Stochastic Analysis of Core Discing for Estimation of In Situ Stress

Extraction of diamond-drilled core from high stress environments can result in the core breaking into discs. Through four decades of research, a variety of computational tools and experimental approaches have been used to analyze the details of the failure processes leading to core-discing-type breakage of diamond-drilled cores. One motivation for this research is to use these calculations for stress estimation based on, for example, the disc length (i.e., thickness) measured in a particular core. This application requires that the core discs lengths do not deviate much from an average or typical value. This paper presents a stochastic approach, based on analyzis of the entire distribution of disc lengths, that provides a basis for application of existing core discing models even when it is unclear how to define a typical disc length, for example when the disc length distribution is highly non-Gaussian. The viability of the stochastic approach is demonstrated by comparison with data collected from 900 m of core extracted from a South Australian granite formation.