CFD modelling and experimental calibration of concentrated windings in a direct oil-cooled segmented stator

Abstract

This paper presents a new approach in performing CFD modelling for electrical machine stator windings. Thermal modelling of the stator windings is important as the winding insulation is rated to a maximum operating temperature which limits the machine life or the torque ratings. While CFD analysis has been applied to investigate the flow path, its application to electrical machine windings has been limited as this is complex and challenging. Electrical machine stator windings are a composite of materials, typically consisting of multiple copper conductors wound around a stator iron bobbin. Each conductor is coated with a two-part insulation of approximately 2-micron thickness. The windings may be sometimes infiltrated with epoxy resin or a process of vacuum pressure impregnation (VPI) to increase mechanical rigidity. Representing the individual insulated conductors in FEA or CFD models requires an inefficient meshing process due to the difference in scale between the main conductor and its insulation. This adds significant complexity and leads to large computational demands and long solution times. To simplify the problem, researchers have often treated the windings as a homogeneous material with equivalent thermal properties. However, this approach lacks detailed information and an accurate physical representation. This paper addresses this problem by simplifying concentrated windings into multi-layered structures on which CFD analysis is performed. A thermal resistance is applied between the winding layers, to represent the thermal contact between the windings. The author believes that this provides a better physical representation of the conductive process in machine windings. The technique enables to achieve a detailed temperature map of the machine windings, thereby providing more information to the thermal designer. The paper uses a direct liquid-cooled axial flux yokeless and segmented armature (YASA) permanent magnet machine as a test case study. Axial flux permanent magnet machines enable higher torque densities and higher efficiencies making them suitable for applications such as road transportation or wind energy generation. This architecture has recently received increasing attention, and several variants can be found in the literature.