- Email: [email protected]
Transferring electrical energy between two dielectric elastomer actuators
Sensors and Actuators A 212 (2014) 123–126
Contents lists available at ScienceDirect
Sensors and Actuators A: Physical journal homepage: www.elsevier.com/locate/sna
Transferring electrical energy between two dielectric elastomer actuators Ho Cheong Lo a,∗ , Todd A. Gisby a , Emilio P. Calius b , Iain A. Anderson a a b
Biomimetics Laboratory, Auckland Bioengineering Institute, University of Auckland, 70 Symonds Street, 1010 Auckland, New Zealand Callaghan Innovation, Brooke House, 24 Balfour Road, 1052 Auckland, New Zealand
a r t i c l e
i n f o
Article history: Received 5 August 2013 Received in revised form 10 March 2014 Accepted 14 March 2014 Available online 24 March 2014 Keywords: Dielectric elastomer Electronics Efficiency Electrode resistance
a b s t r a c t Soft dielectric elastomer (DE) devices used for actuation and power generation are typically operated at kilovolt potential. Effective recovery of the electrical energy inherently stored during operation will greatly improve the efficiency of a DE device. One method of energy recovery involves the use of DC–DC step-down circuits to convert the high voltage energy into a low voltage form. The energy would then go to a storage component and eventually be reconverted back to high voltage for recharging the DE. However, the voltage conversion process incurs significant energy loss. Simply transferring the energy to a capacitor or another DE without large changes in the voltage can achieve better efficiencies. In our experiments, we achieved 51% energy transfer from one DE to another DE of similar size. Improved energy transfer is expected with lower electrode resistance. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Dielectric elastomer (DE) transducers can be used for actuation, electricity generation, strain sensing, and switching [1,2]. They are stretchable capacitors which typically operate at a few kilovolts. Special electronics are required to power these devices [3–5]. This involves electrically charging and discharging the DE at the correct time [6,7]. To achieve good efficiency, discharge of the DE must involve recovery of the stored electrical energy. Step-down circuits can be used to perform the energy extraction and convert the energy into a low voltage form for storage and use [3,4]. However, repeatedly converting electrical energy between low voltage and high voltage forms is not always the best solution due to significant energy loss and the need for complex electronics. It would be more efficient to simply transfer the energy elsewhere while retaining its high voltage form [8]. The energy could be transferred to a capacitor for temporary storage or to another DE for actuation or power generation. One method has been used for DE generators (DEGs) which pumps charge from a low voltage reservoir to a high voltage reservoir [9]. When the voltage across the DEG increases to that of the high voltage reservoir, a connection is made between them. Charge is subsequently pumped into this reservoir as the DEG decreases in
∗ Corresponding author. Tel.: +64 9 9238363. E-mail address: [email protected] (H.C. Lo). http://dx.doi.org/10.1016/j.sna.2014.03.020 0924-4247/© 2014 Elsevier B.V. All rights reserved.
capacitance. However, the connection must be made when their voltages are equal. Failing to do so will cause a large peak current and significant losses through the small resistance of the connecting path, equal to I2 R, where I is the current and R is the resistance of the path which the current flows through. The inclusion of an inductor to the circuit reduces the peak current without adding significant losses, where the current I is related to the inductance L and the integral of the voltage V across it, I = (1/L) V dt. The circuit shown in Fig. 1a would allow efficient transfer of electrical energy from a capacitor at a high voltage to a capacitor at a lower voltage. Furthermore, it is possible for a small capacitor receiving the charge to attain a higher voltage than the larger capacitor being discharged [8]. In this paper, we will briefly assess the practical performance of this circuit when used with DE actuators (DEAs). This will reveal the impact of high series resistance of DEs on the performance of power electronics.
2. Experiment Energy is to be transferred from one charged DE actuator, DEA1, to another uncharged DE actuator of similar size, DEA2. A large inductor of 1 H was used to overcome the detrimental effect of the large electrode resistance [10]. The inductor had the dimensions of 25 mm × 25 mm × 20 mm, had a mass of 24 g, and constitutes approximately 50% of the total circuit mass. Miniature and entirely soft DE systems would enable extremely portable and unobtrusive
124
H.C. Lo et al. / Sensors and Actuators A 212 (2014) 123–126
Fig. 1. (a) Circuit used to transfer electrical energy from one charged capacitor C1 to another uncharged capacitor C2 [8]. (b) Circuit diagram of the experimental setup. Initially, DEA1 was charged to approximately 2.5 kV and DEA2 was discharged to 0 kV. The ‘Transfer Relay’ was turned on to transfer electrical energy from DEA1 to DEA2. The Trek P0865 voltmeter measured the voltages of DEA1 and DEA2 which were recorded using a LabVIEW program.
devices. Many types of electronic components can be made very small, flexible, or soft [11,12]. However, producing physically small or soft inductors with large inductances is a major challenge. A schematic of the experimental setup is shown in Fig. 1b. The electrical efficiency of this process was measured. Equipment and components with low leakage currents were used to minimize energy loss and improve measurement accuracy. Two DAT70510 reed relays were used to establish the initial voltage across DEA1 and DEA2. Another ‘Transfer Relay’ was used in the main circuit to begin the energy transfer.
