A Material Law Based on Neural Networks and Homogenization for the Accurate Finite Element Simulation of Laminated Ferromagnetic Cores in the Periodic Regime

Electromagnetic fields and eddy currents in thin electrical steel laminations are governed by the laws of magnetodynamics with hysteresis. If lamination is large respect to its thickness, field current distributions accurately resolved solving a 1-D finite element magnetodynamic problem hysteresis across half thickness. This model able deliver mesoscopic information be used, after appropriate homogenization, macroscopic modeling an machine or transformer. As each evaluation such homogenized implies simulation at mesoscale, monolithic coupling might very time-consuming. article proposes alternative approach, assuming periodic excitation system, where parameters parametric material law determined neural network. The local can then used as conventional constitutive relationship 2-D 3-D modeling, massive speed-up coupling.