DCT 144 Load Rating of Arch Bridges by MICHAEL CHAJES Department of Civil and Environmental Engineering College of Engineering University of Delaware October 2002 Delaware Center for Transportation University of Delaware 355 DuPont Hall Newark, Delaware 19716 (302) 831-1446 The Delaware Center for Transportation is a university-wide multi-disciplinary research unit reporting to the Chair of the Department of Civil and Environmental Engineering, and is co-sponsored by the University of Delaware and the Delaware Department of Transportation. DCT Staff Ardeshir Faghri Jerome Lewis Director Associate Director Wanda L. Taylor Lawrence H. Klepner Sandi Wolfe Assistant to the Director T2 Program Coordinator Secretary DCT Policy Council Robert Taylor, Co-Chair Chief Engineer, Delaware Department of Transportation Eric Kaler, Co-Chair Dean, College of Engineering The Honorable Tony DeLuca Chair, Delaware Senate Transportation Committee The Honorable Richard Cathcart Chair, Delaware House of Representatives Transportation Committee Timothy K. Barnekov Dean, College of Human Resources, Education and Public Policy Michael J. Chajes Chair, Civil and Environmental Engineering Ralph A. Reeb Director of Planning, Delaware Department of Transportation Stephen Kingsberry Director, Delaware Transit Corporation Shannon Marchman Representative of the Director of the Delaware Development Office Roger Roy Representative, Transportation Management Association Jim Johnson Executive Director, Delaware River & Bay Authority Delaware Center for Transportation University of Delaware Newark, DE 19716 (302) 831-1446 Load Rating of Arch Bridges by MICHAEL J. CHAJES Department of Civil and Environmental Engineering University of Delaware Newark, Delaware 19716 DELAWARE CENTER FOR TRANSPORTATION University of Delaware Newark, Delaware 19716 This work was sponsored by the Delaware Center for Transportation and was prepared in cooperation with the Delaware Department of Transportation. The contents of this report reflect the views of the authors who are responsible for the facts and accuracy of the data presented herein. The contents do not necessarily reflect the official views of the Delaware Center for Transportation or the Delaware Department of Transportation at the time of publication. This report does not constitute a standard, specification, or regulation. Executive Summary Motivation for the Research As the literature contains little information on evaluating the load-carrying capacity of arch bridges, the burden falls to bridge owners to develop their own methods for rating these bridges. Some research has been done on determining how to load rate masonry arches, virtually none has addressed the process for concrete arches. In Delaware, 27 of the 33 arch bridges in the bridge inventory are made of reinforced concrete, while only six are masonry. A majority of the arch bridges in Delaware are at least 75 years old and do not have existing plans or documentation. As a result, additional research was needed to establish a suitable rating methodology for concrete arch bridges. Background on Load Rating Methods Several methods have been explored for load rating of arch bridges, with most of the research being conducted outside the United States. DelDOT is currently using the STAAD-III computer software program to aid in load rating the arch bridges in their bridge inventory. This software package is not specific to arch bridges; rather, it is used to analyze and design a variety of structure types, including those made of truss, frame, and plate/shell elements. When an arch bridge is modeled using this software, it is represented as an assemblage of frame elements. Once the arch is modeled, truck loads representing the various rating vehicles are applied. Then the bridge is analyzed using STAAD-III. Next, three influence lines are determined for the three critical arch sections. These critical arch sections are the midpoint of the arch, the quarter point, and at the springing. Basic principles are used to calculate the properties for the critical sections. The stresses are calculated at the critical sections. Once the influence lines are determined, the load ratings are determined. Then the arch is rated for the controlling sections. This method is the basis of the method developed in this report. The primary difference between STAAD-III and the proposed method is that effective width and live load distribution are altered based on load test results. Bridge 1-136 As a Case Study Bridge 1-136 is located in New Castle County, Delaware, north of Wooddale. It carries Barley Mill Road and crosses over a tributary of the Red Clay Creek. It is a concrete arch bridge with an 18-foot span length and a 9-foot rise. This bridge was built in 1932 and has never been reconstructed. The arch is circular in shape. Bridge 1-136 was chosen for this study because it was typical of the types of concrete arch bridges in Delaware’s bridge inventory. Its 18-ft. span length is close to the average of 25 ft. It is made of concrete, as are the majority of the arch bridges in the inventory. Bridge 1-136 is also posted, which makes it a good candidate for load testing. If the load rating method developed shows that the load posting can be removed, it will indicate that the previous method of rating was conservative. 1 Experimental/Load Test Method To determine whether the analytical method developed in this report is accurate to use when load rating arch bridges, a load test was conducted on Bridge 1-136 using two DelDOT dump trucks crossing the bridge in a variety of combinations, and the results were compared to those from the proposed method. The field tests showed that the governing load location is at the outside edge of the arch at the arch crown. This location sees the highest strains because there is limited ability for transverse load distribution at the edge of the arch. The calculated effective width b (transverse load distribution) value was found to be 8.76 ft. This b value was eff eff also used in the longitudinal direction in the analytical model. The peak strain ε was 20.70 µε, and it occurred at the outside edge of the arch crown (the governing load location). It was also found that higher strains are produced when two trucks are on the bridge at one time. There is a 12% increase in the strain value at the governing load location from when one truck is on the bridge to when two trucks are present. Analytical Method Bridge 1-136 was modeled using ten tapered straight-line beam elements having a unit width. All of the load cases were representative of a single wheel load acting on a unit width of arch. In order to get to the final