In this paper, the best laser powder bed fusion (L-PBF) printing conditions for FeSi steels with two different Si content (3.0% and 6.5%) are defined. Results show very strict processing window parameters, following a lack of fusion porosity at low specific energy values and keyhole porosity in correspondence with high specific energy values. The obtained microstructure consists of grains with epitaxial growth starting from the grains already solidified in the underling layer. This allows the continuous growth of the columnar grains, directed parallel to the built direction of the component. The magnetic behaviour of FeSi6.5 samples, although the performances found do not still fully reach those of the best commercial electrical steels (used to manufacture magnetic cores of electrical machines and other similar magnetic components), appears to be quite promising. An improvement of the printing process to obtain thin sheets with increased Si content, less than 0.5 mm thick, with accurate geometry and robust structures, can result to an interesting technology for specific application where complex geometries and sophisticated shapes are required, avoiding mechanical machining processes for electrical steel with high silicon content.

In this paper, we present a benchmark problem for the validation of numerical models of vector magnetic hysteresis. The problem geometry and the measurement system are described. The materials considered are both oriented grain (OG) and not OG Si-Fe steels. A series of experimental data about magnetic field and magnetic induction loci under rotational magnetization are presented with the corresponding magnetic losses. The measured data can be useful as a reference for the numerical models testing, comparison, and development. In addition, we propose here a comparison of the experimental data with the numerical results computed using a vector hysteresis model based on the extension of the Preisach model in 2-D.

In this paper, we discuss the numerical modeling of hysteresis in magnetic steels with oriented and nonoriented grain. The modeling is fully 3-D, but it is applied here to typical 2-D magnetic problems. Starting from a vector generalization of the Preisach model, we introduce here the concept of an operative magnetic field, function of the state of magnetization of the material. This operative field is introduced to reproduce with more accuracy the peculiar behavior of Si-Fe steels approaching the saturation state. The properties of the model are discussed and some examples are reported.

This paper presents a SPICE circuit able to predict dynamic core-losses in ferrite inductors with non-uniform magnetic field. The circuit is supplied by a current generator and provides the core power loss as output. The paper moves from models available in literature, which has been modified by the introduction of two main improvements allowing to extend the fields of applications. The first one is the design of a circuit able to ensure accurate power losses estimation even in presence of signals affected by noise. The second improvement extends the usability of the model also to those cases where the magnetic material saturates because of high current or temperature effect. Extensive experimental tests to assess the accuracy of the estimations and simulations have been carried out.

This paper deals with a neural network approach to model magnetic hysteresis at macro-magnetic scale. Such approach to the problem seems promising in order to couple the numerical treatment of magnetic hysteresis to FEM numerical solvers of the Maxwell's equations in time domain, as in case of the non-linear dynamic analysis of electrical machines, and other similar devices, making possible a full computer simulation in a reasonable time. The neural system proposed consists of four inputs representing the magnetic field and the magnetic inductions components at each time step and it is trained by 2-d measurements performed on the magnetic material to be modeled. The magnetic induction B is assumed as entry point and the output of the neural system returns the predicted value of the field H at the same time step. A suitable partitioning of the neural system, described in the paper, makes the computing process rather fast. Validations with experimental tests and simulations for non-symmetric and minor loops are presented.