ABSTRACT
The cubic law is commonly used to predict the hydraulic behaviour of rock joints. Yet, test results show that this model can constitute a valuable approximation only in specific conditions of geometry and aperture of the joints. This paper outlines the importance of channelling effects with respect to the relative aperture/roughness size of the joints. Thanks to a set of hydraulic tests carried out on artificial joints with regular and well defined geometry, the validity of the cubic law is discussed. An evaluation of the channel network size and orientation that is established in a real rock joint (granite) after a slight shearing is presented, for different degrees of closure.
