We have studied ways of predicting power losses in soft magnetic laminations for generic time dependence of the periodic magnetic polarization J(t). We found that, whatever the frequency and the induction waveform, the loss behavior can be quantitatively assessed within the theoretical framework of the statistical loss model. The prediction requires a limited set of preemptive experimental data, depending on whether or not the arbitrary J(t) waveform is endowed with local slope inversions (i.e., minor hysteresis loops) in its periodic time behavior. In the absence of minor loops, such data reduce, for any peak polarization value J/sub p/, to the loss figures obtained under sinusoidal J(t) at two different frequency values. In the presence of minor loops of semiamplitude J/sub m/, the two-frequency loss experiment should be carried out for both peak polarization values J/sub p/ and J/sub m/. Additional knowledge of the quasi-static major loop, to be used for modeling hysteresis loss, does improve the accuracy of the prediction method. A more general approach to loss in soft magnetic laminations is obtained in this way, the only limitation apparently being the onset of skin effect at high frequencies.

Magnetic losses under triangular symmetric and asymmetric induction waveforms, like the ones found in power electronics devices, have been measured over a broad range of frequencies and predicted starting from standard results obtained with sinusoidal induction. Non-oriented Fe-Si and Fe-Co sheets, nanocrystalline Finemet ribbons, and Mn-Zn ferrites have been investigated up to f = 1 MHz and duty cycles ranging between 0.5/f and 0.1/f. The intrinsic shortcomings of the popular approach to loss calculation of inductive components in power electronics, based on the empirical Steinmetz equation and its numerous modified versions, are overcome by generalized application of the Statistical Theory of Losses and the related concept of loss separation. While showing that this concept applies both to ferrites and metallic alloys and extracting the hysteresis (quasi-static), excess, and classical loss components, we relate in a simple way the magnetic energy losses under symmetric triangular induction (square wave voltage) and sinusoidal induction. The loss behavior under asymmetric triangular induction is directly retrieved from the symmetric one, by averaging the energy losses pertaining to the two different semi-periods. Good comparison with the experimentally measured energy loss versus frequency behavior is verified in all materials.

We report and discuss significant results on the magnetic losses and their frequency dependence insoft magnetic composites. Two types of bonded Fe-based materials have been characterized atdifferent inductions from dc to 10 kHz and analyzed by extending the concept of loss separation and the related statistical theory to the case of heterogeneous materials. Starting from the experimental evidence of eddy current confinement inside the individual particles, the classical losscomponent is calculated for given particle size distribution. Taking then into account thecontribution of the experimentally determined quasistatic (hysteresis) loss, the excess losscomponent is obtained and quantitatively assessed. Its behavior shows that the dynamichomogenization of the magnetization process with frequency, a landmark feature of magnetic laminations, is restrained in these materials. This results into a partial offset of the loss advantage offered by the eddy current confinement.

A general approach to magnetic losses in soft magnetic composites (SMC) is discussed. It is applied to exciting conditions typical of applications, where the induction waveform is most frequently nonsinusoidal. The standard predicting approach, based on the statistical theory of losses (STL) and the related concept of loss separation, is extended to the case of granular heterogeneous materials. As a starting point, the classical loss component is calculated, taking into account that in small enough samples the measured loss is independent of the specimen cross-sectional area, because of negligible grain-to-grain (macroscopic) eddy currents. The loss separation is then performed under the conventional sinusoidal induction regime and the STL is used to identify a limited set of parameters, associated with the prediction of the excess loss, by which the energy loss under distorted induction can be predicted. Significant predicting examples are provided, regarding two different SMCs, where the induction waveform is either triangular or distorted by introduction of a third harmonic component.

This paper summarizes the main results and contributions of the MagNet Challenge 2023, an open-source research initiative for data-driven modeling of power magnetic materials. The MagNet Challenge has (1) advanced the stateof-the-art in power magnetics modeling; (2) set up examples for fostering an open-source and transparent research community; (3) developed useful guidelines and practical rules for conducting data-driven research in power electronics; and (4) provided a fair performance benchmark leading to insights on the most promising future research directions. The competition yielded a collection of publicly disclosed software algorithms and tools designed to capture the distinct loss characteristics of power magnetic materials, which are mostly open-sourced. We have attempted to bridge power electronics domain knowledge with state-of-the-art advancements in artificial intelligence, machine learning, pattern recognition, and signal processing. The MagNet Challenge has greatly improved the accuracy and reduced the size of data-driven power magnetic material models. The models and tools created for various materials were meticulously documented and shared within the broader power electronics community.

The statistical theory of losses (STL) provides a simple and general method for the interpretation and prediction of the energy losses in soft magnetic materials. One basic application consists, for example, in the prediction of the loss under arbitrary induction waveform, starting from data available from conventional measurements performed under sinusoidal flux. There are, however, persisting difficulties in assessing the loss when the induction waveform is affected by a dc bias, because this would require additional experimental data, seldom available to machine designers. In this paper, we overcome this problem applying, with suitable simplifications, the dynamic Preisach model. Here, the parameters of the STL model are obtained exploiting preemptive conventional measurements only. By this new simplified method, analytical expressions for the loss components are obtained under general supply conditions, including dc-biased induction waveforms.