Dependence analysis on the seismic demands of typical components of a concrete continuous girder bridge with the copula technique

Abstract

The seismic reliability of a bridge system is significantly affected by the dependence among typical bridge components. This study demonstrates the process of using a copula technique to describe the nonlinear dependence among component seismic demands isolated from their marginal probability distributions. A suite of 100 bridge-ground motion samples were developed with the Latin hypercube sampling approach and bin approach. Based on the incremental dynamic analysis, the tail dependence among component seismic demands at different intensity levels was analyzed with the best-fitting copula function selected by the minimum distance method. In the longitudinal direction, the dependence increased first and then decreased with the ground motion intensity, while the dependence slightly decreased in the transverse direction. At low-intensity levels, the upper tail dependence among components was strong in both directions. At high-intensity levels, the upper and lower tail dependences were weak in the longitudinal direction, while the upper and lower tail dependences were strong in the transverse direction. Compared to the linear correlation coefficient, the copula technique provides an efficient way to describe the tail dependence among component seismic demands and can be used extensively in the seismic reliability analysis of the bridge system.