High-Performance Vernier Machines with Halbach Array Permanent Magnets for Direct Drive Applications

This paper investigates the performance enhancements of permanent magnet Vernier machines (PMVMs) that can be achieved using a new structure of Halbach array permanent magnets (HAPMs) for a direct-drive motorcycle application. To start with, size and design specifications of the electric machine are...

Full description

Bibliographic Details
Main Authors: Arash Allahyari, Emine Bostanci, Amin Mahmoudi
Format: Article
Language:English
Published: MDPI AG 2023-05-01
Series:Machines
Subjects:
Online Access:https://www.mdpi.com/2075-1702/11/5/525
Description
Summary:This paper investigates the performance enhancements of permanent magnet Vernier machines (PMVMs) that can be achieved using a new structure of Halbach array permanent magnets (HAPMs) for a direct-drive motorcycle application. To start with, size and design specifications of the electric machine are determined based on the assumed acceleration and drive cycle performance of a motorcycle. Then, five-segment Halbach array permanent magnet Vernier machines (HAPMVMs) with two different slot/pole combinations (24 s/44 p and 12 s/20 p) are suggested and optimized to achieve a high torque density with an acceptable power factor while maintaining a low torque ripple. Two selected designs from optimizations are investigated in the full speed range considering power factor and efficiency maps. Consequently, in order to demonstrate the effectiveness of the proposed five-HAPM structure, the same optimization methods are repeated with three-segment HAPMs as well as with single-piece PMs. The comparisons show a great enhancement in torque and power factor achieved with the use of five HAPMs. For instance, for 22 s/44 p topology, generated torque doubles with the use of five-segment PMs compared to single-segment PMs. Finally, the harmonics of magnetic flux density in the airgap of PMVMs and HAPMVMs are compared and investigated to reveal the reasons behind the superiority of VMs with HAPMs.
ISSN:2075-1702