An anisotropic vector hysteresis model of ferromagnetic behavior under alternating and rotational magnetic field

Vector quantities and tensorial properties rule the magnetization mechanisms in ferromagnetic materials. Every ferromagnetic material is anisotropic to some degree in its magnetic response, so this anisotropy has to be considered for a general model. In this study, a multiscale approach based on a statistical description of the magnetic domain distribution and the knowledge of the crystallographic texture is used to predict the anhys-teretic behavior along an arbitrary space direction. Combined with the vector Bergqvist dry-friction hysteresis model, qualitatively reliable simulation results are obtained under alternating and rotational magnetization. FeSi 3% grain-oriented (FeSi GO) electrical steel is chosen as study material: FeSi GO are widespread so that extensive data are available and strongly anisotropic, forcing the model toward the worst-case scenario from the viewpoint of anisotropy. Moreover, under high excitation of rotational magnetization, losses drop due to the disappearance of the magnetic domains. This behavior is represented correctly by the proposed simulation method. In such conditions, the magnetization behavior is led mainly by the anhysteretic behavior, strengthening the predictive ability of the proposed model. In this manuscript, comparisons between simulations and measurements under many ampli-tudes of alternating and rotational magnetization for different levels of imposed excitation or magnetization are provided and used to validate the simulation method.

Automated summary

The study utilizes a multiscale approach along with statistical description of magnetic domain distribution and crystallographic texture knowledge to predict magnetization behavior in ferromagnetic materials in any direction, using FeSi GO electrical steel as the study material. Simulation results demonstrate that under high excitation, rotational magnetization leads to decreased losses, validating the predictive ability of the model.