Previous work using distinct element (UDEC) models to study failure modes surrounding an underground opening in rock masses of varying quality [1,2] demonstrated that the failure mode and extent of failure observed in the models were highly correlated with rock mass rating (RMR). The previous models were constructed with three different joint orientation combinations, three joint friction angles, four joint spacing values, and two degrees of rounding of block corners. The results, based on limited data (72 models), suggested a drastic difference in behavior between models with symmetric and asymmetric joint orientation conditions. The current study expands this work to include 12 additional joint orientation combinations, a much more robust data set of 360 models, and the combined results do not show the marked difference in behavior between symmetric and asymmetric joint sets, but do display good correlation with both RMR and Q values. The current study also incorporates a suite of models containing rockbolt supports; the results show that standard empirical design guidelines are generally adequate, although special accommodations should be made in cases with steeply dipping joints, and/or low friction angle.
Previous work using distinct element (UDEC) models to study failure modes surrounding an underground opening in rock masses of varying quality [1,2] demonstrated that the failure mode and extent of failure observed in the models were highly correlated with rock mass rating (RMR). The work was done with the related objectives of 1) validating existing empirical design guidelines using the numerical models, and thereby 2) establishing a connection between numerical modeling and empirical design. The overall goal is to encourage the mining industry to accept numerical modeling as a practical design tool. This is important because numerical modeling may provide needed guidance as mines are being forced to operate in lower quality rock masses in which prior experience is not available and empirical rock mass classificationbased designs [3,4] have not been extensively validated. A particular challenge will be design in The previous models were designed to investigate the influence of friction angle, joint spacing, joint orientation, and degree of rounding of block corners. Although these parametric studies (36 and 72 parameter combinations, respectively) were quite comprehensive, the issue has not yet been thoroughly studied. Most significantly, a distinct difference in behavior between the symmetric and asymmetric models was observed in one earlier study , although the data set was too small to draw any clear conclusions. That observation is the primary focus of the current work, investigated by incorporating a much wider range of asymmetric and symmetric joint orientation combinations.
Furthermore, the earlier work concentrated entirely on stability or collapse of unsupported excavations and did not incorporate any ground support features. The current study includes an investigation of the influence of rockbolt supports, thereby making a more direct connection to excavation design