ABSTRACT:
A new methodology for the prediction of jointed rock mass behavior is described that involves the construction and testing of ¡°Synthetic Rock Mass¡± (SRM) samples. SRM samples are three-dimensional and simulate rock as an assembly of bonded spheres (intact rock) with an embedded discrete network of discshaped flaws (joints). The technique brings together two well-established methods, Bonded Particle Modelling (BPM) in PFC3D (Potyondy & Cundall 2004) and Discrete Fracture Network (DFN) simulation. The methodology also employs a new sliding joint model that allows for large rock volumes containing thousands of joints to be simulated in a rapid fashion. The inputs to the SRM are from standard rock mass characterization methods while the outputs are in the form of rock mass properties that may be employed in empirical, numerical or analytical methods of analysis. Of particular interest is the ability to obtain predictions of rock mass brittleness. This is considered a significant step forward as there is no established method for quantifying this property. Validation of the SRM technique has been achieved through comparison of predicted and inferred induced fracturing at a case study block cave mine (Reyes-Montes et al. 2007).
1 INTRODUCTION
Although long recognized as a central concern of rock mechanics, progress in estimating the strength and brittleness of rock masses has been slow, with reliance placed on empirical classification rules and systems derived from practical observations. While, to date, the Hoek-Brown approach has offered the most viable means to estimate rock mass strength, it does not specifically address rock mass brittleness. The unique nature of mining methods such as block, panel and sublevel caving (in which we desire to push the rock mass from peak strength through to full disintegration) renders predictions of rock mass behaviour more sensitive to this property than traditional analysis (e.g. tunnel design).
