On the design of a low cost high performance traction motor with ferrite magnets

Abstract

Permanent magnet motors with rare earth magnets are amongst the best candidates for high performance applications such as automotive. However, due to their cost and risks relating to security of supply, alternative solutions such as ferrite magnets have recently become popular. In this thesis the two major design challenges of using ferrite magnets for a high torque density and high speed application, namely their low remanent flux density and low coercivity, are addressed. It is shown that a spoke type design may overcome the torque density challenge due to a simultaneous flux concentration and reluctance torque possibility. Furthermore, the demagnetization challenge can be overcome through careful optimization of the rotor structure, with the inclusion of non-magnetic voids on the top and bottom of the magnets. To meet the challenges of the high speed operation an extensive rotor structural analysis has been undertaken, during which electro-magnetic as well as manufacturing tolerances are taken into account. In this thesis, the impact of the motor stack length and level of magnetic saturation on the demagnetization risk are studied based on 3-Dimensional Finite Element (3D FE) simulations and a proposed lumped circuit model. It is shown that reducing the stack length can significantly enhance the demagnetization resistance, with the effect being more pronounced for designs with a higher level of magnetic saturation. To benchmark the practicality of the concept, a previously presented high performance ferrite based design is modified by using a 30% weaker grade of ferrite magnet whilst shortening the stack length. It is shown that the demagnetization withstand capability of the design was significantly enhanced and exceeded the short circuit requirement with a good safety margin. The fir tree based spoke type rotor comprises of two sections: a) the ferromagnetic rotor pole to provide the path for the magnetic flux, and b) the non-magnetic rotor support to provide the structural integrity. In this thesis, the Multiphysics and cost implications of the rotor support material, as part of a high performance ferrite magnet traction motor, are analysed, and an optimal selection with respect to those criteria is proposed. The performance of the design based on the proposed rotor support material is validated by electromagnetic and structural testing of three sets of customized prototypes. Based on the analysis, the proposed rotor support material was shown to, significantly, boost the cost competitiveness of a low cost ferrite motor for a high volume production. ii As an alternative design to the proposed fir tree based rotor, a magneto-structurally optimized single piece rotor topology targeting the same EV application requirements is designed and compared against the fir tree based rotor performance. It is shown that an optimally designed single piece rotor design can meet about 80% of the power density of a fir tree rotor design at the cost of ~3 percent lower efficiency. Furthermore, the single piece rotor design may have better demagnetization resistance during severe faults. With regards to the performance per manufacturing costs, it was discussed that the single piece rotor design may match the fir tree solution, and the competitiveness may boost for designs with less severe structural requirements such as those with lower top speed requirement. With regards to the stator design, distributed and concentrated windings may have both advantages and disadvantages when considering manufacturing cost, slot fill factor, the contribution factor of reluctance torque and parasitic effects. Furthermore, the trend toward high speed operation of the traction motors may increase the AC loss effects in the windings, contributing to the motor deficiencies and risk of thermal failure. In this thesis, the performance of a high speed ferrite motor with a distributed and concentrated wound stator, and with regards to torque and power performance as well AC loss effects is assessed. The thermal capability of the windings under peak torque conditions and cyclic loading, as well as the intermittent and continuous performance of the full scale prototype design based on the proposed distributed aluminium wound stator is presented. The theoretical findings have been supported by a series of electromagnetic, thermal and structural testing of a custom built small scale prototype as well as a final full size prototype. The electromagnetic torque and power density is evaluated based on static and full dynamic testing, while the demagnetization withstand capability has been validated using current injection method. The structural testing includes an over-speed rotor spinning at 18000 rpm, as well as a fatigue testing under numerous cyclic loading. The thermal test validations include the evaluation of the aluminium windings temperature rise under the peak load, the reliability assessment under the cyclic load variation, and, finally, an investigation of the intermittent and continuous electromagnetic performance as well as windings temperatures of a fully assembled prototype. To conclude, a comparison of the proposed ferrite traction motor against the industrially available state of the arts is provided, based on which the merits of the PhD thesis findings and the competitiveness of the disclosed design in terms of the performance per cost is highlighted.