We use a model analysis for explaining the fracture movement caused by the thermal expansion of the rock in two configurations. The first is a rock heating experiment, instrumented by both stress and displacement sensors, where the model can partly explain a combination of the sensor expansion together with the compressive stress by shear fracture movements. The second is a tunnel with the seasonal varying temperature, crossed by sub vertical fractures not perpendicular to the axis – different volume of each side of the fracture results in different expansion to the tunnel space. The model is based on the specific use of 2D or 3D finite elements for the rock continuum. Case studies include the computing of the threshold of thermal-induced displacement (or stress) changes in the massif, with a possible comparison with the measurement.
Effect of fractures is one of the difficulties in numerical simulation of phenomena in the rock mechanics. Standard methods evaluate the rock properties in the scale above the fracture scale, allowing use various equivalent continuum methods. Fractures can be included as discrete objects between either rigid or deformable blocks – by means of e.g. discrete/distinct element methods – or be included to an existing continuum discretization in the extended finite element method (XFEM). The former is used by several established rock mechanics simulation codes, with selection of semi empirical stress-strain or strength constitutive relations.
While these are specific features of rock mechanics, they are not easily available in general-purpose multiphysics codes. In this work, we demonstrate a simple approach to include fractures with use of a linear-elastic model – therefore accessible in COMSOL software used for the previous continuum models. The purpose of the model studies is to check the possible effects of fracture – if they can contribute to an explanation of the measured phenomena. It is convenient to start with a simpler extension of the current model and after the evaluation, it can tell if it is worth to consider a use of a specific method or code with realistic constitutive relations.
The work is motivated by two projects in the sites in the Czech Republic. In the first, the coupled thermo-mechanical model without fractures has been developed formerly and verified successfully. The comprehensive case study is nowadays prepared for a separate publication (Rálek et al. 2015).
