Near-fault pulse-like ground motions (NFPLGMs) impose concentrated energy demands that can severely damage bridges, yet their scarcity and the influence of structural parameter uncertainties are often neglected in seismic fragility assessments. This study proposed a synthesis method for NFPLGMs by superposing low-frequency pulse components (extracted via the Gabor wavelet transform and low-pass filtering) with high-frequency stochastic components based on an evolutionary power spectrum. A three-span reinforced concrete bridge was modeled in OpenSeesPy, and Incremental Dynamic Analysis (IDA), together with a quadratic response surface model, were used to plot seismic fragility curves. The damping ratio (
ξ), elastic modulus of steel reinforcement (
Es), yield strength of steel reinforcement (
fy), diameter of longitudinal reinforcement (
D), and peak ground acceleration (PGA) were treated as random variables. Sensitivity indices were computed using Monte Carlo sampling (
n = 10,000). Results show that
ξ most strongly affects the displacement ductility ratio of the bridge pier (
ud) (variation of up to 32.6%), while
Es dominates the shear deformation of the bridge bearing (
d) (variation of up to 43.8%). Neglecting structural parameter uncertainties overestimates median PGA thresholds (
mR) for different damage states by 1.5%–36.1%, and replacing NFPLGMs with ordinary ground motions overestimates seismic capacity by 1.7%–36.6%. The bridge bearing is consistently more vulnerable than the pier, with a collapse probability of 0.9566 at PGA = 1.0 g. These findings highlight the necessity of incorporating both NFPLGM characteristics and structural parameter uncertainties into bridge seismic fragility assessment. On the other hand, when seismic retrofitting of bridges is carried out using coating materials, priority should be given to more vulnerable components, such as bridge bearings, to improve the utilization efficiency of limited resources.
Full article