Prediction of Plastic Damage Index for Assessing Rib Pillar Stability in Underground Metal Mine Using Multi-variate Regression and Artificial Neural Network Techniques

摘要

Large diameter blasthole stoping method has been selected to mine a copper orebody from underground below an open pit mine. The damage induced in the rib pillars as a result of large-scale production is one of the major challenging issues. Three-dimensional numerical modelling techniques with the parametric variation of material of ore and waste, stoping sequence and crown pillar thickness is applied to analyze the stability of stopes at various levels. A total of 138 finite element models are analyzed. In this study, the concept of plastic damage index (eta), defined as the ratio of effective plastic strain to effective total strain, is introduced along with strength reduction ratio (kappa). The parameter kappa is determined during post-yielding phase of rocks tested under uniaxial compression. An analytical equation is derived to relate these two parameters and five damage classes are also developed for predicting the rib pillar stability. Results of the numerical analysis are assessed in terms of the proposed plastic damage index; and, predictive models of eta are developed using multi-variate regression and artificial neural network. These models suggest that stoping sequence is the most crucial parameter for rib pillar stability followed by depth of working, rock material type and crown pillar thickness.