Parametric Optimization of Flux Focusing Type Double Stator and Single Rotor Axial Flux Permanent Magnet Motor

Abstract

Recently, great interest is developing towards axial flux permanent magnet motor (AFPM) for direct-driven in-wheel applications, due to their inherent multipolar disc-type structure and small axial length. Three-disc AFPMs have a high torque density because they effectively utilize the intermediate disc and are compact enough to be easily mounted in the wheel. Mechanical problems are also reduced because an intermediate disc is equally attracted in axial direction by its both sides. The slotted double stator and single rotor (DSSR) AFPM has more power and torque density and less cost, weight, volume, inertia and cooling problems in comparison to the single stator and double rotor (SSDR) AFPM topologies [1]. The flux focusing type slotted DSSR AFPM consumes a less amount of the permanent magnets (PMs) and has more torque density compared to the surface mounted permanent magnet (SPM) type slotted DSSR AFPM [2]. Therefore, in this paper flux focusing type DSSR AFPM is further investigated for parametric optimization. Initial dimensions of the flux focusing type DSSR AFPM are selected using the basic analytical modelling. A 3D finite element analysis (FEA) is utilized for its detailed characteristic analysis. The flux focusing type DSSR AFPM has 24 number of poles and 36 number of stator slots on each stator disc. Although it has a less winding factor (0.866), which decreases the output electromagnetic torque, it has a less total harmonic distortion (THD), zero fundamental or 1st harmonic, which reduces the losses, especially core losses. Due to the symmetry and its high periodicity of 12, 1/24th of each geometrical model of the flux focusing type DSSR AFPM is analysed using a 3D FEA, which decreases the computation time. The design of experiments (DoE) method is used for the parametric optimization of the flux focusing type DSSR AFPM. Although it is time-consuming due to the 3D FEA, it is suitable for the electromagnetic optimization of motor [3]. Initially, the full factorial design (FFD) is applied to analyse the effect of different design variables on the performance of the flux focusing type DSSR AFPM. With the help of the FFD, the significant design parameters can be identified easily. The FFD is very time-consuming, therefore, only the minimum, maximum and mean values of each design variable are considered, which limits the DoE. To extend the DoE and also to reduce the computation time compared to the FFD, the Latin hypercube sampling method (LHS) is used for the detailed characteristic analysis of the flux focusing type DSSR AFPM. The objective is to get best motor performance, such as high electromagnetic torque and back EMF and low torque ripple, cogging torque and total harmonic distortion (THD). The flux focussing type DSSR AFPM has constant outer radius length, current density, airgap, and stator yoke height. The design variables of the flux focussing type DSSR AFPM are shown in Fig. 1, where “A” is the ratio of the stator slot width and the slot pitch, “B” is the height of the stator slot, “C” is the ratio of the PM width and the pole pitch, “D” is the height of the PM, “E” is the ratio of the slot opening width and the slot pitch, “F” is the height of stator tooth tip, and “G” is the rotor’s inner to outer radius ratio. Fig. 2, represents the output torque characteristics of the flux focusing type DSSR AFPM. Initially, a number of one hundred experiments were carried out for the LHS. The experiment “X” has the highest electromagnetic torque, however, it does not have the lowest torque ripple. Similarly, experiment “Y” has the lowest torque ripple but does not have the highest electromagnetic torque. Experiment “Z” has almost same torque ripple, as that of “Y” but has a higher electromagnetic torque. Therefore, the optimal solution is in between “X” and “Z”. Interpolation between both geometrical models will provide an optimal solution of the flux focusing type DSSR AFPM. In the full manuscript, a detailed parametric optimization of the flux focusing type DSSR AFPM will be presented, based on the DoE method coupled with the 3D FEA. The effect of each design variable on the output characteristics, as determined by the analytical modelling and realized by the FFD will be presented and discussed. Along with the FFD, optimal design and a meta-parametric analysis of the flux focusing type DSSR AFPM will be carried out using the LHS.