Influence of grain size or anisotropy on the correlation between uniaxial compressive strength and sound velocity

Uniaxial compressive strength (UCS) is a critical rock strength parameter, and sound velocity tests are often employed to estimate UCS. However, the equations relating UCS to the sound velocity are grain size or anisotropy dependent, and how such equations respond to these factors remains unknown. To provide guidance for using the sound velocity to predict the UCS of heterogeneous rocks, sandstone (fine- and coarse-grain) and gneiss (0°, 45°, and 90° inclined anisotropy) samples are used to explore the effect of grain size and anisotropy, respectively, on correlations between UCS and the sound velocity. Several significant formulas for predicting UCS can be developed via the sound velocity, except in the case of gneiss with vertical anisotropy. The coarse-grain size and 45° inclined anisotropy increase the discreteness of the UCS data but not the sound velocity. The varying grain size and anisotropy orientation exert a negative influence on the correlation degree between UCS and the sound velocity. UCS values estimated using fine-grain rocks have the lowest relative errors, while considerable relative errors in the estimated UCS occur when using the sound velocity from samples with multidirectional anisotropy. Using empirical equations while ignoring grain size and anisotropy may therefore yield considerable discrepancies in the estimated UCS. This study suggests that the P-wave velocity is the best choice for predicting UCS. Tests should take into consideration varying grain sizes and be performed on rocks with unidirectional anisotropy; however, tests should not be conducted parallel to the anisotropy because of the lack of significant correlation.