Role of Strain Localization and Dilatancy on Deformation and Stability of Soil Slopes

George Washington University, Washington DC

Investigators

Abstract

In addition to traditional limit equilibrium methods, the upper bound and lower bound theorems of plasticity are the main tools in evaluation of stability and collapse load of geo-structural systems. A recent study by the PI shows that an analysis that is based on such theorems may significantly over-estimate the collapse load and the factor of safety of slopes and soil embankments. The main cause of such over-estimation is the assumption of normality and the inherent assumption of high dilation angle (equal to friction angle) in the upper bound and lower bound theorems, which implies an unrealistically large dilatational volume change. In order to overcome this deficiency, a proper constitutive model that can accurately model the volume change of the soil should be used. Moreover, since the shear strength of the soil is directly affected by its volume change, an accurate soil model will automatically simulate the softening of dense granular soils. This post-peak behavior of the soil will also affect the stability of a geotechnical system. In the context of numerical simulations, both non-normality and softening cause significant difficulties that are directly related to initiation of localized shear deformations. Here it is proposed to investigate the effects of strain localization and soil volume change on the collapse of soil embankments. The dilatational soil response will be modeled within the framework of critical state soil mechanics. A selection of plasticity models ranging from non-associative Mohr-Coulomb model to a critical state two surface plasticity model will be employed to examine the effect of soil volume change and induced strain softening on the stability of soil slopes. Considering the fact that strain softening will cause significant localization of deformations in certain parts of the systems that are on the verge of instability, enhanced finite element techniques and constitutive models will be developed to handle the post-bifurcation behavior of the systems. The results of the investigation will be presented in form of appropriate charts that may be used as guideline for incorporating the effect of soil dilatancy on slope stability.