Permanent magnet assisted synchronous reluctance motor design

Today, the efficiency of electrical machines is a critical design point. As there are numerous electrical motors and generators throughout the industry and home appliances, a tiny bit of efficiency improvement has critical value. Since the importance of this concern is highligted with the deacreasing amounts of fuel sources and their effects on environment, the attention for high efficiency motors and drives is increasing. Permanent Magnet Assisted Synchronous Reluctance Machines (PMaSynRM) are remarkable alternatives against the Permanent Magnet Machines such as BLDCs or PMSMs and also Induction Machines in this regard. They can be used both in industry and home appliance applications because of their significant advantages such as lower magnet volume, conductor lossless rotor, brushless design, etc. The attention to these machines by motor designers are increasing day by day since J. K. Kostko published the first paper on this type of motor in 1923. The stator structure of the motor is essentially the same as that of an induction machine. Main performance difference comes because of the rotor structure so the design considerations are mainly focused on the rotor. The main torque of this machine is due to the reluctance differences in the rotor geometry so the rotor has to be salient or anisotropic. This anisotropy is introduced to the rotor via flux barriers (air gaps inside the rotor laminations) that guide the flux to provide high and low reluctance magnetic regions under pole faces for the flux to travel hence creating a reluctance torque. The torque actually depends on the rotor saliency ratio. By defining two geometric axes; d-axis on the lower reluctance path and q-axis on the higher reluctance path, performance analysis of these machines is simplified. The induced torque increases with the higher d-axis inductance and the lower q-axis inductance. This property is defined as performance parameters of the saliency ratio (Ld/Lq), and the inductance difference (Ld-Lq). Performance of these machines are affected by these parameters since torque production depends on d-q inductance difference and maximum power factor depends on saliency ratio, so they are desired to be as high as possible. The design parameters such as flux barrier numbers and flux barrier thicknesses also affect d-q inductances. Since the torque ripple is the most important problem in the PMaSynRMs, many studies are conducted to reduce it. These studies show that rotor design parameters such as insulation ratio (the ratio of air to magnetic material along both d and q axes), number of flux barriers, stator slot/rotor flux barrier combinations, magnet positions and magnet amount are playing significant roles on reducing the torque ripple. Also skewing the stator, rotor or both is beneficial in this regard. SynRMs and PMaSynRMs shine out with a lot of advantages when compared to the induction machines. First of all, absence of the conductors in rotor means there are no rotor copper losses. This is the strongest point of SynRMs as there is no excess heat in the rotor. SynRMs are sometimes referred as the cold rotor motors because of this property and they are advantageous when any heat transfer from rotor shaft to load is not desired. The lack of an exciter circuit in the rotor side is also an advantage and reduces the cost of the machine. Because of the flux barriers in the rotor structure, less magnetic material is used in the rotor and this results in lower cost and lower iron losses. The PMaSynRMs also have some drawbacks. One of them is the torque ripple. The reluctance change of the flux paths when iron segments cross under the teeth and teeth openenings results in considerable amounts of torque pulsations. Combined with the air gap MMF distribution harmonics, the torque ripple can be a major problem. There are many studies in the literature to reduce the torque ripple of the SynRMs and the results are promising. The torque ripple can be greatly reduced by choosing the optimum combinations of stator slot numbers and rotor barrier numbers. Another major drawback of these machines is the need of complex control algorithms. Contrary to induction machines, SynRMs and PMaSynRMs cannot be directly started from the source without extra starting cage structures in the rotor. They need a variable frequency drive with the flux weakening capabilities for higher speeds in constant power region. The seperate drive system requirement increases the cost of the motor and maintenance. This drawback is somewhat compensated with today's solid state technology and modern control algorithms. In this study a Permanent Magnet Assisted Synchronous Reluctance Machine (PMaSynRM) is designed in order to replace a Permanent Magnet Synchronous Motor (PMSM) in a washer. Design procedure starts with determining the performance needs and structural dimension limits of the machine since the space where the machine is inserted in the washer is limited and must be kept constant. Also the washing mode and spinning mode has different operation values for the motor operation and they must be met. Structural design starts with the analytical design of the stator. Performance parameters of a stator such as slot number, teeth and yoke width, magnetic material, slot shape and slot opening are investigated to find an optimal point for starting the design. In this regard, pre-designed 12, 24 and 36 slot stators with distributed windings are simulated via ANSYS Maxwell 2D that is a two dimensional Finite Element Method based commercial software. Then these pre-designed geometries are optimized and analyzed again to observe if the results are improved. Also the double layer distributed and concentrated windings are investigated if they are suitable for PMaSynRM applications. After deciding a stator structure due to the performance needs, the rotor design parameters are studied. Initial study for the rotor design was to investigate the effects of flux barrier numbers on the torque production and torque ripple. Effects of barrier numbers were investigated for both 24 and 36 stator slots. An optimal value for rotor flux barrier number is chosen in the light of the FEM analysis results. After choosing the optimal flux barrier number, the effects of insulation ratios are investigated. The optimal starting point for insulation ratios are selected considering the literature data. Firstly, by keeping q-axis insulation ratio constant, the effects of changing d-axis insulation ratio on the average torque and torque ripple are studied. The d-axis insulation ratio is kept constant afterwards and the effects of changing q-axis insulation ratio is studied. After iterating many d-q axis insulation ratios and considering all the the results, a reference d and q axis insulation ratios are chosen fort he prototype. The permanent magnet position is investigated as the last step of design process. Different combinations of magnets are inserted in different locations inside rotor flux barriers for both 24 and 36 stator slot models. First, the 24 slot model is used for magnet insertions of its q-axis center. Magnets are inserted in different barriers and different locations to observe their effects on the average torque and the torque ripple. Magnets are located only on q-axis in the motor models with 36 stator slots. After that magnets on the ends of the flux barriers are placed in the radial direction. Also the models with their magnets filling the whole flux barriers are modeled and analyzed. Considering the average induced torque, torque ripple, efficiency and magnet volume, a final design is chosen. And some refinements are performed on the last design to reduce the magnet size and secure the magnets in place inside the barriers. The optimum motor design is then prototyped and tested, the comparison of test and FEM results are also presented at the end of this thesis.