Project Details
Projekt Print View

Improved numerical modelling and characterization of ferromagnetic materials and their losses

Subject Area Electrical Energy Systems, Power Management, Power Electronics, Electrical Machines and Drives
Term from 2012 to 2016
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 203416626
 
Final Report Year 2017

Final Report Abstract

The emerging market of electrical and hybrid vehicles is boosting intensively the research in highly efficient electrical drive systems, for which in particular electrical machines are demanded with high power densities (which means: high power production with low weight motors) in wide speed ranges, and with elevated operating frequencies, typically within the range 200 Hz to 800 Hz, i.e., well above the power line frequency. For such highly efficient electrical machines there’s a strong need for an accurate estimation of iron losses and an in-depth understanding of the different loss mechanisms at play, in relatively wide operational ranges of frequency f and magnetic flux density B. Such improved understanding of iron losses occurring in the machine’s stator and rotor parts is indispensable, in order to effectively carry out electromagnetic and thermal design of electric machines. This is were the current project starts. For this purpose non-oriented electrical steel sheets are treated as multilevel structures and modeled under arbitrary excitation waveforms by means of a coupled eddycurrent lamination model with magnetic hysteresis. This mesoscale model is transfered to macroscale simulations of realistic applications by a self-developed homogenization technique. The full-fledged models prove to be a suitable tool for estimating and separating the iron losses in electrical machines by the two-dimensional finite element method. The proposed model is readily vectorial and it allows one to systematically identify the model parameters based on standard magnetic measurements, which are bound to be representative of the material, irrespective of any specific experimental condition. This is substantiated by means of comparison with measurements under arbitrary loads at quasi-static and dynamic excitations. Starting from this, the proposed model is ready for further exploitation in the finite element modeling of macroscopic devices by means of the homogenization approach developed in this project. This two-step homogenization procedure is validated for arbitrary excitations and opens up the possibility of accurately evaluating magnetic losses in real-life electrical engineering devices (such as rotating machines, transformers, actuators, and brakes). As an example, the magnetic field and power-density post-processing calculations of the fully-coupled lamination model and its homogenized approximation are compared. It is expected with this mapping of objective features at two different length-scales, it will be possible to elucidate some of the relations existing between the microstructural features of electrical steels and their operational performance. By implementing the homogeneous material model developed in this project with appropriate simulation tools, it will be possible to determine which steel grades are the most appropriate in terms of their characteristics and load of a given application. This accurate matching of microstructural features with the requirements of the final application is at present not possible.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung