Shear strength of rock, rock joints and rock masses - Problems and some solutions

These three important topics, each deserving separate volumes, even when summarized, can only be treated in a single article by linking them. The all-encompassing shear strength of rock masses cannot be described with advanced algebra as in Hoek-Brown, nor as linear Mohr-Coulomb, each of which are a priori estimates rather than the desirable a posteriori based on experience. The highly non-linear shear strength of intact rock, which has finally been defined as strongly deviated from Mohr-Coulomb, and with more curvature than Hoek-Brown, is the component which fails at small strain. Deep in the crust rocks may be ductile or at their critical state. The very different and weaker joints or fractures provide stability problems in civil and mining engineering, and help maintain some permeability in fractured reservoirs. Joints are highly anisotropic features. They exhibit large differences between their high normal stiffness, and their low, scale-dependent shear stiffness. Joints obviously reach their peak shear strength at larger shear strain than intact rock, and their frictional strength 'remains' after cohesion is lost, as in the words of Muller 1966. It is not correct to add the cohesive strength of the intact rock and the shear resistance of the joints, as in c plus sigma-n tan-phi, nor as in the non-linear form of Hoek-Brown. A third shear strength component may kick-in at larger shear strain: the lower frictional strength of clay-filled discontinuities, such as in the neighbourhood of faults. Finally there is the wide-reaching problem of stress transformation, from principal stresses to normal and shear stress components on a plane. Dilation, shearing and the very presence of the plane violates the theoretical assumptions.