Probabilistic Analysis Of Planar Failure In Rock Slope Based On Latin Hypercube Sampling Method

The stability of rock slope is mainly controlled by the structural plane. When the structural plane of a rock slope is not clear, the plane corresponding to the critical inclination angle of the slope can be treated as a sliding surface in the probabilistic analysis of planar failure. The traditional evaluation method for the slope stability, which is based on the index of safety factor, does not take account of the variability of mechanical parameters. Therefore, probabilistic analysis of slope stability is more reasonable. Based on the analytical algorithm of the critical inclination angle, this thesis analyzes the reliability of rock slopes considering both the variability of parameters and the variable critical sliding surface by using the Latin Hypercube sampling (LHS) method, and applies the research results into the engineering practice. The main research work and the conclusions are summarized as follows:(1) The analytical algorithm of critical inclinations in three types of slope models, which is applicable in the evaluation of slope stability, the research of slope reliability and the slope engineering, are given in the thesis.(2) In order to overcome the shortages of the first-order second-moment method and the Monte Carlo method, the LHS method is introduced into the reliability analysis of rock slopes. Through applying and comparing these three methods into an slope example, it shows that the LHS method can improve the efficiency of reliability analysis due to its better performance in the fitting of distribution of the input variables and the convergence speed.(3) The rock slope of San Mau Ping in Hong Kong is analyzed by the LHS modeling by considering the variability of both the parameters and the critical sliding surface. Three different failure probabilities with respect to the critical sliding surface, the critical probabilistic surface and the systematic reliability are calculated, respectively. The results show that the first condition has the minimum probability, while the last condition has the maximum value, which indicates that it is more conservative to adopt the systematic reliability analysis in the rock slope stability analysis.(4) In the systematic reliability calculation, based on the sensitivity analysis of the influences of the parameter variability on the critical inclination value and the safety factor, the shear strength parameters are shown to be most sensitive. In addition, the analysis of the parameter correlation shows that the failure probability decreases with the increase of the negative parameter correlation.(5) The analytical algorithm of critical inclination and the systematic reliability analysis method are finally applied into the stability evaluation of a rock slope surround the electric converter station in Linyi City, Shandong province. Both the safety factor and the failure probability show that the slope is safe.