The Effect of Specimen Size on Strength and Other Properties in Laboratory Testing of Rock and Rock-Like Cementitious Brittle Materials

The effect of specimen size on the measured unconfined compressive strength and other mechanical properties has been studied by numerous researchers in the past, although much of this work has been based on specimens of non-standard dimensions and shapes, and over a limited size range. A review of the published literature was completed concentrating on the presentation of research pertaining to right cylindrical specimens with height:diameter ratios of 2:1. Additionally, new data has been presented considering high strength (70 MPa) cement mortar specimens of various diameters ranging from 63 to 300 mm which were tested to failure. Currently, several models exist in the published literature that seek to predict the strength-size relationship in rock or cementitious materials. Modelling the reviewed datasets, statistical analysis was used to help establish which of these models best represents the empirical evidence. The findings presented here suggest that over the range of specimen sizes explored, the MFSL (Carpinteri et al. in Mater Struct 28:311-317, 1995) model most closely predicts the strength-size relationship in rock and cementitious materials, and that a majority of the empirical evidence supports an asymptotic value in strength at large specimen diameters. Furthermore, the MFSL relationship is not only able to model monotonically decreasing strength-size relationships but is also equally applicable to monotonically increasing relationships, which although shown to be rare do for example exist in rocks with fractal distributions of hard particles.