The optical caustic technique offers unique opportunities for experimental measurement of fracture parameters under conditions where conventional instrumentation either can not be used or is not available. The apparent simplicity of the method has attracted the interest of metallurgists and materials scientists. However, there are subtle problems with the method which complicate the interpretation of results. Some of the problems have not been fully explored and others need clarification before the method can be routinely applied. Of particular concern is the fact that existing theories for caustic behavior apply only in special cases which are rarely, if ever, attained in test specimens of practical dimensions for fracture testing. Nonetheless, the potential promise of the method justifies a systematic study of the influence of various parameters, both physical and optical, on the observed behavior of the caustic.

In previous work, the authors demonstrated that there is a correlation between the measured caustic diameter and the value of the J-integral parameter obtained from mechanical measurements for planar 1T compact tension specimens in various steels, subject to some restrictions. As might be expected, this correlation did not agree with existing theoretical relations since the theories are all based on plane stress assumptions. In the present paper, these correlations are studied further and the influence of imaging distance and specimen thickness examined in detail. In addition, the characterization of caustic behavior in standard bend specimens is investigated.

For these studies, a new four-lens caustic camera was used which permits the simultaneous determination of four caustics with varying image distances at each increment of applied load. With the aid of this camera, sufficient data were obtained to determine empirical relations for the calculation of the J-integral values from optical measurement of the caustic.