A MECHANISTICALLY-BASED MODEL FOR ROUGH ROCK JOINTS

Over the past twenty years, many models have been formulated to predict the behaviour of rough rock joints. While these models have substantially improved our understanding of rock joint behaviour, their limitations must also be recognized. The models currently available are questionable in their application either because they are too simplistic, rely too heavily on empiricism or require complex input parameters that are typically well beyond current capabilities of normal site investigation practice or laboratory procedures. In many cases, the predictive capabilities of the models are limited to peak shear stress determination, giving little or no indication of the displacement to peak. Also, most models are incapable of reproducing even the simplest mechanisms of movement leading to shear displacement. This paper presents a new model that addresses these limitations. It is demonstrated that this new joint model is capable of capturing the basic mechanisms of movement and makes reasonably accurate predictions of shear displacement behaviour. The model relies only on a knowledge of the joint roughness and a few easily determined, basic rock properties.