[IEEE Trans. on Mag.] An Investigation of Zeroth-Order Radial Magnetic Forces in Low-Speed Surface-Mounted Permanent Magnet Machines

AuthorsMostafa Valavi ; Jean Le Besnerais ; Arne Nysveen
StatusPublished in: IEEE Transactions on Magnetics ( Volume: 52, Issue: 8, Aug. 2016 )
Date13 April 2016
KeywordsVibration, Electromagnetic forces, finite-element method, generators, permanent magnet (PM) machines

Abstract

Zero-th mode of vibration is investigated in low-speed high-torque surface-mounted permanent magnet (PM) machines with non-overlapping concentrated windings. Magnetic field distribution in the air gap is computed using time-stepping finite-element analysis, and Maxwell stress tensor is then employed to calculate the radial force density. Based on the time variation of the mean value of the radial force density spatial distribution, the zeroth-order radial forces are studied. Simple structural equations are used to compare the stator deformations for the machines under investigation. Contribution of the zeroth mode to produce vibration is analyzed and compared with the lowest non-zero mode. The influence of pole and slot combinations and the effect of rectifier load are discussed. This paper is carried out to investigate whether or not the zeroth-order radial forces can be of importance in the presented surface-mounted PM machines that are not connected to pulsewidth modulation (PWM) converters.

Preprint and full paper

The full paper can be found on IEEE.

Notes

MANATEE simulation software is an efficient tool to model the electromagnetic and vibroacoustic behaviour of fractional-slot winding permanent magnet synchronous generators for wind turbine applications.
Due to low magnetic symmetry of these machines, finite element electromagnetic models are generally very large and numerical simulation times are too long to carry noise minimization at variable speed. Subdomain models of MANATEE combined with FEA can speed up simulation times, as well as Electromagnetic Vibration Synthesis and spectrogram synthesis algorithms.