The main goal of the thesis is to contribute to a better understanding of the phenomena involved in the process of brittle failure around underground openings, and to develop an appropriate approach to model these phenomena. In deep level civil and mining excavations where stresses easily exceed the strength of the rockmass, the proper consideration of the behaviour of rock during and after failure in constitutive modelling of rock behaviour has been the subject of many theoretical and experimental researches for a long time.; Current empirical and conventional experimental methods for obtaining the deformational behaviours of hard rocks under loading do not lead to results, which can be matched with in situ observations. This study demonstrates that this problem is not a matter of the popular notion of size effects but rather that it is related to the different circumstances under which the cohesive and frictional strength components are mobilized in laboratory compression tests and around underground openings. This difference cannot be captured by laboratory compression tests on cylindrical samples of rock.; Through a rigorous and critical examination and identification of potential failure mechanisms, the thesis introduces a constitutive model in which the process of brittle failure of rock in low confinement environments can be properly simulated in a continuum-modelling framework. The key factors to be considered in the understanding of brittle failure are the micro-mechanical phenomena involved in the brittle failure process. These phenomena can lead to the nonsimultaneous mobilization of the cohesive and frictional strength components during failure, and the thesis treats this in a so-called “cohesion weakening-frictional strengthening” (CWFS) model for brittle failure of rock.; The following summarizes the key conclusions of this study. (1) Constitutive models assuming simultaneous mobilization of the cohesive and frictional strength cannot properly simulate brittle rockmass failure near excavations in areas of low confinement. (2) A bilinear (concave upward) failure envelope is required for brittle failure of rock, such that at low confinement, the cohesive strength initially predominates the mobilized strength and it eventually is replaced by the frictional strength when the cohesion is consumed. The CWFS model can produce such bilinear failure envelopes and captures both the initiation and the arrest of failure around openings in hard brittle rocks. (3) The propagation of the failed zone (depth and extent) is a function of the brittleness index introduced in this study which explicitly considers the relative delay in the frictional strengthening relative to the rate of cohesion loss as functions of plastic strain. (4) Brittleness of rock is the most dominant factor, more than stress, in controlling the breakout shape. This explains the failure of many methods adopted by other researchers in establishing stress related breakout prediction models. (5) Different rocks possess different brittleness in different loading systems (laboratory, in situ) which should be considered in modelling the brittle failure of rock around openings. (6) This study demonstrated that the support pressure tends to reduce the brittleness of rock, which in turn reduces the depth of failure around supported excavations.
展开▼
