In this paper the magnetization processes and power losses in an additively manufactured Fe-Si alloy with 3.7% wt. Si are studied. The material sample of interest is structured with air gaps to limit the dynamic losses by eddy currents. The magnetic behaviour of the material is modelled using an artificial neural network (ANN) trained on the experimental quasi-static hysteresis loops of the material, using a sample with the same geometry but a full section without gaps. The predictions are verified on a test set of experimental loops not involved in the training phase of the ANN. The ANN is later implemented in COMSOL Multiphysics®to simulate in a finite element method (FEM) scheme the dynamic losses of the sample with air gaps and compared with the measurements to validate the results.
Modelling of Magnetization Processes of 3D-Printed Fe-Si Components by Means of an Artificial Neural Network Implemented in a FEM Scheme
Faba A.;Bertolini V.;Cardelli E.
2024
Abstract
In this paper the magnetization processes and power losses in an additively manufactured Fe-Si alloy with 3.7% wt. Si are studied. The material sample of interest is structured with air gaps to limit the dynamic losses by eddy currents. The magnetic behaviour of the material is modelled using an artificial neural network (ANN) trained on the experimental quasi-static hysteresis loops of the material, using a sample with the same geometry but a full section without gaps. The predictions are verified on a test set of experimental loops not involved in the training phase of the ANN. The ANN is later implemented in COMSOL Multiphysics®to simulate in a finite element method (FEM) scheme the dynamic losses of the sample with air gaps and compared with the measurements to validate the results.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
