Testing and numerical simulation of a medium strength rock material under unconfined compression loading

This study aims to numerically and experimentally investigate the response of a medium strength rock material under unconfined compression loading up to failure. The unconfined compressive strength (UCS) is one of the most important parameters in characterising rock material behaviour. Hence the UCS is crucial in understanding the failure mechanism of fractured rocks. An effective approach to determine the UCS and to investigate the behaviours of rock materials under unconfined compression is essential in the majority of research fields of rock mechanics. The experimental configuration for the unconfined compression test, suggested by the protocols of the ASTM standard, has some limitations which affect the accuracy in determination of the real UCS. Among several alternative configurations proposed, the Mogi's configuration seems to be the most appropriate one. Therefore, the ASTM and Mogi's configurations were used to perform the tests on a medium strength rock material, i.e. Pietra Serena sandstone. The results using two configurations were discussed in terms of the differences. The tests were also replicated in LS-DYNA using a finite element method (FEM) coupled smooth particle hydrodynamics (SPH) technique. This technique is employed in this study due to its capabilities to cope with large deformation issues related to the rocks. An advanced material model, called the Karagozian and Case Concrete (KCC) model, is implemented in the numerical simulations. The KCC model consists of three independent fixed failure surfaces and it can consider the damage accumulation based on the current state of stress among these failure surfaces. An equation-of-state (EOS) is used in conjunction with KCC material model for decoupling the volumetric and deviatoric responses. The numerical and experimental results were finally compared with the focus on the stress–strain diagram and the failure patterns. The comparison shows that the numerical results are in good agreement with the experimental results.