Homogenization of composite vicinal surfaces: evolution laws in \(1+1\) dimensions (Q449059)

The main goal of the paper is to elucidate how the microscale dynamics of constituent atom defects influences crystal surface (and surface structures) evolution at large scales. In particular, in order to find the macroscale description of surface relaxation that is consistent with the microstructure inhomogeneities of a composite stepped surface, the theory of singular perturbations has been invoked, i.e., that of a classical homogenization and multiscale expansions in one space dimension. The Fick law for the surface diffusion is established, the role of time scales and the mesoscale are quantified. An analysis is formal, leaving open questions. For example, the continuum limit is assumed to exist, its very existence has not been put under scrutiny. Certain regularity assumptions are known not to obeyed by some coefficients of the diffusion equation, leaving the description mathematically inadequate. The main goal of the authors was rather to give some intuition and guidance for mathematically more rigorous analysis, by exemplifying physical assumptions and the core elements of the homogenization process.