The paper discusses in detail the construction of the Core Less Axial Flux Permanent Magnet generator simulation model. The model has been prepared in such a way that full compatibility with the elements of the SimPowerSystem library of the Matlab/Simulink package is preserved, which allows easy use of the presented simulation model for testing the work of the generator as part of a larger system. The parameters used in the model come from the MES 3D calculations performed in the Ansys/Maxwell software, for a machine prototype with a rated power of 2.8 kW, which was then used to experimentally verify the correct operation of the presented model of machine.
In this study, the optimization of air gap magnetic flux density of open slotted axial flux permanent magnet (AFPM) machine which was developed for wind turbine has been obtained using the Taguchi experimental method. For this, magnetic analyzes were performed by ANSYS Maxwell program according to Taguchi table. Then the optimum values have been determined and the average magnetic flux density values have been calculated for air gap and iron core under load and no-load conditions with ANSYS Maxwell. Traditionally, 15625 analyzes are required for 6 independent variables and 5 levels when experimental method is used. In this study, optimum values are determined by 25 magnetic analyzes, which use L25 orthogonal array. For this purpose, both factor effect graph and signal to noise ratios are used, according to the factors and levels which are obtained from the factor effect graph and the signal to noise ratio. Parameters are re-analyzed by Maxwell. The optimum factors and levels are determined. For optimized values, the air gap magnetic flux density is improved by 65.7% and 173.26%, respectively, according to the average value and the initial design. Therefore, the variables are optimized in a shorter time with Taguchi experimental design method instead of the traditional design method for open slotted AFPM generator. In addition, the results were analyzed statistically using ANOVA and Regression model. The variables were found to be significant by ANOVA. The degree of influence of the variables on the air gap magnetic flux density was also determined by the Regression model.