The origin of Griffith cracks.

The presence of any pre-existing Griffith crack or a pore is not necessarily to be expected in metals as a result of the extremely high interatomic forces. It seems likely that pores and cracks may not be created intrinsically by the atomic mechanisms involved in the formation of a solid by solidification from liquid, or condensation from vapour phases nor, probably, by mechanisms of plastic deformation. It is proposed here that initiation sites for pores and cracks for most, if not all failures of metals, can only be introduced into metals via extrinsic (entrainment) mechanisms resulting from production processes, particularly melting and casting, but also spraying and powder metallurgy processes. It seems probable that only entrainment processes can create unbonded interfaces that can explain microstructures containing cracked (apparently 'brittle') intermetallics and decohered phases, and the initiation of tensile fracture and fatigue.