Efficient NVH Analysis of Electric Motors by Means of 2D Spectral Decomposition of Airgap Forces

Abstract

The growing popularity of hybrid and electric vehicles has created a need for quieter electric motors. This has led to a focus on designing electric machines with reduced audible noise. Estimating the emitted noise requires complex calculations, but analyzing electromagnetic forces alone can provide valuable insights. Electric motors’ simple cylindrical structure allows for decoupling excitation forces and mode shapes, which are closely related. By representing the internal forces in a complex-valued matrix and calculating its 2D Fast Fourier Transform (FFT), we can obtain a comprehensive 2D spectrum that reveals temporal and spatial frequencies. The method enables individual analysis of unique harmonic components. This paper discusses the mathematical representation of forces, their spectra, and the relationship between force harmonics and mode shapes. Industrial examples of electric power-assisted steering systems demonstrate the practical application of this approach.