The purpose of this paper is to focus on the loss separation of non-grain-oriented electrical steels used for speed-variable rotating electrical machines. The impact of laser-cutting, used in prototype manufacturing and of flux density harmonics, occurring locally in the lamination, on the loss distribution is studied in detail. Iron losses occurring under operation can physically be separated in different loss components. In this paper, a frequency-based loss model with parameters identified for single-sheet tester specimens, cut in strips of different widths, is therefore used. Moreover, a time-domain approach considers loss distributions occurring from higher harmonics. Hysteresis losses having high sensitivity to cut edge effects are calculated by the well-known Jiles-Atherton model adapting the frequency-based loss parameters. The model is validated by free-curve measurements at a single-sheet tester. It has been shown that the studied elliptical hysteresis model becomes inaccurate particularly for specimens with small strip widths with similar dimensions as teeth of electrical machine laminations. The incorrect mapping of losses occurring from minor hysteresis loops due to higher harmonics is concluded. The results showconsequently that both, the impact of a cut edge effect and local distributions of flux density harmonics need to be considered in terms of accurate iron loss prediction of electrical machine design.
The ways of the improvement of the method for the determination of steel losses in the electrical devices of basic types are substantiated. The method is refined by taking into account the magnetic system properties at high saturation. The presence of the interrelation between the special features of the domain structure movement and the shape of the hysteresis loop is proved for laminated cores. It enabled the explanation of the causes for the abnormally high values of the losses in the steel and the atypical shapes of the hysteresis loops at its high saturation. The empiric dependence for the determination of steel losses is obtained. It provides for the high convergence of the calculated and experimental data at the actual degree of saturation and can be used in the direct-current operation of the analyzed devices.