Development of an improved semi-analytical model for describing the shear strength of rock joint discontinuities

The shear strength of rock joints is known to be scale dependent, which is often referred to as the “scale effect”. A new method has recently been proposed to solve the long-known issue of scale effect on the shear strength of large discontinuities. This method, stochastic in nature, uses the roughness information available from visible traces to create N synthetic rock surfaces from a rigorous random field model. The shear strength of these N surfaces is then estimated using a semi-analytical model and a distribution of shear strength, with a mean and a standard deviation is produced. This approach was validated at small scale but it was found that the semi-analytical model under-estimates the shear strength at low values of normal stress, which is explained by a number of simplifying assumptions made in regard to asperity shearing. This paper explores strategies to improve the model without compromising its computational efficiency and the resulting improvement in shear strength prediction.