DECOVALEX

TC 2

Problems and research performed Main achievements and outstanding issues

Problems and research performed

TC2 was defined so as to simulate an intermediate scale in-situ heating test of a host rock at the Fanay-Augères test site, France, with coupled thermo-mechanical processes, for the unsaturated in-situ condition and small permeability of the fractures. The objective was to develop validating the abilities of numerical modeling to simulate 3D fractured rocks under combined heating and mechanical loading/unloading conditions. Figure 6 shown the layout of the test site geometry (Fig. 6a), locations for installing extensometers for monitoring the displacement field (Fig. 6b), and the results of evolutions of normal displacements at an extensometer borehole EF2 (Fig. 6c).

Three teams participated in TC2, using different numerical methods and computer codes: INERIS (supported by ANDRA, using code 3DEC of DEM), ENSMP (supported by IPSN, using code VIPLEF (with different elements (4-noded and 10-noded) of FEM), and CLAY (supported by SKB, using code ABAQUS of FEM) and all models were set-up in 3D, concerning coupled TM processes.

Figure 6: a) Test layout at Fanay-Augères test site, France; b) arrangements of borehole extensometer locations; c) part of the measured normal displacements from extensometers EF2 vs. time.

Main achievements and outstanding issues

It was found that both modeling approaches of the equivalent continuum (FEM models) and the discrete system (DEM model) were able to simulate the coupled TM processes in a fractured rock mass. However, the main difficulty was being able to adequately represent the 3D problem geometry (the mesh generation being a main challenge) and properly developing the constitutive models of the fractures. There were divergent results, such as extension of fractures, which diverged due to the different assumptions and model parameters applied, despite the fact that the calculated temperature field agreed well with measured data, and the predicted floor heaving by the models had a qualitative agreement with the measured data. The results indicated that more comprehensive and reliable constitutive models of rocks incorporating fully coupled THM processes are needed for future research and more improvement of computational capacity for 3D problems at site scales.