This action might not be possible to undo. Are you sure you want to continue?

Welcome to Scribd! Start your free trial and access books, documents and more.Find out more

com/0332-1649.htm

**Electromagnetically excited audible noise – evaluation and optimization of electrical machines by numerical simulation
**

¨ C. Schlensok, B. Schmulling, M. van der Giet and K. Hameyer

Institute of Electrical Machines, Aachen, Germany

Abstract

Purpose – Disturbing vibrations and noise of electrical machines are gaining impact. The paper aims to focus on the necessity of estimating the electromagnetic, structure-dynamical, and acoustic behaviour of the machine during designing and before proto-typing. Design/methodology/approach – An adequate tool is numerical simulation applying the ﬁnite-element method (FEM) and the boundary-element method (BEM) allowing for the structured analysis and evaluation of audible noise also caused by manufacturing tolerances. Findings – The simulated results show good accordance to measurement results. The methods and simulation tools allow the analysis and evaluation of every type of energy converter with respect to its electromagnetic, structure-dynamical and acoustic behaviour. Originality/value – The methods developed and proved can be applied to any electromagnetic device in general. Keywords Simulation, Electric machines, Audibility, Noise control Paper type Research paper

Electromagnetically excited audible noise 727

1. Introduction The coupled physics of a complete acoustic simulation – starting from the electromagnetic force excitation, computing the mechanical deformation of the electromagnetic device and concluding in the estimation of the radiated audible noise – is a multi-physics problem. The central part of the computational chain is the electromagnetic ﬁeld simulation of which the surface-force density-waves are derived. These excite the stator of the machine resulting in vibrations. The periodical oscillation of the machines surface is decoupled and radiated as disturbing audible noise. For the numerical structure-dynamical and acoustic tool, a number of software programs have to be coupled, trimming the interfaces in a reasonable manner. This demands for a broad knowledge of the applied methods and their implementation. Therefore, an own tool box has been developed at the Institute of Electrical Machines (IEM) compatible to all types of electromagnetic devices such as transformers, rotating electrical machines (DC-, AC-machines, switched reluctance machines), actuators, and others (Arians et al., 2006; van Riesen et al., 2004; Zienkiewicz and Taylor, 1989; Kost, 1994; Bathe, 1986). The structure of an entire acoustic multi-physics simulation-chain is shown in Figure 1 (Schlensok et al., 2006). In a ﬁrst step a FEM-model of the electromagnetic device

COMPEL: The International Journal for Computation and Mathematics in Electrical and Electronic Engineering Vol. 26 No. 3, 2007 pp. 727-742 q Emerald Group Publishing Limited 0332-1649 DOI 10.1108/03321640710751181

1984.3 728 is simulated. The computation time of electromagnetic model rises disproportionately with Nel. In the following the three links of the simulation chain are discussed and examples show the efﬁciency of the applied tools. the sound-particle velocity. Piriou and Razek. Structure of the entire acoustic simulation-chain Multi – physics acoustic simulation . For the solving of the problem the FEM is applied. Often the number of elements Nel is more than ten times lower as in the 3D model. Therefore. Dziwniel et al. In the mechanical model various materials and geometries can be considered to analyse the vibration.COMPEL 26. 2D models result in least computational costs. A systematic. If.. 1984. In general. the ﬂux paths are in the orthogonal plane to the rotor axis. An aimed systematic optimisation of the machine is performed based on this multi-physics problem.. Using the same discretisation (element size) in the 3D model results in a non-reasonable high Nel. In the last step the audible noise of the multitude of variants is estimated and analysed with an acoustic Boundary-Element (BEM) model (Brebbia. time harmonic or transient. 2004). The resulting quantities are the sound pressure. 1998. Brebbia et al. Brebbia. a 2D FEM-model can be applied. 2. The electromagnetic model provides the normal component of the surface-force density on the elements used as excitation for the structure-dynamical model. 1990. 2D FEM-models should be applied whenever possible.. Hence. operation on the grid or on a converter).. Hartmann.. Williamson et al. various modes of operation can be considered. the variants of the electromagnetic model are considered as parameters. Dziwniel et al. Ho and Fu. 1989). electromagnetic field computation mechanical deformation calculation Figure 1. an electrical machine is studied. parameter-oriented model allows for a large number of geometry variations. furthermore. For special 3D geometry effects such as skewing 2D models can be applied as well by interconnecting several 2D models (Boualem and Piriou. the acoustic power of the sound pressure. Gyselinck et al. These can be static. 1990. Here. the effects of manufacturing tolerances can be taken into account as well as the designed geometry of the device (Schlensok and Henneberger. the type of ﬁeld problem has to be deﬁned ﬁrst. Electromagnetic ﬁeld computation As example for the electromagnetic ﬁeld problem. Furthermore. 2001. Therefore. The FEM can be used for the simulation of 2D as well as 3D electromagnetic ﬁeld problems. and the sound intensity. Therefore. 1999. Williamson et al. the discretisation of the geometry is ﬁner and hence the accuracy increases since there are more elements in the cross section of the geometry than in the case of a 3D model. 1978. The acoustic simulation chain allows for the a-priori analysis of several machine variants at various modes of operation (e. The simulation of 2D FEM-models shows some signiﬁcant advantages over 3D models: Two-dimensional models are much smaller.. all the other cases require 3D models. de Oliveira et al. In the case of a long electrical machine the end effects can be neglected.g. 2000. 2004). In general. 1995. 1997..

