Simulating tunnel support integrity using FEM and FDEM based on laboratory test data

•Shale was characterized to be weak and moderately to fairly anisotropic.•Laboratory tests numerically calibrated to simulate tunnel models in shale.•Tunnel liner stress and strain asymmetric when explicitly simulating fracturing.•Substantial increase in liner stress and strain from large rock mass deformations. The West Vaughan Sewer System (WVSS) tunnel, at the time of writing, is planned to be constructed in Toronto, Ontario, Canada. A geotechnical program was conducted to characterize the mechanical properties of the anisotropic bedrock, Georgian Bay shale, and was determined to exhibit weak rock strength properties, moderate to fair anisotropy, and extremely low abrasivity. Following characterization, the shale was numerically calibrated using the hybrid finite-discrete element method (FDEM) and used to qualitatively simulate the WVSS tunnel. The major advantage of FDEM is its ability to explicitly simulate fracturing in geomaterials allowing for the excavation damaged zone (EDZ) to be estimated. In addition, anisotropic elastic and fracture models were used to appropriately simulate shale. The objective of the simulations is to explicitly compare the difference between the finite element method (FEM) and FDEM in estimating the deformation and stress in the rock mass and support liner. In FDEM simulations where the virgin stress conditions did not cause the rock to fracture, the FEM and FDEM results are very similar. However, in simulations where the rock is fractured, predominantly consisting of bedding plane slippage due to the orientation of the maximum in-situ stress relative to the strata, several key differences are evident. The liner simulated in FDEM has an asymmetric distribution of stress and strain, liner deformation is approximately 60% greater, stresses can be up to four times larger, and the stresses in the liner are purely compressive.