Slope failure analysis using the random material point method

Abstract

The random material point method (RMPM), which combines random field theory and the material point method (MPM), is proposed. It differs from the random finite-element method (RFEM), by assigning random field (cell) values to material points that are free to move relative to the computational grid rather than to Gauss points in a conventional finite-element mesh. The importance of considering the effects of both large deformations and the spatial variability of soil strength properties in slope stability analyses is highlighted, by comparing RMPM solutions with RFEM and deterministic MPM solutions for an idealised strain-softening clay slope characterised by a spatially varying undrained shear strength. The risks posed by potential slides are quantified by the extent of retrogressive failure – that is, due to the tendency for secondary failures to be triggered by the removal of support from the remaining soil mass caused by the initial failure. The results show that RMPM provides a much wider range of solutions, in general increasing the volume of material in the failure compared with the RFEM solutions, which are usually limited to the initial slide. Moreover, the anisotropic nature of soil heterogeneity is shown to have a significant influence on the nature and extent of failure