Modeling of the Geometry Effect on the Core Loss and Verification with a Measurement Technique Based on the Seebeck Effect and FEA

Abstract

In this paper, the effect of the core geometry of nontoroidal magnetic cores on the magnetic loss is investigated. A frequency dependent core material-geometry loss factor is developed. This factor is function of the change in the non-toroidal core section and the Steinmetz parameter "β". In addition, the temperature effect is included in the developed loss model for wide range of frequency and magnetic flux density. The model is applied for ER core and 3C92 ferrite material. The core loss measurements are performed using a Peltier cell. The principle of operation of the Peltier cell is based on the Seebeck effect, which convert the heat flow due to the temperature difference into electric power. The calibration of the Peltier cell is validated with a resistive load and a relative error lower than 1% is achieved. The accuracy of the developed model is assessed with FEA and the experimental results. A maximum error of 10% is registered of the developed core loss model.

Core Loss Calculation of Symmetric Trapezoidal Magnetic Flux Density Waveform

“…A fast-calorimetric measurement technique that uses a thermoelectric device to measure power loss is given in [21] and [23][24]. This method is based on Peltier effect where the heat flow can generate electric power that can be accurately measured with a high precision digital meter.…”

“…This method is based on Peltier effect where the heat flow can generate electric power that can be accurately measured with a high precision digital meter. The reader is referred to [21] and [24] for more details about the calibration of the Peltier device and the set-up of this technique. Valid for trapezoidal flux waveform of Fig.…”

“…The core loss depends also on the core geometry which needs to be accounted. A geometry factor correction is also applied to model (19) to take into consideration the geometry effect of the core due to the non-homogenous distribution of B as stated in [24]. Finally, the model to be implemented and verified is the following.…”

High Performance Planar Magnetics Based on an Unbalanced-Flux Approach

“…13). The reader can refer to [11][12][13] about the principle of this technique and its application for core loss measurement. Each core was tested at the following frequencies: [500, 600, 700, 800] kHz.…”

“…This model could not be applied to the unbalanced-flux design method because the magnetic flux density is not uniform. Thus, it is required to use the superposition technique through which the core loss is calculated for each part and then sum up all the loss terms [11]. Consequently, the total core loss can be expressed as follows:…”