Leakage inductance and frequency-dependent resistance of windings are key parameters for the design of magnetic devices. This report focuses on analytical or semi-analytical calculation techniques to be integrated into a medium frequency transformer optimization process. To this extend, a review of currently most used method to calculate inductance and resistance of windings transformer is established. Based on this review, new calculation methods are developed. Magnetic field cartography is evaluated on the basis of two-dimensional structures thanks to a combination of Ampère's law and image method. From this magnetic field map, leakage inductance can be calculated. The obtained accuracy with this method is equivalent to three-dimensional finite element modeling. For frequency-dependent resistance evaluation, new methods based on two-dimensional or one-dimensional approach are compared to finite element modeling and reviewed models. An analysis of behavior of each model is done, showing that newly developed models are more suitable in the case of medium frequency transformer design in terms of accuracy, robustness and calculation time. This work could be useful to magnetic devices designers because newly developed models could easily be adapted to various geometries, such as toroidal structures or even inductors.

In power electronics applications, magnetic components are often subjected to non-sinusoidal waveforms, variable frequencies and DC bias conditions. These operating conditions generate different losses in the core compared to sinusoidal losses provided by manufacturers. In the conception and design stage, lack of precise losses diagnosis has unacceptable effects on system's efficiency, reliability and power consumption. Since virtual prototyping is used to predict and improve system's behavior before realization, losses and behavior prediction of components is possible. Circuit simulators and their compatible components models are required. This work is summarized by proposing non-linear dynamic model of powdered material magnetic core for use in circuit simulators. It includes the material's nonlinear hysteresis behavior with accurate winding and core modeling. The magnetic component model is implemented in circuit simulation software "Simplorer" using VHDL-AMS modeling language. Waveforms and losses of a powder core inductor in a buck converter application are simulated and compared to measured ones. The model is validated for different ripple currents, different loads and a wide frequency range. DC bias is taken into account in both continuous and discontinuous conduction modes.

Increasing electrification in transport sectors, from automotive to aerospace, highlights the need for low size and high power density components. The recent advent of planar technology theoretically allows to reduce considerably the size of the magnetic components. This article focuses on the design of a high frequency planar transformer intended to be used in a 4 kW 500 kHz DAB converter. In particular, the inter-winding capacitances are assessed, as they have a strong influence on the behaviour of the DAB, and in some extreme cases may impede operation entirely. Analytical and finite element models are used to evaluate the stray elements of the transformer (resistance of the conductors, inter-winding capacitance and leakage inductance), and the resulting circuit model is compared with experimental measurements. This work focuses on influences of design parameters on the transformer stray elements.

Some soft magnetic materials are strongly dependent of the temperature, because of their low Curie temperature. To be able to predict their behaviour in electrical devices, hysteresis models able to take into account the temperature are needed. The vector play static hysteresis model is an interesting hysteresis model that can be modified to be able to take into account the thermal behaviour. In this paper, the temperature is taken into account in this last model through its parameters. The variations of some parameters with temperature are mainly issued from numerical interpolation and specific assumptions. Simulation results are compared to measurements and discussed. KEYWORDSCurie temperature, soft magnetic materials, static hysteresis, vector play model 1 Int J Numer Model. 2018;31:e2221.wileyonlinelibrary.com/journal/jnm