An Evaluation of the Double Torsion Technique

Double Torsion (DT) is a powerful testing technique for fracture mechanics characterisation of brittle materials as, in principle, it provides a crack length independent test configuration. However, several corrections have been proposed to address variations of experimental results reported from various laboratories. These correction factors address the validity of the DT configuration and its crack length independent stress intensity. Never the less, there seems to be no consensus in literature on the various corrections and the reason of reported variations. This paper presents firstly a critical review of the DT technique, followed by proposed corrections through an experimental analysis using the proposed corrections, a Finite Element model of the geometry and the use of Digital Image Correlation to measure out-of-plane surface deformations. It focuses on the validity of the constant stress intensity regime and the independence of crack length in a critical evaluation using Polymethylmethacrylate test specimens. Assessment of three un-grooved specimen geometry configurations demonstrated the apparent regime of approximately constant stress intensity, although a small but clear dependence of the stress intensity on crack length was observed in all specimen configurations. This dependence is attributable to significant load-point deflections and out-of plane deformations that are not accounted for in the DT analysis. Revisions of the proposed analysis methodologies show that a crack length independent specimen geometry can be achieved, however at the cost of less accurate data. Reliable and accurate data can be achieved with a DT testing configuration using an optimum specimen configuration.