The authors have developed a theoretical model for the shear behavior of rock joints under several confining conditions. The model considers the influence of the normal confining conditions, the material strength of the intact rock, the basic friction angle, and joint surface roughness. If such information can be obtained, the model can express typical shear behavior, namely, the shear stress increases, reaches the peak shear strength, decreases (softening), and then gradually reaches the residual state. Comparing the analytical and the experimental results, a good agreement can be found. In the residual state, however, there are some differences. The differences in the residual state cause the influence of the powder produced by the deleted asperities. The asperities are shaved and the powder plays the role of a ball bearing. In this research work, the theoretical model is improved in consideration of the influence of the powder which plays the role of a ball bearing. Then, comparing the analytical and the experimental results, the validity of the improved theoretical model is discussed and confirmed.
In order to clarify the mechanical behavior of the rock joints, direct shear tests must be conducted on the joints. It is difficult, however, to totally express a mechanical model for rock joints since the shear behavior of rock joints is controlled by certain factors, namely, the confining conditions, the material strength of the intact part, the joint surface roughness, the basic friction angle, etc. Estimations of the shear behavior of rock joints have been performed by many researchers, such as Patton (1966), Ladanyi and Archambault (1981), Barton (1973), Barton and Choubey (1977), Bandis et al (1981). They are representative works of the estimations of the shear properties of rock joints. No research, however, has ever been able to present the entire shear behavior, namely, the increase in the shear stress, the decrease (softening), and then the gradual arrival at the residual state. The authors previously performed direct shear tests on rock joints under constant normal confining conditions and developed a theoretical model which could systematically express the shear behavior of rock joints by inputting the normal confining stress, the uniaxial compressive strength of the intact material, the joint surface roughness digital data, and the basic friction angle (Kishida and Tsuno 2001; Kishida et al. 2001). The model was used to estimate the results of direct shear tests under constant normal stiffness conditions. And, in fact, it successfully presented the shear behavior of rock joints (Kishida et al. 2002). However, this previously proposed model had the following problems, namely, 1) in the residual state, the analytical shear stress sometimes seemed to be larger than the experimental shear stress since the influence of the shaved-off powder material was not considered in the shear process and 2) there was a dependency for the peak shear strength of the analytical interval to appear clearly in the first step of the analytical process since no consideration was given to the deformation of each asperity.