Investigations into the mechanism of fracture onset and growth in layered rock using physical and numerical modelling

Abstract

One of the major impediments in the field of numerical modelling in rock mechanics is limited knowledge of the mechanisms of fracture and failure of brittle rock. One important tool for improving the understanding of rock behaviour is the use of laboratory experiments under controlled conditions. The Displacement Discontinuity Method, capable of fracture growth simulation (DIGS), has been used to model fracturing in samples under punch loading. A Finite Difference Method, capable of plastic deformations due to its explicit time marching scheme (FLAC), has also been used to model the punch tests. By comparing numerical simulations with results from laboratory experiments of punch tests, it has been possible to define the basic failure mechanism for pillar foundation failure. Two different test set-ups were used namely, steel jacketed axisymmetric punch tests and long strip punch tests in the triaxial cell which is built for these specific tests. The layered structure of the test specimens and in the test procedure had significant effects on the fracture pattern as well as the failure load. When the layer is near to the punch area, then both the layer and the layer conditions had a strong effect on the failure load. When the layer was frictionless, the failure stress dropped by about 20 percent. The same result occurred in both the axisymmetry and strip loading tests. When shear fractures intersect a layer with either low or high friction it terminates. This is not the case for the tensile fractures, which can pass through the layer media. However, it is important to note that the tensile fractures which originate from near the cone area can not pass through the layers. They stop at the interface.