Proceeding of International Conference on Electrical Machines and Systems 2007, Oct. 8~11, Seoul, Korea A Study on the Design and the Characteristics in Single-phase Line-start Permanent Magnet Motor 1 Dae-Sung Jung 1, Seung-Bin Lim1, Jin-Hun Lee1, Sang-Hoon Lee1, Hyung-Bin Lim1, Youn-Hyun Kim2, Ju Lee1 1 Department of Electrical Engineering, Hanyang University, Korea Department of Electrical Engineering, Hanyang University, Seoul, 133-791, Korea 2 Department of Hanbat National University San16-1, Duckmyoung-dong, Yuseong-Gu Daejeon 305-719, Korea E-mail: jungds61@hanyang.ac.kr Abstract — The purpose of this paper is the optimal design of a single-phase LSPM (Line-Start Permanent magnet Motor). A single-phase LSPM has a permanent magnet in the rotor that is same as induction motor. For that reason, the transient characteristics are very different from an induction motor. Therefore, we need the PM design to consider transient and steady-state characteristics. In this paper, we designed a singlephase LSPM that has a high power and a good starting characteristic, and analyzed the performance of designed LSPM by varying each parameter. Finally, we examined and compared the simulation results and the prototype motor`s characteristics. I. INTRODUCTION In recent years, high-efficiency motors have been needed in a large variety of industrial products in order to save electrical energy. For many applications, a permanent-magnet synchronous motor can be designed which is smaller in size but more efficient as compared to the asynchronous machines such as induction motors [1]. Also the decreasing price of permanent magnets (PMs) and their improved performances make PM motors even more interesting in an industrial point of view as there is an increasing demand for high efficiency motors, such as compressors for house appliances. Single phase line-start synchronous motor with interior permanent magnet and flux barrier inside the rotor has high efficiency and almost unity power factor in the steady state, and robust starting performance caused by induction cage [2]. In this paper, for precise analysis and design by coupled Finite-Element Analysis (FEA), Design of Experiments and electric circuit analysis in steady state, the variation of d-q axes inductances by load condition are calculated by FEM then applied to performance analysis. Moreover, the results are verified as to compare with the experimental results. II. transient state. Therefore, a careful study on the parameters should be done to make a structure that has good starting characteristics. Additionally, during start-up, the machine should show the characteristic of the induction motor, however, because of the permanent magnet generated breaking torque and cogging torque, there are cases for starting motor to have difficulties with starting. In short, a trade off exists between the permanent magnet size and power generated and the starting characteristics. Therefore, a design considering both starting and power should be done [3], [4]. ANALYSIS MODEL The LSPM motor, which is used in this paper, has a rotor structure of interior type PM with a higher efficiency, as shown in Fig. 1. Single phase LSPM has a permanent magnet inserted rotor which generates reluctance torque caused by the magnetic resistance difference. The reluctance torque also varies due to the inductance variance caused during the Fig. 1. Structure of the single-phase LSPM motor A. Magnetic circuit analysis of LSPM The equivalent magnetic circuit is used to design the permanent magnet of LSPM, and we assumed the flux path as Fig 2(a). In Fig 2(a), five flux paths are split by cage bars, and some flux paths may be saturation. So, we assumed the reluctance of a path 1 and a path 5 is saturation, and the reluctance of others in rotor are not a saturation. A Fig 2(b) is a complete equivalent magnetic circuit by the Fig 2(a). We simulated many models by varying the geometrical component such as an angle or a radius of magnetic pole. And we can get some information from simulations that some models do not make desired torque though same size of magnetic pole and others have problems to self start. So, we used the D.O.E to make self starting and desired torque. The main specifications of the single phase LSPM is shown in table I - 878 - Authorized licensed use limited to: ULAKBIM UASL ISTANBUL TEKNIK UNIV. Downloaded on March 19,2010 at 05:21:28 EDT from IEEE Xplore. Restrictions apply. TABLE I The Specification of the Single LSPM Stator Rated torque[Nm] 0.34 Rated speed [rpm] 1800 Frequency[Hz] 60 Rotor Number of pole 4 Air gap Length[mm] PM Residual flux density[T] 0.15 1.12@20 C o Fig.3 Tendency analysis of the design factors (a) Flux pattern of LSPM C. Flux barrier design of LSPM Fig 4 shows non barrier, Barrier I and Barrier II. (b) Magnetic circuit of LSPM Fig. 2. Simple magnetic circuit analysis of LSPM B. Optimum rotor design by DOE The permanent magnet to be inserted in the rotor has impact not only on the output power but also on the starting time. For the design, thickness, insert angle, insertion position and the rotor bar resistance are selected as the main design parameters. Based on this assumption, the analysis models are minimized by DOE and a FEA was done for the selected models. The design factors and their level are shown in table 2. Fig 3 shows the influence and tendency of the starting time and torque depending on the selected four parameters. TABLE II Design Level and Design Factor of Each Factor Min Max Design level Magnet arc[deg] 30 35 2 Magnet thickness[mm] 1 2 2 Magnet deep[mm] 14 15 2 Resistant 1 2 2 Design factors Fig. 4. Design of barrier The flux distribution of the non barrier model with a magnet installed in a arc shape is shown in Fig 5(a). Permanent magnet design is an important design factor for efficiency and synchronization of the single phase LSPM. With a magnet showing high density, the efficiency may increases but can how loss of synchronization. With a magnet with low density, the motor can