Under high deviatoric stresses, rock salt shows shear failure, which is accompanied by microdamage. Microscopically, damage results from the accumulation of microcracks, leading to mechanical softening, increased permeability and a volume increase, known as dilatancy, which is a convenient damage indicator since it is directly accessible in the laboratory. On the other hand, it is well-known that rock salt can heal itself: Cracks can close due to viscous creep, and cohesion is restored by physicochemical processes on contact surfaces, such that over time, the damage induced by shear failure is completely healed in suitable stress conditions. This implies that softening, and the associated dilatancy, should decrease to zero. However, the understanding of healing processes is not yet as good as e.g. of creep, both experimentally and theoretically.
The IfG has developed a phenomenological approach which treats healing as a viscous two-component process, comprising a crack closure and a resealing part, such that the former dominates for large values of dilatancy with open cracks, i.e. in the postfailure regime, while the latter becomes important at lower dilatancy, where new cohesion is developed on closed crack surfaces. The approach has been included in the constitutive models of the IfG, i.e. the advanced strain-hardening model of Günther and Salzer, the elasto-visco-plastic model of Minkley and the discontinuous approach of Knauth et al. It is implemented numerically in the codes UDEC, 3DEC, FLAC and FLAC3D.
In this paper we introduce the approach and implementation, and present a few simulations to validate the models. At the present stage, this should be considered a proof of principle, and further modelling and experimental efforts are certainly required.
For the planning of the extraction of salt resources as well as for the safe long-term disposal of hazardous waste in deep salt formations, constitutive models are needed which comprehensively consider the visco-plastic deformation behaviour of rock salt.