Trends in Dimensioning PM and Reluctance Machines

Transcription

1 17 Trends in Dimensioning PM and Reluctance Machines Trends in Dimensioning PM and Reluctance Machines Tim Miller FEMAG Anwendertreffen Oktober Retrospeed 1

2 Dimensions Size + Shape + Drawing Dimensions + Control Parameters How big? Or rather, Shapes Features Details Diameters Lengths and widths Angles Voltage Current Phase angles How small? Priority, hierarchy, dependency, weighting Refinement Optimization Materials 2015 Retrospeed 2

3 Motor chart PM DC WF AC IPM Synchronous Reluctance Spoke Universal Induction SPM Switched Reluctance Classical Electronically controlled AC drives 2015 Retrospeed 3

4 Synchronous reluctance, PM-assisted synchronous reluctance, IPM machines 2015 Retrospeed 4

5 Synchronous reluctance motor in smartfem Things to discuss Dimensioning principles Historical review Some comments about Time-simulation Performance calculations We don't need FEA to determine the number of poles! 2015 Retrospeed 5

6 Brushless synchronous reluctance / PM motor rotor lamination According to the theory We must maximize the saliency d What the electric machine designer sees: A 4-pole rotor Paths of easy magnetization defining d-axis Relatively low flux per pole Saturates very easily Delicate mechanical configuration What we need Finite-element analysis Mechanical and electromagnetic Time simulation (for controller design) A way to evolve an optimum design Tooling for manufacture d 2015 Retrospeed 6

7 Saturation of Ld and Lq due to the currents in their own axes We often see these curves Retrospeed 7

8 Saturation and cross-saturation... but!! It can be an advantage to use flux-linkage, and avoid Inductance completely. See Vagati et al [1997,1998] Calculate the mean electromagnetic torque If current is known, calculate the voltage If voltage is known, calculate the current 2015 Retrospeed 8

9 Energy-conversion diagram for machine with sinusoidal current & flux-linkage Flux is limited by cross-sectional dimensions of magnetic core, and by saturation Current is limited by cross-sectional dimensions of winding, and by temperature rise Flux or flux-linkage is useful in sizing, but Inductance is not very helpful in the sizing process. Magnetization curves 2015 Retrospeed 9

10 Relationship between energy-conversion diagram and dimensions (SR motor) Y Flux is limited by cross-sectional dimensions of magnetic core, and by saturation Aligned inductance depends on airgap length and pole arc dimensions again Current is again limited by cross-sectional dimensions of winding, and by temperature rise Unaligned inductance depends on space around wound poles in unaligned position dimensions again 2015 Retrospeed 10

11 Features of the simple energy-conversion diagram It can be used for preliminary design or sizing It is directly related to the machine dimensions and material properties It requires only a magnetostatic calculation No time-stepping simulation required Current waveform is assumed known Suitable for 2-D FEA, but end-effects must be added It defines the maximum torque per ampere The required voltage can be deduced from it It shows several figures of merit It can easily be scaled according to the number of turns Pre-calculated sets of magnetization curves can be used for dynamic simulation (see next slide) 2015 Retrospeed 11

12 Time-stepping simulation (one phase only) Voltage equation (1 phase only) Stepwise integration... Simple enough by Euler or Runge-Kutta Numerical model of flux-linkage curves is much more complex and we need the inverse Similar numerical approximation is required to calculate the torque Retrospeed 12

13 Electric and magnetic loading Torque / Rotor Volume TRV is proportional to the electric and magnetic loadings. This is an engineering formulation of "torque = flux current". Electric loading A common design parameter (How many amps (really, ampere-turns) can we pack into the stator?) Magnetic loading A common design parameter Average flux-density Over stator bore circumference (How much flux can we pack into the stator?) In AC machines, we need this relationship because the flux is sinusoidally distributed and the induced voltage depends on the fundamental, not the average Retrospeed 13

