INTRODUCTION
The behavior of rock joints (including all discontinuities in the rock mass) under shear loading depends upon not only the friction between joint surfaces and the joint material strength, but also the geometrical profile (roughness) of the joint surface. Two mechanisms usually occur during shear failure of a rock joint: (1) dilation by sliding over the asperities of the joint surface, and (2) shearing through the asperities. The shear strength of a rock joint is the combination of the shear resistances provided by these two mechanisms. The shear resistance in the first mechanism depends on the basic friction resistance between joint surfaces, and the dilation during shearing. The shear strength of the joint material provides the shear resistance in the second mechanism. The dilation of a rock joint depends on the roughness of the joint surface, and is also a function of the joint material strength and the normal stress on the joint surface. The strength of the joint material and the basic friction resistance between planar joint surfaces can be readily obtained through shear strength tests in the laboratory and in the field (e.g., Brown 1981). On the other hand, the relationship between the roughness and the shear strength of rock joints is usually expressed by empirical relations, for example, the joint roughness coefficient (JRC) proposed by Barton (1973), the inclination angle by Dight and Chiu (1981), or the roughness angle by Turk and Dearman (1985). The descriptions of roughness in these empirical relations usually lack clear physical meanings, cannot be determined objectively, or are not pure geometrical characteristics but must be reduced from the shear strength test results. In this paper, the relationship between the shear strength of a matched clean joint and the geometry of the joint surface profile is established. Results of laboratory shear tests on various joint surfaces were compared with the calculated shear strengths.