A group of researchers, engineers, and government transportation officials have teamed up to design two bridges with simply-supported FRP composite structural beams. The Toms Creek Bridge, located in Blacksburg, Virginia, has been in service for six years. Meanwhile, the Route 601 Bridge, located in Sugar Grove, Virginia, has been in service for two years. Researchers have conducted load tests at both bridges to determine if their performance has changed during their respective service lives. The key design parameters under consideration are: deflection, wheel load distribution, and dynamic load allowance. The results from the latest tests in 2003 yield little, yet statistically significant, changes in these key factors for both bridges. Most differences appear to be largely temperature related, although the reason behind this effect is unclear. For the Toms Creek Bridge, the largest average values from the 2003 tests are 440 me for service strain, 0.43 in. (L/484) for service deflection, 0.08 (S/11.1) for wheel load distribution, and 0.64 for dynamic load allowance. The values for the Route 601 Bridge are 220 me, 0.38 in. (L/1230), 0.34 (S/10.2), and 0.14 for the same corresponding paramters. The recommended design values for the dynamic load allowance in both bridges have been revised upwards to 1.35 and 0.50 for the Toms Creek Bridge and Route 601 Bridge, respectively, to account for variability in the data. With these increased factors, the largest strain in the toms Creek Bridge and Route 601 Bridge would be less than 13% and 12%, respectively, of ultimate strain. Therefore, the two bridges continue to provide a large factor of safety against failure. / Master of Science

The Tom's Creek Bridge is a small-scale demonstration project involving the use of fiber-reinforced polymer (FRP) composite girders as the main load carrying members. The project is intended to serve two purposes. First, by calculating bridge design parameters such as the dynamic load allowance, transverse wheel load distribution and deflections under service loading, the Tom's Creek Bridge will aid in modifying current AASHTO bridge design standards for use with FRP composite materials. Second, by evaluating the FRP girders after being exposed to service conditions, the project will begin to answer questions about the long-term performance of these advanced composite material beams when used in bridge design. This thesis details the In-Service analysis of the Tom's Creek Bridge. Five load tests, at six month intervals, were conducted on the bridge. Using mid-span strain and deflection data gathered from the FRP composite girders during these tests the above mentioned bridge design parameters have been determined. The Tom's Creek Bridge was determined to have a dynamic load allowance, IM, of 0.90, a transverse wheel load distribution factor, g, of 0.101 and a maximum deflection of L/488. Two bridge girders were removed from the Tom's Creek Bridge after fifteen months of service loading. These FRP composite girders were tested at the Structures and Materials Research Laboratory at Virginia Tech for stiffness and ultimate strength and compared to pre-service values for the same beams. This analysis indicates that after fifteen months of service, the FRP composite girders have not lost a significant amount of either stiffness or ultimate strength. / Master of Science

more funding must be allocated for their rehabilitation or replacement. The Federal Highway Administration's (FHWA) Long-Term Bridge Performance (LTBP) Program has been developed to help bridge stakeholders make the best decisions concerning the allocation of these funds. This is done through the use of high quality data obtained through numerous testing processes. As part of the LTBP Pilot Program, researchers have performed live load tests on the U.S. Route 15 Southbound bridge over Interstate-66. The main performance and behavior characteristics focused on are service strain and deflection, wheel load distribution, dynamic load allowance, and rotational behavior of bridge bearings. Data from this test will be used as a tool in developing and refining a plan for long-term bridge monitoring. This includes identifying the primarily loaded girders and their expected range of response under ambient traffic conditions. Information obtained from this test will also aid in the refinement of finite element models by offering insight into the performance of individual bridge components, as well as overall global behavior. Finally, the methods and results of this test have been documented to allow for comparison with future testing of this bridge, which will yield information concerning the changes in bridge behavior over time. / Master of Science

The Route 601 Bridge in Sugar Grove, Virginia spans 39 ft over Dickey Creek. The Bridge is the first to use the Strongwell 36 in. fiber reinforced polymer (FRP) double web beam (DWB) in its superstructure. Replacement of the old bridge began in June 2001, and construction of the new bridge was completed in October 2001. The bridge was field tested in October 2001 and June 2002. This thesis details the field evaluation of the Rt. 601 Bridge. Using mid span deflection and strain data from the October 2001 and June 2002 field tests, the primary goal of this research was to determine the following AASHTO bridge design parameters: wheel load distribution factor g, dynamic load allowance IM, and maximum deflection. The wheel load distribution factor was determined to be S/5, a dynamic load allowance was determined to be 0.30, and the maximum deflection of the bridge was L/1500. Deflection results were lower than the AASHTO L/800 limit. This discrepancy is attributed to partial composite action of the deck-to-girder connections, bearing restraint at the supports, and contribution of guardrail stiffness. Secondary goals of this research were to quantify the effect of diaphragm removal on girder distribution factor, determine torsion and axial effects of the FRP girders, compare responses to multiple lane symmetrical loading to superimposed single lane response, and compare the field test results to a finite element and a finite difference model. It was found that diaphragm removal had a small effect on the wheel load distribution factor. Torsional and axial effects were small. The bridge response to multilane loading coincided with superimposed single lane truck passes, and curb-stiffening effects in a finite difference model improved the accuracy of modeling the Rt. 601 Bridge behavior. / Master of Science