Constitutive Model for Thermal Compaction of Clayey Geomaterials and Application to COx Claystone

International audienceClay formations present very favourable conditions for the long-term containment of radioactive waste due to their low hydraulic conductivity, high radionuclide retention capacity and limited fracturing in their natural state. But it is also due to the very low permeabilities of clays and claystone that drained conditions occur only at very slow temperature increases. The response is then partially drained or undrained with the consequence of thermal pressurisation induced by the thermal gradient. Furthermore, thermo-hydro-mechanical characterisations under drained and undrained conditions carried out on samples of the Callovo-Oxfordian (COx) clay formation and in situ experiments performed by Andra at the Meuse/Haute-Marne underground research laboratory (LS M/HM) have also highlighted the strongly coupled and complex thermo-hydro-mechanical processes of this material. In particular, the volume changes of the COx claystone, subjected to a temperature rise under a constant isotropic stress close to the in situ conditions of the LS M/HM laboratory in saturated and drained conditions, were highlighted. A behaviour comparable to that of over-consolidated clays depending on temperature is then experimentally observed, i.e. thermoelastic expansion followed by thermoplastic contraction: the transition temperature between these two volumetric deformation mechanisms would correspond to the maximum temperature value undergone by the COx claystone during its geological history. Based on the extensive literature on the thermomechanical behaviour of clays and clayey soils, recent thermo-hydro-mechanical tests conducted on COx samples, as well as the database on the instantaneous behaviour of COx claystone, a constitutive model for the thermomechanical behaviour of COx claystone is proposed, then implemented in a commercial computation code. The transverse isotropic elastic and then elastoplastic instantaneous behaviour (beyond the elastic limit) up to the peak strength of the COx claystone, thermoplastic and hydrostatic compactions are taken into account. The model is first validated on triaxial and hydrostatic paths. The simulation of one of the drained hydrostatic stress heating tests performed by Braun et al. verified that the proposed model was able to reproduce the thermal compaction phenomenon highlighted by the authors and to show its operational character