Proceedins of the World Conress on Enineerin 9 Vol I WCE 9, July 1-3, 9, London, U.K. FE Manetic Field Analysis Simulation Models based Desin, Development, Control and Testin of An Axial Flux Permanent Manet Linear Oscillatin Motor 1 Govindaraj T, Dr. Debashis Chatterjee, 3 Prof. Dr. Ashoke K. Ganuli Abstract- Development, finite element(fe) analysis of manetic field distribution, performance, control and testin of a new axial flux permanent manet linear oscillatin motor (PMLOM) alon with a suitable speed and thrust control technique is described in this paper. The PMLOM can perform precision oscillation task without exceedin the iven limit on allowable averae power dissipation. The use of new powerful permanent manet materials such as Neodymium-Iron-oron alloys can reatly improve the performance of electrical machines. Also its performance parameters, such as the force, current etc. are experimentally assessed. The objective of this paper is to determine the forces for aluminium mover embedded with rare earth permanent manet experimentally and analytically throuh FEMM software and develop a microcontroller based IGT Inverter for its control. Index Terms- Axial flux machine, finite element analysis, microcontroller based IGT inverter, permanent manet linear oscillatin motor, rare earth permanent manet. I. INTRODUCTION Many analytical and numerical techniques have been developed for the analysis of the end zones of electrical machine. The -D finite element method involves important computational means many efforts have been undertaken in order to use -D finite-element method FEMM.This paper presents different methodoloies based on -D eometries usin analytical solutions, This method has been implemented in conjunction with various eometry optimization techniques as it provides very fast solutions and has exhibited very ood converence with radient free alorithms. Manuscript received March 15, 9. 1 Govindaraj T is Ph.D. Research scholar with the Department of Electrical Enineerin, Jadavpur University, Kolkatta 7 3, India (Phone 91 9865494551, E-mail ID: ovindarajthanavel@mail.com, ovindarajt@yahoo.com) Dr. Debashis Chatterjee is with the Department of Electrical Enineerin, Jadavpur University, Kolkatta 73, India (Email ID: debashisju@yahoo.com) 3 Prof. Dr. Ashoke K. Ganuli is with the Department of Electrical Enineerin, Jadavpur University, Kolkatta 73, India (Email ID: ashoke_anuli@yahoo.com) Interior permanent manet motors are widely applied to the industry because of many advantaes. Also the characteristics of manetic materials are important to the performance and efficiency of electrical devices. Tradeoffs between accuracy, robustness and speed are central issues in numerical analysis, and here they receive careful consideration. The principal purpose of the work is to evaluate the performances of the PMLOM models when implemented in the FEM analysis of electrical machines. The developed methods are applied in a -D in-house FEM code, specialized for the desin and analysis of electrical machines. The FEM simulations and the analysis on axial flux PMLOM, and the numerical results are validated experimentally. The techniques developed for the calculation of interal parameters involve particular assumptions and simplifications and present specific advantaes. LINEAR motors are findin increasin applications in different specific areas like hih-speed transport, electric hammers, looms, reciprocatin pumps, heart pumps etc.[1]-[7]. They are also well suited for manufacturin automation applications. Therefore, desin of enery efficient and hih force to weiht ratio motors and its performance assessment has become a research topic for quite a few years. The permanent manet linear oscillatin motors (PMLOMs) are one of the derivatives of the linear motors in the low power applications havin the advantae of hiher efficiency. They can be supplied with dc or ac voltaes [4]-[7] of which, the dc motors are havin better efficiency due to the absence of the core losses. The motor desined and analyzed in this paper finds the suitability of application in the loads havin low frequency and short stroke requirements. One such application is the artificial heart pump, where frequency of oscillation is to be adjusted between.5 to 1.5 Hz, with the requirement of variable thrust dependin on the condition of the heart under treatment. For analysis of such motors the main task is to determine the essential equivalent circuit parameters, which are its resistances and inductances. The resistances, for the machine, thouh vary with operatin conditions due to temperature, do not affect much on its performance assessment. However, the inductances for these machines are mover position dependent and mostly affect the machine performance. Therefore, determination of these parameters is essentially required for analyzin the machine model. There are several works [6], [9] executed which assumes the machine inductance to be constant for simplicity of the model ISN: 978-988-171-5-1 WCE 9
