Fast NVH Prediction for E-Motor Noise and Validation 2023-01-1037

Rapid electrification in the automotive industry leads to increasing demand for electric motors. These must satisfy multiple requirements, including performance, efficiency, cost, size, weight, while minimizing material use, and being designed and developed rapidly. It is crucial to limit customer annoyance from motor noise, vibration and harshness (NVH). This is often considered separately from the main design process, or through experience based rules. Issues are often found at a late stage, when resolution can be time consuming and expensive. Changes affecting motor noise can also change performance and rotor strength, delaying development if considered separately. Fast NVH prediction from the start of the motor design is important to speed up development by avoiding major NVH issues later. In this paper, different methods for electric motor noise and vibration predictions are proposed for different stages in the design process. The results are compared to previous experimental validation (Chen, et al). The first method combines electromagnetic finite element analysis to calculate the forces, with analytical models for the stator vibration and acoustic response (Gieras, et al). This requires minimal structural design data, and can be carried out rapidly, allowing the exploration of many design possibilities, and is implemented in Ansys Motor-CAD. The second method uses frequency response functions (FRF) to calculate the noise response from the predicted excitation forces, and is then compared to the full structural finite element analysis (FEA) (Zhang). It is shown that the fast analytical NVH process provides a good prediction for the motor noise level early in the design process, and that additional detail based on the overall system structural response can then be added using the FRF or full FEA process.