The electromagnetic simulation is performed in the time domain. 2000. By this. 2D model of the IM . All other parts of the IM are not modelled since they have no signiﬁcant impact on the electromagnetic ﬁeld in the IM. the copper winding of the stator. Furthermore. The boundaries of the IM at the shaft and the outer circumference are assigned with a Dirichlet boundary condition (tangential magnetic ﬁeld). 2006. Electromagnetic. Albertz and Henneberger. and the air gap. the aluminium rotor bars. 1989). the mounting kerfs on the outer circumference top layer bottom layer Z U Z V stator Z V V V V V X W W rotor rotor bar W X X Z X X X mounting kerf Electromagnetically excited audible noise 729 phase slot opening air gap Y Y Y U Y W Y Y W W U U Z U Z U Figure 2. The ﬂux-density distribution in Figure 4 shows that the highest values of B are reached in the tooth tips. At each time step the rotor is rotated depending on the rotor speed n and the time-step Dt: 180+ ð1Þ Da ¼ + ðnDtÞ: 60 After rotating and re-meshing the air gap an appropriate stator current is assigned to the model and the ﬂux-density distribution of the previous time step is taken into consideration for the right-hand side of the equation system. The four magnetic poles of the machine can be easily recognised.In this paper an induction machine (IM) with squirrel-cage rotor and a switched reluctance machine (SRM) are studied. Both electrical machines are simulated electromagnetically applying appropriate 2D FEM-models.. Hereby. shown in Figure 2 consists of the statorand rotor laminations. 2004). From each of the ﬁeld problems the magnetic vector potential results. Each of the models (one FEM-model per time step) represents a static electromagnetic ﬁeld problem. At IEM electromagnetic simulations are usually performed using the Galerkin scheme by setting u ¼ 2/3. the transient relaxation factor u weights the previous solution (Zienkiewicz and Taylor. Figure 3 shows the magnetic vector potential for one time step. The ﬂux-density distribution is given by: ~ ~ B ¼ rot A ð2Þ ~ with A being the magnetic vector potential. De Gersem et al. A quasi-static model of the IM computed for a ﬁxed number of time steps N applying a transient solver (Arians et al.. The process is exemplarily described for the IM in the following. The electromagnetic FEM-model of the IM. fast convergence is assured.

00E–02 8. The preferred resolution of the spectrum of the surface-force density used as excitation for the structure-dynamical simulation has to be considered here.00E–03 –8.00E-00 6. the resulting values are below the saturation point of the iron lamination this has no signiﬁcant effect on the motor’s operational behaviour.00E-01 2.00E–03 6. 1999): .COMPEL 26.00E–03 –1.00E–02 Flux Density B in T 2.3 Vector Potential A in Vs/m 1.00E–03 0. the cut-off frequency fco of the spectrum and the resolution Df depend on Dt and N with Shannon’s sampling theorem ¨ (Luke. The number of simulation time steps N has to be chosen depending on the resolution of the spectrum of all required quantities.00E+00 1. Both.00E–03 730 4.80E+00 1. Magnetic vector potential in the IM for one single time step –6.00E–03 Figure 3.40E+00 1.00E–03 –4.00E+00 8.00E–03 2.00E-01 0. constricting the ﬂux and increasing the ﬂux density slightly.00E+00 –2.60E+00 1.20E+00 1.00E-01 Figure 4. Since. Flux-density distribution in the IM for one single time step 4.00E+00 of the IM have some impact on the ﬂux paths.

