Comparison of seismic demands and fragilities of bridge columns using equivalent-linear and nonlinear site response analyses

One-dimensional site response analyses have been widely employed to estimate the amplification of ground motions as the motions propagate from the bedrock to the ground surface. The two primary approaches for site response analysis are equivalent-linear and nonlinear analyses. The former is commonly utilized in practice because of its simplicity and low computational cost, whereas the latter is more complex, but better at capturing realistic soil behaviors under strong ground motions. Thus, the results of equivalent linear and nonlinear site response analyses are expected to diverge under high-intensity ground motions. This study aimed to compare the seismic demands of bridge columns subjected to free-field motion generated using the two site response analyses. To this end, a series of nonlinear dynamic analyses of a bridge system was performed using 319 ground motions obtained from each site response analysis method. Subsequently, the maximum and residual drift ratios of the bridge column were compared, and their fragility curves were developed in terms of the maximum drift ratio. For site classes D and E, the average maximum drift ratios of the bridge column under the ground motions obtained from the equivalent linear analysis were approximately 1.40-1.49 times higher than those from the nonlinear approach. Similarly, the average residual drift ratios obtained using equivalent-linear-based ground motions were approximately 1.53-3.77 times higher than those obtained using nonlinear-based ground motions. Column fragility curves subjected to equivalent-linear-based ground motions showed 7%-12% higher exceedance probabilities compared with the nonlinear-based ground motions.