Experimental and numerical modeling of the shear behavior of filled rough joints

The shear behavior and mechanisms of clean and filled rough joints under direct shear tests were comparatively studied using experimental and numerical approaches. Concrete was used for casting laboratory specimens with different roughness joint surface profiles, and clay was used as an infill material. In addition, a modified shear box genesis method was employed to reproduce the shear behavior of clean and clay-filled joints under different normal stress by two dimensional Particle Flow Code (PFC2D). The results show that as the normal stress increases, the shear stress increases, and the normal displacement decreases. Due to the presence of clay, the shear strength and the normal displacement of the clean joint are larger than those of the clay-filled joints, and the influence of the joint roughness on the shear stress is significantly reduced. The mechanism of the joint roughness on the shear stress and shear dilatancy mode during the shear process becomes more complicated. Furthermore, according to the bond fracture mode in the numerical shear tests, it can be concluded that the cracks of the clean joints are mainly caused by parallel-bonded shear cracks and parallel-bonded tensile cracks at an asperity, but for filled joints, the cracks are mainly caused by smooth-jointed shear cracks between block and filling, and parallel-bonded tensile cracks between the filling particles.