Recently, the double-torsion (DT) technique has been widely used for fracture mechanics studies of brittle materials. The technique is popular for a number of reasons. First, the compliance analysis of a precracked DT specimen indicates the applied stress intensity factor, KI, is independent of crack length. This allows tests to be run conveniently on opaque samples and in hostile environments. In addition, specimen and loading geometries are simple and easily adapted to most testing machines. Furthermore, specimens generally are precracked easily and slow crack growth data can be obtained using simple load relaxation tests. As in most testing techniques, however, experimental conditions may arise under which the validity of the data may be in question. For instance, the independence of KI on crack length in a DT specimen occurs over only a limited range of crack lengths. The crack propagation characteristics have also been shown to be dependent on the specimen dimensions. Finally, irreproducibility of slow crack growth data occurs under certain testing conditions.
In this paper these and other experimental aspects of the DT test are examined in order to define the conditions under which the data are valid. Among the topics discussed are specimen and loading geometry, precracking techniques, data reproducibility, slow crack growth measurements, and comparisons of DT data with other techniques.