An efficient slope stability algorithm with physically consistent parametrisation of slip surfaces

TL;DR

This paper introduces an optimized slope stability algorithm that uses a physically consistent parametrisation of slip surfaces, significantly reducing computation time and improving accuracy over traditional methods.

Contribution

The paper presents a novel physically based parametrisation of slip surfaces combined with a hybrid optimisation routine, enhancing efficiency and accuracy in slope stability analysis.

Findings

Computational time reduced by up to 92% compared to traditional methods.

Achieved about 5% improvement in accuracy over grid-based approaches.

Enables efficient analysis of slopes with uncertain properties.

Abstract

This paper presents an optimised algorithm implementing the method of slices for analysing the stability of slopes. The algorithm adopts an improved physically based parameterisation of slip lines according to their geometrical characteristics at the endpoints, which facilitates the identification of all viable failure mechanisms while excluding unrealistic ones. The minimisation routine combines a preliminary discrete calculation of the factor of safety over a coarse grid covering the above parameter space with a subsequent continuous exploration of the most promising region via the simplex optimisation. This reduces computational time up to about 92% compared to conventional approaches that rely on the discrete calculation of the factor of safety over a fine grid covering the entire search space. Significant savings of computational time are observed with respect to recently published heuristic algorithms, which enable a continuous exploration of the entire parametric space. These efficiency gains are particularly advantageous for numerically demanding applications like, for example, the statistical assessment of slopes with uncertain mechanical, hydraulic and geometrical properties. The novel physically based parametrisation of the slip geometry and the adoption of a continuous local search allow exploration of parameter combinations that are necessarily neglected by standard grid-based approaches, leading to an average improvement in accuracy of about 5%.