14 Internal power factor and type of rotor In general the flux and the ampere-conductor distribution are not orthogonal. There is an angle between them, which can be regarded as an internal power-factor angle. It gives rise to an internal power-factor IPF. The value of IPF depends on the type of rotor Surface PM motor : IPF can be as high as 1 Interior PM motor : 0 8 Induction motor : 0 8 Synchronous reluctance motor : 0 6 We should de-rate the initial design of the motor by this factor Or for the same power output, it will need a bigger inverter than the PM motors or the induction motor 2015 Retrospeed 14

15 Machine configurations; historical progress Cruickshank, Anderson, Menzies 1971 Cruickshank A.J.O., Anderson A.F. and Menzies R.W. Theory and performance of reluctance motors with axially laminated anisotropic rotors Proceedings IEE, Vol. 118, No. 7, July 1971, pp Retrospeed 15

16 Lawrenson 1965; segmental rotor (line-start) 2015 Retrospeed 16

17 Essential flux-barriers (Fong and Htsui [1970]) 2015 Retrospeed 17

18 Essential flux-barriers (Fong and Htsui [1970]) 2015 Retrospeed 18

19 SynchroTek 1984 Bak D.J. Rotor Design Minimizes Magnetic Leakage Design News, May 21, 1984, pp (A brief review of the Synchroloc motor of Bogue Mfg. Co., developed by Shekar Rao.) Rao S.C. Dynamic performance of reluctance motors with magnetically anisotropic rotors IEEE Transactions, Vol. PAS- 95, No. 4, July-August 1976, pp Retrospeed 19

20 El-Antably and Hudson 1985 : 6-pole axially-laminated rotor El-Antably A. and Hudson T.L. The Design and Steady-State Performance of a High-Efficiency Reluctance Motor IEEE IAS, 1985, pp Retrospeed 20

21 Mayer and Weh 1986 ; Boldea 1994 : 2-pole axially-laminated rotor 21/16 Saliency ratio, no-load/full-load Mayer R., Mosebach H., Schröder U. and Weh H. Inverter-Fed Multiphase Reluctance Machine with Reduced Armature Reaction and Improved Power Density ICEM 1986, Pt. III, pp Boldea I., Fu Z.X. and Nasar S.A. High Performance Reluctance Generator IEE Proceedings-B, Vol. 140, No. 2, March 1993, pp Retrospeed 21

22 Soong 1995 : 4-pole axially-laminated rotor Soong W.L., Staton D.A. and Miller T.J.E. Design of a new axially-laminated interior permanent-magnet motor. IEEE Transactions on Industry Applications, Vol. 31, No.2, March/April 1995, pp Retrospeed 22

23 Fratta, Vagati, et al 1993 Fratta, Vagati et al 1993 : Effects of positions of ends of flux-barriers in generating flux and torque pulsations (including harmonic leakage inductance) Oscillation between Figs. 2 and 3 causes flux pulsation which causes torque ripple, core loss, and harmonic leakage Analytical model Flaring and fairing It seems Vagati et al established the viability of the transverse-laminated synchronous reluctance motor. After this, the axially laminated concept faded away. Fratta A., Troglia G.P., Vagati A. and Villata F. Evaluation of Torque Ripple in High Performance Synchronous Reluctance Machines IEEE IAS 1993 pp Retrospeed 23

24 Okuma 1998 Okuma synchronous reluctance motor, Nikkei Mechanical, No. 523, pp Retrospeed 24

25 ABB ABB 2013 A promising technology for efficient fan systems is the synchronous reluctance motor, which has a winding-free rotor, thus avoiding rotor losses and keeping the motor relatively cool. Shaped flux-barriers in transverse lamination, (and not too many of them) Vagati et al, 1995 and earlier Vagati et al, 1997 Drives and Controls, July/August Retrospeed 25

