ABSTRACT
The paper investigates the stability of slopes with spatially variable and cross-correlated shear strength parameters in 3D. The influence of various cross-correlation coefficients between these parameters on the probability of 3D slope failure has been considered for different levels of anisotropy of the heterogeneity in the shear strength. Specifically, 3D random fields of cohesion and friction angle were generated using the Local Average Subdivision method, and these were correlated with eachother by various degrees. The fields were then linked to finite element analyses within a Monte Carlo framework. The results indicate that a positive cross-correlation between the parameters reduces the slope reliability, whereas a negative cross-correlation between the parameters increases the reliability.
