Statistical coarse-graining as an approach to multiscale problems in magnetism

Abstract

Multiscale phenomena which include several processes occuring simultaneously at different length scales and exchanging energy with each other, are widespread in magnetism. These phenomena often govern the magnetization reversal dynamics, and their correct modeling is important. In the present paper, we propose an approach to multiscale modeling of magnets, applying the ideas of coarse graining. We have analyzed the choice of the weighting function used in coarse graining, and propose an optimal form for this function. Simple tests provide evidence that this approach may be useful for modeling of realistic magnetic systems.