Effects of anisotropy in correlation structure on reliability-based slope stability analysis of a landfill

In this study, a reliability-based analysis of slope stability in a landfill considering the spatial variability of shear strength (c and φ) and unit weight ( γ) of municipal solid waste was performed using the random finite difference method. The effect of the anisotropic correlation (AC) structure of random variables including c, φ, and γ on mean and coefficient of variation (CoV) of safety factor values was investigated using Monte Carlo simulation. Also, probability of failure was determined through probability distribution fitting to a series of safety factor data. The results showed that the AC of the random variables has a significant effect on the stochastic safety factor of landfill slope. In general, the mean of the stochastic safety factor decreases with increasing horizontal and vertical correlation lengths (CLs). Also, the CoV of the safety factor increased with increase in horizontal or vertical CLs or both. Moreover, it was found that failure probability increases with an increase in the horizontal and vertical CLs and assuming isotropic structure for the correlation of random variables leads to the underestimation of failure probability. Overall, the results indicate that assuming an AC structure results in different failure mechanisms in the landfill slope, which are different from the deterministic cases.