Iron-Loss Modeling Based on a Loss-Separation Approach in Modelica

Abstract

In many electromagnetic energy conversion processes, transformers, or electric machines are controlled with modern power converters at high frequencies up to a range of several 10 kHz. Such operation modes lead to not negligible, distinct iron losses in the flux-leading laminated core. For estimating iron losses, no universally valid equation is available covering transient effects and nonlinear magnetization processes at the same time for a wide frequency range. Sophisticated 2-D and 3-D models are available but these hysteretic models are computationally very intensive when applied in a finite-elements method. This paper presents a 1-D iron-loss model, which allows estimating the static and dynamic hysteresis losses during the electromagnetic energy conversion processes with low computational effort. The proposed model can be used as lumped circuit element in a spatial magnetic equivalent network representing an electromagnetic device e.g., an induction machine. The model is based on a loss-separation approach, in which the different magnetization-loss mechanisms are interpreted corresponding to an analogous mechanical model. The proposed approach is implemented in Modelica; the model setup and the simulations are performed in Dymola. The validation of the developed iron-loss model is performed with measurements conducted on an Epstein frame.