Contactless Energy Transfer to a Moving Actuator - CiteSeerX

1 downloads 0 Views 764KB Size Report
provides long-stroke contactless energy transfer capability in a plane and a ... energy transfer system is proposed for energy transfer to a planar ..... RMS Value.
1

Contactless Energy Transfer to a Moving Actuator Jeroen de Boeij, Student Member, IEEE, Elena Lomonova, Jorge Duarte Member, IEEE and Andr´e Vandenput, Senior Member, IEEE

Abstract— In this paper a new topology for contactless energy transfer is proposed and tested that can transfer energy to a moving actuator using inductive coupling. The proposed topology provides long-stroke contactless energy transfer capability in a plane and a short-stroke movement of a few millimeters perpendicular to the plane. In addition, it is tolerant to small rotations. The experimental setup consists of a platform with one secondary coil, which is attached to a linear actuator and a 3-phase brushless electromotor. Underneath the platform is an array of primary coils, that are each connected to a halfbridge square wave power supply. The energy transfer to the electromotor is measured while the platform is moved over the array of primary coils by the linear actuator. The secondary coil moves with a stroke of 18 cm at speeds over 1 m/s, while up to 33 W power is transferred with 90% efficiency. Index Terms— Contactless energy transfer, inductive coupling, moving load.

I. I NTRODUCTION Most high-precision machines are positioning stages with multiple degrees of freedom (DOF), which often consist of cascaded long- and short-stroke linear actuators that are supported by mechanical or air bearings. Usually, the long stroke actuator has a micrometer accuracy, while the submicron accuracy is achieved by the short-stroke actuator. To build a high-precision machine, as much disturbances as possible should be eliminated. Common sources of disturbances are vibrations, Coulomb and viscous friction in bearings, crosstalk of multiple cascaded actuators and cable slabs. A possibility to increase throughput, while maintaining accuracy is to use parallel processing, i.e. movement and positioning in parallel with inspection, calibration, assembling, scanning, etc. To meet the design requirements of high accuracy while improving performance, a new design approach is necessary, especially if vacuum operation is considered, which will be required for the next generation of lithography machines. A lot of disturbance sources can be eliminated by integrating the cascaded long- and short-stroke actuator into one actuator system. Since most long-stroke movements are in a plane, this can be done by a contactless planar actuator. A contactless planar actuator or planar motor is supported by magnetic bearings that levitate the actuator platform, while controlling all six DOF of the platform. Long-stroke linear movement in 2D is also provided by the magnetic bearing This research is sponsored by SenterNovem. SenterNovem is an agency of the Dutch Ministry of Economical Affairs. J. de Boeij is with the Department of Electrical Engineering of the Eindhoven University of Technology, Eindhoven, The Netherlands; (email:[email protected])

while small translations in height and small rotations remain possible. Magnetic bearings can also operate in vacuum. Parallel processing requires power on the platform to drive the actuators on the platform. In order to remove as much disturbances as possible, the power transfer needs to be contactless, i.e. without wires from the ground to the platform. A coil topology and geometry for a contactless energy transfer system is proposed for energy transfer to a planar moving platform. The platform is equipped with permanent magnets and is levitated and propelled by a matrix of coils, which are fixed to the ground. Such a planar actuator is currently under investigation at Eindhoven University of Technology [13]. The aim of this research project is to transfer energy to the moving platform continuously and at every position in order to enhance the functionality of the platform, while maintaining the advantages of operating without contact and cables slabs. When energy is transferred to a moving load (i.e. an electromotor) by means of inductive coupling, one has to deal with a changing coupling between a primary and secondary coil. The change in coupling results in a different characteristic of the energy transfer capability of the system as is shown in [1], [2], [3], [4] and [5]. Most of these systems can only transfer energy at certain positions ([1], [2]) or they suffer from large changes in power transfer capability throughout the range due to the changing coupling ([4], [5]). Another solution for transferring energy to a moving load is using elongated primary coils in combination with elongated cores [7]. This results in a stable energy transfer but the stroke is limited by the size of the primary coil. If long strokes are required, the system becomes heavy and bulky. The topology proposed and tested in this paper provides longstroke contactless energy transfer (CET) in a plane with only small changes in power transfer capability. II. CET T OPOLOGY The design of the primary and secondary coil is optimized to get a coupling that is as constant as possible for a sufficiently large area. This area should be large enough to allow the secondary coil to move from one primary coil to the next one without a large reduction in coupling. If this can be achieved, the power can be transferred by one primary coil that is closest to the secondary coil. When the secondary coil moves out of range the first primary coil is turned off and the next one will be energized. To ensure a smooth energy transfer to the moving load, the position dependence of the coupling should be minimized, while keeping the coupling high enough to get a high-efficiency energy transfer.

