An approach to the assessment and modelling of rock mass damage around underground excavations

Abstract

The work described in this thesis aims to develop methods for the assessment, description and modelling of damaged and disturbed rock around underground excavations. Rock mass damage is defined as the creation of additional fractures and the disturbance or loosening of existing discontinuities. Two processes contribute to damage around excavations in 'hard rock'; stress redistribution (as a result of creating the excavation), and disturbance by the excavation process. Where mining is conducted using blasting methods the potential for significant additional damage and disturbance is significant. A constitutive model, called the rock fabric model, has been developed to describe the stress-strain behaviour of fractured rock masses. The model accounts for the non-linear load-displacement behaviour of unfilled discontinuities in rock as well as their geometry. The in-situ geometry of rock discontinuities is complex and a number of models exist which attempt to describe this geometry. The assumed fracture geometry has a significant effect on the deformational behaviour of rock masses, this effect is has been included into the rock fabric model by the application of a 'stress concentration factor'. Coupled with the finite element method the model allows rigorous back analyses, as well as forward analyses to be performed. The assessment of damage is based on the back analysis or 'characterisation' of the rock mass structure based on in-situ measurements of rock mass displacements and linear fracture frequencies. An investigation into the ability of back analysis, based on in-situ measurements of stress and displacement, to characterise the rock mass structure has been conducted, and the method applied successfully to a set of data pertaining to an in-situ block test conducted at the Colorado School of Mines (CSM) experimental room, Edgar mine, Idaho Springs. Two characterisation studies have been conducted on data sets obtained from operating mines in Australia. These studies show clearly the potential of the back analysis method and highlight a number of areas in which further refinements are necessary.Finally, two tensor descriptions of rock mass structure are proposed. The tensor description allows the anisotropy of structure, as well as the quantity, to be conveniently described. The extension of these descriptions to damage is a simple matter. The relationships between the proposed tensors and rock mass strength and deformability have been investigated and an approach to estimating the in-situ strength of anisotropic rock masses described.Application of the rock fabric model to the CSM experimental room allows the damaged and disturbed zones to be identified and compared with in-situ measurements made at the site. The comparisons between the computed and actual results show reasonable agreement and as such are seen as providing some validation for the model.