recommended bridge model, a series of load cases was explored. The refinement of the load cases was based on the comparison of the calculated strains and the measured strains on the bridge. Once the STAAD analysis was run for each live load case, the axial forces and bending moments were converted into strains so that they could be compared to the strains obtained in the experimental load test. To determine the load rating for Bridge 1-136, the moment and axial force capacities at each node of the bridge model were calculated. Once this was done, the load rating factors were determined. Since DelDOT uses the load factor rating (LFR) method, that was the rating method used to rate Bridge 1-136. The load rating factors were calculated based on the axial loads and the bending moments for all load distributions and for concrete strengths of both 3,000 psi and 7,000 psi. It was determined that the rating factors due to bending moment and concrete strength of 3,000 psi governed. Inventory and operating ratings were determined. It was also necessary to determine the posting rating factor. Based on the recommended load distribution of 2V:4H, the posting load rating factor is 1.71 for a S335 truck type (similar to the test truck). According to this rating factor it would be possible to remove the load posting sign for Bridge 1-136. Recommendations and Conclusions The recommended load rating method for the arch bridges in the DelDOT inventory is basically the same as the one currently being used. The only difference is that the recommended method uses a different load distribution for the longitudinal and transverse directions. DelDOT is currently using the AASHTO load distribution of 2 units vertical to 1 unit horizontal, as opposed to the recommended distribution of 2 2 units vertical to 4 units horizontal. (This recommended distribution applies only to Bridge 1-136.) Similarly, the AASHTO load distribution was used as a starting point for the analysis described in this report. The posting level obtained using the AASHTO distribution was roughly half the value currently being used by DelDOT. DelDOT’s current posting rating factor for Bridge 1-136 is 0.66. This research yielded a rating of 0.31. It appears that the difference is due to the fact that the edge section, which has half the effective width, governs. When the recommended distribution of 2V:4H was used, a posting factor of 1.71 was determined. Since the recommended load rating method was based on only a single experimental load test, further verification is needed. Additional load tests can be performed on other arch bridges to determine whether this method is satisfactory. 3 Table of Contents Abstract..................................................................................................2 Executive Summary................................................................................3 Motivation for the Research....................................................................3 Background on Load Rating Methods........................................................3 Bridge 1-136 As a Case Study.................................................................3 Experimental/Load Test Method ..............................................................4 Analytical Method..................................................................................4 Recommendations and Conclusions..........................................................5 Chapter 1: Introduction.........................................................................6 1.1 Research Motivation .....................................................................6 1.2 Research Objective.......................................................................7 1.3 Research Plan..............................................................................7 Chapter 2: Background Information......................................................8 2.1 Prior Research .............................................................................8 2.1.1 Boothby’s Method...................................................................8 2.1.2 Walrath’s Method ................................................................. 10 2.1.3 Modes of Failure of Masonry Arch Bridges................................ 11 2.1.4 Spandrel Wall Assessment..................................................... 11 2.1.5 Experimental Load Tests ....................................................... 12 2.2 Arch Load Rating Procedures in Use Today..................................... 13 2.2.1 MEXE Method....................................................................... 13 2.2.2 STAAD-III Method................................................................ 16 2.3 DelDOT’s Current Load Rating Method........................................... 17 Chapter 3: DelDOT Arch Bridge Inventory...........................................18 Chapter 4: Experimental Method/Load Test ........................................21 4.1 Method Description..................................................................... 21 4.2 Experimental Load Test Example.................................................. 21 4.2.1 Bridge Description .................................................................... 21 4.2.2 Setup of the Experimental Load Test....................................... 22 4.2.3 Running the Experimental Load Test....................................... 25 4.2.4 Experimental Load Test Results.............................................. 25 4.3 Summary of Results ................................................................... 31 Chapter 5: Analytical Method...............................................................35 5.1 Method Description..................................................................... 35 5.2 Analytical Method Example.......................................................... 35 5.2.1 Bridge Model ....................................................................... 35 5.2.2 Load Cases.......................................................................... 36 5.2.3 Analytical Test Results .......................................................... 40 5.2.4 Load Rating......................................................................... 45 5.3 Summary of Results ................................................................... 49 Chapter 6: Recommendations and Conclusion.....................................51 6.1 Recommendations for Load Rating an Arch Bridge........................... 51 6.2 Conclusion ................................................................................ 