Abstract

The bridge network is an essential part of the transportation infrastructure for mobility. However, bridges are susceptible to seismic activity that produces damages and affects the entire transportation system. Their fragilities have been studied based mostly on analytical models but neglecting empirical data from reports. Few studies compared fragility curves from different development methods, analyzing their similarities, advantages, and limitations for a better risk assessment. This article aims to develop and compare fragility curves for bridges based on empirical and analytical methods. The empirical approach is based on damage reports from the 8.8Mw Maule earthquake, while the analytical approach is based on numerical simulations of the seismic response using ground motion components from stations that recorded the Maule earthquake. The curves are developed for bridges with steel and concrete beams, representative of 57% of Chilean bridges. The comparative Kolmogorov–Smirnov test revealed significant compatibility in moderate and severe damage curves when comparing these approaches, unlike the slight damage curves. The lognormal mean parameter varies between the methods by 36%, 1.5%, and 7.5% for the slight, moderate, and severe damage curves respectively for a particular bridge configuration, showing considerable differences in the slight damage curve between methods.