Safe Wireless Power Transfer to Moving Vehicles: Design of ...gcep.stanford.edu/pdfs/TechReports2013/5.5_Fan... · Safe Wireless Power Transfer to Moving Vehicles: Design of Radiationless

Safe Wireless Power Transfer to Moving Vehicles: Design of Radiationless Antenna

Investigators Prof. Shanhui Fan, Electrical Engineering, Stanford; Dr. Sven Beiker, Center for Automotive Research, Stanford; Dr. Richard Sassoon, Global Climate and Energy Project, Stanford; Dr. Xiaofang Yu, Post-doctoral researcher, Stanford; Dr. Sunil Sandhu, Post-doctoral researcher, Stanford; Dr. Peter Catrysse, Senior research engineer, Stanford. Abstract This project aims at the feasibility of wirelessly charging the electric vehicles cruising on the high way. We explored a wireless power transfer mechanism utilizing high quality magnetic resonances. We demonstrated numerically and experimentally that the energy could be transferred efficiently between two magnetically coupled resonating coils in a complex electromagnetic environment.

1. Introduction

Eco-friendly vehicle is the international trend in the automobile industry. Advances in batteries and electronics have revived electric vehicles (EVs). EVs offer superior energy efficiency and reduce CO2 emissions if the electricity is supplied from a renewable or nuclear source. However, they are presently neither range- nor cost-competitive compared to conventional vehicles, due to limited options for recharging, and expensive energy storage (batteries). Batteries take up large space and weight portion in vehicles, and require frequent and long hour charging time.

Figure 1: Magnetic resonance power transfer to a moving vehicle.

This project aims at overcome these challenge by extending the wireless power transfer to moving vehicles. We envision a system in which a sequence of coils placed beneath the roadway will deliver the power wirelessly to cars moving above (See Figure 1). In this project, we have demonstrated numerically and experimentally that the energy could be transferred efficiently between two magnetically coupled resonating coils in a complex electromagnetic environment associated with the automobile and the highway.

Our study provides an indication that the wireless power transfer scheme can be extended to moving vehicles as well, since the vehicle moving timescale is much smaller than the energy transfer timescale. As a result of dynamic charging, the batteries size of EVs can be greatly reduced and the driving range can be increased to unlimited.

2. Background

There has been much previous work aiming to achieve efficient power transfer to both stationary (Delphi, HaloIPT, Evantran, WAVE, Waseda WEB) and moving vehicles1,2,3,4,5,6, some dating back over twenty years. The majority of these studies used the inductive power transfer schemes. However, it is well documented that there are substantial limitations to an inductive power transfer scheme. The transfer distances are typically below 20cm. This has become a substantial issue. For safety reasons, and in order to ensure that the road can still be used for other kinds of vehicles, the source needs to be buried below the pavement. Thus the transfer distance in the inductive power transfer scheme is in fact not sufficient. The lateral tolerance of these schemes is also quite stringent, typically on the order of 10cm. Such a stringent lateral tolerance may become a limiting factor for power transfer to a moving vehicle.

In contrast to the more conventional inductive power transfer scheme, our approach here is closely associated with a resonant power transfer scheme. Similar to inductive power transfer scheme, resonant power transfer occurs through magnetic field coupling. However, in resonant power transfer scheme, both the source and the receiver sides consist of resonant circuits. Efficient power transfer occurs when the two circuits have the same resonant frequency, and when the coupling constant of the two resonances, (which is related to the mutual inductance between the inductors), dominates over the intrinsic loss rates of the resonators. Using this scheme, a recent experiment, which was conducted at MIT 7 , has demonstrated highly efficient power transfer over a distance of approximately 1m. There are also indications that the resonant power transfer scheme can be far more robust in lateral tolerance compared with the inductive power transfer scheme. However, the MIT experiments were carried out in free space, whereas to achieve wireless power transfer to a moving vehicle, more complex electromagnetic environment has to be taken into account. In particular, the body of the car is typically metallic that may significantly influence the power transfer characteristics.

3. Results 3.1 Numerical study

With the finite-difference-time-domain (FDTD) method and the coupled mode theory analysis, we numerically demonstrated that efficient wireless energy transfer could be achieved between two high Q resonators in a complex electromagnetic environment. In particular, in the close proximity of metallic planes, efficient wireless energy transfer can be achieved with proper system designs.

