Brittle to ductile transition in cleavage fracture

In many cleavable solids, cleavage cracks can propagate at steady state by laying down a trail of dislocations emanating from the crack tip and lying on planes inclined to the crack front. These crack-tip initiated dislocations produce shielding at the crack tip that reduces both the crack tip tensile stresses and the shear stresses on the inclined planes. They also blunt the crack. Cleavage cracks, can nevertheless, still propagate under appropriately increased stress intensity conditions to keep the crack tip tensile stress constant. A condition is reached in the propagation of such slightly blunted cracks where a small increment in temperature or a decrement in the crack velocity permits the nucleation of a new set of dislocations that produce additional shielding and blunting which tip the balance against the crack-tip tensile stresses. This results in a transition from brittle cleavage to ductile behavior. The steady state specific plastic work that can just be tolerated by a propagating cleavage crack before it catastrophically blunts is calculated to be only of the order of 10% of the specific surface energy. Although most geometrical details of the dislocation emission process are adequately modeled, the calculated brittle to ductile transition temperatures are found to be more than an order of magnitude higher than those that have been experimentally measured. This discrepancy is a result of the present inadequate methods of modeling activation configurations by considering the dislocation loop radius as the only activation parameter, while proper modeling of such configurations must consider also the Burgers shear displacement of the loop as an activation parameter. Such two parameter analyses, however, require accurate information on interlayer atomic shear resistance profiles for specific crystals which are presently not available. The analysis furnishes ready explanations of the toughening effects of so-called "ductilization" treatments, and embrittling effect of aging and dislocation locking, as well as the relatively large difference between the lowest levels of toughness between fracture in polycrystals and in single crystals.