Yannis Dafalias

Abstract:

Critical State Theory (CST) in soil mechanics considers conditions on stress and void ratios for Critical State (CS) failure without reference to orientation of particles, contact normal or void vectors at grain level, collectively called granular fabric and measured by an appropriately defined fabric tensor F. Yet it is an experimentally verified fact that at CS there is a strong fabric that induces a significant anisotropic pre-failure response of the medium. A recently developed Anisotropic Critical State Theory (ACST) identifies the norm and direction of the fabric tensor all the way to CS and introduces an additional CS condition related to the relative orientation of the fabric tensor and loading direction by means of a Fabric Anisotropy Variable (FAV) that affects the location of the dilatancy state line in the void ratio-pressure space. ACST can address better the granular response where the classical CST fails to do so. The Discrete Element Method is used to achieve the identification of the fabric tensor and its evolution, since actual experimental measurements are tedious. Yet the leap from discrete to continuum description for incorporation into the continuum ACST is not trivial. Appropriate rate equations of evolution of F must satisfy objectivity, and the continuum version of F must be thermodynamically consistent in regards to per volume dissipation that requires special normalization of the DEM measurements. Practical questions on the initial value of the fabric tensor in real soil deposits remain a challenge that is discussed. Within the new ACST framework, soil plasticity constitutive models can be constructed that obey the classical CS conditions and in addition reflect realistically the strong anisotropic response before failure. It is expected to change the way CST is presented and taught. Keywords: Critical State; Fabric; Discrete Element Method; Anisotropy; Soil Mechanics; Granular Mechanics; Soil Plasticity