A MODEL FOR GRAVITY FLOW OF FRAGMENTED ROCK IN BLOCK CAVING MINES

Abstract

The work in this thesis is directed towards the development of a model for gravity flow prediction that can assist in the design of a drawpoint layout and draw schedule that maximizes recovery of ore from block and panel caving mines. Experience suggests that maximum recovery and minimum dilution are achieved most easily if one is able to achieve relatively uniform downward movement of all fragmented rock over the area under draw. A critical review of existing experimental and simulation data demonstrated that this is encouraged through overlap of the movement zones (IMZs) associated with each individual drawpoint and that the far-field shape, eccentricity and internal velocity profile of an IMZ are functions of the mean fragment size and friction angle of the caved rock. A series of DEM and continuum modelling studies were used to isolate shear banding as the mechanism controlling these behaviours. Incremental equations were developed to embody these mechanisms at the scale of a disk-shaped volume, representing a horizontal slice through an IMZ, and embedded within a cave-scale simulator (REBOP). It is shown that the resulting model provides predictions of far-field material movements (including free surface rilling) and extraction zone shapes that are consistent with observations from scaled physical models. It is also shown that the model is capable of providing reasonable predictions of drawpoint grade-tonnage trends when applied at the mine scale. This thesis also aimed to isolate the mechanisms controlling flow under interactive draw conditions, in which stress-driven yielding and flow of stagnant material between IMZs can result in uniform drawdown prior to overlap. A series of DEM and continuum modelling studies demonstrate that the stress distribution around a single IMZ is characterized by a plastic zone whose thickness is a function of IMZ radius, caved rock friction angle and overburden height. It was also demonstrated that when a group of IMZs are spaced such that their plastic zones overlap, vertical stresses above the yield limit of the rock in the overlapping region are capable of arching into surrounding stagnant material, thereby inhibiting interactive draw mechanisms. These findings suggest that the relation between stress and strength in the stagnant zone exerts more fundamental control over interactive draw than the ratio between IMZ diameter and drawpoint spacing, as has been suggested previously by other researchers. Logic was developed for the prediction of stresses and yielding inside the volume under draw that accounts for arching effects and can be used to infer where interactive draw conditions may exist. Comparison with the results of numerical and physical experiments suggests that the predictions are reasonable but would benefit from further testing and validation. A series of drawpoint-scale simulations were conducted using DEM to explore the mechanisms controlling fines migration in a caving environment. It was demonstrated that fines migration can occur inside the IMZ and that it is induced by the shearing that accompanies movement toward the drawpoint. It was also found that lateral movement of finer particles is restricted during percolation and that this can lead to stagnation and accumulation of fines near the base of an IMZ. A series of shear box tests were simulated via DEM to quantify percolation rates for caved rock materials. When embedded within REBOP, these were shown to reproduce the percolation rates observed in scaled physical models of draw. Secondary fragmentation within caving is commonly attributed to a combination of splitting (bulk fracture or crushing) and rounding (abrasion). A critical review of available data suggested that splitting occurs via compression in stagnant zones and via shearing inside the IMZ at high stress:strength ratios. Rounding is most likely to occur via shearing inside the IMZs at low stress:strength ratios. The factors controlling these mechanisms were isolated, including shear strain, stress, fragment shape, size distribution, intact strength and initial porosity. An existing empirical model for shearing-induced fragmentation of manufactured materials was adapted for embedment within REBOP and tested via comparison with changes in drawpoint fragmentation observed at an operating mine. The comparison suggests that the logic may overpredict the degree of secondary fragmentation due to the assumption of an initially uniform fragment size in the empirical model. New DEM techniques were developed for the study of secondary fragmentation and were shown to offer promise as a means to extend and improve existing empirical models for both shearing and compression-induced fragmentation. Ore grades and fragmentation measured in drawpoints of the 7700 Level panel cave at Henderson Mine were used to test the newly developed model on a full-scale problem. Good matches were obtained between predicted and measured drawpoint grades where accurate measures of initial fragment size were available. In other areas, poor matches between measured and expected grades could be accounted for in the new model via differences in fragmentation and by the tendency for materials to rill long distances horizontally beneath uncaved ground. Fines migration and secondary fragmentation mechanisms were predicted to have a measurable, but secondary influence on recovery and dilution.