Coupled mechanical shear and hydraulic flow behavior of natural rock joints

Abstract

Abstract A comprehensive summary of the coupled mechanical hydraulic (MH) behavior of natural rock joints as investigated in two different laboratory experiments (test cases) under the international cooperative project DECOVALEX is presented. The first test case experiment (TC1) was conducted by the Norwegian Geotechnical Institute (NGI) using their coupled shear-flow test apparatus. With this apparatus, joints can be closed and sheared under stress controlled conditions while fluid is injected into the fracture. Two different tests were conducted on joints of varying roughness denoted as TC1:1 and TC1:2. Numerical simulation of both TC1:1 and TC1:2 was also conducted using several computer codes with different constitutive relations for the joint. The second test case experiment (TC5) was conducted at the Center for Nuclear Waste Regulatory Analyses (CNWRA) using a direct-shear apparatus modified to include fluid flow within the joint. In this apparatus, the fluid is injected along one edge (inlet) of the rectangular joint and simultaneously collected form the opposite edge (outlet), while the remaining two joint edges are sealed to prevent leakage. Linear flow measurements were conducted under normal load as well as under combined normal and shear loading on a rock comprised of two blocks of rock bounding a naturally fractured, welded tuff joint. Numerical simulation of both test cases required assumptions to be made with respect to the material properties and specifically the boundary conditions. For TC1:1 and TC1:2, numerical modeling showed some deviation between the Barton-Bandis joint model and the experimental results. The deviation might be caused by the joint itself, or the way the joint behaves in the apparatus. With regard to the shear deformation behavior, the modeling of both TC1:1 and TC1:2 shows that the dilation and changes in hydraulic aperture are not well modeled. However, some of the numerical results did tend to agree with the upper bound of the measured hydraulic aperture as well as the residual hydraulic aperture. The experimental results for TC5 show that the hydraulic conductivity of the joint can change by up to a factor of 3 under shear deformation. However, it was determined that additional tests and improvements to the experimental apparatus are necessary in order to make more quantitative assessments of changes in such properties during shear and production of gouge within the joint.