1996).80E+05 4. 1998. magnetostriction and Lorentz forces are neglected and only s is taken into consideration further on. 2002.. For the acoustic simulation chain T and F are not of interest and therefore not regarded here. s is calculated for each time step on the stator teeth using the Maxwell – stress tensor – method (Ramesohl et al.40E+04 1. 2002).00E+05 3. The harmonics depend Surface-Force Density [N/m-2] 8. some speciﬁc electromagnetic devices such as transformers (Kubiak and Witczak. Bauer and Henneberger.Df ¼ f co ¼ 1 2DtN 1 : 2Dt ð3Þ ð4Þ Electromagnetically excited audible noise 731 From the ﬂux-density distribution the torque T.. Witczak et al. the surface-force density s. Surface-force density-distribution on the stator teeth of the IM for one single time step . The included zoom of one of the stator teeth shows.00E+04 0. The skewing of the rotor is reﬂected by the skewed force distribution on the teeth. The frequencies detected result from the fundamental and the harmonic air-gap ﬁeld-components of the stator which interact with the fundamental and harmonic components of the rotor (Jordan.20E+05 6. Belmans and Hameyer. that at motor operation the highest magnitudes appear on the up-running edge of the teeth. The surface-force density of the marked ﬁrst element (Figure 5) of the tooth is transformed to the frequency domain and analysed. 2002) afford the consideration of magnetostriction (Delaere. 1989). In addition to. s. Figure 6 shows the resulting spectrum.40E+05 5. and other quantities are derived.60E+00 8. 1999). Therefore.20E+05 2. In regular rotating electrical machines s predominates in general (Jordan. Timar. Figure 5 shows the resulting force excitation of a 3D IM-model for one time step.60E+05 4. Timar. the net force F. 1950.00E+05 7. 1989.00E+00 tooth 1 passing edge first element up-running edge Figure 5. 1950.

and the number of phases m. These frequencies are considered in the following structure-dynamical simulation of the IM described in the next section. Spectrum of the surface-force density of a single stator-tooth element on the point of operation deﬁned by the speed n. Parameters and point of operation of studied IM Table II. Table I collects these parameters for the studied IM and the studied operation point.COMPEL 26. the slip s. 1950. p NS NR m n f1 s Pole-pair number Number of stator slots Number of rotor slots Number of phases Speed Stator frequency Slip 2 36 26 3 1.138 Hz . 1989) listed in Table II. and the stator frequency f1 as well as the winding arrangements of rotor and stator described by the number of pole pairs p.183 Table I. Signiﬁcant harmonic frequencies 2 f1 NR n 2 NR n NS n NR n þ 2 f1 NR n 2 2 f1 2 NR n þ 2 f1 2 NR n 2 2 f1 Double stator frequency 1st rotor-slot harmonic 2nd rotor-slot harmonic 1st stator-slot harmonic Modulated 1st rotor-slot harmonic Modulated 1st rotor-slot harmonic Modulated 2nd rotor-slot harmonic Modulated 2nd rotor-slot harmonic 98 Hz 520 Hz 1. the stator and rotor slot numbers NS and NR.3 40 35 30 s Element (f) 732 s in N / cm2 25 20 15 10 5 0 0 100 420 520 620 940 1040 1140 f in Hz 1460 1560 1660 Figure 6.040 Hz 720 Hz 422 Hz 618 Hz 942 Hz 1.96 Hz 0. Timar. From these parameters the signiﬁcant harmonic frequencies are calculated (Jordan.200 rpm 48.

