A mine shaft case study on the accurate prediction of yield and displacements in stressed ground using lab-derived material properties

摘要

Continuum models provide a useful tool for the prediction of stress re-distribution due to excavation and induced yielding, and are used as a key analysis tool in the design of many underground excavations. Recent developments in the study of rock strength and post-yield behaviour have played a key role in improving our understanding of how plastic constitutive models can also be used to practically replicate observed phenomena in brittle rocks. In particular, new models for rock dilatancy can help to improve the applicability of plastic constitutive models as a predictive tool for excavation design. In this study, laboratory data for a heterogeneous, brittle, conglomerate unit from a mine shaft has been analysed. Using parameters from this analysis, brittle strength and dilatancy models have been implemented in a finite-difference code to predict not only stress re-distribution and yield around the shaft, but also to obtain realistic displacement values. Comparison of the modelling results to displacements measured using borehole extensometers show that the constitutive model and lab-derived parameters used were effective in predicting the rockmass behaviour. Parameters were further optimized through back analysis. One interesting finding of this analysis is that the in-situ rockmass dilation decay rate (as a function of plastic strain) appears to be faster than estimated based on laboratory data, which may be indicative of the influence of rockmass-scale natural fractures and other geological structures on the dilation decay process. It also appears possible to model the in-situ dilation decay rate using a single parameter, instead of separate parameters for unconfined and confined conditions. To conclude the study, more numerical results obtained using alternative dilatancy models are presented to illustrate the problem of non-uniqueness in plasticity back analyses. (C) 2015 Elsevier Ltd. All rights reserved.