keyways sUGGesTions For FUTUre research Responses to the survey for this synthesis provided the following topics for future research and development programs. design Design guidelines could be developed for the connection between adjacent box beams to prevent longitudinal cracks, reflective cracks, and subsequent leaks. Items to be evalu- ated include the number and location (vertically and hori- zontally) of transverse ties, sequence to connect adjacent beams, magnitude of the post-tensioning, optimum keyway configuration, and types of grout. A research problem state- ment addressing this topic is included in Appendix D. durability Methods could be identified to improve the long-term durability of adjacent box beam bridges, including the use of stainless steel stirrups, corrosion protection of the tie rods, waterproofing membranes that can accommodate dif- ferential deflection of beams, and long-term maintenance procedures. repair Practical methods could be developed to replace interior beams, restore load sharing between beams, repair dete- riorated or damaged box beams, and install transverse ties under staged construction. Construction Practices• Using stay-in-place expanded polystyrene to form – the voids Sandblasting the longitudinal keyway surfaces of – the box beams immediately before shipping to pro- vide a better bonding surface for the grout Cleaning the keyway surfaces with compressed air – or water before erection of the beams to provide a better bonding surface for the grout Grouting the keyways before transversely post-ten- – sioning to ensure compression in the grout Using a grout that provides a high bond strength to – the box beam keyway surfaces to limit cracking Providing proper curing for the grout to reduce – shrinkage stresses and ensure proper strength development Wet curing of the concrete deck for at least 7 days to – reduce the potential for shrinkage cracking and to provide a durable surface Maintenance Practices• Sealing longitudinal cracks as soon as they occur to – prevent salt and water penetration Washing the decks on an annual basis to remove – chlorides Cleaning drain holes on a regular basis to prevent – water accumulation in the boxes practices to avoid The following practices were identified as ones to avoid: Design Practices• Using nontensioned transverse ties because they do – not prevent cracking Construction Practices • Using an asphalt wearing surface unless a water- –