We simulate the energy transfer systems shown in Fig. 2. The resonators are all identical and consist of a single turn loop and a capacitor formed with two paralled metal

planes. One resonator serves as the source and the other serves as the receiver. The big metallic plane is a rough representation of the car body. All materials are assumed to be copper. The central operation frequency is 10MHz.

In the simulation we excite a magnetic dipole source at the middle of the source coil. The source has a gaussian-like profile in time and its central frequency is the resonant frequency of the resonators. We record the magnetic field (with a direction perpendicular to the coil plane) at the monitor points which are placed near the source and receiver coil in a symmetrical way. From which we extract the instrinsic parameters (loss rate, coupling coefficient, resonant frequency) of the system and calculate the transfer efficiency.

Figure 2: A variety of the resonator structures for the wireless energy transfer. (a) Our current optimal design with a “twisted” geometry. The distance between the coils ~ λ /15 . Two metallic ground planes are placed near the system symmetrically. (b) Same system as (a) without the ground plane at the source coil side. (c) Same system as (a), but with the coil planes perpendicular to the ground planes. (d) Same system as (a) but with the capacitors at a different orientation.

Structure (a) in Fig. 2 turn out to be our optimal design. For this unique optimized geometry (Fig. 2), the full-field three-dimensional simulations have demonstrated an efficiency exceeding 97%, when the two resonators are separated by two meters, with an operating frequency of 10MHz. And the transfer time scale is shown to be much shorter than the time scale of a moving vehicle.

!"#$

!%#$

!&#$

!'#$

In Figure 3 we show the transfer efficiencies for the structures shown in Figure 2 at different transfer distance. The efficiency of our optimal configuration (red curve in Fig. 3) decays slowly as a function of distance, and remains above 90% for a distance shorter than 2.8m (if it operates a 10MHz frequency). These results have recently been published

Figure 3: Transfer efficiencies as functions of the coil-to-coil distance for different structures shown in Fig. 2, assuming a working wavelength of 30m.

in Applied Physics Letters8, and a patent application has been filed based on this work. This work was supported in a previous GCEP seed project that concluded in 2012. We reproduced these results since they form the theoretical foundation for the work supported under this particular GCEP project.

3.2 Experimental study In this GCEP project, motivated by the theoretical work outlined above, we have focused on the experiment demonstration of efficient wireless power transfer systems in the presence of metallic planes and confirmed our numerically studies.

To illustrate the experimental setup and to establish a reference for comparison, the setup shown in Figure 4 is used to measure the power transfer between two identical resonators, which consist of a 2 turn copper ribbon coil with a diameter of about 60cm and a 5 − 70pF adjustable high voltage capacitor. The copper ribbon is 3cm wide and 0.14mm thick. Our measurement shows that the quality factor of the resonator is 1338 with the absence of the metallic planes and 1329 in the presence of a metallic plane.

1 1.5 2 2.5 30

10

20

30

40

50

60

70

80

90

100

Distance (m)

Tran

sfer

effi

cien

cy (%

)

(a)(b)(c)(d)

Figure 4. Photos and schematics of the wireless power transfer experiment. The system consists of two high Q resonators and two coupling coils. The coupling coils are placed right next to the source and receiver resonator serving as the input and output port.

Figure 5. Maximum transfer efficiency as a function of transfer distance. The red squares are directly measured efficiency. The blue squares are the theoretical result obtained from experimental extracted parameters (coupling coefficients and intrinsic loss rates).

Figure 5 shows the transfer efficiency measurement result. The red line is the direct experiment result and the blue line is the coupled mode theory model result from the experimentally extracted system parameters (intrinsic loss rate, coupling coefficients). The experimental results demonstrate high efficiency wireless power transfer. A 96% maximum transfer efficiency of an optimized wireless transfer system (Figure 4) over a transfer distance of 60cm was demonstrated in the absence of a metallic object nearby. The small discrepancy between the theory and the experiment is likely due to the non-continuous tuning of the coupling loops (optimal operation point isn’t reached).

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS 2

8.3 8.35 8.4 8.450

0.2

0.4

0.6

0.8

1

Frequency (MHz)

Tra

nsm

issi

on (

a.u

.)

Fig. 2. (a)Schematics of Q measurement setup. (b) The transmission spectrum(red dots are the experiment data and blue line is the Lorentzian fit). The insetis the photo of the resonator being characterized.

