A theory is proposed for cleavage cracking surrounded by pre-existing dislocations. Dislocations are assumed not to emit from the crack front. It is argued that the pre-existing dislocations, except for occasional interceptions with the crack front, are unlikely to blunt the major portion of the crack front, so that the crack front remains nanoscopically sharp, advancing by atomic decohesion. The fracture process therefore consists of two elements: atomic decohesion and background dislocation motion. An elastic cell, of size comparable to dislocation spacing or dislocation cell size, is postulated to surround the crack tip. This near-tip elasticity accommodates a large stress gradient, matching the nanoscopic, high cohesive strength to the macroscopic, low yield strength. Consequences of this theory are explored in the context of slow cleavage cracking, stress-assisted corrosion, fast running crack, fatigue crack growth, constraint effects, and mixed mode fracture along metal/ceramic interfaces. Computational models and experiments to ascertain the range of validity of this theory are proposed.
