The aim of this research is to increase the current density of electrical machines by improving the heat transfer from the stator. Hence, this research investigates key heat transfer parameters that limit convective and conductive heat transfer. The current density is interdependent on temperature and parameters governing heat transfer. Therefore, thermal analysis of electrical machines is important to design high current density electrical machines. This research starts by investigating the role air-cooled axial flux machines in the context of electric transportation. These are found to suffer from thermal limitations, forcing the propulsive power to be distributed among several wheels. The machine topology is found to play an important role in the heat transfer limits. The internal rotor topology suffers from heat transfer limits from the casing while the internal stator topology suffers from heat transfer in the rotor-stator gap. Addressing the latter is more challenging. This research does this by investigating a novel evaporative cooling mechanism to transport heat from the machine's internal stator to the outer rotor. A proof of concept was experimentally established and the challenges for adopting this mechanism to an electrical machine are investigated. The research focus is turned to direct oil-cooled machines. These do not suffer from the same thermal limits as they use an external radiator to expel heat. However, direct liquid cooled machines suffer from a non-uniform flow distribution, which affects the stator temperature distribution. To investigate this problem, an efficient thermo-fluid model was developed to predict the flow and temperature distribution in an oil-cooled stator. This was compared to CFD models and validated to within 6% of experimental results. The stator temperature distribution is improved by carefully controlling the flow distribution. The hot spot temperature is reduced by 13 K, doubling the insulation lifetime, or for the same hot spot temperature increasing the current density by 7%. The heat transfer coefficient an oil-cooled machine was measured by adapting the double layer thin film heat flux gauge technique. Correlations for the heat transfer coefficient on the pole piece surfaces are established and compared with analytical and CFD predictions. Finally the focus is turned to conductive heat transfer in concentrated windings. These are shown to suffer from a severe temperature gradient. Heat is transferred from one winding layer to the next and a hotspot is formed on the layer with the longest thermal path. The hotspot limits the current density of the machine. A lumped parameter thermal model was developed to predict the value and location of the hotspot in concentrated windings. To shorten the thermal path of the windings, a heat sink was interleaved between the windings. The new construction offers a reduction in hotspot temperature by up to 70 K. For the same maximum temperature the current density is increased by 30%. This thesis revisits flat windings and addresses their manufacturing challenges. Lastly, the relevance of thermal contact resistances is broadened to the general thermal design of electrical machines. This research shows that modeling the thermal resistance at the interface of concentric geometry by a constant parameter is an oversimplification. This was experimentally demonstrates to change with heat flux, contact pressure and material properties.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:730465 |
| Date | January 2016 |
| Creators | Camilleri, Robert |
| Contributors | Howey, David ; McCulloch, Malcolm |
| Publisher | University of Oxford |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | https://ora.ox.ac.uk/objects/uuid:2541b67a-bea8-4496-8410-ca1e88bf56e2 |
Page generated in 0.0022 seconds