Spline Search for Slip Surfaces in 3D Slopes

Abstract

A novel method involving the transformation of spline surfaces is introduced to search for the critical slip surface in a three-dimensional (3D) slope, which corresponds to the minimum limit equilibrium method factor of safety for overall slope stability. A slipping surface in a slope can be represented as the intersection of any continuous geometrical entity with the slope topography, over which the mass of sliding soil is discretized to solve for the factor of safety satisfying given equilibrium conditions. Traditionally, many researchers have searched for a critical ellipsoidal or spherical surface, or surfaces formed using other simple shapes. However, the critical slip surface in complicated cases, for example, a landslide, is seldom purely ellipsoidal or spherical, which leads to overestimations of the true factor of safety in a slope. To provide greater flexibility for transforming the shape of the slip surfaces during a global search, the geometry representing the slip surface is assumed to be in the form of a nonuniform rational basis spline (NURBS) surface in this paper. The proposed method involves varying the parameters of a parametric exponential function, which spawns control points within its domain to create NURBS surfaces. The parameters in the exponential function are varied to transform the slipping surface using a metaheuristic search algorithm, such as particle swarm optimization. A major advantage of the proposed method is that the final spline surface in the search is formulated such that it can then be locally optimized using surface altering optimization methods by adjusting the locations of its control points.