Prediction of Uncertain Failed Slope Shear Strength Parameters by Stochastic Back-Analysis

The shear strength parameters of a slope failure which determine from the laboratory may be misleading. As a solution, the back-analysis method can be estimated much reliable mobilized in-situ parameters. On the other hand, the conventional approaches for back-analysis cannot reflect the uncertainty of the soil parameters. While the characterization of soil is subjected to uncertainties due to the inherent variability of properties within it. In this paper, the application of the stochastic back-analysis in an arbitrary slope is investigated. For this purpose, cohesion, internal friction angle, and specific gravity are considered as stochastic variables. In this way, a coded computer program in MATLAB is developed to couple the back -analysis with Monte Carlo Simulations (MCS). For each set data the slip line with the minimum factor of safety is determined by Particle Swarm Optimization (PSO) technique and the simplified Janbu’s method. Based on stochastic back-analysis uncertainties of developed shear strength parameters with respect to the depth of the slip line and probability density function of these parameters are obtained.

کلیدواژه ها:

Back Analysis ، Slope Stability ، Stochastic Analysis ، UncertainParameters