Design and Optimization of 4.8kW Permanent MagNet Brushless Alternator for Automobile G. Sabaresh (PG Scholar) Kongu Engineering College-Perundurai Dept. of EEE sabareshgs@gmail.com 45 Dr. N. Senthilnathan (HOD) Kongu Engineering College-Perundurai Dept. of EEE nsenthilnathan@gmail.com Abstract This paper presents the optimized design of 4.8kW, 184V at 3600 RPM- Outer rotor permanent magnet brush less alternator (PMBLA) for recreational vehicles. The finite element analysis is carried out using MagNet 2D/3D FEA- package for fine tuning the design and performance evaluation of the alternator under no-load and full-load conditions for saturation considerations. The analysis results demonstrate the effectiveness of proposed machine design methodology Index Terms Finite element analysis, Permanent magnets (PMs), Permanent magnet brush less alternator (PMBLAs), Ferrite (Fe), Neodymium Iron Boron (NdFeB) 1. INTRODUCTION The recent trend in the automotive field is adding new features improve the performance in terms of fuel efficiency and also in comfort of the customers. The power demand in the vehicles is increasing gradually in terms of kilowatts due to the addition of several electric systems like air conditioning, automatic parking system, traction control etc. The development of electronics field has simultaneously enabled improvement of the performances and customer satisfactions in the automobile application like cruisecontrol, Auto-Parking facilities, air conditioning or GPS. To drive all these a high power alternator is necessary. The Permanent magnet brush less machines have more advantages over other electrical machines such as brushed DC machines, synchronous machines, induction machines and switched reluctance machines. Because the field winding is replaced by permanent magnets, these machines exhibit high efficiency in operation due to electronic commutation. As a result, PMBL machines have been increasingly used in small motor drives for automobiles, recreational vehicles, and aircrafts. Recent researches in the field of high speed PMBL machine (around 2000-4500 RPM) for automotive applications and low speed PMBL machines (around 200-1500 RPM) for wind power generations,have introduced different types of methods, algorithm and model types improve the efficiency of the PM machines. The losses which include copper loss, eddy current loss, hysteresis loss, friction and windage losses were taken in to account in these methods The main objective of this work is to design an efficient three-phase PMBLA supplying 184V, 4.8kW at 3600 RPM. The converter topology of alternator is uncontrolled three phase bridge rectifier is shown in Figure.1 This paper presents the design algorithm of the three phase PMBLA. The Finite element analysis is carried out using Magnet 2D and 3D FEA- package. Fig.1. PMBLA with three phase converter. 2. MAGNET FINITE ELEMENT ANALYSIS (FEA) TOOL The MagNet tool is An electromagnetic field simulation software, an Electromagnetic field simulation software designed by Infolytica Corporation. This tool helps to predict the performance of any electromagnetic or electromechanical devices like Electric motors / Generators, Transformers, Actuators, Induction heating s and more in different conditions. It helps to model the machines in both 2D and 3D for to solve static 2D/3D, time harmonics, transient 2D / 3D (or) transient 2D / 3D with motion analysis and it provides accurate field results. 3. ALTERNATOR DESIGN 3.1 Topology of alternator The surface mounted permanent magnet outer rotor configuration is selected for the proposed machine and shown in Figure.2. Based on various parameters the 16Pole-48Slot double layer distributed winding configuration was selected for proposed PMBLA. The cogging torque in the outer rotor is lesser than inner rotor configuration for the same performance level and it is can further be minimized by skewing the stator or rotor. It has a lower audible noise due to greater inertia and less cogging makes the machine very suitable for use in quiet applications.