29 Greuel A., T.M. Baseheart, B.T. Rogers, R.A. Miller, and B. Shahrooz, “Evaluation of a High Performance Concrete Box Girder Bridge,” PCI Journal, Vol. 45, No. 6, Nov.– Dec.2000, pp. 60–71. Gulyas, R.J., G.J. Wirthlin, and J.T. Champa, “Evaluation of Keyway Grout Test Methods for Precast Concrete Bridges,” PCI Journal, Vol. 40, No. 1, Jan.–Feb. 1995, pp. 44–57. Hanna, K.E., G. Morcous, and M.K. Tadros, “Non-Post-Ten- sioned Transverse Design and Detailing of Adjacent Box Beam Bridges,” NCBC Concrete Bridge Conference, Proceedings, St. Louis, Mo., May 5–7, 2008. Hanna, K.E., G. Morcous, and M. K. Tadros, “Transverse Design and Detailing of Adjacent Box Beam Bridges,” PCI National Bridge Conference, Proceedings, Phoenix, Ariz., Oct. 22–24, 2007. Harries K.A., “Full-scale Testing Program on De-commis- sioned Girders from the Lake View Drive Bridge,” Report No. FHWA-PA-2006-008-EMG001, Pennsylvania Depart- ment of Transportation, Harrisburg, 2006. Hawkins, N.M. and J.B. Fuentes, “Structural Condition Assessment and Service Load Performance of Deterio- rated Pretensioning Deck Beam Bridges,” Civil Engi- neering Studies, Transportation Engineering Services, No. 124. Illinois Cooperative Highway and Transporta- tion Series No. 285, Report No. FHWA-IL-UI-285, 2003. Hlavacs, G.M., T. Long, R.A. Miller, and T.M. Baseheart, “Nondestructive Determination of Response of Shear Keys to Environmental and Structural Cyclic Loading,” Transportation Research Record 1574, Transportation Research Board, National Research Council, Washing- ton, D.C., 1997, pp. 18–24. Huckelbridge A.A. and H.H. El-Esnawi, “Evaluation of Improved Shear-Key Designs for Multi-Beam Box Girder Bridges,” Final Report No. FHWA/OH/97-009, Depart- ment of Civil Engineering, Case Western Reserve Uni- versity, Cleveland, Ohio, 1997. Huckelbridge, A.A., H.H. El-Esnawi, and F. Moses, “Shear Key Performance in Multibeam Box Girder Bridges,” Journal of Performance of Constructed Facilities, Vol. 9, No. 4, Nov. 1995, pp. 271–285. AASHTO, Standard Specifications for Highway Bridges, 17th ed., American Association of State Highway and Transportation Officials, Washington, D.C., 2002. AASHTO, AASHTO LRFD Bridge Design Specifications, 4th ed. (plus 2008 Interim Revisions), American Asso- ciation of State Highway and Transportation Officials, Washington, D.C., 2007. ACI Committee 222, “Protection of Metals in Concrete Against Corrosion,” ACI 222R-01, American Concrete Institute, Farmington Hills, Mich., 2001. ACI Committee 228, “Nondestructive Test Methods for Evaluation of Concrete in Structures,” ACI 228.2R-98, American Concrete Institute, Farmington Hills, Mich., 1998. Ahlborn, T.M., C.G. Gilbertson, H. Aktan, and U. Attanayake, “Condition Assessment and Methods of Abatement of Prestressed Concrete Box-Beam Deterioration,” Research Report RC-1470, Michigan Technological University, Houghton, 2005. Anderson, R.E., “New PPG Deck Beam Designs and Details on State System,” Memorandum 07.2, Illinois Depart- ment of Transportation, Springfield, Aug. 29, 2007. Attanayake, U. and H. Aktan, “Issues with Reflective Deck Cracks in Side-by-Side Box Beam Bridges,” NCBC Con- crete Bridge Conference, Proceedings, St. Louis, Mo., May 3–5, 2008. Badwan, I. and R. Y. Liang, “Transverse Post-tensioning Design of Precast Concrete Multibeam Deck,” PCI Jour- nal, Vol. 52, No. 4, July–Aug. 2007, pp. 84–92. Dong, H., L. Yue, and T.M. Ahlborn, “Performance of Joint Connections between Decked Prestressed Concrete Bridge Girders,” PCI National Bridge Conference, Pro- ceedings, Phoenix, Ariz., Oct. 22–24, 2007. El-Remaily, A., M.K. Tadros, T. Yamane, and G. Krause, “Transverse Design of Adjacent Prestressed Concrete Box Girder Bridges,” PCI Journal, Vol. 41, No. 4, July– Aug. 1996, pp. 96–107. Grace, N.F. and E. Jenson, “Use of Unbonded CFCC for Transverse Post-Tensioning of Side-by-Side Box-Beam Bridges,” Research Report RC-1509, Lawrence Techno- logical University, Southfield, Mich., 2008. reFerences

