An Efficient Slope Stability Algorithm with Physically Consistent Parametrisation of Slip Surfaces

Abstract

This paper presents an optimised algorithm implementing the method of slices for slope stability analysis. The algorithm features a novel physically based parameterisation of slip surfaces, defined by their geometric characteristics at the endpoints, capturing all and only valid failure mechanisms. The algorithm also employs a hybrid discrete–continuous search strategy to identify the critical slip surface, beginning with a discrete computation of the factor of safety over a coarse grid spanning the entire parameter space, followed by a focused continuous exploration of the most promising region via a simplex optimisation. This methodology reduces computational time up to 92% compared to conventional fully discrete algorithms that compute the factor of safety over a fine grid spanning the entire search domain. The novel physically based parametrisation and the hybridisation of the search algorithm allow consideration of slip geometries that are necessarily neglected by conventional fully discrete approaches, leading to an accuracy gain of about 5%. Furthermore, compared to recent fully continuous algorithms, which perform an unbroken exploration of the entire search domain, the computational effort is reduced by up to 33 times while preserving accuracy. These efficiency gains are particularly advantageous for numerically demanding applications like the statistical assessment of slopes with uncertain hydro-mechanical and geometrical properties.