Slope Reliability Analysis Based on Nonlinear Stochastic Finite Element Method

Abstract

AbstractIn slope stability reliability analysis, the deterministic analysis method is usually used to calculate the safety factor to measure the stability of the slope, but the traditional deterministic analysis method cannot fully consider and describe the natural spatial variability of soil, which leads to the failure probability calculation of the slope is not accurate enough. Aiming at the problem of spatial variability of soil mechanical parameters in slope stability analysis, this paper proposes a stochastic finite element method for calculating the distribution of FS (factor of safety) of dam slopes, and MC (Monte Carlo) strength reduction combined method and MC direct method are proposed to calculate the reliability of slope. Taking isotropic two-dimensional slope as an example: firstly, the random field is sampled to get the corresponding random field of material properties, and then the slope displacement, stress and plasticity zone results are calculated; then on the basis of NMC times sampling of random field, there are: (i) Combined method (M1): the strength reduction method is used to get the reduction coefficient of each sample, and then its distribution, slope failure probability and reliability index are calculated; (ii) MC direct method (M2): using the viscop lastic method to solve and judge the instability of slopes, and the instability cases under all sample conditions are counted to obtain the failure probability and reliability index of slopes. The results show that the slope stability analysis considering the random field of material properties can obtain the real and reliable slope stability analysis results by comprehensively evaluating the slope safety through the mean value, variance, distribution and reliability index of the slope safety factor.