Bathe.. The surface-force density on the stator teeth. 1997). it is not possible to ﬁnd an exact analytical solution... ð7Þ where H is Hooke’s matrix. the calculation of the periodic. The determination of its parameters relies on material identifying algorithms. Owing to the large numbers and complexity of its components. is the threshold-accepting method (Ramesohl and Henneberger. 1996. the laminated sheet stack of the stator is modelled as such an equivalent material with an anisotropic Hooke’s matrix. Ramesohl et al. a structure-dynamical simulation is performed to determine the deformation or the oscillations. 2006. The structure-dynamical simulation is performed by means of FEM (Arians et al. Zienkiewicz and Taylor.. 1986). . As in the electromagnetic case. The deformation of the machine is represented by the displacement of the individual nodes of the mechanical FE-model (Schlensok et al. it is necessary to use equivalent materials in the simulation.3. this is expressed as: s ¼ H ·1. Delaere. Its entries are deﬁned by Young’s modulus E and Poisson’s ratio m of the corresponding material. For example. is used as excitation. which is obtained from the electromagnetic simulation. van Riesen et al. For the case of isotropic and homogenous bodies. Strain and deformation are related by: ~ 1 ¼ S·u 2 › ›x Electromagnetically excited audible noise 733 ð5Þ › ›y › ›x 0 › ›y 0 0 › ›z 0 › ›z › ›y 6 6 with S ¼ 6 0 4 0 › 3T ›z 0 0 › ›x 7 07 7 5 ð6Þ The correlation between the strain 1 and the tension s is given by Hooke’s law. it has the following form: 3 2 m m 0 0 0 1 12m 12m 7 6 m m 6 12m 1 12m 0 0 0 7 7 6 7 6 m m 6 12m 12m 1 0 0 0 7 7 6 Eð1 2 mÞ 7 6 122m ð8Þ H¼ 6 0 0 0 2ð12mÞ 0 0 7 7 ð1 þ mÞð1 2 2mÞ 6 7 6 7 6 122m 6 0 0 0 0 0 7 2ð12mÞ 7 6 5 4 122m 0 0 0 0 0 2ð12mÞ To decrease the computational effort to a reasonable measure. One algorithm. 2006. which is used to obtain the material parameters of an electrical machine’s stator. Neglecting initial strain and tension. 2004. mechanical deformation of electrical machines requires a numerical simulation. 1989. Structure-dynamic calculation After the electromagnetic simulation of an electrical machine. 2002).

The kinetic energy ~ of an oscillating body reads: Z r ·2 ~ · u dV. T¼ ð10Þ 2 V with the mass density r of the body. together with a subsequent modal super-position. For the structure-dynamical simulation of electrical machines.e. C is the damping matrix and F is the excitation force (Schlensok et al.3 The potential energy of a body due to strain reads: Z Z ~ ~ Pp ¼ 1 T ·H ·1dV 2 u·s ·1d›V.. The appropriate approach for solving the equation (15) depends on the application. Then. i. The deformation-solver formulation is constructed using Hamilton’s principle. the structure-dynamical simulation is performed. The mechanical model (Figure 7) consists of all components of the electrical machine.e. i. or by performing a modal analysis. the coupling between structure-dynamical and acoustic simulation can be considered a numerically weak coupling. This can be done either by solving equation (15) directly for each individual frequency separately. 2006). the inﬂuence of the air on the deformation of the solid body is negligible. . 2006). with: dD _ ¼ jvD D¼ dt equation (12) becomes: ðK þ jvC 2 v 2 M Þ·D ¼ F: ð15Þ ð14Þ The complex surface force density F is transformed from the electromagnetic simulation to the mechanical model for each frequency to be analysed. 0¼ d dt by ð11Þ ð12Þ ›L ›L ›F · 2 ›u þ · . since for the expected small deformations of maximally a few micrometers. D is the vector of the node deformation. The surrounding air is not included in the model. ~ ~ ~ ›u ›u where F represents a damping function. minimising the Lagrange function: L ¼ T 2 Pp . Thus. solving equation (15) for each frequency individually has turned out to be advantageous in practice (Schlensok et al.. the following oscillation equation is obtained: _ € K·D þ C·D þ M ·D ¼ F: ð13Þ K is the global stiffness matrix. V ð9Þ ›V 734 where ss is the surface force density and V the volume of the body. After discretising. Owing to harmonic analysis.COMPEL 26. ﬁnding the Eigen values and eigenvectors of the corresponding Eigen problem.

manufacturing failures and asymmetries have a large impact on the deformation modes. As in the electromagnetic case. for example. shaft cap B–side Electromagnetically excited audible noise 735 bearing rubber ring A–side housing Figure 7.73E-10 5.16E-09 8. speed. Mechanical model (exploded view) To achieve accurate values of the deformation of the entire machine structure. Deformation of the stator at f ¼ 618 Hz . extensive experience is required to set up a correct structure-dynamical model.04E-09 1. has to be modelled with much more details. A small deformation will enhance the acoustic behaviour. The modes of oscillation allow conclusions concerning the mechanical strength.62E-09 2.82E-10 2. Small mode numbers r are considered to be critical. certain modes of oscillation ensue. mode number 1 is considered very critical. Deformation [m] 2.75E-09 1. For example.33E-09 2. Figures 8 and 9 show. The rotor. the number of pole pairs and the winding arrangement. since it leads to large force excitations on to the bearings of the machine. may be represented as a solid cylinder. which depend on stator frequency. however. the mechanical structure becomes stiffer.91E-09 2. With an increasing mode number.45E-09 1. there are different requirements concerning the detailing of the individual parts. for example.00E+00 Figure 8. The deformation of an electrical machine can be evaluated as follows: According to Jordan (1950). numbers of slots. As studies have shown. and hence cannot be strongly deformed. a deformation of the stator of an IM. rotor. and therefore on the acoustic behaviour of electrical machines.91E-10 0. with mode number r ¼ 6. The housing. it is possible to determine the expected single tones.clamping bolt stator. In addition.