Eq. 5 shows that the resonant transfer scheme requiresresonance frequency matching, resonant modes of high qualityfactor, and a fast coupling rate, in consistency with Ref. [2]–[4]. The optimal operation regime is the “strong coupling”regime when ⇤/

⇥�10�20 > 1. Moreover, to maximize the

transfer efficiency, an optimal configuration consists of the tworesonators having the same electromagnetic characteristics, i.e.⌅1 = ⌅2 = ⌅0, �1 = �2 = �, �10 = �20 = �0. For suchan electromagnetically symmetric configuration, the transferefficiency is maximized when the operating frequency is ⌅0

and � =p

⇤2 + �20 :

⇥max

= | ⇤

�0 + �|2 (6)

The aim of this experiments is to achieve such an elec-tromagnetically symmetric configuration, either by enforc-ing a geometric symmetry in the system, or by controllingthe resonators such that their electromagnetic properties aresymmetric, even when the structure itself lacks geometricsymmetry.

III. EXPERIMENTAL TECHNIQUE AND COMPONENTCHARACTERIZATION

A. Resonators and their quality factorsOne of the essential factors in an efficient wireless power

transfer system is the quality factor of the resonators. Thequality factor (Q) is related to the intrinsic loss rate (�0) ofthe resonator by Q = ⌅/2�0. This intrinsic quality factor ismeasured with the experimental setup shown in Fig. 2(a).A vector network analyzer (VNA) is connected with twoprobing coils and measures the transmission spectrum. The

60cm60cm

!"#$

Fig. 3. Photos and schematics of the wireless power transfer experiment. Thesystem consists of two high Q resonators and two coupling coils. The couplingcoils are placed right next to the source and receiver resonator serving as theinput and output port. Both coupling coils are connected to a vector networkanalyzer.

probing coils are placed near the resonator and are oriented tominimize the direct coupling between them. The quality factoris extracted from the line-shape of the transmission spectrum.This quality factor is for the resonator loaded with the probingcoils. By increasing the distance between the probing coilsand the resonator, the coupling between the probing coils andthe resonator reduces. When further reducing the couplingdoesn’t change the line-shape of the transmission spectrum,the intrinsic Q of the resonator is obtained.

Fig. 2 is an example of the Q measurement. The resonatorshown in the inset of (b) consists of a 2 turn copper ribboncoil with a diameter of about 60cm and a 5�70pF adjustablehigh voltage capacitor. The copper ribbon is 3cm wide and0.14mm thick. Fig. 2(b) shows the transmission spectrum.The red dots are the transmission measurement and the blueline is the Lorentzian fit, from which a resonance frequencyof 8.38MHz and a quality factor of 1338 are extracted. Thisis in the regime discussed above where the probing coils aresufficiently far away from the resonator. Hence the Q-factorhere is the intrinsic quality factor of the resonance.

B. Transfer efficiency measurement and optimizationIn order to illustrate the experimental setup and to establish

a reference for comparison, the setup shown in Fig. 3 is usedto measure the power transfer between two of the resonatorscharacterized in Sec. III. A without the metallic plates. Theseparation between the resonators is the transfer distance. Twosingle-turn wire loops serving as the input and output ports areattached to the VNA and placed next to the source and receiverresonator. The transmission spectrum is directly read out fromthe VNA.

From the coupled mode analysis, a high transfer efficiencyin such a wireless power transfer system can be achievedby doing the following : 1) tune the resonators to resonatetogether (⌅1 = ⌅2); 2) operate at the optimal frequency;3) maximize the intrinsic quality factor of the resonators(minimize �0); 4) orient the resonators along the same axis tomaximize the coupling (⇤); 5) tune the coupling between thecoupling coil and the resonator on each side by adjusting thecoupling loop size with the optimal efficiency reached when�1,2 = ⇤� �0.

