The quality and usefulness of a DFN model is highly contingent on its ability to realistically capture the type, intensity, persistence, orientation and termination characteristics of rock discontinuities. As such, the DFN modelling process involves multistaged calibration and validation procedures. This iterative process is highly dependent on the availability and quality of discontinuity data to calibrate and validate the DFN model against. The paper presents a DFN model calibration and validation workflow for an over-station development design in Sydney, Australia. A calibration tool was created using Grasshopper VPL within Rhinoceros to understand the range of fracture intensities (P10, P21, P32) for Class I to Class IV Sandstone in the Sydney region. This tool incorporates the typical fracture type, orientation and spacing criteria published for the region in a Monte Carlo type simulation. This tool provides guidance on the ranges of volumetric fracture intensities used as input parameter for DFN generation within the different classes of Hawkesbury Sandstone. The generated DFN models using FracMan® software and Arup workflow are validated against the areal and volumetric fracture intensities (P21 and P32) obtained from site-specific mapping data and from the calibration tool. Validation results show discrepancies in P21 values between mapping records and the generated DFN models. Several reasons are presented to explain the validation results.
DFN modelling involves an iterative, multi-staged calibration and validation process to achieve sufficient realism in the DFN model. Discontinuities types and attributes including: orientation, spacing, intensity, persistence, termination, aperture, etc. should be adequately captured by the DFN model.
This paper presents a DFN model calibration and validation workflow for an over-station excavation design in Sydney, Australia.
A calibration tool was created using Grasshopper VPL (Robert NcNeel & Associates 2019, Rhinoceros 6.0) to investigate the ranges of fracture intensities (P10, P21, P32) for Class I to Class IV Sandstone in the Sydney region.