Depending on the technical question. ecc. however. Both eccentricities can occur together as a combined static-dynamical eccentricity. ecc. Therefore. stator air gap rotor Figure 10. the right evaluation criterion is to be chosen. Figure 10 shows the different rotational eccentricities. which may be identical to measuring positions. For the machine design. the body sound-level does not provide conclusions about the behaviour of the complete structure. the centres of the rotor. of the stator and of the rotation are located at the same position. Using this index. In the centric case. it is possible to globally compare the deformation for different excitations. Therefore. rotors of electrical machines are always mounted eccentrically. the body sound-level can be determined at ﬁxed locations. Owing to manufacturing tolerances. which are due to manufacturing tolerances. .COMPEL 26. For the eccentric cases.3 736 Figure 9. its disadvantage. it is necessary to estimate the effects of the eccentric support. the bearing is considered ideal. is its local scope. or the centres of the rotation and of the stator differ (static eccentricity). the body-sound index represents an integral quantity for the complete body. The advantage of this method is its comparability. stat. Different rotational eccentricities of the IM centrical dyn. and centric support is assumed. either the centres of the rotation and of the rotor (dynamical eccentricity) differ. Contrary to the body sound-level. Mode of deformation at f ¼ 618 Hz Alternatively.

The basic principle of the BEM is the solution of the Helmholtz differential equation: Dp þ k 2 p ¼ 0 _ _ ð16Þ Electromagnetically excited audible noise 737 120 100 80 L Sm in dB 60 40 reference dyn. Comparison of the body-sound index for eccentric models .-dyn.Figure 11 shows the results for selected frequencies of an induction motor at different eccentricities of the rotor compared to the centric machine. 2000). 4. An alternative is offered by the BEM (Schlensok et al. The static eccentricity leads to the largest body sound-level. stat. which generates single tones in the range larger than 3 kHz. and has to be regarded very critical (Jordan. Friction of the bearings of electrical machines is a further sound source. since the entire calculation area has to be discretised. for calculation of air-borne noise this method is unfavourable. only the surface of the area is discretised. However. the mechanical deformation of the machine is converted to the velocity. The presented calculation method only discusses the noise radiation generated by the electromagnetic deformation (vibration). 2006. In principle. It shows that for an eccentric machine an increase in deformation has to be expected. calculation of acoustic ﬁelds is possible with the FEM. For the acoustic simulation. eccentr. eccentr.. 20 stat. which consists of single tones in the entire range of audibility. Here.000 Hz) results in air turbulences generated by the rotating motor. Mai and Henneberger. the largest deformation occurs for NS n ¼ 720 Hz (NS ¼ 36) despite very small force magnitudes (Figure 6). 1950). For example. eccentr. 0 20 100 420 520 620 720 940 1040 1140 ¶ in Hz Figure 11. Housing vibration excited by the electromagnetic ﬁeld of electrical machines generates magnetic noise. It further shows that it is not possible to conclude immediately the oscillation of an electrical machine from the excitation force density. Acoustic simulation The acoustic noise radiated from electrical machines consists of three parts: the broad-band fan and ventilation noise (500-1.