In the experiment, the intrinsic decay rates �10,20 of theresonator are fixed by construction. The coupling coefficient

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS 3

8.2 8.4 8.60

0.2

0.4

0.6

0.8

Frequency (MHz)

Tra

nsf

er

effic

ien

cy

8.2 8.4 8.60

0.2

0.4

0.6

0.8

Frequency (MHz)

Tra

nsf

er

effic

ien

cy

8.4 8.5 8.6 8.7 8.80

0.2

0.4

0.6

0.8

Frequency (MHz)

Tra

nsf

er

effic

ien

cy

1

1

1

(a)

(b)

(c)

Fig. 4. Transfer efficiency spectra for diferent coupling loops when thetransfer distance is 60cm. Red dots are experimental data and the blue line isthe coupled mode theory model fit. From (a) to (c), the size of the couplingcoils are incresed.

⇥ is determined once the transfer distance is chosen and theresonators are aligned for maximum coupling. To optimize thetransfer efficiency, only the size of the coupling coil needs tobe adjusted, to affect the input and output coupling rates (�1,2),and until the optimal operating frequency in the transmissionspectrum is found.

Fig. 4 illustrates how to optimize the transfer efficiency byadjusting the size of the coupling loops. The figure shows howthe transfer spectra vary as the size of the coupling loops aregradually increased for a transfer distance of 60cm. The reddots are the experimental data and the blue lines are the datafitting based on Eqn. (5). In the case where �0 + � ⇥ ⇥, thefrequency splittings in the transfer efficiency spectra (Fig. 4(a),(b)) can be clearly seen, but the maximum transfer efficiencyis small; when �0 + � � ⇥, there is no frequency splitting inthe transmission spectrum (Fig. 4(c)), instead there is a widebandwidth peak. A maximum transfer efficiency of 96% wasachieved when the transfer distance is 60cm in the experiment.

The red squares in Fig. 5 shows our experimental measure-

60 80 100 120 1400

20

40

60

80

100

Distance (cm)

Tra

nsf

er

effic

ien

cy (

%)

experimentcoupled mode theory

Fig. 5. Maximum transfer efficiency as a function of transfer distance. The redsquares are directly measured efficiency. The blue squares are the theoreticalresult obtained from experimental extracted parameters (⇥ and �0).

!"#$%&% !"#$%'%

!"#$%&% !"#$%&%%()%*+,-".%

!"#$%&%()%%/+01$%*$10+%

!"#$%'% !"#$%'%%()%*+,-".%

!"#$%'%()%%/+01$%*$10+%

!"%23456% 7897% 7897% 787'% 78:;% 78:'% :8<9%

=% &997% &'7<% &9':% >?&% <;:% :9?%

216% 2@6%

2A6%

Fig. 6. Photos and schematics of coils with (a) and without (b) an attachedcapacitor. (c) Resonance frequencies and quality factors of coils in differentenvironment.

ment of the maximum transfer efficiency versus distance. Theblue diamonds are the theoretical predictions of the maximumtransfer efficiency at each transfer distance from experimentalextracted parameters (⇥, �0) assuming the optimized condi-tions (⇥ = � + �0). The result shows that the experimentaldata is always a few percent lower than the theoretical optimalprediction, which is possibly due to the noncontinuous tuningof the coupling loops (optimal operation point isn’t reached).

IV. POWER TRANSFER IN THE PRESENCE OF METALLICPLATES: EXPERIMENT

In the previous section, the wireless power transfer in freespace is demonstrated. In practice, there are many applicationsthat require the system to work in a rather complex environ-ment, such as in a close proximity of a metallic object, insea water, etc. In this section, as an important example ofa complex environment, the influence of a metallic plate onthe resonators performance and hence the performance of awireless power transfer system is studied.

A. Resonance properties as influenced by external environ-ment.

The resonator (Coil 1 in Fig. 6) characterized in the previoussection, differs from some of the resonator geometries that

In the following experiment, an aluminum plate is placed at a distance of 20cm from one of the resonators (as shown in Figure 6(a)) in the optimized system, i.e. close enough to interact with the weak electric fringing fields from the nearest resonator. Since the distances between the plate and the two resonators are different, the self-resonance frequencies of the resonators shift differently and no longer match. According to the coupled mode theory model, the transfer efficiency will drop as well. The experiment results do confirm that the transfer efficiency drops (Figure 6(b)). The maximum transfer efficiency drops from 96% to 37%.

