Performance-based seismic analysis of bridges

Codes and specifications are in continuous development to improve the performance of structures against seismic loading. There is an increasing interest in performance-based design for structures, which can show the situation of the structure and its components under different load intensities. Linear procedures are not capable of working efficiently with performance-based design. If the structure responds beyond the elastic limits, linear analyses may indicate the location of first yielding but cannot predict failure mechanisms and account for redistribution of forces during progressive yielding. Nonlinear (static and dynamic) procedures are the solutions that can overcome this problem and show the performance level of the structures under any loading level. Nonlinear time-history analysis is not the panacea for seismic performance problems because it is too sophisticated and time consuming. Recently there have been quite a few studies in simplified nonlinear analysis, or what is referred to as the pushover analysis, that can give acceptable results with less effort. Nonlinear Static (Pushover) Procedure (NSP) is specified in the guidelines for seismic rehabilitation of buildings presented by FEMA-273 as an analytical procedure that can be used in systematic rehabilitation of structures. However, those guidelines were presented to apply the Displacement Coefficient Method (DCM) only for buildings. Capacity Spectrum Method (CSM) that was developed by the Applied Technology Council (ATC), also applies the NSP only for buildings. This study is intended to evaluate the applicability of NSP to bridges. For comparison purposes, the nonlinear time-history analysis is also performed. A three-span bridge of 97.5 m (320 ft) in total length was analyzed using both the NSP-DCM and NDP. Nine time-histories were implemented to perform the nonlinear time-history analysis. Three load patterns were used to represent distribution of the inertia forces resulting from earthquakes. Demand (target) displacement, base shear, and deformation of plastic hinges obtained from the NSP are compared with the corresponding values resulting from the nonlinear time history analysis. Analysis was performed using two levels of seismic load intensities (Design level and Maximum Considered Earthquake (MCE) level). Performance of the bridge was evaluated against these two seismic loads. Comparison shows good agreement between the results of the NSP and NDP in the design level, while the NSP gives conservative results, compared to the NOP, in the MCE level. Another case study is associated With Hsin Shi-Nan bridge which suffered some damage in the transverse direction during the Taiwan Chi-chi earthquake in 1999. The bridge was analyzed using the seismic record obtained from local stations. Analysis results proved that the rotation of plastic hinges for the damaged column was beyond the limit, which agrees with the observed damage.