2

r2 r1 c h c t

Fig. 1.

c w c w

Coil dimensions TABLE I D IMENSIONS OF PRIMARY AND SECONDARY COIL Parameter cw ct ch r1 r2

Primary coil Value Dimension 60.0 mm 10.0 mm 10.0 mm 1.0 mm 11.0 mm

Secondary coil Value Dimension 130.0 mm 30.0 mm 2.0 mm 1.0 mm 31.0 mm Fig. 3.

Secondary coil above a matrix of nine primary coils

Surface fitted through measurements

0.3 0.28 0.26 0.24 0.02 0 −0.02 y [m]

Fig. 4.

Surface fitted through FEM simulation

0.32 Coupling [−]

0.32 Coupling [−]

A lot of systems use 2D spiral coils for the primary and secondary coil, since the spiral coil geometry allows relatively high coupling (up to 60 %) and some tolerance for misalignment of the coils [8], [9]. However, to allow the secondary coil to move from one primary coil to the next, the tolerance for misalignments should be increased. In the proposed system this is done by using a 3D geometry for the primary coil. This results in a fairly constant B-field around the primary coil, which accommodates good coupling in a large area. Furthermore, since the system is supposed to transfer power to a load moving in a plane, it is convenient to use a shape that is symmetrical in 2D for both the primary coil and the secondary coil: a square for instance. The geometry of the primary and the secondary coils are optimized with FEM using Maxwell 3D 10 Optimetrics. The resulting geometry of the coils is shown in Fig. 1 and 2 and the dimensions are listed in Table I. More information on the considerations for the topology can be found in [10]. The drawing in Fig. 3 shows one secondary coil above nine primary coils. The black square shows the area in which the center of the secondary coil can move while maintaining good coupling with the middle primary coil. The secondary coil is situated in the bottom-left corner of the area of interaction with

0 −0.02 x [m]

0.02

0.3 0.28 0.26 0.24 0.02 0 −0.02 y [m]

0 −0.02 x [m]

0.02

Coupling between primary and secondary coil

the middle primary coil. The coupling between the primary coil and the secondary coil within that area is calculated with Maxwell 3D 10 Optimetrics and measured. The results are shown in Fig. 4, which show that the FEM predictions are very close to the measured values. The coupling k is fairly constant within most of the area, only on the outer edges it drops fast. However, the ripple defined by Eq. 1 is 25%, which is quite small considering the large displacement of the secondary coil: ripple =

max(k) − min(k) · 100% max(k)

(1)

Although this system is designed with square shaped coils, it is also possible to design a system with similar characteristics using rectangular coils. III. S TEADY-S TATE E LECTRIC C IRCUIT A NALYSIS

Fig. 2.

Primary and secondary coil

Since the system will be used in a maglev application based on repulsive forces between coils and permanent magnets, the use of iron or ferrites is prohibited. In addition, the use of cores will limit the stroke of the system. Therefore, a coreless or aircore inductive coupling is used to transfer the energy. To keep the efficiency of an aircore inductive coupling high a resonant capacitor is used for both the

3

chosen: C V

+

Fig. 5.

