Random Material Point Method for Large-Deformation Analysis of 3D Pile-Reinforced Slopes with Spatially Variable Soils

Antislide piles are widely utilized for disaster prevention and slope stabilization. However, previous studies have often neglected the influence of pile design on both slope destabilization and postfailure mitigation, focusing predominantly on deterministic analyses aimed at increasing safety factors while inadequately addressing large deformations. To address this issue, this study adopts an Euler–Lagrange algorithm, integrating the spatial variability of three-dimensional (3D) soil parameters with the large-deformation random material point method (RMPM). The mitigation of landslide disasters is investigated for slopes reinforced with antislide piles at various spacings in 3D spatially heterogeneous soils, and the postfailure behavior of the slopes is quantitatively characterized. Results from 3D RMPM analyses, considering both self-instability and seismic-induced landslides, indicate that under uncertainty conditions, the runout distance of slopes reinforced with piles may exceed that of unreinforced slopes when pile spacing is inappropriate. In homogeneous soils, larger pile spacing is required, leading to more conservative mitigation designs. Under seismic loading, slopes with 3D spatial variability experience more severe damage at the identical pile spacing compared to homogeneous slopes, thereby reducing the effectiveness of antislide piles. These findings highlight the necessity of maintaining larger safety margins than those suggested by deterministic analyses to ensure the protection of surrounding infrastructure.