A step towards the end of the scale effect conundrum when predicting the shear strength of large in situ discontinuities

The shear strength of rock discontinuities is known to be scale dependent, and past research has revealed that both positive and negative scale effects could be observed. It is far from trivial to predict the occurrence of the scale effect and, to date, there is still no consensus on how to satisfactorily predict the shear strength of large discontinuities. A new stochastic approach was proposed and validated at laboratory scale by the authors. The approach consists of (1) using the information available from visible traces to create synthetic surfaces via a random field model and (2) estimating the shear strength of each one of the synthetic surfaces in order to obtain a distribution of shear strength and a mean shear strength. This paper presents the first application of this new approach to a large discontinuity that was surveyed with a resolution of 1 mm and an accuracy in the order of 150-230 µm in the Pilkington reserve of Newcastle, Australia. The paper first confirms that a scale effect does exist for the surface tested, before demonstrating that the new stochastic approach produces a strength envelope that is very close to the deterministic failure criterion of the whole surface. The key conclusion of this research is that there is enough information on visible traces, if surveyed accurately, to obtain an estimate of the shear strength of the discontinuity.