Comparison of Joint Shearing Resistance Obtained with the Barton and Choubey Criterion and with Direct Shear Tests

In most rock engineering projects, shear strength of discontinuities plays a critical role. Empirical shear models are often used for assessing shear strength of rock joints. They may show some limitations if applied normal load does not match their effective normal load range. Barton's shear model (JRC-JCS model) has been initially proposed for discontinuties under high normal loads. This model is now used for many rock mechanics applications under low normal loads, such as dam foundations without paying attention to its applicability. This paper aims at assessing the applicability of this model for predicting shear strength parameters of rock joints under low normal loads. For this purpose, over 20 rock joint samples were collected from a dam foundation and their shear strength was measured from laboratory direct shear tests. The shear strength of these samples were also calculated using Barton's shear model. The input parameters for Barton's shear model were obtained from scanning the morphology of joint surface for JRC, direct shear testing of saw cut joints for φb and performing Schmidt hammer for JCS respectively. Results show that discrepancies may reach 50% for rough joints (JRC=11) and are minimized to 17% for low roughness joints (JRC=4). An analysis of the model input parameters and equivalents thereof, derived from back analysis of direct shear tests, results in interesting findings which may explain the discrepancies. Actually, the basic friction angle of a rock joint is not constant as stated by the model and varies according to potential second order asperities and magnitude of the low normal load. Roughness acquired by digitization of the joint surface is different from actual roughness, which contributes to dilation angles during the shearing process.