Influence of data analysis when exploiting DFN model representation in the application of rock mass classification systems

Abstract

Discrete fracture network (DFN) models have been proved to be effective tools for the characterisation of rock masses by using statistical distributions to generate realistic three-dimensional (3D) representations of a natural fracture network. The quality of DFN modelling relies on the quality of the field data and their interpretation. In this context, advancements in remote data acquisition have now made it possible to acquire high-quality data potentially not accessible by conventional scanline and window mapping. This paper presents a comparison between aggregate and disaggregate approaches to define fracture sets, and their role with respect to the definition of key input parameters required to generate DFN models. The focal point of the discussion is the characterisation of in situ block size distribution (IBSD) using DFN methods. An application of IBSD is the assessment of rock mass quality through rock mass classification systems such as geological strength index (GSI). As DFN models are becoming an almost integral part of many geotechnical and mining engineering problems, the authors present a method whereby realistic representation of 3D fracture networks and block size analysis are used to estimate GSI ratings, with emphasis on the limitations that exist in rock engineering design when assigning a unique GSI value to spatially variable rock masses.