TC1 was based on a laboratory experiment, conducted at NGI (Norwegian Geotechnical Institute, Oslo, Norway), for coupled stress-flow processes in a rock fracture (Fig. 5a). The sample of the fracture was created as a large diameter core containing a single fracture under multiple stepped loading–unloading cycles while water was conducted through the fracture. The normal and shear stresses and displacements, as well as flowrate, were measured during the test. Modeling of TC1 was conducted in two steps: TC1:1 and TC1:2. TC1:1 modeling was first carried out using the simplified model set-up of the test system, as shown in Fig.5b, but it was found out that the stiffness of the test frame adopted in the TC1 model was not adequate to produce reliable results. Therefore, the model set-up was developed to represent the complete test frame including all parts, as shown in Fig. 5a, called TC1:2.
Four research teams participated in TC1:1: CNWRA, ITASCA (Sweden branch) and NGI using code UDEC of DEM, and LBL using code ROCMAS of FEM. Three teams participated in TC1:2: AECL using code MOTIF of FEM, LBL using the code ROCMAS of DEM, and NGI using the code UDEC of DEM.
The results demonstrated that TC1:2 much improved the model development, with reasonable agreement between normal stress-normal displacement for loading paths, but less agreement for the unloading path, indicating inadequate ability to represent the stress-path dependence of the constitutive models of the fractures in the codes applied, which, in turn, led to inadequate agreement between measured and calculated hydraulic aperture, mechanical and hydraulic apertures, measured and calculated shear dilation. These findings caused significant concern in the project— such that coupled THM processes of single rock fractures became a continued issue of research and discussions in the subsequent phases of the DECOVALEX projects.