Recent developments in the understanding of the structure and properties of grain boundaries in metals

The basic physical problems of the geometric models and structural models of grain boundaries based on computer simulations are considered. By comparing the approximations made in these models with the general solution of the problem based on the theory of metallic cohesion, we are led to conclude that i geometric grain boundary models (coincidence models) may be used to derive certain geometric features of a boundary (e.g. the boundary periodicity). However, due to the fact that interatomic forces are neglected, it seems not possible to deduce the actual atomic structure and properties of grain boundaries on the basis of these models. ii Dislocation models of grain boundaries seem to be of physical significance only for low energy boundaries. In all other boundaries, misfit dislocation are theoretically and experimentally found to be delocalized in the sense that their core is smeared out in the plane of the boundary. iii Structural boundary models deduced by computing the minimum energy atomic configuration using pairwise interatomic potential functions seem to represent a reasonable approximation of the actual atomic structure. iv The experimental and theoretical evidence presented indicates that grain boundary properties can, in general, not be derived from computer models, as these properties depend partially on electronic effects. v Further developments of the physical understanding of grain boundaries seems to require the incorporation of electronic effects in the theory of interfaces.