Failure in Transversely Isotropic Rock Discs Subjected to Diametrical Loads and a Range of Load Contact Configurations

Abstract

Many rocks exhibit a type of anisotropy, meaning their mechanical properties are directionally dependent. Transversely isotropic rocks are very common, arising from the presence of predominant layers. Understanding the effects of transverse isotropy on strength and deformation is critical. This study focuses on failure of transversely isotropic rock in Brazilian tests. The load may be applied through flat platens, curved jaws or thin strips at diametrically opposite positions on the discs. The different load contact configurations give rise to subtle but important differences to stress distributions throughout specimens during testing and at the onset of failure. In this thesis, diametrical loading on transversely isotropic rock discs is studied, analytically and using experimental methods, to identify failure conditions under different orientations of the anisotropic planes, load configurations. Both static (i.e. with loads applied slowly) and dynamic conditions are considered. An analytical technique is adopted to assess the effects of the load contact configurations on the stress distributions in transversely isotropic discs. The stress distributions are coupled with a transversely isotropic failure criterion to infer the locations where crack initiations occur. In the experimental part Hawkesbury sandstone is studied. A set of static and dynamic Brazilian tests are performed on samples with three different anisotropic plane (bedding) orientations with respect to the x-axis. For the static tests, three types of load contact configurations are used. For the dynamic tests a split Hopkinson pressure bar system is used. It is concluded that, when the contact load lengths are large, failure and cracking initiates near the discs centers meaning that the Brazilian test is suited to determining an indirect tensile strength. Also, the anisotropy angles have an influence on the magnitude of the strengths determined. It is also concluded that the anisotropy angle has a considerable influence on the results of static Brazilian test when the loads are applied across small lengths. However, for bigger contact lengths, the effects become negligible. Also anisotropy angle does not significantly affect the dynamic results. This is because fast fragmentation does not allow the rock anisotropy show significant reaction, especially for low degrees of anisotropy.