Seismic fragility and demand hazard analyses for earth slopes incorporating soil property variability

Fragility functions and demand hazard curves are important metrics for assessing the seismic damage and risk of an engineering system. This study aims at conducting seismic fragility and demand hazard analyses for earth slopes incorporating soil property variability. The soil parameters of cohesion, friction angle, and initial shear modulus are taken as random variables following a lognormal distribution. Based on a suite of 100 input motions selected, nonlinear dynamic analyses are conducted for a 10-m-high earth slope implemented in FLAC. The corrected cloud analysis is used to describe the relationship of ground-motion intensity measure and engineering demand parameter (EDP) of the slope. The response surface method in conjunction with Monte Carlo simulation is employed to efficiently develop fragility functions accounting for the variability of soil property parameters. The seismic demand hazard curves implemented by incorporating the soil property variability of the slope are subsequently developed. Comparative results indicate that: (1) considering the soil property variability would shift the median and increase the dispersion of the slope fragility functions; (2) neglecting the soil property variability would significantly underestimate the seismic demand hazard, especially for relatively high EDP levels; (3) increasing the variability of soil property parameters would increase the dispersion of fragility functions, inevitably escalating the seismic demand of the slope system. The proposed numerical-based approach provides an effective way of conducting the fragility and demand hazard analyses for earth slopes.