Proposal for Standardization of Pull Tests on Rock Joints

This note intends to approach the mechanical characterization of rock joints, emphasizing the application of pull tests, which are an economic and simple type of shear experiments. Explanation is given on the drawbacks resulting from the classical testing procedures which do not control the sample's middle plane position, leading to normal scale effects (represented by negative exponential regressions), where small sample results are misleading in terms of safety. In effect, pull tests allow a very easy measurement of several joint properties, such as the so-called joint roughness coefficient, the slope of joint roughness, the roughness anisotropy, and joint strengths under low normal stresses. Some rules for their standardization are suggested by pointing out the systematic leveling of the middle joint sample plane, thus rehabilitating the interest on small samples, because of the inverse scale effects on strength that are obtained, represented by positive exponential regressions.