Experimental measurements and numerical modelling of additively manufactured Fe-Si cores

This paper presents experimental measurements and numerical computations for toroidal Fe-Si cores made by means of additive manufacturing (AM). The production technique of the samples is based on laser powder bed fusion (LPBF). Two different approaches, the Jiles–Atherton (JA) model and an artificial neural network (ANN), have been implemented and compared to represent the hysteretic behaviour of the samples made. The models are identified using experimental measurements at quasi-static rate for a bulk toroidal core. Their accuracy and reliability are evaluated by means of a comparison between simulations and measurements for a different data set. Moreover, a specific JA approach is implemented on the finite element scheme to evaluate its usefulness in the simulations of dynamic magnetization processes with not negligible eddy current effects.