2010 年 59 巻 5 号 p. 336-341
This paper presents a probabilistic approach to evaluating the geotechnical stability problem by incorporating the stochastic spatial variability of soil strength within the numerical limit analyses. The undrained shear strength is treated as a random field which is characterized by a log-normal distribution and a spatial correlation length (i.e., isotropic correlation structure). The current calculations use a Cholesky Decomposition technique to incorporate these random properties in numerical limit analyses. The Random Field Numerical Limit Analyses are applied to evaluate the effect of spatial variability of soil strength on the slope stability and the failure mechanism. Monte Carlo simulations are then used to interpret the failure probability of slope for selected ranges of the coefficient of variation in undrained shear strength and the ratio of correlation length to slope height. The results show how the failure probability of slope is related to the average strength, the coefficient of variation and correlation length scale in the shear strength of slope. Based on the result, the conventional safety factor of slope stability is evaluated to obtain a target probability of failure.