Finite element modelling of rock cutting using a polycrystalline diamond compact cutter

Abstract

This thesis focuses on three-dimensional finite element modelling of the process of rock cutting using a single polycrystalline diamond compact bit. The modelling is carried out using a modified and enhanced version of the Imperial College Geomechanics Toolkit (ICGT). First, local continuum damage mechanics (CDM) approaches are reviewed and implemented. In these approaches, cracks are not explicitly represented or modelled, but their effect is represented by a degradation in the elastic moduli, with the help of a local damage parameter. It is clearly shown, however, that although such local models perform relatively well in two dimensions, in three dimensions, fracture growth and deformation localisation becomes inhibited, and the energy needed in order propagate a fracture becomes unrealistic, as the damage is spread over too many elements. Consequently, local continuum damage mechanics is discarded, and attention is directed at models in which discrete fractures are propagated. In this approach, stress intensity factors are computed at each crack tip, and crack growth criteria are used to determine which cracks will grow, and in which directions. In order to be used for the specific problem of rock cutting, several modifications were required to the ICGT. Specifically, the geometric representation of the growing fracture is modified to account for the model boundary, namely, the top of the rock block, and the contact between the cutter and the rock. This allows for an accurate meshed representation of the geometry, a good representation of the fractures faces, and prevents fractures from propagating into the cutter. To validate, understand, and inform the numerical rock cutting simulations, physical scratch tests were performed. The novelty of these tests resides in the use of chamfered cylindrical cutters, measurement of both forces and groove widths for tests at depths of cut as high as 4 mm, and the establishment of correlations between groove dimension, cutting efficiency, and cutting aggressivity. In the last chapter, the developed modelling tool is applied to numerically model the rock cutting process. First, elastic contact between the cutter and the rock is studied. The main conclusion is that use of a low local friction coefficient results in a more accurate ratio between the predicted vertical to horizontal forces. Then, initial flaws are inserted into the rock matrix to allow for rock degradation. Only a vent crack develops, which appears to shield the other flaws from propagating, although further loading would lead to the propagation of other fractures.