Experimental and Analytical Seismic Studies of a Four-span Bridge System with Composite Piers

Report No. CCEER-13-17Funded by the National Science Foundation through the Network for Earthquake Engineering Simulation (NEES) research program, a major multi-university research project has been in progress at the University of Nevada, Reno. This study describes the study of one of the three large-scale bridge models that were tested to failure on three shake tables system. This model was supported on fiber-reinforced polymer (FRP) composite piers implementing accelerated bridge construction (ABC) techniques. The bridge was a quarter scale model of a 4-span bridge with continuous reinforced concrete superstructure and a drop cap, two-column pier design. Each pier utilized different unconventional FRP details. The purpose of using these innovative details was to improve the seismic performance of the bridge. The first pier consisted of cast-in-place concrete-filled glass FRP tubes with ±55 degree fibers. The second pier consisted of two segmental reinforced concrete columns wrapped with layers of unidirectional carbon FRP sheets to provide confinement and shear reinforcement. Only nominal hoops were used to hold the longitudinal reinforcement, as FRP jacket and tube were sufficient in providing confinement and shear required reinforcement. The third pier had the same configuration as that of pier 1 but the columns and footing were precast. The top connections in piers 1 and 3 consisted of pipe-pin joints to facilitate ABC and provide hinge behavior. The objectives of the study presented in this document were to evaluate the biaxial seismic performance of this bridge system incorporating composite piers, investigate the performance of each detail and compared them to each other and to conventional ones, determine the influence of abutment-superstructure interaction on the response, assess the performance of a bridge model incorporating ABC techniques, evaluate sufficiency of analytical modeling of the performance of composite material and details, and to conduct parametric study of different variations of the bridge model to study the effect of several important factors such as near-fault earthquake effects and the variations in the configuration of the bridge model. A large-scale 4-span bridge model was designed, constructed, and subjected to simulated earthquake loading on three shake tables. The simulated shake table motions were the modified 1994 Northridge, CA ground motion recorded in Century City and were applied to the bridge model in ten runs with increasing amplitudes. Over 380 channels of data were collected. Compared to conventional reinforced concrete bridges, experimental results showed superior performance under extreme seismic loading even under lateral drift ratios exceeding 9%. Extensive post-test analytical studies were conducted and it was determined that a computational model of the bridge that included bridge-abutment interaction using OpenSees was able to provide satisfactory estimations of key structural response parameters such as superstructure displacements. The analytical model was also used to conduct parametric studies on response of the bridge model and its variations under near-fault excitations. The effects of changing the column section properties were also explored. It was found that concrete-filled FRP tube piers and CFRP wrapped post-tensioned segmental piers reduce residual displacements compared to their conventional reinforced concrete counter parts even under impulsive near-fault motions