30E+01 8.20E+01 8. The mechanical velocity is transferred to this acoustic mesh. The used program was the procedure presented here.47E-03 5.79E-05 .COMPEL 26. sound pressure and sound particle velocity are evaluated on predetermined points or surfaces (Figures 12 and 13). the BEM is applied.3 with the sound pressure p and wave number k ¼ v/c.10E+02 1.60E+01 8. 2006) and the velocity vector ~ serving as the v excitation value.95E-03 1.53E-03 3.89E-03 3.04E+02 1. After further calculations the following equation system results: ð17Þ H ·p ¼ G·v: _ _ ! 738 H and G are system matrices (Schlensok et al. Therefore.90E+01 8.50E+01 9. Sound pressure distribution on the analysis sphere sound particale velocity [m/s] 6. v is the angular frequency and c the sound velocity.07E+02 1.18E-03 4. For the use of this method a third. Since. A numerical solution of equation (17) results in the sound pressure p. As further quantities acoustic power (Kollmann.80E+01 9. sound pressure [dB] 1. Here. The results are available for discrete frequencies. Sound particle velocity ﬁeld on the analysis sphere 1.24E-03 2. acoustic model of the electrical machine is needed (additional to the electromagnetic and mechanical model)..60E-03 Figure 13.82E-03 5.01E+02 9. 2000) and sound intensity of the machine are calculated.63E-04 1.00E+01 Figure 12. This model consists of the outer surface mesh only which represents the noise radiating area of the motor.31E-03 6. there is no solution for the air volume surrounding the machine.

Therefore. cast iron Electromagnetically excited audible noise 739 Figure 14. The aluminium housing produces more acoustic noise than the cast iron housing. The presented methods and simulation tools allow the analysis and evaluation of every type of energy converter with respect to its electromagnetic. The simulated sound pressure and sound particle velocity are shown in Figure 11. In the given example the noise radiation is also affected by the current waveform. The results agree good to sound measurements. 5. The analysis also provides the possibility of detecting manufacturing faults in electrical machines. At f ¼ 4. structure-dynamical and acoustic behaviour. This is due to the higher material density of cast iron and the difference of the Young’s modulus. By means of measurement devices it is also possible to verify the simulation results. Conclusions Numerical simulations allow for the consideration of the structure-dynamical and acoustic behaviour of an electrical machine a-priori during the phase of design. validation and optimisation of electrical machines. The proceeding of the numerical simulation is explained based on the considered machines (IM and SRM). Two current waveforms and two housing versions (aluminium and grey cast iron) are calculated for a selected operating condition. The direction of noise radiation and acoustic power are determined by these quantities. Acoustic power for the studied operational condition of the SRM . aluminium current 1. the simulated results show good accordance to measurement results. 80 70 60 accoustic power in dB 50 40 30 20 10 0 1200 1400 1600 1800 2000 2200 frequency in Hz 2400 2600 2800 3000 current 1. cast iron current 2.Here.200 to 3. On the one hand. the numerical acoustic simulation is an essential tool for the design. aluminium current 2. Waveform 2 lowers the radiated noise. on the other hand the high optimisation potential concerning vibrations and noise of the analysed machines is presented.400 Hz the SRM radiates the most acoustic noise in axial direction. Figure 14 shows the results of both housing versions in the frequency range from 1. the evaluation of the acoustic simulation is presented for an SRM.000 Hz.

