Monte carlo simulation for stability of finite slope subjected to earthquake loading

Abstract

Analysis of finite slopes is a common practical problem which faces the geotechnical engineering. The aim of the slope analysis is to predict the safety factor against shear failure of the slope which is the ratio between the resisting and the disturbing forces developed along an assumed slip surface. Presence of seismic forces due earthquake lead to more critical situation. The deterministic approaches are employed for this purpose. These approaches do not take into account the uncertainty related to soil properties and/or seismic loads in their procedures. In the present work, a probabilistic approach is adopted using the Monte Carlo simulation technique to study the effect of uncertainty due to soil properties (unit weight, angle of internal friction and cohesion) and seismic accelerations (vertical and horizontal). The results of the work demonstrate that the use of probabilistic approach will yield more accurate results for estimating the safety factor and then economic design of the slope. Also, the results reveal that the statistical properties for the soil shear strength parameters have significant impact on the standard deviation of the output (factor of safety), while the soil unit weight and horizontal acceleration have small effect. However, the vertical acceleration has no effect on it. On other hand, the mean of the output distribution does not affect by the statistical parameters of all input variables. All cases simulated in this study give reliability index (RI) less than 3. Hence, RI values indicate that there are unsatisfactory level of safety for the slope subjected to seismic loading. Finally, the soil shear strength parameters have the same positive significant impact on the safety factor. But, the soil properties and the coefficient of the horizontal acceleration should be selected carefully in the stability analysis due to the effect of their uncertainty on the analysis results.