This work reports the results of an experimental validation study of a recently developed analytical solution, by the group authors, for the measurement of rock tensile strength using diametrical point load strength index. Results, obtained from a series of more than 200 point load and Brazilian tests with a wide range of rocks, suggest that the proposed equation is in good agreement with the laboratory measurements. When direct or other indirect methods are less applicable, it is concluded that the proposed equation can be conveniently adopted for the assessment of rock tensile strength.
Unlike metals, rocks are brittle and fracture without significant plastic deformation prior to failure. Rocks have also a high ratio of compressive to tensile strength, known colloquially as a high brittleness index. This makes rocks (especially hard ones like granite, basalt, and the like), extraordinarily weaker under tensile stresses than under compressive or shear forces. Therefore, tensile fracturing of rocks can be abrupt and violent, leading to shattering without any pre-warning. To avoid such instantaneous and unwanted failure modes, reliable prediction of tensile crack initiation in rocks is much needed, although this has always been a major uncertainty. The rock mechanics literature is therefore replete with direct and indirect testing methods to estimate this critical design parameter. The direct test methods, however, are difficult and expensive in practice, hence indirect tests have been reported to offer the most desirable alternative. Among the indirect methods, the Brazilian test (Carneiro, 1943), and the Ring test (Ripperger and Davids, 1947; Serati and Williams, 2015; Serati et al., 2016b), are perhaps the most widely utilised methods. Both tests simply involve compressing a disc or cylinder until it splits into two halves under tension. However, according to Griffith's strength criterion, the tests are strictly valid only if the rupture forms a single straight tensile crack initiated at the centre of a Brazilian test (BT) specimen, or at the hole periphery in the Ring test. Nevertheless, reports indicate that when hard rocks (like granite and basalt), or any type of materials with a high brittleness index (e.g. ceramic, glass or diamond composites) are tested, the expected tensile breakage is often characterised by multiple cracks, inverse shear conical plugs and triple-cleft fractures (Serati, 2014; Serati et al., 2015). That is, test results deviate significantly from the standard recommendations and produce erroneous estimations of the actual tensile strength of the target materials, for which there is no definitive interpretation available, and they are frequently overlooked.