Proceedins of the World Conress on Enineerin 9 Vol I WCE 9, July 1-3, 9, London, U.K. althouh different other works [4], [7], and [8] dynamically estimate the inductance throuh FEM and field analysis [1] for ettin the correct results. In this paper, the machine under consideration is an axial flux machine and the mover is havin a non-manetic structure, which is aluminium. Also the rare earth permanent manets used in the mover are havin a relative permeability nearly equal to unity and therefore the manetic circuit under consideration will be unsaturated due to major presence of air in the flux path. Hence, consideration of constant inductance is quite errorless for such kind of machines, which also conforms to the experimental data shown later. Finally the machine is analyzed with the help of the field equations and solved for forces and resultant flux densities throuh FEMM [1] backed by suitable experimental results. A controller usin PIC16F877A microcontroller has been developed for its speed and thrust control for successful implementation in the proposed application. II. MACHINE CONSTRUCTION The construction of the prototype PMLOM is shown in Fi. 1 below. Also the dimensional details of the motor are shown in Fi.. There are two concentric coils on the surface of the stators connected in such polarities that the fluxes for both the coils aid each other to form the poles in the iron parts. The formation of the N and the S poles of the electromanet of the stator are shown in the Fi.. The mover consists of aluminium structure embedded with rare earth permanent manets with the polarities as shown. The force developed will be attractive on one side and simultaneously repulsive on the other side. These two forces act in the same direction to enhance the total force on the mover, assistin the linear oscillation of the mover cyclically. Fi.. Dimensional details of the developed PMLOM Al Aluminium material PM-N4 Permanent Manet Attraction Force F and Repulsion Force F III. A Coil 1 aa and bb Coil cc and dd DETERMINATION OF TOTAL FORCE Determination of the Attraction Force Consider the Electromanet and permanent manet system in Fi., Let us find the manetic force between the electromanet and permanent manet poles, F= Force between the Electromanet and permanent manet, N; R = Flux density in the airap, Wb m ; Fi. 1. Construction details of the developed PMLOM (i) Stators to be mounted on both sides of the mover and (ii) the mover (iii) the PMLOM machine A= Area of manetic pole, m ; If one of the poles is moved by a distance of dx, Work done = F dx.this work done is equal to the chane of enery stored in manetic field. Chane in enery stored in manetic field = enery density chane in volume = 1 A dx µ = 1 A dx µ (1) ISN: 978-988-171-5-1 WCE 9
Proceedins of the World Conress on Enineerin 9 Vol I WCE 9, July 1-3, 9, London, U.K. 1 F dx= A dx () µ 1 or F = A N =.51 A k (3) µ µ Hence from above, Force per unit area 1 Pm = N m =.51 k m µ µ The flux density in the airap,, depends upon the mmf of the excitin windin and the permanent manet. Let AT be the total mmf of the excitin windin. A portion of this mmf is required for the airap and the rest for the iron parts of the manetic circuit. Let AT be the mmf required for the airap and AT for the iron parts. i permanent manet. where Now, H c (4) Hcl m be the mmf of the µ AT =, (5) l = Coercivity of the Rin type Rare Earth Permanent Manet L =axial lenth of permanent manet m 7 = 4 1 H m µ π l = Axial airap lenth x = displacement of the mover at any point of time From the (3), the Force ( ) AT A AT 1 µ 1 F = = µ A N l µ l If there is no saturation in the iron parts, the mmf required for them is small and therefore AT = AT (7) This ives: ( ) ( ) AT = AT + H 1 cl m + AT + H cl m (8) 1 AT AT FA = µ A N =.51µ A k (9) l l Referrin to Fi. (a) and Fi. (b), when the polarities are opposite, the MMFs of primary coil and secondary PM assist each other, but when the polarity is the same, the MMfs of primary coil and secondary PM oppose each other. In these (6) machines the flux density, can be increased to 1 tesla without