Multiscale modeling of heterogeneous magnetic materials.

ABSTRACT This paper analyzes and discusses some computational methods for the simulation of the electric and magnetic behavior versus frequency of heterogeneous soft magnetic materials (such as soft magnetic composites, soft ferrites, and polymer composites). All these media are characterized by a granular structure, with magnetic grains separated by high resistivity layers, so that the study involves phenomena occurring at both macroscopic (sample size) and microscopic scale (grain size). The main purpose of the analysis is the prediction of the effective magnetic and electric properties, and of the relative energy losses starting from the known properties of the constituents, as well as the extrapolation of the constituent parameters from the measured global properties of the heterogeneous medium (inverse problem). The adopted computational methodologies include both analytical mixing rules and numerical approaches on the basis of homogenization and multiscale methods. The paper reports some examples illustrating advantages and drawbacks of these techniques, giving also an outline of nonlinear media homogenization. Copyright © 2013 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]