Modelling Progressive Failure of Good Quality to Massive Rock Masses

  • Author / Creator
    Rafiei Renani, Hossein
  • The behavior of excavations in rock masses with different degrees of fracturing can be fundamentally different, yet commonly used constitutive models and failure criteria assume similar failure mechanisms and strength relationships. This work was dedicated to evaluating current methods and developing improved approaches to modelling the behavior of good quality to massive rock masses with Geological Strength Index (GSI) greater than 60. Starting with the simplest case of intact rock, the well-established Hoek-Brown criterion was compared with the recently proposed Christensen criterion. Results of triaxial compression, triaxial extension, and polyaxial compression tests on different rock types were used for quantitative comparison. It was shown that while the Christensen criterion has several attractive characteristics and has been successfully applied to other materials, on average it leads to 65% higher errors compared to Hoek-Brown criterion when applied to intact rock due to mathematical limitations on the slope of failure envelope. The behavior of a good quality rock mass with GSI≈60 was examined using the measured displacements around a deep shaft. While a single three-dimensional numerical model could not accurately match the readings of all extensometers, the results of using strain softening material with empirically estimated parameters gave a reasonable match to measurements of two extensometers. Higher readings of the third extensometers could be reproduced with a weaker model which provided a possible range of in situ parameters. A two-dimensional model with longitudinal displacement profile was also used for back analysis. While this model gave apparently better overall match to measurements of all extensometers, it was shown to be the result of compromising three-dimensional effects around an advancing face. It was also shown that longitudinal displacements profile may not adequately capture three-dimensional effects in an anisotropic in situ stress fields. The progressive failure of brittle rocks was studied using the results of damage-controlled tests on samples of granite and limestone under triaxial compression. It was confirmed that increasing damage causes degradation of cohesion and mobilization of friction. Furthermore, while cohesion degradation appears independent of confining stress, mobilized friction angle is reduced at higher confinements. Theoretical Cohesion-Weakening Friction-Strengthening (CWFS) models were proposed capturing the progressive failure in the laboratory and in situ. The proposed model closely matched the laboratory stress-strain curves. The model also predicted the in situ behavior of four tunnels in a massive rock mass with GSI≈100. It was shown that the proposed model eliminated the problematic characteristics of the current CWFS model. General guidelines were given for estimating the parameters of the proposed CWFS model for practical applications. The data base of observed pillar behavior in hard rock mines was used to validate the developed modelling approach. The documented behavior of 85 pillars in two mines was used to examine the accuracy of the proposed model for rock masses with GSI≈80. It was shown that the model could separate stable pillars from unstable/failed pillars. While pillars with width-to-height ratio, W/H˂2 showed peak strength at relatively low strains and subsequent strain softening behavior, wider pillars could continue to sustain increased stresses at higher strains. The model also captured the observed mechanism of pillar failure in situ starting from minor spalling to slabbing, formation of hour glass shaped pillars and final shear failure. It was shown that current perfectly-plastic and strain softening models can significantly overestimate the strength of deep pillars while empirical strength formulas systematically underestimate the strength of wide pillars.

  • Subjects / Keywords
  • Graduation date
    Spring 2017
  • Type of Item
  • Degree
    Doctor of Philosophy
  • DOI
  • License
    This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.