1

R 1

I -

R 1

L 1

L k

1

I 2

2

R 2

L

V 1

+

R 1

I -

Z

1

1

L 1

V 1

+

I -

1

Z 1

R

a

C2

=

C1

=

1 ω02 L2 1 2 ω0 L1

(5) (6)

This choice of the resonant capacitors ensures that the impedance of the secondary circuit Z2 , the reflected impedance of the secondary circuit to the primary circuit ZR and the impedance seen by the power supply Z1 are purely resistive at ω = ω0 : 1 Z2 = R2 + RL + j(ω0 L2 − ) ω0 C2 = R2 + RL (7) ω02 M 2 ω02 k 2 L1 L2 = (8) ZR = Z2 R2 + RL 1 ) + ZR Z1 = R1 + j(ω0 L1 − ω0 C1 ω 2 k 2 L1 L2 = R1 + 0 (9) R2 + RL By substituting Eq. 5 into Eq. 4 for ω = ω0 the relation between I1 and I2 becomes:

Electric circuit of contactless energy transfer system

C

Fig. 6.

C 2

b

Simplified electric circuit of contactless energy transfer system

I2 K

= KI1 ω0 M = j R2 + RL

(10) (11)

primary and the secondary coil. Moreover, due to the position dependent coupling, a series resonant capacitor is used for both coils to ensure that the resonant frequency of the circuit does not depend on the coupling as discussed in [6].

where |K| is a gain relating the current I2 to I1 . Eq. 11 shows that I2 leads I1 by 90 degrees. Now the power transferred to the load Pout , the power supplied by the power supply Pin and the efficiency of the total system η are calculated:

The electric circuit of the CET system is shown in Fig. 5, where V1 is the RMS voltage of the power supply, I1 the RMS current supplied by the power supply, I2 the RMS current induced in the secondary circuit. C1 and C2 are the series resonant capacitors in the primary and secondary circuit, respectively, R1 the resistance of the primary coil, R2 is the resistance of the secondary coil, L1 and L2 are the self inductance of the primary and secondary coil, respectively, k is the inductive coupling factor between the primary and secondary coil and RL is the resistance of the load. The load RL represents the rectifier and additional power electronics. Simplified versions of the circuit are shown in Fig. 6 a and b, where ZR is the reflected load of the secondary circuit on the primary circuit and Z1 is the load seen by the power supply.

= I12 Z1 (12) 2 2 = |K| I1 RL (13) Pout |K|2 RL η = = (14) Pin Z1 From Eq. 5 and 6 it is clear that the resonant frequency of the circuit does not depend on the coupling, since the choice of the resonant capacitors only depends on the inductance of the coils. In reality the two series resonant capacitors will not cancel the inductances of the primary and secondary coil completely. Therefore, the load seen by the power supply will not be purely resistive. The load seen by the power supply Z1 does depend on the coupling. This implies that the power transfer capability of the system depends on the coupling as well.

The equations for this system with an AC voltage source with angular frequency ω [rad/s] are: M

 = k L1 L2

(2)

V1

= R1 I1 + j(ωL1 −

(3)

=

(4)

jωM I1

1 )I1 − jωM I2 ωC1 1 (R2 + RL )I2 + j(ωL2 − )I2 ωC2

where M is the mutual inductance of the primary and sec0 ondary coil. To obtain the desired resonant frequency f0 = w 2π [Hz], the resonant capacitors C2 and C1 must be suitably

Pin Pout

IV. E XPERIMENTAL S ETUP An experimental setup was built to test the CET design, which consists of an array of three stationary primary coils that are fixed in a row on top of a ceramic structure. The ceramic structure is used to allow heat from the coils to be conducted to the iron base frame and at the same time to prevent eddy current losses in the iron base frame. The primary coils are made of litz wire. Each bundle of litz wire consists of 60 strands of 71 µm and the strands are wrapped together with a layer of cotton. The strand size has

4

V DC bus

VB HO VS

IN __ SD

VCC LO COM

VCC L O A D

IR 2104

Ground

Fig. 7.

