Probabilistic stability analysis of reinforced soil slope with non-circular RLEM

Abstract

Probabilistic analysis is an important step in assessing the reliability of geosynthetic reinforced slopes. In this paper, a practical procedure is developed to implement probabilistic stability analysis for reinforced slope considering both the spatial variability of soil strength and the uncertainty of geosynthetics. In the proposed procedure, the random limit equilibrium method (RLEM) is used to calculate the factor of safety (FOS) and the simple genetic algorithm (SGA) is used to find the non-circular critical slip surface. A comparison is elaborately conducted between the procedure and random finite difference method (RFDM), verifying the feasibility of the proposed procedure in probabilistic stability analysis of reinforced soil slopes. Based on the proposed procedure, a series of probabilistic analyses are performed to study the effects of uncertainties of some key parameters on FOS and slip mass volume. The results show that the soil variability, correlation length and geosynthetic length are significant factors, while geosynthetic variability shows limited effect on FOS. The geosynthetic length plays an important role within the statistical characteristics of FOS and slip mass volume by differentiating the failure mechanism into internal failure mode and external failure mode.