Triaxial strength and deformability of intact and increasingly jointed granite samples

Triaxial strength tests were performed on artificially jointed standard-size granite specimens with two sub vertical and three sub-horizontal (2+3) joints. A fully servo-controlled press was used to load the samples and attain residual strength. These tests add to an existing test database for intact and artificially jointed (1+2 joints) samples of the same rock. The main parameters from each test were used to analyse tendencies, in particular, changes in elasticity and peak and residual strength in line with increased jointing. Peak strength and apparent Young's modulus decreased from the intact to the jointed specimens, reflecting a tendency analogous to that deriving from the use of classification systems to characterize rock masses. Mohr-Coulomb and Hoek-Brown failure criteria were fit to peak and residual strength tests results. In the case of the Hoek-Brown criterion, it seems that an equivalent of the geological strength index (GSI), which reflects structure at the specimen scale, could be of use in estimating the decrease in strength of increasingly jointed samples. Residual strength envelopes were practically the same for the intact and jointed specimens. Our results for granite would suggest that the so-called rock mechanics scale effect at large (i.e. the relevant change in material properties from rock samples at lab scale to that of the rock mass including the occurrence of natural discontinuities at the engineering work scale), is more due to structure (number, orientation, spacing and features of occurring joints), than to sample scale effect (i.e. the variation in properties when testing different size, from a few mm to some decimetres, unjointed intact rock samples at the laboratory).