Figure 6. (a)Schematics of the wireless power transfer system in the close proximity of an aluminum plate. (b)Transfer efficiency spectrum of the wireless transfer system with a metal plate on one side. (c)Transfer efficiency spectrum of the wireless transfer system with an aluminum plate on one side after optimization. (d)Schematics of the wireless power transfer system with two aluminum plates. (e)Transfer efficiency spectrum of the wireless transfer system with aluminum plates on both sides. (f)Transfer efficiency spectrum of the wireless transfer system with aluminum plates on both sides after optimization. Red dots are the experimental data and the blue line is the coupled mode theory model fit.

In consistent with our numerical prediction, one could use a symmetric configuration by placing a metal plate in the vicinity of the source resonator. Alternatively, high efficiency power transfer can also be accomplished for a system lacking spatial symmetry, by adjusting the resonance frequency and the input impedance of the source resonator. In both cases, the demonstrated maximum efficiency exceeds 94%, over a distance of 60cm, for a resonator with a coil radius of 30cm, operating at 8.4MHz. The

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS 4

8.4

8.6

8.8

f 0 (

MH

z)

20 40 60 80 100

1350

1400

1450

1500

Distance (cm)

Q

Fig. 7. Resonance frequency and quality factor change with the distancebetween Coil 1 and the aluminum plate.

were previously used in wireless power transfer experiments.The experiments in Ref. [4], [8] used a self-resonant coilwith distributed self-capacitance. For comparison study in thepresent section, a similar self-resonant coil, which consists ofa 6-turn copper wire that is left open on both ends (Coil 2 inFig. 6), is also characterized. The coil diameter is 60cm andthe wire diameter is 3mm.

To compare these two resonators, Fig. 6 lists the resonancefrequency and quality factor measurement results under dif-ferent conditions: free space clear of external objects, with alarge lossy dielectric nearby (e.g. a person standing by), andwith a metal plate in a close proximity. When an aluminumplate is placed (parallel to the coil plane) 20cm away from theresonator, the resonance frequency of Coil 1 increases from8.38MHz to 8.82MHz and the resonance frequency of Coil2 increases from 8.92MHz to 9.43MHz; the quality factordoesn’t changes much for Coil 1 but increases by 175 forCoil 2.

When a lossy dielectric comes close to the resonator (suchas a person), the resonant frequencies shift as well and thequality factors decrease. From the table in Fig. 6, for Coil 1,neither the resonance frequency nor the quality factor changesmuch when a person is standing right next to the resonator.However, for Coil 2, its resonance frequency shifts and thequality factor goes down drastically. It turns out that Coil 1performs better in terms of the robustness due to the betterconfinement of electric field. Thus, in our experiments, Coil 1is used in the following power transfer experiment.

A more detailed study on the influence of the aluminumplate on the resonator is shown in Fig. 7. The resonancefrequency doesn’t change much for Coil 1 unless a metallicobject is close (< 40cm) to the resonator. The quality factorstays above 1300 for all measured distances from 20cm to100cm.

B. Wireless power transfer under the influence of environmentchanges.

As shown in the previous section, both the resonancefrequency and the quality factor of the resonator are influencedby the environment. The coupled mode theory analysis showsthat the performance of the wireless power transfer system will

8.8 8.9 9 9.10

0.2

0.4

0.6

0.8

Frequency (MHz)

Tra

nsf

er

effic

ien

cy

8.7 8.8 8.9 9 9.10

0.2

0.4

0.6

0.8

Frequency (MHz)

Tra

nsf

er

effic

ien

cy

60cm60cm 60cm60cm

8.6 8.8 90

0.2

0.4

0.6

0.8

Tra

nsf

er

effic

ien

cy

8.7 8.8 8.9 90

0.2

0.4

0.6

0.8

Frequency (MHz)

Tra

nsf

er

effic

ien

cy

Frequency (MHz)

1 1

1 1

Fig. 8. (a)Schematics of the wireless power transfer system in the closeproximity of an aluminum plate. (b)Transfer efficiency spectrum of thewireless transfer system with a metal plate on one side. (c)Transfer efficiencyspectrum of the wireless transfer system with an aluminum plate on one sideafter optimization. (d)Schematics of the wireless power transfer system withtwo aluminum plates. (e)Transfer efficiency spectrum of the wireless transfersystem with aluminum plates on both sides. (f)Transfer efficiency spectrumof the wireless transfer system with aluminum plates on both sides afteroptimization. Red dots are the experimental data and the blue lines are thecoupled mode theory model fit.

change as well. In particular, the close proximity of a metallicplate will greatly influence the power transfer system.

