Numerical Simulation of Laboratory Strength Tests Using a Stochastic Approach

Abstract

Heterogeneity and discontinuity significantly affect rock strength. For accurate stability prediction, intact rock behavior is imperatively included in rock mass behavior. However, past research largely used arbitrary scaling approaches to produce rock strength. This paper considers a stochastic approach in order to determine the strength of a rock. Based on the laboratory data, a MATLAB® with extreme value stochastic model generates a database for each physico-mechanical property. Then FLAC® simulates laboratory-sized rock specimens. The grids developed in the numerical model can in turn develop random material properties in MATLAB®, which researchers then applied to the final FLAC® model. Model runs simulate the approach performed in the laboratory. The results from the model indicate that a stochastic approach produces strengths that are lower than a deterministic approach. Failure modes for each specimen are different, similar to observations in the laboratory. In addition, random density also influences the failure mode, highlighting the importance of stochastic analysis in rocks.