Exploring Heterogeneity Consideration in Bonded Block Analyses of Laboratory and In Situ Tests on Opalinus Clay

The impact of heterogeneity on the failure of rocks has prompted researchers to perform numerical investigations. Heterogeneity can be incorporated into simulations through the implicit heterogeneity approach via cohesive zone modeling or the explicit heterogeneity approach by employing either grain-based modeling or the random distribution of properties technique. In this study, implicit and explicit heterogeneity schemes are thoroughly compared. Accordingly, bonded block models are developed using the Hybrid Lattice/Discrete Element Method and applied to laboratory test simulations and a microtunnel excavation of Opalinus Clay, an anisotropic shale. The laboratory models were used to simulate unconfined compression and direct tension tests, from which stress and strain behavior, failure modes, and acoustic emission (AE) activity were observed. The microtunnel model was based on the HG-A gallery in Mont Terri laboratory, and the simulations included the manifestation of the highly damaged and excavation damage zones as well as AE activity during the excavation phase. In conclusion, the implicit heterogeneity approach yielded easy-to-calibrate laboratory models that accurately captured anisotropy in deformability and peak strength properties. However, they exhibited linear-elastic behavior and lacked AE activity prior to peak load. In contrast, the explicit heterogeneity approach resulted in models with realistic anisotropic nonlinear behavior, but they were challenging to calibrate. The highly damaged zone in the implicit HG-A microtunnel model corresponded well with breakouts observed in situ. Nevertheless, it underestimated the excavation damage zone. Conversely, the explicit model presented a realistic excavation damage zone but lacked a well-defined highly damaged one.