In the previous section, a 96% maximum transfer efficiencyof an optimized wireless transfer system (Fig. 3) over atransfer distance of 60cm was demonstrated in the absenceof a metallic object nearby. In the following experiment, analuminum plate is placed at a distance of 20cm from one of theresonators (as shown in Fig. 8(a)) in the optimized system, i.e.close enough to interact with the weak electric fringing fieldsfrom the nearest resonator. Since the distances between theplate and the two resonators are different, the self-resonancefrequencies of the resonators shift differently and no longermatch. According to Eqn. (5), the transfer efficiency will dropas well. The experiment results do confirm that the transferefficiency drops (Fig. 8(b)). The maximum transfer efficiencydrops from 96% to 37% .

There are two approaches to compensate for the effects ofthis aluminum plate:

1) Re-tune the resonance frequencies such that the tworesonators resonate at the same frequency by adjusting thecapacitor. In addition, adjust the coupling coils size to have theimpedance match and reach the optimal operating point again.The transfer efficiency spectrum after the re-optimization isshown in Fig. 8(c). With this approach, the transfer efficiencyincreases back up to 94%.

2) Place another aluminum plate in a symmetrical way,

experiments show that the key for high efficiency transfer in both cases is to achieve symmetry in the electromagnetic parameters of the source and receiver resonators.

4. Conclusions In conclusion, a numerical and experimental study of a resonant wireless power transfer system is reported. A coupled mode theory analysis is used to optimize the system parameters and proved to be in excellent agreement with the measured data. A transfer efficiency of 96% over a distance of 60cm was demonstrated. Restoring the electromagnetic symmetry of the system can overcome the system performance degradation when it is placed in a complex electromagnetic environment.

5. Publications and Patents: X. Yu, S. Sandhu, S. Beiker, R. Sassoon, and S. Fan, “Wireless energy transfer with the presence of metallic planes”, Applied Physics Letters 99, 214102 (2011). X. Yu, T. Skauli, B. Skauli, S. Sandhu, P. B. Catrysse, and S. Fan, “Wireless power transfer in the presence of metallic plates: experimental results” (under review). U. S. Provisional Patent Application 61/542,667, filed 10/3/2011. 6. Contacts:

Shanhui Fan: [email protected]; Sven Beiker: [email protected]; Richard Sassoon: [email protected]; Xiaofang Yu: [email protected]; Sunil Sandhu: [email protected]; Peter Catrysse: [email protected]. References 1 Covic, G. A., Boys, J. T., Kissin, M. L. G., and Lu, H. G., “A three-phase inductive power transfer system for Roadway-Powered Vehicles”, IEEE trans. on Industrial Electronics, vol. 54, No. 6, 3370-3378, 2007. 2 Elliott, G. A. J., Raabe S., Covic, G. A., and Boys, J. T., “Multiple pickups for large lateral tolerance contactless power-transfer systems”, IEEE trans. on Industrial Electronics, vol. 57, No. 5, 1590-1598, 2010. 3 Huh, J., Lee, S. W., Lee, W. Y., Cho, G. H., and Rim, C. T., “Narrow-Width Inductive power transfer system for online electrical vehicles”, IEEE trans. on Power Electronics, vol. 26, No. 12, 3666-3679, 2011. 4 Suh, I. S., and Gu, Y., “Application of shaped magnetic filed in resonance (SMFIR) technology to future urban transportation”, CIRP Design Conference, 2011 IEEE. 5 Lee, S., Huh J., Park, C., Choi, N., Cho, G., Rim, C., “On-line electric vehicle using inductive power transfer system”, Energy Conversion Congress and Exposition (ECCE), 2010 IEEE. 6 “Roadway powered electric vehicle project parametric studies: phase 3D final report”, California Partners for Advance Transit and Highways Research Reports, Oct. 1996. (http://www.path.berkeley.edu) 7 Kurs, A., Karalis, A., Moffatt, R., Joannopoulos, J. D., Fisher, P., and Soljačić, M., “Wireless power transfer via strongly coupled magnetic resonances”, Science, vol. 317, No. 5834, 83-86 (2007) 8 X. Yu, S. Sandhu, S. Beiker, R. Sassoon, and S. Fan, “Wireless energy transfer with the presence of metallic planes”, Applied Physics Letters 99, 214102 (2011).