DEM analysis of failure mechanisms in the intact Brazilian test

A comprehensive numerical study is conducted to examine the failure behaviors in the intact Brazilian test. Brazilian test is considered a robust laboratory test to obtain an indirect measure of the uniaxial tensile strength of quasi-brittle materials such as rocks. Though validity of the Brazilian test is based on the premise that the diametrical splitting failure is caused by tensile crack growth from the center of the specimen (Scenario I), an alternative indentation-type of failure mechanism (Scenario II), where the splitting failure pattern forms as a result of cracks emanating from the crushed zones adjacent to the loading areas, is often observed experimentally. In this work, how the failure mechanisms and consequently the Brazilian tensile strength are affected by the material properties and the sample size are investigated using the DEM code PFC2D/3D. A novel displacement-softening contact model is implemented in PFC2D/3D so that materials with realistic uniaxial compressive over tensile strength ratios as high as ∼30 could be modeled. Formulation of the softening contact model and the effect of the micro-scale softening parameter on the macro-scale mechanical properties are first described. Intact Brazilian test is modeled in both 2D and 3D. We show that the failure scenarios transition from Scenario I to Scenario II, if the compressive over tensile strength ratio decreases or the sample size increases. For a low strength ratio material, if the failure mechanism follows Scenario II, the nominal Brazilian tensile strength could potentially underestimate the intrinsic tensile strength. Implications of this numerical analysis to laboratory testing and to the calibration of material properties in DEM modeling in general are also discussed.