Core-loss prediction is an important issue in design and analysis of permanent-magnet (PM) motors. Because of the diverse structure, flux distribution, and rotational variation of flux, it is difficult to predict the core loss in a machine exactly. In this paper, a core-loss model for the PM motor is introduced in which flux variation loci in different parts of the motor are predicted by carrying out a finite-element transient analysis. Since the flux variation pattern is complicated, an improved equation based on the conventional three-term expression is used for core-loss calculation. The core-loss model is developed totally in ANSYS parametric design language as a parametric model and it can be used easily for different types of PM motors. Calculation and experiments on a high-speed PM motor have shown that the model can produce results that agree with the experimental ones.

A three-dimensional analytical modeling of the magnetic field of the stator-ironless axial flux permanent magnet (AFPM) machine under open-circuit condition is presented in this paper. It involves the analytical solution of the governing field equations in the region between back-irons in the cylindrical coordinate, in which the magnets are assumed to be axially magnetized and have constant relative recoil permeability. The proposed modeling method is applied to a specific AFPM machine, and the analytical results are in good agreement with those of three-dimensional finite element analysis (FEA).

In a permanent magnet synchronous generator (PMSG) with modular winding, significant eddy current may be induced in the rotor magnets due to asynchronous rotating stator magneto-motive forces (MMFs), and a rectifier load may signify the situation further. The eddy-current loss prediction in the rotor magnets of a PMSG with modular winding feeding a rectifier load is described. An analytical method considering the stator current harmonics and stator MMF spatial harmonics and a time-stepping, coupled-circuit, 2-D finite-element method (FEM) are adopted. The predicted losses obtained from these two methods are compared and investigated.