Modelling brittle damage to tunnel excavations in a block cave operation using linear elastic damage criteria

Large scale underground bulk mining methods such as Block Caving allow for mass extraction of large, low grade ore bodies in a cost effective manner. They involve de-stabilising large volumes of rock to induce significant stress changes in the surrounding rock mass. When caving operations are established in deep, high stress environments in hard competent rock, the potential for brittle damage to tunnel excavations exists. Whilst high stress to strength ratio, large scale caving mines set in hard competent rock are few, the industry is trending towards such operations. Having tools which can assist site based engineers in observing, measuring, and forward predicting the location and depth of brittle damage to tunnel excavations can lead to less disruption to mining activities and risk mitigation in these mines.This thesis presents a methodology to assist site based engineers in observing, measuring, and forward predicting the location and depth of brittle damage to tunnel excavations using linear elastic modelling in caving operations. The thesis demonstrates that existing brittle damage empirical relations derived by Diedrichs (1999), Martin et al. (1999), and Kaiser et al. (2000), that assume the rock mass strength is governed by cohesion, are applicable for modelling the location and depth of damage associated with the on-set of damage to tunnel excavations in brittle rock that exhibit randomly orientated, low persistence, jointing. The thesis also presents a methodology for forming damage criteria that can predict the location and depth of damage associated with more significant brittle damage that results in spalling and notch formation to tunnel excavations; assuming that some of the cohesive strength of the rock mass has been lost and some frictional strength has been mobilised as the damage increases.A key finding of the thesis is that a lower bound 95% confidence fit was the best approach to use when forming the damage criteria from the stress state data. Another key finding was that reasonable correspondence was achieved between the inflexion points in the extensometer displacement vs time curves and the transition of the rock mass from its elastic to in-elastic response