Loss Prediction in DC-Biased Magnetic Sheets

Abstract

Power losses in soft magnetic materials can be solidly assessed by the statistical theory of losses (STLs), which provides physical foundation to the concept of loss separation. The theory is, however, limited to the conventional case of symmetric hysteresis loops and cannot be straightforwardly applied for a magnetic core operating under a dc bias. We show, in this paper, that such constraint can be released by combining the STL with a simplified approach to the dynamic Preisach model. This approach leads to the more affordable static Preisach model with largely reduced computation time. In this way, the hysteresis and excess loss components, with and without dc bias, are identified and calculated starting from a minimum set of experimental data. We provide a wide-ranging experimental validation of the theory, which is applied to the behavior of the energy loss versus frequency, measured up to $f = 1$ kHz, in non-oriented and grain-oriented iron-silicon sheets and in iron-cobalt alloys, subject to different polarization bias levels.