Modeling and Experimentation of Power Magnetic Components at Temperature Units

点击翻译按钮获取中文摘要

Modeling and Experimentation of Power Magnetic Components at Temperature LimitsWith the availability of semiconductor devices that operate at temperatures up to 300'C, there is a growing need in the formulation of a design paradigm for power converters in high-temperature, high-power-density applications. Since Power Magnetic Components (POMADES) are an integral part of power converters, their engineering for temperatures that are as high as possible, or "temperature limits," is an essential part of the said paradigm.Thus, an interdisciplinary research program is proposed that will take a refreshing look at core material and component modeling, simulation, configurations, experimental measurements, and design optimization. The proposed work is "refreshing" in that temperature will be of first-order importance, and will not be conveniently neglected. The results are expected to be useful for not only high-, but also low-temperature magnetic designs.The core materials will be modeled by temperature-dependent "measured-core models" that characterize both the intrinsic magnetic properties and the geometry/dimensional effects. Temperature will be emphasized by temperature-induced, field-dependent functions that modify the applied magnetic fields, thereby reshaping the static hysteresis loops as temperature changes. The frequency and geometry dependencies of the hysteresis phenomenon will be modeled by filters, or by the equivalent circuits that result from finite-element discretization of Maxwell's equations and from model-reduction techniques.POMAC configurations will be identified that are suitable for high-temperature operation. The effectiveness of "matrix" configurations, planar configurations, "hybrid cores," and "heat spreaders" in distributing the electromagnetic and thermal variables uniformly will be investigated as means to extend the temperature limits.POMAC models will be developed for efficient coupled nonlinear electro-magneto-thermal simulation of POMACs with the other converter components. The models can predict flux and temperature nonuniformities due to frequency-, geometry-, and self-heating effects that, when coupled with the degradation of material parameters near temperature limits, ultimately lead to thermal runaway. A POMAC model will comprise coupled electrical and thermal subcircuits representing electrical and thermal substructures of weakly nonuniform distribution of flux and temperature. The subcircuit topology will be the high-order (in the geometry and frequency sense) generalization of the measured-core model. The subcircuit components are extracted from the electrical and thermal finite-element matrices by substructuring and model-reduction techniques.Experimentally-measured data for the core materials and POMACs at temperature limits and in the multimegahertz frequency range will be obtained and published. These data are not widely available and difficult to find in the literature.The temperature-sensitive design optimization will start with the derivation of analytical relationships between the set of performance parameters (e.g. temperature limits, power density, efficiency, and bandwidth) and the set of input parameters (e.g., temperature, dimensions, and excitation waveshapes) using the multiple response surface methodology. Design guidelines, performance boundaries, and design trade-offs unique to operation near temperature limits will be established.