A new theory was developed which attributed sharp brittle-ductile transitions to a cooperative Kosterlitz-Thouless instability in dislocation generation. Simulations which were based upon the model showed that the new theory did not predict a strain-rate dependent temperature for the sharp transition. Instead, it predicted a sharp transition that was independent of the strain-rate. In the quasi-brittle regime, it predicted a strain-rate dependent gradual transition. Thus, the new theory did not explain the strain-rate dependent sharp transitions that were observed experimentally in Si. The results of experiments and simulations indicated that this transition was due essentially to the non-homogeneous emission of dislocations from the crack tip. Emission began at certain points along the crack tip. This generated a strongly shielding plastic zone which covered the entire length of the crack tip at the transition temperature; before the stress reached that for brittle fracture. For a given strain-rate, the transition temperature was therefore controlled by the dislocation velocity and a length which depended upon the original source distribution. This model, unlike the Khantha-Pope-Vitek theory, correctly predicted the strain-rate dependence of the sharp transition, and explained the fact that it was structure-sensitive.

P.B.Hirsch, S.C.Roberts: Acta Materialia, 1996, 44[6], 2361-71