Analysis of a Modular LPTN for an Oil-Cooled Hairpin Traction Motor With HTC From CFD

The thermal analysis of oil-cooled traction motors with hairpin winding is crucial due to their intricate geometries and nonuniform cooling. This article proposes a modularized lumped parameter thermal network (LPTN) approach using heat transfer coefficients (HTCs) derived from computational fluid dynamics (CFD), thus dividing the motor into modules based on end winding cooling performance for more focused and efficient analysis. By comparing the modularized LPTN results with those from finite-element analysis (FEA) and experimental data, traditional LPTN models, which use a single module to represent the entire motor, are inadequate in predicting hotspot temperatures. In scenarios with minimal HTC variation or more uniform cooling, increasing the number of modules enhances the model’s ability to distinguish between well-cooled and less-cooled areas, thus resulting in more accurate predictions of average and maximum temperatures. Conversely, in cases with significant HTC variation or less uniform cooling, fewer modules may suffice to predict average and hotspot temperatures effectively. The findings highlight the effectiveness of the modular LPTN model in providing reliable thermal analysis of oil-cooled hairpin motors. This approach, combining LPTN with CFD, leverages both analytical and numerical methods to create a relatively simple and accurate thermal model.