Optimization of PM Segments Shift Angles for Minimizing the Cogging Torque ‎of YASA-AFPM Machines Using Response Surface Methodology

• Main Author: S.J. Arand
• Format: Article
• Language: English
• Published: University of Mohaghegh Ardabili 2021-12-01
• Series: Journal of Operation and Automation in Power Engineering
• Subjects: multi-objective design; cogging torque; rsm; afpm machines; magnet segmentation
• Online Access: http://joape.uma.ac.ir/article_1098_e13247bdfc014e7974eeb84e9b7f0af8.pdf

Mitigating the cogging torque is an important issue in designing the YASA machines. The main aim of the paper is to optimize an efficient technique to mitigate the cogging torque of YASA machines. In the suggested technique, the permanent magnets (PMs) are segmented into several segments in the radial direction, and then these PM segments are shifted at appropriate angles in the peripheral direction. The proposed PM segmentation method is compared with the conventional PM segmentation as well as the conventional PM skewing approaches in terms of the amount of cogging torque reduction and the amount of negative impact on the generator load-ability. It is shown that compared to the other two studied approaches, the proposed method is more effective in reducing cogging torque and at the same time, has a less negative impact on the generator output power. Using the suggested technique and via several finite elements based simulations, it is shown that without causing a significant negative impact on the generator load-ability, the generator cogging torque can be reduced considerably (about 90%). By implementing the RSM (Response Surface Methodology), optimal shift angles of the PM segments (factors) are determined to mitigate the cogging torque and maintain the generator load-ability. The experiments are carried out based on the RSM, as an important topic in the statistical DOE (Design of Experiments) approach, to study the impacts of PM segments shift angles on the output power and cogging torque of the YASA-AFPM generator. All of the experimental samples are extracted via the FEA simulations. Also, some of the simulation results are verified using the experimental tests.