ABSTRACT
Although slope failure is inherently three dimensional, slopes have typically been analysed using 2D limit equilibrium approaches, largely due to computational limitations. This is changing as commercial 3D limit equilibrium computer programs become available. A parametric study was used to shed light on how slope heterogeneity influences safety, and whether it is safe to simply extrapolate (or extrude) a 2D cross-section to carry out a 3D analysis. It was found that critical 3D slip surfaces were dependent on toe conditions in the central zone of the surface and crest conditions in the side zones (of the critical slip surface). In some cases, the critical factor of safety of a 3D heterogeneous slope was lower than the 2D factor of safety computed for the weakest cross-section. Consequently, even the weakest 2D cross section may not reflect the safety of a slope, with this being even more so if a 3D analysis is carried out on an extrapolated 2D section. If 3D approaches are to be used, site investigation work will need to be more extensive than currently utilized for establishing 2D sections.
