ABSTRACT
In the early period of scienti‰c hot working around 1960 [1-4], it had already been substantiated that subgrains are formed at the beginning of, and remained stable throughout, steady-state creep of Al and many other metals (Chapter 9). The subgrain diameter ds increased as temperature T rose or strain rate ε declined, being uniquely related to the reciprocal of stress σ [5-22]. The grains elongated while the substructure formed within and remained equiaxed; Al alloys generally did not recrystallize after industrial processing, but might do so when cooled slowly; the need for quenching to preserve the true structure was slow to be appreciated. If stress were changed during steady state, the subgrain size would change gradually to the appropriate dimensions [17,22]. Likewise, substructures in cold-worked specimens could be altered to creep types with inhibition of SRX [23-25]. The creep rate dependence on T and ε [6-8] had led to theories in which dislocation climb was rate controlling. Limited measurements of subgrain misorientation (some on anodized Al, by POM [26]) indicated that it increased with strain well into mechanical steady state and led to the theory that the dislocation network internal to the cells was strength de‰ning while the subgrain boundaries (SGBs) with their very short range stress ‰elds (only if equilibrium) were inconsequential [12,19]. An important characteristic of high-temperature deformation was the formation of serrations in GB as they migrated for short distances, absorbing SGBs; the apex pointed down an SGB in one grain and the concave side contained a subgrain in the other grain (Section 4.6) [27,28].
