ABSTRACT
Mining activities and the excavation of open pits in harder rock environments inherently leads to damage of the rock mass that will form the pit walls. Both blasting and stress relaxation, due to unloading, contribute to damage mechanisms which typically include crack extension and tensile failure. This damage factor (D) is used as a strength reduction in the Hoek-Brown criterion, and how D is spatially defined has been a long running issue in current rock mechanics practice.
It is widely accepted in the literature that the degree of damage is higher at the slope face and decays logarithmically deeper into the slope. Carvalho [7] developed a method to incorporate a gradational D factor which can be applied within standard modeling packages as a shear-normal function. The rate of decay is important as the level of damage due to blasting is more significant but only impacts the near surface, typical estimates are usually ≤30m behind the face. The damage caused by stress relaxation is proportional to the depth of the total excavation and extends further into the slope at lower levels than blast damage. Commonly accepted estimates for D are 1/3 of the slope height, which is supported by modelling work carried out by Guzman and Perez [4].
The addition of the variable D parameter by depth within the Generalised Hoek-Brown criterion in Rocscience Slide2 (V9.012) uses a linear decay function to represent the reduction in damage from the slope surface. This does not account for the values of D closer to the slope face decaying at a faster rate than those at depth and as such cannot represent D for both blasting and stress relief within the same constitutive model.
Applying a gradational damage factor to a shear-normal function can be onerous and has some limitations when applied in common numerical analyses. For example, the shear strength reduction method in Rocscience’s RS2 (v11.0) does not accommodate use of shear-normal functions and as such inhibits using this gradational D approach.
