Laboratory Investigation of Shear Behaviour in Rock Joints Under Varying Boundary Conditions

Abstract

Discontinuities (joints, fractures, faults, bedding planes etc.) in rockmasses are an important mechanical component of a rockmass. In low to moderate in situ stress conditions where gravity-driven failure dominates, these discontinuities act as planes of weakness and govern the material response. Discontinuity behaviour can have a large impact in geotechnical design, and therefore it is essential to determine their geomechanical properties in order to predict rockmass behaviour and prevent any potential failure that would pose a threat to personnel safety or damage property. A common method of determining the mechanical properties (e.g. stiffness, strength, and dilation) of rockmass discontinuities, which are used as inputs for numerical geotechnical design, is through laboratory direct shear testing. Direct shear testing is typically conducted using two different boundary conditions: constant normal stress (CNL*) and constant normal stiffness (CNS). The CNL* boundary condition represents fracture behaviour in slopes and near surface conditions where gravity loading dominates, while the CNS boundary condition is considered to better reflect behaviours of fractures in underground excavations where in situ and induced stresses control rockmass stability and joint opening (dilation) is suppressed during shear. This research addresses the further refinement of direct shear laboratory testing protocols (sample selection, test program creation, and test procedure) and understanding of the behaviour of rock joints under constant normal stiffness boundary conditions. Furthermore, a machine normal stiffness estimation tool for CNS testing is developed in order to improve the reliability and applicability of CNS direct shear testing programs for tunneling applications. Finally, intact geomechanical laboratory testing was completed (unconfined compressive strength and Brazilian indirect tensile strength testing) was completed to determine intact geomechanical properties. A new secant dilation angle is created to characterize the post-yield behaviour of rock joints to be used an input to account for post-yield behaviour in numerical models. Measured geomechanical properties (fractures and intact rock) were then used to complete a preliminary numerical modelling program of laboratory scale direct shear tests.