saturatin the core, in which case the force density 5 becomes 57lb in or 4 1 N m ). Clearly this level of force density is very hih. Permanent manets, as an MMF source, yield a hih air ap flux density and simplify the power supply system. Determination of Repulsion Force Now, considerin the repulsion force, we have the polarities of the currents reversed as shown in Fi. (b). The fluxes in the airap are predominately radial. Qualitatively, it may be seen from Fi. (b) that we now have a force of repulsion between the coil and the permanent manets. For a small airap and for a uniform flux density in the airap a simple manetic circuit approach yields the manetic stored enery. Now ( ) ( ) AT = AT H 1 cl m + AT H cl m (1) F 1 AT µ x R = A N IV. SIMULATION AND EXPERIMENTAL RESULTS (11) The proposed scheme is simulated under FEMM4. environment, which provides a finite element analysis. The machine specification used for both simulation and experiment is iven in Table I. The maneto static fields are provided by Maxwell s equations H = J (1) = (13) where H is manetic field intensity, is manetic flux density and J is the current density of the manetic field. Subject to a constitutive relationship between and H for each material: = µ H (14) Where µ denotes material permeability. oundary conditions that must be satisfied at the interface between two materials havin finite conductivities are, ( 1 ) ( ) nˆ H H = (15) nˆ = (16) 1 Since the diverence of the curl of any vector must always be zero, it follows from (13) that there exists a so-called manetic vector potential A such that, = A (17) Substitutin (14) and (17) into (1) and takin a curl on both sides yields If 1 A = J µ (18) J = Jzˆ (19) ISN: 978-988-171-5-1 WCE 9
Proceedins of the World Conress on Enineerin 9 Vol I WCE 9, July 1-3, 9, London, U.K. Then, A= Azˆ () Thus, (18) reduces to, 1 A = J µ (1) The above equation (1) may be written in the expanded form as, 1 A 1 A + = J x µ x y µ y This equation () represents the scalar Poisson equation. () FEMM retains the form of (18), so that manetostatic problems with a nonlinear -H relationship can be solved. The advantae of usin the vector potential formulation is that all the conditions to be satisfied have been combined into a sinle equation. If A is found, and H can then be deduced by differentiatin A. The form of (18), an elliptic partial differential equation, arises in the study of many different types of enineerin phenomenon. Fi. 3(a) shows the FEM mesh confiuration for the PMLOM Prototype. Fi. 3(b) shows the Manetic flux plottin of PMLOM while mover is oscillatin within Stator 1, at 1 Hz, 4Amps. Fiure 3(c) is the correspondin flux plottin of the machine while mover is oscillatin within Stator 1, at Hz. Fi. 3(d) illustrates the finite element analysis of the PMLOM at the axial airap. Thus, the eometries of the mover and stator have been accurately discretized with fine meshes. Symmetry was exploited to reduce the problem domain to half of the axial cross section of the overall motor. The halved lonitudinal cross section of the motor has created the calculation area, with Dirichlet boundary conditions (Fi. 3(d)). Thus, the manetic field has been analyzed. For the calculations, material linearity of the NdFe permanent manet ( µ =1.48) was supposed. Its coercive force was r assumed to be H c = 95 KA/m and the manetization vector direction were adopted for the calculations. Very small air aps compared with the main motor dimensions between permanent manets and ferromanetic rins were nelected due to very small manetic permeability of the permanent manets, it is acceptable. The finite-element mesh (Fi. 3(d)) is dense in the air ap between stator cylinder sleeve and the mover. The fine mesh is also used near the edes of ferromanetic parts where the manetic field is expected to vary rapidly (Fi. 3(d)). In order to predict the interal parameters of the PMLOM, it is necessary to analyze the manetic field distribution in the stator and mover. Obviously, it is possible to optimize the construction by makin chanes in the stator and mover eometries. The improvements of the structure result from knowlede of the manetic field distribution. The presented results have been obtained for one variant of the motor construction. The control block diaram alon with the experimental set-up power electronic control circuit is shown in Fi.4. Here the thrust control is provided with the help of phase controlled ac supply which can vary the input voltae. The frequency control is provided with the help of a low cost and commercially available microcontroller PIC16F877A. The set-up is reliable and provides a scope for portability to any remote place. Fi. 5 shows the plot of the input voltae and current of the machine at 5 Hz. From which the assumption of constant inductance for the machine can be well validated. Fi.6 shows the characteristics plot of input power, voltae and force as a function of current for the machine taken at a frequency of 1 Hz. Fi. 7 shows Force at different axial airap lenth. The Force observed by measurement is compared with the theoretical Force value and shown in fi. 8. Fi. 3(a). Finite element mesh of PMLOM while mover is oscillatin with in Stator 1. Fiure. 