A number of studies have investigated the size effect of intact rock mostly focusing on the uniaxial compression test. The most widely used being that by Hoek and Brown . The authors have conducted a suite of uniaxial compressive and point load tests on some igneous and metamorphic rocks to assess the effect of size on the results. The results illustrate that the trend of the Hoek and Brown  size relationship is not applicable to the tested rock samples particularly at small diameters. It was also found that the size effect relationship differed for the point load and unconfined compressive tests. The authors present the results and provide a discussion as to the potential reasons for the varying size effects at different stress paths. The results have implications for designers when considering potential testing programs and estimations of strength parameters for their designs in rock.
The effect of sample size on the behavior of quasibrittle and brittle materials such as rock has been studied by a number of authors [2, 3, 4, 5, 6]. Hoek and Brown  provided an empirical size effect model that is often cited. This relationship is similar in format to that suggested by Weibull theory  to explain the size effect in brittle materials.
Hawkins  conducted a number of UCS tests on seven different sedimentary rocks. Interestingly, he observed that at small diameters (less than 54 and 38mm) the trend of the size effect was not in agreement with the Hoek and Brown equation and was in fact reversed (Figure 2). This behavior was also reported by some other researchers [13, 14]. Recently, Masoumi et al.  reported similar results for Gosford sandstone (Figure 3).