Probabilistic assessment of seismic performance of slopes considering the sensitivity of sliding surface to input motion

Abstract

Seismic fragility assessments that predicts the exceedance probabilities of predefined limit states of a structure subjected to earthquakes are widely used to reduce economic losses and casualties. This paper presents a probabilistic assessment of slopes based on the permanent displacements calculated from a suite of two-dimensional (2D) nonlinear dynamic analyses. The slopes are categorized into four groups based on static factor of safety (FSstatic). The centrifuge test measurements of a slope composed of granular soil are utilized to validate the 2D finite element numerical model. The Newmark displacements are calculated by integrating the stresses imposed on the sliding surface. The sliding surface is determined from the maximum accumulated strain contours, thereby accounting for the slope and input motion characteristics. The Newmark displacement is correlated with various earthquake intensity measures (IMs). The optimal IMs are selected based on the goodness of fit, efficiency, practicality, and proficiency. The fragility curves are developed using selected optimal IMs for three sliding displacement-based damage levels, which are minor, moderate, and severe. The results show that FSstatic, which contains information on both the slope geometry and the shear strength of soil, is an effective parameter to estimate the Newmark displacement. In addition, a suite of seismic fragility surfaces are developed using dual IMs.