Probabilistic Assessment of Soil–Rock Mixture Slope Failure Considering Two‐Phase Rotated Anisotropy Random Fields

ABSTRACT Soil–rock mixture (SRM) slopes consist of soils and rocks and are widely distributed globally. In addition to heterogeneity and discontinuity within SRM slopes, the inherent spatial variability can be observed in soil and rock properties. However, spatial variability in rock and soil properties and layouts has not been well considered in the stability analysis of SRM slopes. Additionally, SRM slopes commonly show a rotated anisotropic fabric pattern, while such fabric has rarely been accounted for in SRM slope stability analysis. In this study, a two‐phase rotated anisotropy random field simulation method is proposed to model these spatial variations simultaneously. The proposed approach is then integrated with the finite element method (FEM) to study the impacts of soil volume fraction and bedding dip angle (i.e., rotated anisotropy) on the probability of failure (pf) and failure mode of SRM slopes. It is found that considering only spatially varying layouts can underestimate pf by up to 97% compared to considering both spatially variable properties and layouts. The increase in soil volume fraction significantly improves pf and the likelihood of deep failure. The bedding dip angle greatly influences pf, yet deep failure remains dominant across different bedding dip angles. Furthermore, the failure mode of SRM slopes is more sensitive to the changes in soil volume fraction than to bedding dip angle.