show better synchronization, however the efficiency will be degraded. Therefore, optimal magnet design is necessary. Even with optimal designs, inserting a permanent magnet without a flux barrier, as in Fig 5, the flux does not link with the windings. This increases flux leakages at the end of the permanent magnet which degrades the back-emf characteristics. The back-emf waveform is as in Fig 5(b) and the maximum value is 194[V]. - 879 - Authorized licensed use limited to: ULAKBIM UASL ISTANBUL TEKNIK UNIV. Downloaded on March 19,2010 at 05:21:28 EDT from IEEE Xplore. Restrictions apply. preventing heat demagnetization caused by high temperature and reverse magnetic field. Additionally, reviewing heat demagnetization is essential for a reliable motor design. The demagnetization characteristic of the NdFeB with different temperatures is shown in Fig 8. The knee-point exists in the second quadrant at normal temperature. However, with higher temperature, even though the knee-point is in the second quadrant, the residual flux density and coercive force decreases. A finite element analysis supported by the B-H data showed that a B lower than the knee-point does not exist. The results show that non-irreversible demagnetization does not happen until 140 C. (a) Vector of magnetic flux density (b)Result of EMF Fig. 5. Magnetic flux density and EMF of Non barrier model To prevent the flux leakage at the ends of the permanent magnet, a flux barrier can be installed besides the permanent magnet as shown in model a of Fig 6(a). As it is shown, the flux leakage at the end of the permanent magnet are decreased, however, the flux shows a leak through the other poles. The maximum value of the back-emf is 195[V]. Fig. 8. Influence of temperature of the demagnetization curve (a) Vector of magnetic flux density (b)Result of EMF Fig. 6. Magnetic flux density and EMF of barrier model I III. Model II of Fig 7(a) extended the flux barrier to the point where the squirrel cage were located, with the bars excluded from the model. This is to prevent the flux leakage to the adjacent poles. Compared to the non barrier model and the barrier model 1, the flux leakage has decreased and the flux path has been made for the flux to link through the armature windings. This shows that considering flux barriers is important for designs with optimal permanent magnet designs. Additionally, this design can reduce the magnet size which reduces the production cost of the motor. The maximum value of the back-emf is 263[v] which is higher than the previous two models also showing a more sinusoidal waveform. COMPARISON BETWEEN THE ANALYSIS RESULTS AND THE EXPERIMENTAL RESULT A. EMF The back-EMF of simulation and experiment are shown in Fig 9. The result of main winding is 157 and 168. As it is shown in Fig 9, the results of simulation and experiment match well. The motor was connected to the dynamometer to rotate at the rated speed, 1800rpm (a) Vector of magnetic flux density (b)Result of EMF Fig. 7. Magnetic flux density and EMF of barrier model I D. Analysis of LSPM considering demagnetization NdFeB is known to have outstanding magnetic characteristics. However, carefulmagnet design is required for Fig. 9. EMF - 880 - Authorized licensed use limited to: ULAKBIM UASL ISTANBUL TEKNIK UNIV. Downloaded on March 19,2010 at 05:21:28 EDT from IEEE Xplore. Restrictions apply. B. Speed-Time response The starting characteristics at no-load is shown in Fig 10. It can be seen that simulation and experimental results match. A large overshoot is observed when the motor enters steady state from transient state. This is because of the initial position of the rotor is different. A breaking torque is made when the motor enters synchronization. This breaking torque can be a positive value or a negative value depending on the initial position of the rotor. If the breaking torque has a value larger than 0, overshoot is observed. If the breaking torque has a value lower than 0, a synchronization without overshoot will be observed. IV. CONCLUSION In this paper, a single phase LSPM was designed for energy saving and high efficiency. The stator, bearing, housing and shaft of the single phase induction motor was adopted and only the rotor was optimally redesigned. The permanent magnet inserted to the rotor was selected by DOE which optimizes the starting time, steady state power output. The insertion position and permanent magnet size was selected. For an efficient use of the permanent magnet flux, a flux barrier was designed and the optimal position of the squirrelcage was selected. Additionally, the possibility of thermal demagnetization was also reviewed. To verify the design results, an experiment was done with a prototype. Through this paper, the size and the inserted position of magnet which makes synchronization possible were predicted. V. [1] [2] [3] [4] REFERENCES Andrew M. Knight and Catherine I. McClay, "The Design of HighEfficiency Line-Start Motors," IEEE Trans. Ind. Applicat., vol. 36, no. 6, pp.1555-1562, Novemver/December 2000. T.J.E Miller, "Single-phase permanent-magnet motor analysis.", IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. IA21, NO.4, pp.651-658, May 1985. K.J.Minns and M.A.Jabbar, "High-filed self-start permanent magnet synchronous motor," Proc. IEE, vol. 128, no.3, pp.157-160, 1981. Byung-Taek Kim, Young-Kwan Kim and Duk-Jin Kim, "Analysis of Squirrel Cage Effect in Single Phase LSPM", KIEE International Transactions on EMECS, Vol. 4-B No.4, pp. 190~195, 2004 Fig. 10. Simulation and measured speed-time responses C. Result of efficiency and power factor Fig 11 shows the power factor and efficiency characteristics of the designed single phase LSPM. PF above 0.95 was achieved for the entire drive range and the efficiency was 72.3% maximum at load angle 72. Fig. 11. Measured results of load performance characteristics - 881 - Authorized licensed use limited to: ULAKBIM UASL ISTANBUL TEKNIK UNIV. Downloaded on March 19,2010 at 05:21:28 EDT from IEEE Xplore. Restrictions apply.