The Alternator is designed by using two different types of magnets Ferrite and NdFeB for different machine dimensions with the same pole slot combination. The grades of the magnets is shown in Table 1 S. No Description Fe-Magnet NdFeB-Magnet 1 Magnetic field density 0.45 T 1.1 T 2 Coersivity 342 KA/mm?^2 769 KA/mm?^2 Table 1 Grades of the magnets. Fig.4-2D model of NdFeB machine Fig.2 Surface Mounted outer rotor PMBLA 3.2 MagNet model Item No Description 1 Yoke 2 Magnets 3 Coils 4 Stator Core 5 Rotor The initial wire frame design of alternator is drawn in Auto- CAD software, imported and modelled in MagNet tool. The design has several constraints that it should satisfy. Like slot fill factor, current density, demagnetization of magnets, cogging torque after the selection of poles and slots. The Ferrite and NdFeB MagNet 2D model are shown in Figure.3 and Figure 4. The dimension details are as shown in Table 1. Table.2 Machine dimensions. The initial design start-up was with 200mm outer-dia and the axial length of 50mm. Subsequently this design is optimized to the final level by varying the axial length, changing slot area and shape, tooth width, back iron portion that satisfy the output requirement, with adequate flux density to avoid saturation and current density issues to ensure high efficiency of the machine. While designing and optimizing the machine size and volume should be take into account because it plays a major role in the automotive field in terms of cost and also in weight. The Flux density and current density values are shown in Table 3 and the Flux density saturation in tooth and yoke were shown in Figure 5 Figure 5 Flux density saturation 3.3 ANALYSIS OF ALTERNATOR The proposed model is analysed in various stages to check the performances. The types of analysis are ` Fig.3-2D model of Ferrite machine 46 1. Static 2D analysis 2. Transient 2D with motion and
3. Static 3D analysis 3.3.1. Static 2D Analysis A static analysis calculates the magnetic fields created by magnets. The loss computations cannot be predicted by using the static analysis. Initially the static analysis is carried out to check the current delivering capability of machine. This method is relatively easy and gives quick results. The machine is model is created. Because the model exhibits axial symmetry analysis need only be done in two dimensions. The machine is analysed in two different conditions under the static 2D. No-Load analysis and Load analysis. [1] No-Load analysis It is carried out for one-pole pair rotational angle and it is divided in fine steps to get better results. Fig 6 Flux Linkage between rotor and stator One pole pair rotational angle = Where Pole-pair = The rotational angle (0 45 ) is divided into fine steps to get better results. 1. To calculate the phase voltage, Induced emf, E = (pψ * ω) Where pψ =, ω = - To calculate the line voltage, - N = Speed (in RPM) dψ = Change in flux linkages (in Wb) dθ = Change in angle (in deg) The difference between phase voltages is line voltage. V RB = V R-V B (in Volt) Where R indicates R Phase Voltage B indicates B Phase Voltage. [2] LOAD Analysis Fig 7 Three phase flux linkage Load analysis is carried for load current of 50A. The response surface method used for carrying out load analysis gives the output of flux linkage values at three phase terminals after solving the problem, for different rotor positions and load current array given as input. With the help of flux linkage data, the output current produced by the machine under different speed for the corresponding regulating voltage is easily calculated by another tool called Claw Pole Circuit Simulator. The output performance of the machine is easily identified by plotting the graph of DC output current produced by the machine for different speed. The output load performance curve for the alternator is shown in Figure.8 The No-load flux linkage from rotor to stator is shown in Figure 6 and three phase flux linkage waveform is shown in Figure 7. 47
Response Surface Method INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY Fig.8 Load performance Curve In statistics, response surface methodology (RSM) explores the relationships between several explanatory variables and one or more response variables. The method was introduced by G. E. P. Box and K. B. Wilson in 1951. The main idea of RSM is to use a sequence of designed experiments to obtain an optimal response. Box and Wilson suggest using a second-degree polynomial model to do this. They acknowledge that this model is only an approximation, but its usage is plenty because such a model is easy to estimate and apply, even when little is known about the process. 