Schematic of the primary coil power supply

TABLE II E LECTRIC CIRCUIT PARAMETERS OF THE THREE PRIMARY CIRCUITS IN THE EXPERIMENTAL CET SETUP Inductance Resistance Capacitance Resonance

Primary Circuit 1 970 µH 4.6 Ω 0.72 nF 191 kHz

Primary Circuit 2 937 µH 4.5 Ω 0.74 nF 190.5 kHz

Primary Circuit 3 864 µH 4.1 Ω 0.81 nF 190 kHz

been chosen after examining the AC losses using the method described in [12]. The turns of the coil are fixed by glue that has been applied during the winding process. Approximately 120 turns fitted in the cross-section, resulting in a 0.3 filling factor. Each primary coil is connected in series with a resonant capacitor. Each resonant circuit is driven by a separate halfbridge power supply, that applies a square wave voltage of 191 kHz over the resonant circuit. The schematic of the halfbridge power supply is shown in Fig. 7. An overview of the primary coils and the corresponding series capacitors is shown in Table II. The secondary coil is fixed onto a ceramic plate that is bolted to the mover of a linear actuator. Again ceramic material is used for heat conduction and the minimization of eddy current losses. The linear actuator can move the secondary coil over the three primary coils. The position of the secondary coil with respect to the array of primary coils is measured by the encoder of the linear actuator. A picture of the experimental setup is shown in Fig. 8.

Fig. 9.

CD electrical drive and rectifier connected to secondary coil

The secondary coil is connected in series with a resonant capacitor. The circuit is then connected to a full-bridge diode rectifier to generate a DC output. The DC output of the rectifier is connected to the load, which is an electromotor of a CD drive running at 12 VDC as shown in Fig. 9 All subsystems are connected to a ds1103 dSpace system running the control program at 8 kHz. This way the DC bus voltage of the primary coil power supplies is controlled as well as which of the primary coil power supplies is enabled. The position of the linear actuator is controlled using a PID controller running on the dSpace system. Depending on the position of the linear actuator the dSpace system enables the primary coil that is completely overlapped by the secondary coil. The primary coil activation is controlled by a multi-port switch. The multi-port switch has four active coil states, state 1 enables the power supply of the first primary coil, state 2 and 3 enable the power supply of the second and third primary coil, respectively. State 4 disables all power supplies and this state is used for switching from one power supply to the next. When the secondary coil moves out of range of primary coil 1 (active coil state 1), the active supply is switched off (active coil state 4) and one sample time later the second supply is switched on (active coil state 2). For one sample time none of the power supplies is active (active coil state 4), which is necessary to allow the triac in the power supply that is switched off (see Fig. 7) to block the circuit after the current in the resonant circuit is damped. There is no other control mechanism in the power electronics, and the system operates without any measurement on the secondary site, except for the position of the secondary coil.

V. R ESULTS

Fig. 8.

Picture of experimental CET setup

An electromotor of a CD drive that runs on 12 VDC is connected to the rectifier. The voltage and current from the DC bus supply as well as the voltage and current to the CD drive are measured and shown in Fig. 10 and 11. The secondary coil is moving over all three primary coils following a sinusoidal position reference, which represents a total displacement of 18 cm (i.e. the amplitude of the sine wave is 9 cm). The frequency of the sinusoidal position reference is 2 Hz, so in one second the secondary coil makes

5

TABLE III VALUES OF VOLTAGES , CURRENTS , POWER AND EFFICIENCY OF THE CET

Power supply voltage Voltage [V]

52

TO A

51 50 49 0

0.2

0.4 0.6 Time [s] Power supply current

0.8

1

0.2

0.4 0.6 Time [s] Active Coil State

0.8

1

0.15 0.1

RMS Value 51.1 76.6 3.84 12.5 276 3.44 89.4

0 0

0.4

4

0.2

3.5

3 Current [A]

Active Coil

4.5 Primary coil 1 Primary coil 2

0.3

4

2 1 0

0.2

0.4 0.6 Time [s]

0.8

1

0.1 0

2

−0.2

1.5

−0.4 0

Measured voltage, current and active primary coil for the DC bus

3 2.5

−0.1

−0.3

Fig. 10. supply

Dimension [V] [mA] [W] [V] [mA] [W] [%]

0.05

Active coil

Current [A]

CD DRIVE ELECTROMOTOR

Variable VDCbus IDCbus Pin Vload Iload Pout η

1 1

2 Time [s]

3

4 −4

x 10

0.5 0

1

2 Time [s]

3

4 −4

x 10

Fig. 12. Plot of current in primary coil 1 and 2 during switching and the active coil command. Voltage [V]