paper presented at 1st International Seminar on Vibrations and Acoustic Noise of Electric Machinery. T. 3. pp. Sadowski. Piriou. ““Comparison between two approaches to model induction machines with skewed slots”. (2000).A. N. W. Maschinenakustik. pp. 2265-73. (1998).J. Vol. J. Vol. “A time-stepped 2D-3D ﬁnite element method for induction motors with skewed slots modelling”. IEEE Transactions on Magnetics. Vol. (2004).... 35 No. Springer-Verlag. Berlin. 9-23.. Berlin. Boualem. 4. “Electrical machine analysis considering ﬁeld – circuit – movement and skewing effects”. The Boundary Element Method for Engineers. 818-22. available at: www. Belmans. COMPEL.. “Multi-slice FE modelling of electrical machines with skewed slots – the skew discretization error”. J. IEEE Transactions on Magnetics. Ducreux. Kuo-Peng. (2002). S. C. B. T. de Oliveira. pp. Monzel. G.L. Bethune. C. “Comparison of sliding-surface and moving-band techniques in frequency-domain ﬁnite-element models of rotating machines”. and Fu. F.. Gerauscharme Elektromotoren. “On the use of the new edge based A-A. Kaehler.J. COMPEL. W. K.. De Gersem. stationary and transient eddy current ﬁeld problems”. P. C.N. K-J. 1598-601. P. and Thomas. Topics in Boundary Element Research. Dular. 36 No. pp. 23 No. P. Katholieke University. van Riesen. Springer-Verlag. “Computational and experimental analysis of electric machine vibrations caused by magnetic forces and magnetostriction”. 3202-5. Kost. Boundary Element Techniques. Springer-Verlag. Springer-Verlag. P. Telles. and Wrobel. C. “Impact of inverter supply and numerical calculation techniques in audible noise problems”. Giradet. Brebbia. R. IEEE Transactions on Magnetics. Vol. (1978). (2006). pp. (1997). (1950). Vol. F. pp. T formulation for the calculation of time-harmonic.M. Gyselinck. pp.. and Henneberger. Arians. 5. Berlin. Springer-Verlag. Delaere. IEEE Transactions on Magnetics. Ho. Introduction to Boundary Elements Method. G. (1999). K. (2000)... J. PhD-thesis. P. ¨ Jordan. 33 No.. Bauer. 34 No. 36 No. 3. IEEE Transactions on Magnetics. and Dular. (2000). Hartmann. Piriou. P. Ducreux. 3. Kollmann. and Henneberger.C. F. Numerische Methoden in der Berechnung elektromagnetischer Felder. Boualem. “Three-dimensional calculation and optimization of the acoustic ﬁeld of an induction furnace caused by electromagnetic forces”. Vol. Bauer. T. Vol. pp. “Numerical models for rotor cage induction machines using ﬁnite element method”. P. A. H. J-P. IEEE Transactions on Magnetics. B. J. pp. (1994). (Ed.. and Schlensok. K.3 740 References Albertz. G. Essen.C. . Springer-Verlag. (2002). 3. (1984). Finite Elemente Methoden. P. pp. and Weiland. F. Vol. IEEE Transactions on Magnetics. Berlin. Vandevelde. L. 1453-7. Berlin...COMPEL 26. and Hameyer. H. (2001). 1080-91. Bathe. 23 No. 37 No. D. 4. 35 No. A. Berlin. Dziwniel. 3233-7. Verlag W. Pentech Press.) (1984). Leuven.A. “Innovative modern object-oriented solver environment – iMOOSE”. F. 5. C.G. D. Hameyer. Dziwniel. and Thomas.de. imoose. “A comprehensive approach to the solution of direct-coupled multi slice model of skewed rotor induction motors using time-stepping eddy-current ﬁnite element method”. Brebbia. Gyselinck. Mai. London. (1989)... L. Brebbia. and Piriou. Antunes. (1986).. C. and Melkebeek. 1007-14. Vol. VANEM.. 1262-5. (2004). J-P. (1999). 3.

and Smith. van Riesen. and Henneberger. 26 No. 2. W. G. Vol. and Witczak. pp. Compumag. 57-61. 26 No. 32 No. and Henneberger. Rio de Janeiro.. C. pp. J. W. 1096-100. Henneberger. and Szulakowski. C. pp. IEEE Transactions on Magnetics. VANEM. G. T. Kubiak. (1999). Vol. 1009-15.. Springer-Verlag. 2005 he has worked as a researcher at the Institute of Electrical Machines (IEM) at RWTH Aachen University. paper presented at 1st International Seminar on Vibrations and Acoustic Noise of Electric Machinery.de ¨ B. and Henneberger. From 2001 to 2006 he has been researcher at the Institute of Electrical Machines (IEM) at RWTH Aachen University. COMPEL. 1390-3. pp. Piriou. “A model for coupled magnetic-electric circuits in electric machines with skewed slots”. Schlensok. Lodz. Signalubertragung. O. paper presented at 3rd International Seminar on Vibrations and Acoustic Noise of Electric Machinery. Schlensok is the corresponding author and can be contacted at: Christoph.H. pp. 2. IEEE Transactions on Magnetics. 71-5. A. P. pp. Schlensok received the MSc degree in electrical engineering in 2000 from the Faculty of Electrical Engineering and Information Technology at RWTH Aachen University. New York. “iMOOSE – an open-source environment for ﬁnite-element calculations”. Zienkiewicz. 1677-81. 3.. Kaehler. and Razek. G. Lim. 2. IEEE Transactions on Magnetics. 2. Berlin. T. I. 40 No. S. Mai. Since. G. and Volschenk. and Henneberger. COMPEL. C.D. C. (1997). The Finite Element Method. G. 3. P.J. ¨ Ramesohl. C. pp.. (1998). and Hadrys. Ramesohl. 1685-8. P. Kuppers. (2004). VANEM. “Magnetostriction vibration of transformer core”. Elsevier. (2000). “Transient analysis of cage-induction motors using ﬁnite elements”. Vol. A. W. 1070-9. Vol. Schlensok. and Henneberger. G.. In 2005. (1989). 941-4. Bethune. and Taylor. 36 No. van Riesen.Kubiak. “Acoustic simulation of an induction machine with squirrel-cage rotor”.F. 25 No. 5. D. Noise and Vibration of Electrical Machines. W. L. I. (1989). Kuest.L. S. Mlotkowski. paper presented at 11th Conference on the Computation of Electromagnetic Fields. Williamson.Schlensok@iem. Witczak. “Calculations of local magnetic forces in electric machinery”. Vol. (2004). Vol. Williamson. pp. dynamical and static-dynamical eccentricity in induction machines with squirrel cages using 2D-transient FEM”. NY.Vol.. 1. Vol. S. Monzel. (2006).. C. “Three dimensional calculation of magnetic forces and displacements of a claw-pole generator”. C. 23 No. (1990). IEEE Transactions on Magnetics.. pp. Flack. “Object-oriented design of ﬁnite element calculations with respect to coupled problems”. T. H. pp. Oxford. “Comparison of static. 31 No. (1996). At the beginning of 2007 he moved to industry and is now a System Engineer at Bosch Rexroth AG in Lohr am Main. “Representation of skew in time-stepped two-dimensional ﬁnite-element models of electrical machines”. 475-86. IEEE Transactions on Magnetics. A. (1995). Electromagnetically excited audible noise 741 . he obtained the PhD degree at the Faculty of Electrical Engineering and Information Technology at RWTH Aachen University with a thesis on numeric simulation and optimisation of induction machines. McGraw-Hill.. Timar. ¨ Schlensok. (2002). Schmulling received the MSc degree in electrical engineering in 2005 as Engineer from the Faculty of Electrical Engineering and Information Technology at the University of Dortmund. 4. “Automatic structural material identiﬁcation of electrical machines using the threshold accepting procedure”. IEEE Transactions on Magnetics.. (1990). D. F. ¨ ¨ Luke. About the authors C. Vol..2.rwth-aachen.

