Dynamic Reliability Analysis of Three-Dimensional Slopes Considering the Spatial Variability of Soil Parameters

This research aims to investigate 3D slope dynamic reliability considering the spatial variability of soil parameters. In this paper, practical expressions for the cohesion and friction angle in each part of the soil are simulated as two lognormal random fields by using the spectral representation method. Then, the finite element method is introduced to simulate the conditions and obtain the safety factor of the 3D slope. An efficient probabilistic analysis method for stochastic seismic analysis named the generalized probability density evolution method is performed, and the failure probability is provided. This technique, designed to use less data to obtain more accurate results, is helpful for analyzing high-dimensional stochastic problems. The results demonstrate that the dynamic failure probability of the slope can be reduced by appropriately increasing the variation coefficient of the parameters over a suitable range. In addition, the spatial variability of the friction angle has a strong influence on the slope reliability. Thus, the significance of the dynamic reliability analysis of 3D slopes is indicated and the realistic impact of the spatial variability of soil parameters is highligted.

Keywords: Three-dimensional slope, Generalized probability density evolution method (GPDEM), Random field, Reliability analysis, Spatial variability