Load voltage 14 12

Current [A]

10 0

0.2

0.4 0.6 Time [s] Load current

0.8

1

0.2

0.4 0.6 Time [s] Active Coil State

0.8

1

0.5

0 0

Active Coil

4

In Fig. 12, the transient behavior is shown when the secondary coil is moving from primary coil 1 to primary coil 2. It is clearly visible that all power supplies are switched off when the active coil state has value 4. There is also some delay between the active coil state switch and the response from the power electronics, which is caused by a slow rising edge of the enable signal and by delay in the power electronics. In Fig. 10 and 11 the switching is also visible in the current waveforms, since no current is drawn from the DC bus supply and no current is available for the electromotor of the CD drive.

3 2 1 0

0.2

0.4 0.6 Time [s]

0.8

1

Fig. 11. Measured voltage, current and active primary coil for the CD drive electromotor load

two cycles (one cycle implies moving from primary coil 1 over primary coil 2 to primary coil 3 and back). The cycle is clearly visible from the Active Coil plot in Fig. 10 and 11, which represents the state of the active coil multi-port switch. The secondary coil reaches a maximum speed of 1.1 m/s over the second primary coil. Due to this speed the secondary coil is in range of the second primary coil for only 60 ms. By calculating the RMS values of the voltages and currents the power from the DC bus supply Pin as well as the power to the CD drive load Pout and the efficiency η according to Eq. 14 can be calculated. This calculation includes losses in the power electronics. The values are listed in Table III.

The ripples visible in the voltage and current waveforms from the DC bus power supply and to the CD drive are related to the changing coupling. However, since the CD drive does not represent a purely resistive load, the ripple is somewhat smoothed by the inductance of the load. This effect is more visible when a purely resistive load will be connected to the system. In addition, the CD drive does not need much power to operate and a resistive load can be operated at higher power levels. Therefore, a 50 Ω resistive load is used at a higher power level. The same trajectory is used for the secondary coil. The measured voltage and current waveforms of the DC bus supply and the load are shown in Fig. 13 and 14 respectively. The RMS values of voltage, current and power as well as the efficiency are shown in Table IV. The variation in coupling is now clearly visible in the current and voltage waveforms of the load. This suggests that the power transfer can be further smoothed by measuring the coupling and changing the voltage of the DC bus supply accordingly, as is discussed in [11]. The results are very similar to the results of the CD drive. Higher power levels have not been tested using the linear actuator, since the capacitors in

6

the resonant circuit cannot operate above 800 V. Operating at higher power requires new capacitors which have not been realized yet. It is expected that power transfer up to 300 W is feasible.

Power supply voltage Voltage [V]

160 150 140 0

0.2

0.4 0.6 Time [s] Power supply current

0.8

1

0.2

0.4 0.6 Time [s] Active Coil State

0.8

1

0.2

0.4 0.6 Time [s]

0.8

1

Current [A]

0.5

0 0

Active Coil

4 3 2 1 0

Fig. 13. Measured voltage, current and active primary coil for the DC bus supply (resistive load)

Voltage [V]

Load voltage 50

Current [A]

0 0

0.2

0.4 0.6 Time [s] Load current

0.8

1

0.2

0.4 0.6 Time [s] Active Coil State

0.8

1

0.2

0.4 0.6 Time [s]

0.8

1

1 0.5 0 0

Active Coil

4 3 2 1 0

Fig. 14. Measured voltage, current and active primary coil for the 50 Ω load

TABLE IV VALUES OF VOLTAGES , CURRENTS , POWER AND EFFICIENCY OF THE CET TO A

Variable Vdcbus Idcbus Pin Vload Iload Pout η

50 Ω RESISTOR

RMS Value 150 239 35.9 40.8 807 32.9 91.8

Dimension [V] [mA] [W] [V] [mA] [W] [%]