His research interests are numerical ﬁeld computation. van der Giet received the MSc degree in electrical engineering in 2004 as Engineer from the Faculty of Electrical Engineering and Information Technology at RWTH Aachen University. Professor Hameyer is the Director of the “Institute of Electrical Machines” and holder of the chair “Electromagnetic Energy Conversion” of the RWTH Aachen University in Germany. To purchase reprints of this article please e-mail: reprints@emeraldinsight.com Or visit our web site for further details: www.3 742 M. Berlin.com/reprints . Since. Germany. the design of electrical machines.emeraldinsight. 2004 he has worked as a researcher at the Institute of Electrical Machines (IEM) at RWTH Aachen University.COMPEL 26. Hameyer received the MSc degree in electrical engineering from the University of Hannover. In 1988 he became a member of the staff at the University of Technology. as a Design Engineer for permanent magnet servo motors and board net components. Germany. induction machines and numerical optimisation strategies. in particular permanent magnet excited machines. He received the PhD degree from University of Technology Berlin. K. After his university studies he worked with the Robert Bosch GmbH in Stuttgart. Germany. Until February 2004 Professor Hameyer was a full Professor for Numerical Field Computations and Electrical Machines with the KU Leuven in Belgium. Currently. Germany.

- Articol Model IM
- 2510-ff-011541
- Washing Machine Motor Controller
- Artificial Killing Machine
- Synchronous Motor
- 0014-ff-004723
- StateMachinesInVerilog
- Forming Machine Fm 250
- Turning
- Fitz_ch05
- AC Motor Diagrams - Basic Stator and Rotor Operation
- TechEd Sample
- Screw Machinist
- Ground Cleaning Report
- 3133-40 usuario
- Indian Electricals Limited
- ZAAZ 20K User Manual1
- Duomatic En
- Multi Material Stereot
- CoCom Lists - 1985
- A Simple Vending Machine
- Applied Mathematical Modelling Volume 34 Issue 4 2010 [Doi 10.1016%2Fj.apm.2009.07.019] v. Deljoo; S.M.J. Mirzapour Al-e-hashem; F. Deljoo; M.B. Aryanez -- Using Genetic Algorithm to Solve Dynamic Cel
- synopsi
- Vibro-Hydro Brick Block-Semi Auto
- mikrotron_RUMUL
- Manual Xerox 252
- RTGTR.PDF0131
- CAD Mechanical Product Designer in Denver CO Resume Loren McGilvrey
- Space Phasor Theory and Control of Multiphase Machines
- Mill 55

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.

scribd