26 ABB Retrospeed 26

27 Stator : typical integral-kw WEG 2015 Retrospeed 27

28 Vagati's rule n r = ns ± 4 Vagati US Patent 5,818,140 Oct Retrospeed 28

29 Magnetic equivalent circuit The R-wave needs to follow the F- wave as closely as possible, to minimize the introduction of spurious harmonics. See Vagati et al [1997,1998] This also leads to the principle of equal permeances for the fluxbarriers Retrospeed 29

30 Synchronous reluctance motor with different flux-barrier angles 12th-order torque harmonic from poles 1 & 3 is 180 out of phase with the 12th-order torque harmonic from poles 2 & 4. Angles A, B chosen to cancel a given torque harmonic, e.g. the 24th. Angles a,b also chosen to cancel the same torque harmonic. Poles 1&3 (AB) and 2&4 (ab) combine to compensate the 12th-order torque harmonic. Resulting torque ripple < 5% (peak-to-peak). This is about 1/3 1/2 the value in a similar IPM designed as a "control" experiment. An analytical formulation of the torque ripple is also presented and tested. Bianchi's "Machaon" rotor combines two different pole geometries in a single lamination. Bianchi N., Bolognani S., Bon D. and Dai Pré M. Rotor Flux-Barrier Design for Torque Ripple Reduction in Synchronous Reluctance and PM-Assisted Synchronous Reluctance Motors IEEE Transactions on Industry Applications, Vol. 45, No. 3, May/June 2009, pp He also presented a rotor with 2 symmetrical halves having different pole geometries ("R & J" type) Retrospeed 30

31 Pellegrino's dimensioning for design optimization Simple flux-barrier shapes minimize the number of dimensions Achieved similar torque to that of a motor optimized with 20 parameters, but reduced the torque ripple significantly No. of fluxbarriers No. of dimensions Pellegrino G., Cupertino F. and Gerada C., Automatic Design of Synchronous Reluctance Motors focusing on Barrier Shape Optimization, Transactions IEEE, Industry Applications, Vol. 51, No. 2, March/April 2015, pp Retrospeed 31

32 Basic dimensioning for parameterized CAD No. of fluxbarriers No. of dimensions Even this is not enough: Many details are missing, and the poles may not all be identical Retrospeed 32

33 Detail : end of flux-barrier 2015 Retrospeed 33

34 Systematic design of rotor (a) (b) (c) (d) (e) (f) 2015 Retrospeed 34

35 It would seem that single-tooth windings are not appropriate for synchronous reluctance motors, because of the high harmonic content in the stator ampere-conductor distribution. Synchronous reluctance motor with single-tooth windings? 6 stator slots? However, this technology is well established and automated, so we may see attempts in future to try to develop the synchronous reluctance motor with such windings. An obvious problem is the winding factor: Grundfos PM motor from the Green Book, used here only to show an example of singletooth windings. It is not otherwise related to this presentation. 2-pole pole Retrospeed 35

36 Synchronous reluctance motor with single-tooth windings Reported by Mecrow: Cruciform rotor has poor saliency Single-tooth winding has high space-harmonic content Shown with the d-axis aligned with phase C, at the instant of peak current in phase C. Unsaturated saliency ratio 4 4 Power-factor 0 51 Torque ripple 44% Efficiency 91% (but uses segmented core with 58% slot fill) Again we see high copper content in the stator, required to allow the high electric loading needed to compete with the induction motor. See Spargo et al for details of the flux-barrier geometry Spargo C.M., Mecrow B.C., Widmer J.D., Morton C. and Baker N.J., Design and Validation of a Synchronous Reluctance Motor with Single Tooth Windings, IEEE Trans. on Energy Conversion, Vol. 30, No. 2, June 2015, pp Rotor concepts "outside the box" of conventional practice Retrospeed 36

37 Thank you! Thank you! Most of the ideas in this presentation are described in detail in this book: "Reluctance Machines", 2015, 185 pages Switched and synchronous reluctance machines, plus IPM and PM-assisted synchronous reluctance machines and certain flux-switching machines and variants. Extensive bibliographies are included for both types of machine, with full references to the published works mentioned in this presentation Retrospeed 37