52 References ...........................................................................................54 Appendix: Photographs of Bridge1-136...............................................55 Abstract In the Delaware Department of Transportation (DelDOT) Bridge Inventory, there are 33 arch bridges, most of which were built over 75 years ago. Due to the age of these bridges, it is necessary to make sure that they are strong enough to carry the ever-increasing loads and volumes of traffic on today’s roadways. In order to do this, accurate methods for evaluating and rating arch bridges are needed. Since there is currently no designated method of load rating arch bridges at DelDOT, a standard method needs to be established. The research conducted in this project aids in the development of an analytical method to load-rate arch bridges. In order to establish this new method, analytical modeling and experimental load testing were performed on DelDOT’s Bridge1-136. From the results, recommendations are made regarding future methods for evaluating and rating DelDOT’s arch bridges. 2 Executive Summary Motivation for the Research As the literature contains little information on evaluating the load-carrying capacity of arch bridges, the burden falls to bridge owners to develop their own methods for rating these bridges. Some research has been done on determining how to load rate masonry arches, virtually none has addressed the process for concrete arches. In Delaware, 27 of the 33 arch bridges in the bridge inventory are made of reinforced concrete, while only six are masonry. A majority of the arch bridges in Delaware are at least 75 years old and do not have existing plans or documentation. As a result, additional research was needed to establish a suitable rating methodology for concrete arch bridges. Background on Load Rating Methods Several methods have been explored for load rating of arch bridges, with most of the research being conducted outside the United States. DelDOT is currently using the STAAD-III computer software program to aid in load rating the arch bridges in their bridge inventory. This software package is not specific to arch bridges; rather, it is used to analyze and design a variety of structure types, including those made of truss, frame, and plate/shell elements. When an arch bridge is modeled using this software, it is represented as an assemblage of frame elements. Once the arch is modeled, truck loads representing the various rating vehicles are applied. Then the bridge is analyzed using STAAD-III. Next, three influence lines are determined for the three critical arch sections. These critical arch sections are the midpoint of the arch, the quarter point, and at the springing. Basic principles are used to calculate the properties for the critical sections. The stresses are calculated at the critical sections. Once the influence lines are determined, the load ratings are determined. Then the arch is rated for the controlling sections. This method is the basis of the method developed in this report. The primary difference between STAAD-III and the proposed method is that effective width and live load distribution are altered based on load test results. Bridge 1-136 As a Case Study Bridge 1-136 is located in New Castle County, Delaware, north of Wooddale. It carries Barley Mill Road and crosses over a tributary of the Red Clay Creek. It is a concrete arch bridge with an 18-foot span length and a 9-foot rise. This bridge was built in 1932 and has never been reconstructed. The arch is circular in shape. Bridge 1-136 was chosen for this study because it was typical of the types of concrete arch bridges in Delaware’s bridge inventory. Its 18-ft. span length is close to the average of 25 ft. It is made of concrete, as are the 3 majority of the arch bridges in the inventory. Bridge 1-136 is also posted, which makes it a good candidate for load testing. If the load rating method developed shows that the load posting can be removed, it will indicate that the previous method of rating was conservative. Experimental/Load Test Method To determine whether the analytical method developed in this report is accurate to use when load rating arch bridges, a load test was conducted on Bridge 1-136 using two DelDOT dump trucks crossing the bridge in a variety of combinations, and the results were compared to those from the proposed method. The field tests showed that the governing load location is at the outside edge of the arch at the arch crown. This location sees the highest strains because there is limited ability for transverse load distribution at the edge of the arch. The calculated effective width b (transverse load distribution) value was found to be 8.76 ft. eff This b value was also used in the longitudinal direction in the analytical model. eff The peak strain ε was 20.70 µε, and it occurred at the outside edge of the arch crown (the governing load location). It was also found that higher strains are produced when two trucks are on the bridge at one time. There is a 12% increase in the strain value at the governing load location from when one truck is on the bridge to when two trucks are present. Analytical Method Bridge 1-136 was modeled using ten tapered straight-line beam elements having a unit width. All of the load cases were representative of a single wheel load acting on a unit width of arch. In order to get to the final recommended bridge model, a series of load cases was explored. The refinement of the load cases was based on the comparison of the calculated strains and the measured strains on the bridge. Once the STAAD analysis was run for each live load case, the axial forces and bending moments were converted into strains so that they could be compared to the strains obtained in the experimental load test. To determine the load rating for Bridge 1-136, the moment and axial force capacities at each node of the bridge model were calculated. Once this was done, the load rating factors were determined. Since DelDOT uses the load factor rating (LFR) method, that was the rating method used to rate Bridge 1- 136. The load rating factors were calculated based on the axial loads and the bending moments for all load distributions and for concrete strengths of both 3,000 psi and 7,000 psi. It was determined that the rating factors due to bending moment and concrete strength of 3,000 psi governed. Inventory and operating ratings were determined. It was also necessary to determine the posting rating factor. Based on the recommended load distribution of 2V:4H, the posting load rating factor is 1.71 for a S335 truck type (similar to the test truck). According to this rating factor it would be possible to remove the load posting sign for Bridge 1-136. 4