A Comparison of Tensile Testing Techniques and Predictive Correlations Using Simulated Rock

ABSTRACT: The mechanical properties of rock are important parameters in the structural analysis of dams as they help inform an understanding of how a structure is likely to behave under loading. One commonly utilized but difficult-to-obtain engineering parameter is the tensile strength of rock. Direct tensile strength (DTS) testing of rock is rarely performed due to cost considerations and the complexity associated with the testing. The majority of testing is performed using indirect tensile methods, such as the Brazilian Tensile Test (BTS), akin to the splitting tensile strength test typically performed on concrete, or by using correlations to other mechanical properties like the crack initiation strength, mi from Hoek-Brown, or the Mode I fracture toughness. While the BTS is easier to prepare and more economical, significant uncertainty remains in how the test results correlate to the DTS and the actual behavior in situ. All the major organizations that standardize rock testing are in broad agreement on how to run the BTS however, there is still some room for more detailed guidance about platen shape, the use of a cushion, stress concentrations, and how to use the BTS to estimate the DTS. The goal of this research is to perform laboratory testing on stiff and soft simulated rock to improve understanding about the DTS and BTS tests, how different methods compare, and refine correlations between mechanical properties and the DTS. 1. INTRODUCTION In order for a robust model to be developed, the anticipated range of behavior of rock when subjected to a variety of loading conditions must be understood and quantified. This range of behavior can be defined by a full strength envelope developed from laboratory testing and up-scaled to the problem of interest. A full strength envelope for rock is provided in Figure 1, with the typical laboratory tests used to define the various portions of the curve given on the figure (Hoek and Martin, 2014). The tensile strength of rock defines a cutoff value in the overall shear strength envelope, but as indicated above, this value can be difficult to determine. The significantly more common uniaxial and triaxial compression tests are typically used to define the remainder of the envelope past the tensile cutoff.