3(b). Manetic flux plottin of PMLOM while mover is oscillatin with in Stator 1, at 1 Hz, 4 Amps ISN: 978-988-171-5-1 WCE 9
Proceedins of the World Conress on Enineerin 9 Vol I WCE 9, July 1-3, 9, London, U.K. Fi. 3(c). Manetic flux plottin of PMLOM while mover is oscillatin with in Stator 1, at Hz. Fi. 5. Current waveform of PMLOM taken from Tektronics make Storae Oscilloscope Fi. 3(d). Finite element Manetic flux plottin at upper and lower part of the airap while mover oscillates within stator 1.Now Mover is attracted to the Stator 1 Fi. 6. Measured Coil current versus Power (W).Voltae (V), Force (N) Characteristics of PMLOM Axial Airap versus Force Force(N) 18 16 14 1 1 8 6 Fi. 4. Power Circuit of PMLOM 4 3 3.5 4 4.5 55.5 6 6.5 7 7.5 88.5 9 9.59.5 Airap (mm) Fi. 7. Axial Airap Lenth versus Force 9 8.5 8 7.5 76.5 6 5.5 5 4.5 43.5 3 ISN: 978-988-171-5-1 WCE 9
Proceedins of the World Conress on Enineerin 9 Vol I WCE 9, July 1-3, 9, London, U.K. V. CONCLUSIONS A simple control method alon with the development of an axial flux PMLOM suitable for low frequency and short stroke application is presented. Analytical solution to the forces and determination method of the interal parameters of a PMLOM are shown. Finite element method with FEMM4. is used for the field analysis of the different values of the excitin current and for variable mover position. Computer simulations for the manetic field distribution, forces are iven. To obtain experimentally the field distribution and its interal parameters, a physical model of the motor toether with its electronic controller system has been developed and tested. The Prototype has been operated in the oscillatory mode with small loads at low frequency up to 5 Hz. The theoretically calculated results are compared with the measured ones and found a ood conformity. Fi. 8. Comparison of measured Force Versus Theoretical Force TALE I. PMLOM DESIGN PARAMETERS Rated Input Voltae 7V Rated input power 175 watts Stroke lenth Outer Diameter (Stator) Stator core type Thickness of lamination Stator lenth 1 mm 85 mm CRGO Silicon Steel.7 mm 6 mm Number of turns in Coil aa,cc 1 Number of turns in Coil bb,dd 5 Coil resistance Slot depth Permanent Manet Type Permanent Manet Lenth Coercivity Remanence Outer diameter (Mover) Shaft diameter Coil Inductance 18 ohms 5 mm Rare Earth N4, Nd-Fe- mm 95 A/m 1.3 T 65 mm 8 mm.18 Henry REFERENCES [1] M. Athans and P. L. Falb, Optimal Control. New York: McGraw Hill, 1966. [] H. D. Chai, Electromechanical Motion Devices. Upper Saddle River, NJ: Prentice Hall, 1998. [3] W. H. Hayt, Enineerin Electromanetics. New York: McGraw Hill, 1989. [4] G. Kan, J. Hon, and G. Kim, Desin and analysis of slotless-type permanent-manet linear brushless motor by usin equivalent manetizin current, IEEE Trans. Ind. Appl., vol. 37, no. 5, 1, pp. 141 147. [5] S. A. Nasar and I. oldea, Linear Electric Motors. Enlewood Cliffs, NJ: Prentice-Hall, 1987. [6]. Tomczuk and M. Sobol, Influence of the supply voltae on the dynamics of the one-phase tubular motor. with reversal motion, in Proc. 39th Int. Symp. Electrical Machines SME 3, Gdansk/Jurata, Poland, Jun. 9 11, 3, pp. 417 46 [7] N. Sadowski, R. Carlson, A. M. eckert, and J. P. A. astos, Dynamic modelin of a newly desined linear actuator usin 3D ede elements analysis, IEEE Trans. Man., vol. 3, no. 3, May 1996, pp. 1633 1636. [8] D. G. Taylor and N. Chayopitak, Time-optimal position control of electric motors with steady-state temperature constraints, in Proc.IEEE Int. Symp. Industrial Electronics, Montreal, QC, Canada, Jul. 6, pp. 314 3146. [9] S. Vaez-Zadeh and A. Isfahani, Multi-objective desin optimization of air-core linear permanent-manet synchronous motors for improved thrust and low manet consumption, IEEE Trans. Man., vol.4, no. 3, 6, pp. 446 45. [1] Finite element Method Manetics, Version 4. User s Manual, May 15, 8. ISN: 978-988-171-5-1 WCE 9