A Fast and Accuracy 3-D Transient Thermal Modeling for Surface Water Cooling Permanent Magnet Machine

This article proposes a network-based 3-D thermal network (TN) model for permanent magnet (PM) machines with the surface-attached cooling structure (SACS). This model is developed based on the network topology method. It combines detailed radial and axial thermal information and discretizes the motor structure into highly discrete thermal components. In this way, highly accurate results are obtained. First, an algorithm for automatically constructing the TN model is designed. Then, the computational fluid dynamics (CFDs) is used to obtain the precise convective heat transfer coefficients (CHTCs). Afterward, the coefficients are integrated into the TN model for transient thermal analysis. Finally, the accuracy of the model is validated through finite element method (FEM) and temperature rise experiments.