Application of StADSS to large in-situ rock discontinuities: practical challenges and theoretical developments

<p dir="ltr">Predicting discontinuity shear strength remains challenging due to scale effects, with no clear consensus on how to manage them. This thesis advances the Stochastic Approach for Discontinuity Shear Strength (StADSS), an innovative method that bypasses scale effects by deriving predictions from full-scale surface roughness data. Building on previous laboratory validations, this work develops and applies StADSS to an in-situ discontinuity through three key investigations.</p><p dir="ltr">First, a field-based method was established to digitise discontinuity seed traces from photographs, quantifying errors from camera orientation, image resolution, and post-processing. Acceptable accuracy (≤ ±10% error) was achieved with a ground sampling distance <1.4 mm/pixel and near-perpendicular camera alignment.</p><p dir="ltr">Second, the variability of rock strength was examined through unconfined, triaxial, Schmidt hammer, and tilt tests on limestone, sandstone, and granite. Variability decreased with confinement, and simpler tests were shown to provide conservative estimates of strength variability. An algorithm was developed to incorporate these variations into StADSS via a Monte Carlo simulation.</p><p dir="ltr">Finally, StADSS was applied to a large-scale limestone discontinuity, with predicted peak and residual shear strengths of 442.9 kPa and 16.0 kPa, respectively, compared with an in-situ measured strength of 27.1 kPa. The residual prediction aligned with observed stability, confirming model feasibility.</p><p dir="ltr">This research enhances StADSS, demonstrating its potential for field-scale application and identifying future developments needed for practical engineering use, including improved digitisation of natural surfaces and further in-situ validation.</p>