A Review and Detailed Investigation of Continuum-Based Numerical Modelling Methods for the Simulation of Brittle Rock Failure Around Underground Excavations

Abstract

Brittle failure occurs around underground excavations when the rock being excavated is competent and the induced stresses are high. Predicting and modelling this type of failure requires special care, as traditional methods of rock strength assessment do not apply directly to this type of ground. Therefore; alternative methods have been proposed to adequately predict the behaviour of brittle rock around excavations. A literature review of current methods was performed where 21 different methods were identified and detailed. These methods can be subdivided into four general categories: empirical methods, continuum numerical modelling methods, discontinuum numerical modelling methods, and hybrid continuum-discrete numerical modelling methods. The list proposed is not exhaustive, and no single method can be suggested to be the best, instead, the user has to determine which method is more suitable for their purpose and experience. Two of the most prominent continuum numerical modelling methods were further investigated, these were the Cohesion Weakening Friction Strengthening (CWFS) method and the Damage Initiation Spalling Limit (DISL) method. Both methods were applied to a common case study, the Mine-by Experiment (MBE) at the Underground Research Laboratory. Then all the input parameters for each method were investigated to determine the effect of each parameter on the simulation results. Alternative stress scenarios were investigated using the rock at the MBE to determine each method's capabilities of modelling brittle failure in different ground stress conditions. From this in-depth investigation of the CWFS and DISL methods, it was found that the CWFS method is a more robust method that can be implemented by rock mechanics practitioners and is ideal for parametric studies and construction monitoring. On the other hand, the DISL method produces hard-to-interpret results which make the method’s interpretation subjective and viable for rock mechanics engineers who are experts in numerical modelling. Once noted that the CWFS is a more robust method was applied to a different case study, the Qirehataer Diversion Tunnel (QDT) in Western China. This application to a new case study increased the understanding of the CWFS method and demonstrated that it can be feasibly used to perform a parametric calibration to match field observations. The last stage of this dissertation was the development of a novel method to model brittle failure around underground excavations. The method developed consisted of using the IMASS constitutive model in FLAC3D to apply the theory of the DISL and CWFS methods. This new method was developed using the MBE as a case study. The results showed that, by modifying the IMASS input parameters, failure around underground excavations can be predicted. To aid future users of the method a preliminary set of guidelines was proposed. Then, using these guidelines, the IMASS method for brittle rock modelling was applied to the QDT. The results matched the field observations from this study validating both the IMASS method and the proposed guidelines. The investigation of the CWFS and DISL methods expanded on the understanding of both methods from the systematic sensitivity study. This research can be used by future users of the methods as guidance for parameter selection. The development of the IMASS methods provides another tool that rock mechanics practitioners can use for brittle rock failure around underground excavations which has a simpler parameter selection process than other continuum based methods.