30 Report No. 06-13. Lehigh University, Bethlehem, Pa., 2006. Needham, D. and D. A. Juntunen, “Investigation of Condi- tion of Prestressed Concrete Bridges in Michigan,” Research Report R-1348, Michigan Department of Trans- portation, Lansing, 1997. Nottingham, D., Discussion of “Evaluation of Keyway Grout Test Methods for Precast Concrete Bridges,” by R.J. Gulyas, G.J. Wirthlin, and J.T. Champa, PCI Journal, Vol. 40, No. 4, July–Aug. 1995, pp. 98–103. PCI, PCI Bridge Design Manual, Precast/Prestressed Con- crete Institute, Chicago, Ill., 1997 (updated July 2004). PCI, The State-of-the-Art of Precast, Prestressed Adjacent Member Bridges, Precast/Prestressed Concrete Institute, Chicago, Forthcoming, expected in 2009. Ralls, M.L., B. Tang, S. Bhidé, B. Brecto, E. Calvert, H. Capers, D. Dorgan, E. Matsumoto, C. Napier, W. Nickas, and H.G. Russell, Prefabricated Bridge Elements and Systems in Japan and Europe, Report No. FHWA- PL-05-003, Federal Highway Administration, Washing- ton, D.C., 2005. Rao, C. and G.C. Frantz, “Fatigue Tests of 27-Year-Old Pre- stressed Concrete Bridge Box Beams,” PCI Journal, Vol. 41, No. 5, Sep.–Oct. 1996, pp. 74–83. Shahawy, M.E., “Feasibility Study of Transversely Pre- stressed Double Tee Bridges,” PCI Journal, Vol. 35, No. 5, Sep.–Oct. 1990, pp. 56–69. Shenoy, C.V. and G.C. Frantz, “Structural Tests of 27-Year- Old Prestressed Concrete Bridge Beams,” PCI Journal, Vol. 36, No. 5, Sep.–Oct. 1991, pp. 80–90. Whiting, D. and B.G. Stejskal, “Field Studies of Corrosion in Prestressed Concrete Bridges,” In Philip D. Cady Inter- national Symposium on Concrete Bridges in Aggressive Environments, ACI SP-151, ed. R.E. Weyers, American Concrete Institute, Farmington Hills, Mich., 1994, pp. 73–93. Yamane, T., M.K. Tadros, and P. Arumugasaamy, “Short to Medium Span Precast Prestressed Concrete Bridges in Japan,” PCI Journal, Vol. 39, No. 2, March–April 1994, pp. 74–100. Illinois DOT, “Concrete Deck Beams,” Guide Bridge Spe- cial Provisions (GBSP), No. 62, Illinois Department of Transportation, Springfield, June 13, 2008. Iqbal, M.A., “Davis Narrows Bridge in Brooksville, Maine: Fast Track Solution to an Environmentally Sensitive and Tourist Location,” PCI National Bridge Conference, Pro- ceedings, Grapevine, Tex., Oct. 23–25, 2006. Issa, M.A., C.L. Ribeiro do Valle, H.A. Abdalla, and S. Islam, “Performance of Transverse Joint Grout Materials in Full-Depth Precast Concrete Bridge Deck Systems,” PCI Journal, Vol. 48, No. 4, July–Aug. 2003, pp. 92–103. Lall, J., E.F. DiCocco, and S. Alampalli, Full-Depth Shear- Key Performance in Adjacent Prestressed-Beam Bridges, Special Report No. 124, Transportation Research and Development Bureau, New York State Department of Transportation, Albany, 1997. Lall, J., S. Alampalli, and E.F. DiCocco, “Performance of Full Depth Shear Keys in Adjacent Prestressed Box Beam Bridges,” PCI Journal, Vol. 43, No. 2, March–April 1998, pp. 72–79. Macioce, T.P., H.C. Rogers, R. Anderson, and D.C. Puzey, “Prestressed Concrete Box Beam Bridges—Two DOTs’ Experience,” PCI National Concrete Bridge Conference, Proceedings, Phoenix, Ariz., Oct. 22–24, 2007. Miller, R.A., G.M. Hlavacs, T. Long, and A. Greuel, “Full- Scale Testing of Shear Keys for Adjacent Box Girder Bridges,” PCI Journal, Vol. 44, No. 6, Nov.–Dec. 1999, pp. 80–90. Miller, R.A., K. Parekh, and J.T. Halsey, Destructive Load Testing of a Damaged and Deteriorated Prestressed Concrete Box Beam, Report No. UC-CII 93/03, Cincin- nati Infrastructure Institute, University of Cincinnati, Ohio, 1995. Modeer, V.A. and R.E. Anderson, “Damage Inspections and Monitoring for Deteriorated Precast Prestressed Con- crete Deck Beam Bridges,” Memorandum, Illinois Department of Transportation Springfield, 2005. Naito, C., R. Sause, I. Hodgson, S. Pessiki, and C. Desai, “Forensic Evaluation of Prestressed Box Beams from the Lake View Drive Bridge over I-70,” Final Report, ATLSS

31

32

33

34

35

36

37