Generalized Spectral Approach of Magnetic Field Prediction and Slotting Effect Modeling

Presence of slots/teeth in slotted surface-PM machines creates a non-uniform permeability distribution; therefore, in the modeling process, the governing PDE is no longer Poisson equation and the magnetic field distribution is not the combination of trigonometric harmonics. In this paper, the governing PDE is solved through separation of variables method (SVM) and it is discovered that the resulting magnetic field is the combination of a set of hitherto anomalous eigenfunctions. In short, slotting effect is modeled using a generalized spectral analysis rather than the common Fourier analysis. The superiorities of this approach are demonstrated with a numerical case-study.