Abstract

In this project, 139 uniaxial compressive strength tests on specimens of Olkiluoto gneissic rocks at scales varying from 14 mm to 100 mm diameter were carried out. The aim of the study was to produce relevant information concerning the scale effect of the observed strength of Olkiluoto rock mass. Before the compressive tests, porosity and density were measured and geophysical measurements of P- and S-wave propagation velocity were performed for every specimen to obtain dynamic elastic properties of the rocks. Wave propagation measurements were performed for all specimens. Obtained results were in line with previous results for Olkiluoto gneissic rocks. No relevant scale effect was observed for wave propagation velocities. Differently oriented tests on the same specimen showed that wave velocities, particularly of S-waves, are subject to a relevant level of directional variability associated with the anisotropic nature of rock. During the project, strain measurements were planned to be done by means of LVDTs for the compressive strength tests. Initial results showed that axial strain measurements of the UCS tests based on LVDTs of axial strain were inaccurate. An energy-based corrective approach was developed to correct these strains. Values of UCS were obtained as well as values of the elastic parameters (E and n). CI values were obtained based on the crack volumetric approach and a version of the ITLS method, and CD values based on the total volumetric approach. Analysis of results showed that scale effect exists on UCS and on absolute values of CI and CD. This scale effect produces an increase in strength and crack damage parameters from small 14 mm diameter specimens up to 50 mm diameter specimens. Then a mild decrease is observed in line with traditionally considered scale effects. There was also a scale effect associated with the Young’s modulus of Olkiluoto VGN specimens: the stiffness tends to grow with specimen size. Poisson’s ratio and crack damage parameters in terms of UCS did not seem to be affected by the size of the specimen. A comparison between geomechanical and geophysical parameters was carried out. Traditional formulae to obtain the static Young’s modulus starting from the dynamic one and bulk density may produce reasonable results only for 50 mm and 60 mm diameter specimens, but not for the smaller ones. Recently developed formulae adjusted by the authors were able to produce reasonably accurate results of static Young’s moduli of rock specimens starting from dynamic modulus, density, porosity and UCS of the specimen.