Resource Efficient Determination of Electrical Machine Thermal Parameters

Improvements in electrical machine efficiency, power density and cost are driving the development of accurate multi-physics design tools, integrating Electromagnetic (EM) and thermal analysis to model the dissipation of internally generated losses. Lumped-Parameter (LP) and Finite Element Analysis (FEA) methods rely on time consuming experimental verification in the form of machine or motorette testing to determine unknowns resulting from manufacturing processes and incomplete material data. A resource efficient series of experiments is proposed, involving representative material samples to emulate the thermal behaviour of the stator-housing and stator-winding regions, both of which lie within the main heat extraction path of a machine. Comprehensive investigations are performed to concurrently determine the thermal conductivity and thermal contact conductance of these regions with trends compared to previous work to confirm method suitability. A further comparison is conducted against a conventional motorette testing rig incorporating compressed windings. This work aims to create a library of thermal data to be incorporated into design software increasing the validity of prototype data and enabling more effective investigation into the impact of different material combinations, packing factors and conductor dimensions on electrical machine thermal behaviour.