VI. C ONCLUSION A new topology for contactless energy transfer (CET) to a moving load has been proposed, built and tested. The CET topology allows for a long-stroke movement in a plane and a short-stroke movement of a few millimeters perpendicular to the plane. In addition, it is tolerant to small rotations. The power electronics consist of a half-bridge square wave power supply for each primary coil and series resonant capacitor and a full-bridge diode rectifier at the load. Power transfer up to 33 W with resistive load of 50 Ω has been demonstrated The CET system was used to power a 3-phase brushless electromotor of a CD drive and showed stable power transfer of 3.44 W. The power was transferred at approximately 90 % efficiency, while the secondary coil was moving with speeds up to 1.1 m/s over the primary coils. R EFERENCES [1] G.A. Covic, G. Elliott, O.H. Stielau, R.M. Green, J.T. Boys, The Design of a Contact-less Energy Transfer System for a People Mover System, Proceedings IEEE International Conference on Power System Technology, PowerCon 2000, Vol 1, December 2000, pp. 79-84. [2] H. Ayano, K. Yamamoto, N. Hino, I. Yamato, Highly-Efficient Contactless Electrical Energy Transmission System, 28th Annual Conference of the IEEE Industrial Electronics Society, IECON 2002, Vol 2, November 2002, pp. 1364-1369. [3] S. Adachi, F. Sato, S. Kikuchi: Considerations of Contactless Power Station with Selective Exitation to Moving Robot, IEEE Transactions on Magnetics, Vol. 35, No. 5, September 1999, pp. 3583-3585. [4] F. Sato, J. Murakami, H. Matsuki, S. Kikuchi, K. Harakawa, T. Satoh: Stable Energy Transmission to Moving Loads Utilizing New CLPS, IEEE Transactions on Magnetics, Vol. 32, No. 5, September 1996, pp. 5034-5036. [5] F. Sato, H. Matsuki, S. Kikuchi, T. Seto, T. Satoh, H. Osada, K. Seki: A New Meander Type Contactless Power Transmission System - Active Excitation with a Characteristics of Coil Shape, IEEE Transaction on Magnetics, Vol. 34, No. 4, July 1998, pp. 2069-2071. [6] C. Wang, G.A. Covic, O.H. Stielau: Power Transfer Capability and Bifurcation Phenomena of Loosely Coupled Inductive Power Transfer Systems, IEEE Trans. on Industrial Electronics, Vol. 51, No. 1, February 2004, pp. 148-157. [7] G.L.M. Jansen: 2-Dimensional Displacement Device, International Patent Application, WO 2005/013464 A1. [8] C. Fernandez, O. Garcia, R. Prieto, J.A. Cobos, S. Gabriels, G. Van der Borght: Design Issues of a Core-less Transformer of a Contact-less Application, 17th Annual IEEE Applied Power Electronics Conference and Exposition, APEC 2002, Vol.1 pp. 339-345 [9] R. Mecke, C. Rathge: High Frequency Resonant Inverter for Contactless Energy Transmission over Large Airgap, 25th Annual IEEE Power Electronics Specialist Conference, 2004, pp 1737-1743. [10] J. de Boeij, E. Lomonova, A.J.A. Vandenput: Contactless Energy Transfer to a Moving Load Part I: Topology and FEM Simulation, International Symposium on Industrial Electronics, Montreal, Canada, July 2006. [11] J. de Boeij, E. Lomonova, J.L. Duarte, A.J.A. Vandenput: Contactless Energy Transfer to a Moving Load Part II: Simulation of Electrical and Mechanical Transient, International Symposium on Industrial Electronics, Montreal, Canada, July 2006. [12] C.R. Sullivan: Computationally Efficient Winding Loss Calculation with Multiple Windings, Arbitrary Waveforms and Two-Dimensional or Three-Dimensional Field Geometry, IEEE Trans. on Power Electronics, Vol. 16, No. 1, January 2001, pp. 142-150. [13] J.W. Jansen, E.A. Lomonova, A.J.A. Vandenput, C.M.M. Van Lierop: Analytical Model of a Magnetically Levitated Linear Actuator, IAS 2005 Annual Meeting, Hong Kong, China, October 2005. [14] Ansoft: Maxwell 10 User’s Guide, Pittsburgh PA, USA.