Author(s)

M. A. Hicks, J. Chen & W. A. Spencer

Abstract

Influence of spatial variability on 3D slope failures M. A. Hicks, J. Chen & W. A. Spencer School of Mechanical, Aerospace and Civil Engineering, The University of Manchester, UK Abstract This paper considers the influence of spatial variability of undrained shear strength (cu) on the stability of long slopes cut in clay. Random fields of cu are mapped onto finite element meshes used in Monte Carlo analyses, and slope performance is quantified in two ways: (a) reliability is computed as a function of the global factor of safety, F (based on the mean property value); and (b) volumes of material associated with potential slides are discussed with respect to the probability of failure. By plotting contours of displacement and shear strain invariant at failure, slide volumes are shown to be influenced by the depthdependency of the statistics of cu in 2D analysis and, additionally, by the type of failure mode in 3D analysis. Spatial variability can occasionally result in slope failure at relatively high values of F, although the associated risk may then be relatively low due to the greater likelihood of more localised failure. Keywords: finite elements, reliability, risk, slope stability, spatial variability, stochastic analysis. 1 Introduction Spatial variability of material properties affects soil behaviour and geo-structural performance [1]. It also causes uncertainty about actual ground conditions, and leads to the need for probabilistic analysis and measures of response [2]. These include reliability, which is the probability of failure not occurring, and risk, which is the probability of failure × consequence of failure. This paper considers reliability and risk for the problem of 3D slope stability. In particular, it uses finite elements and stochastic analysis to investigate the influence, on slope performance, of the spatial variability of undrained shear strength (cu).

Keywords

finite elements, reliability, risk, slope stability, spatial variability, stochastic analysis.