THE AUTHORS of the above referenced paper have recently attempted to modify the strain energy density criterion as applied to predict the onset of rapid fracture [l]. A mean value of the strain energy density factor 3 obtained by averaging S around the crack tip from 0 = 0" to 360" was used to determine the onset of fracture while the direction of crack initiation is still governed by the minimum value of S or Smin. The results based on 9 led to the false conclusion that /I = 72" gave the lowest critical stress rather than /I = 90" when the applied stress is normal to the crack plane. The discrepancy is due to Mode I and III interaction of the experimental data and not that of Mode I and II as concocted by the authors in their modified version of the strain energy density criterion. A discussion of the work in [l] can be found in 121. The same senior author used the strain energy density in its original form for the same angle crack problem and claimed agreement between theory and experiment [3].
The present discussion is concerned with the further modification of the strain energy density criterion made by the same authors who now concoct that the direction of crack initiation should coincide with the maximum value of the strain energy density function (d W/d V),,, calculated along the elastic-plastic boundary when plasticity prevails near the crack tip region. This is contrary to their previous assumptions used in [l, 31. The authors have not only failed to recognize that the strain energy density criterion is independent of material behavior regardless of whether yielding takes place or not but are also inconsistent in their lines of thought. The location of (d W/d V),,, has been shown [4] to correspond with the direction of maximum yielding and not the direction of crack initiation as claimed incorrectly by the authors. The paper also contains numerous unsupported assumptions, statements and claims that require clarification.
First of all, it is preposterous to make the general claim that the condition of diminishing Mode II stress intensity factor governs the final direction of crack propagation. In real structures such as aircrafts, ships, off-shore rigs, etc., load direction may change continuously and the crack may always be in a state of mixed mode loading. There is no reason for Mode II stress intensity factor to vanish.+ Even in the case of an angle crack in a finite size specimen under monotonic loading, Mode II prevails while the crack turns and may not vanish at the instance when the specimen separates into two pieces. The burden should rest on the authors to explain why so much effort has been spent on duplicating and incorrectly modifying the research activities of others without an understanding of the underlying principles. Remember that crack initiation is important because it represents the beginning of the fracture process followed by slow crack growth and final termination. These three stages should be addressed simultaneously and consistently by a single criterion under general conditions.