3.3.2. Transient Analysis The transient analysis is mainly used for loss computation. The types are 1. Transient 2D and 2. Transient 2D with motion. The Transient 2D is similar to Static 2D it gives the additional information about losses. In Transient 2D with motion the loss data for specific speed can be obtained. The rotor components are created as 'Motion components' and the stator winding are connected with diode bridge rectifier. The battery is connected in the rectifier end and it acts as an electrical load. Then the machine is made to run at 3600 RPM loss data has been computed and load current value was verified. The winding connected with rectifier and battery is shown in Figure 9. Fig 10. 3D model of Ferrite machine Fig 11. 3D model of NdFeB machine The machine is modelled for one-pole pair with appropriate 'Boundary conditions' in order to reduce time. The grey section in the model that represents as 'Boundary conditions' its periodicity of 450 and it applied electrically to full machine. The machines were modelled for actual axial distance is shown in Figure 12 and Figure 13 Battery. Fig 9 Three phase winding with diode bridge rectifier and 3.3.3 Static 3D Analysis This Static 3D is used to verify the results obtained from the above analysis and it is the final verification of results. In this the machine is modelled like the actual machine and the Static 2D procedure is followed to analysis. The 3D model of Fe and NdFeB machine is shown in Figure.10 and Figure 11. 48
Fig 12 Actual axial length of Ferrite machine organization and in making necessary arrangements for successfully developing the machine designed Fig 13 Actual axial length of NdFeB machine IV SHORT CIRCUIT ANALYSIS AND COGGING. The machine winding terminals are shorted and the maximum short circuit current is calculated. That current is fed in to the winding and the machine rotates for one pole pair angle. The demagnetization properties of the magnets were studied. These properties should take into account while modelling the PM machines because the magnets plays a vital role in the operation The Cogging Torque is minimised by using Skewing Technique. V RESULTS AND DISCUSSIONS S. No Description Ferrite NdFeB Units 1 No-Load voltage (@ 3600 RPM) 440 446 V 2 Load current ( For 184 V, 3600 RPM) 28.15 31 A 3 Output power ( For 184 V, 3600 RPM) 5.15 5.6 kw The above table shows that the proposed design meets the requirement. The machine that designed and optimized in FEA software is being developed practically for testing and verifying the results and performance as already achieved through simulation. References [1] J. R. Hendershot and T. J. E. Miller, Design of Brushless Permanent Magnet Motors. Oxford, U.K.: Magna Physics Publishing and Clarendon Press, 1994. [2] M. Comanescu, A. Keyhani, and Min Dai, "Design and Analysis of42v Permanent-Magnet Generator for Automotive Applications," IEEE Transaction on Energy Conversion, vol. 18, no.1, March 2003. [3] E. Spooner and A. Williamson, "Direct-coupled, permanent magnet generators for wind turbine applications," IEE Proceeding of Electric Power Applications, 1996, vol. 143, no. 1, pp. 1-8. [4] B. H. Bae, and S.K. Sul, Practical design criteria of interior permanent magnet synchronous motor for 42V integrated startergenerator, Electric Machines and Drives Conference. IEMDC 03, Vol. 2, No., pp. 656-662, 1-4 June 2003. [5] MIT/Industry Consortium on Advanced Automotive Electrical/Electronic Components and systems, Discussion with Automotive OEMs about Typical Future Vehicle Requirements, 1996 [6] Edward Carl Francis Lovelace, Optimization of a Magnetically Saturable Interior Permanent-Magnet Synchronous Machine Drive, Doctoral dissertation at MIT Jun/2000 [7] B. Stumberger, G. Stumberger, D. Dolinar, etc., Evaluation of saturation and cross-magnetization effects in interior permanentmagnet synchronous motor, IEEE Trans. Ind. Appl., Vol. 39, No. 5, Sept./Oct. 2003. [8] J. Y. Lee, S. H. Lee, G. H. Lee, and J. P. Hong, Determination of parameters considering magnetic Non linearity in an interior permanent magnet synchronous motor, IEEE Trans. Magn., Vol. 42, No. 4, Apr. 2006. [9] J. Chen, C. Nayar and L. Xu, "Design and finite-element analysis of an outer rotor permanent-magnet generator for directlycoupled wind turbine applications," Proceedings of the IEEE Trans. on Magnetics, vol. 36, no. 5, September 2000, pp. 3802-3809. [10] S. A. Papathanassiou, A. G. Kladas and M. P. Papadopoulos, "Direct coupled permanent magnet wind turbine design considerations," Proceedings of the European Wind Energy Conference (EWEC'99), Nice, France, 1999 AUTHOR PROFILE: G. Sabaresh is currently pursuing masters degree program in power electronics and drives in Kongu Engineering College, Perundurai, Tamil nadu, India, PH- +91 96553 98553. E-mail: sabareshgs@gmail.com ACKNOWLEDGMENT The corresponding author hereby expresses deep gratitude to Lucas-TVS Ltd. & its G. Vigneshwaran (Development Engineer- Advanced Engineering department- Lucas-TVS Ltd) for his constant guidance and support to carry out this project in their prestigious 49