were approximately 25 k for the configuration in Fig. 2a and 5 k for the configuration in Fig. 2b. 2.3. Method Using relays, DEA1 was charged to approximately 2.5 kV using a Biomimetics Laboratory EAP Controller while DEA2 was fully discharged to ensure it holds no energy initially. The relays were then turned off, disconnecting both the voltage source from DEA1 and the discharge path from DEA2 so no more energy would enter the system. The energy transfer would then occur by switching on the
2.1. DE capacitances DEA1 and DEA2 were similar dot actuators of 30 mm diameter, fabricated using 3 M VHB 4905 with 3-by-3 equi-biaxial pre-stretch and Nyogel 756G carbon grease electrode. The capacitances of the charged DEA1 and the uncharged DEA2 were determined to be approximately 0.96 nF and 0.73 nF respectively. These capacitance values were assumed to have remained unchanged when the voltages were measured immediately after the energy transfer. This assumption can be made because the mechanical time constant of the DEA was several orders of magnitude greater than the electrical time constant of the circuit. The time to transfer the electrical energy is governed by the damped natural frequency of the series RCL system as shown in Eq. (1). Given that the series resistance R was measured to be approximately 25 k, C1 and C2 were 0.97 nF and 0.73 nF, and L was 1 H, the electrical energy would be transferred within 74 s. The much slower mechanical response time of VHB is in the order of 1 s [13]: t=
2
(1)
(((C1 + C2 )/(LC1 C2 )) − (R/L) )
2.2. Electrical connections Two different methods of establishing an electrical connection to the DEAs were used to investigate how the length of the carbon grease connection affects the performance of the circuit. A shorter electrode path would have lower series resistance. Fig. 2 illustrates the two configurations. The measured equivalent series resistance
Fig. 2. Cross-sectional diagrams of the DEA configurations used in the experiment. (a) High resistance: a thin carbon grease track establishes an electrical connection to the DEA’s active area. (b) Low resistance: an aluminium foil track establishes an electrical connection to the centre of the DEA’s active area.
H.C. Lo et al. / Sensors and Actuators A 212 (2014) 123–126
125
Fig. 3. Images showing the changes in DEA deformation before and after energy transfer for the low resistance configuration shown in Fig. 2b. The voltage was measured before deformation while the images show the DEA after deformation. Only a small amount of deformation occurred in DEA2 because not all the energy was transferred. Regarding the results for ‘After Transfer’, the voltages were measured before deformation occurred, while the images show the DEAs after deformation occurred.
‘Transfer Relay’. The voltages of DEA1 and DEA2 immediately before and after switching on the ‘Transfer Relay’ were recorded using a high-impedance Trek P0865 electrostatic voltmeter and a LabVIEW program. The voltages were measured before considerable changes in capacitance arising from voltage-induced deformation had occurred. A relatively low voltage was used so that the deformations were small yet observable, minimizing the changes in DEA capacitances. 3. Results and discussion 3.1. Voltage measurements Experimental results show that highly resistive electrical connections greatly reduced the amount of energy transferred. When resistance was large (Fig. 2a), charge was distributed almost equally between the two DEAs, with each DEA reaching approximately 1.4 kV. When resistance was small (Fig. 2b), DEA1 discharged to 1.2 kV and DEA2 charged to 2.0 kV. The equal distribution of charge when the resistance was large implies the resistance was so large that the inductor in the circuit had negligible effect on the amount of energy transferred. Reducing the resistance increased the energy transfer. These measurements were consistent between consecutive tests, with a variability of <1% in a series of three tests. However, performance was poorer a week later where DEA2 was only charged to 1.8 kV (down from 2.0 kV) using the configuration shown in Fig. 2b. This was likely due to the degradation of the VHB dielectric and carbon grease electrode.
incurred from energy transfer. Partial discharge may also be useful for DE generators (DEGs). In theory, a DEG achieves maximum energy generation when it is fully discharged every oscillation cycle [7,14]. However, in practice the energy consumed to fully discharge and recharge the DEG may exceed the amount of energy gained. 3.3. Energy measurements The distribution of electrical energy was determined using the measured capacitances and voltages. When series resistance was large, DEA1 and DEA2 ended up with 26% and 25% of the energy respectively. In theory, an equal energy distribution of 25% would occur if an inductor was not used. The results are similar to this, hence the inductor had negligible effect when series resistance was large. When series resistance was small, DEA1 and DEA2 ended up with 24% and 51% of the energy respectively. The allocation of electrical energy is depicted in Fig. 4. In this study, only the electrical efficiency of the energy transfer was measured. Since energy was converted between electrical and mechanical forms when deformation occurred, both the electrical and mechanical energy must be measured to determine the
3.2. Observed actuation After transferring the energy, both DEAs became partially charged. However, neither were at a high enough voltage to undergo significant actuation as shown in Fig. 3. Nevertheless, we now require less input energy to actuate DEA2. This would also be true for the partially charged DEA1, requiring less input energy to recharge and actuate. The partial discharge of DEA1 may seem undesirable, but it can also be an advantage. Since little actuation occurs below 2 kV, it would be more efficient to deactivate the DEA by partially discharging it to 2 kV and avoid some of the losses
Fig. 4. Electrical energy distribution after discharge of DEA1. Simply discharging DEA1 would result in losing 100% of the stored energy. With large series resistance (Config-1), 25% of the energy was transferred to DEA2. With small series resistance (Config-2), 51% of the energy was transferred to DEA2.
126
H.C. Lo et al. / Sensors and Actuators A 212 (2014) 123–126
overall efficiency of the energy transfer. The stored mechanical energy is difficult to measure and will depend on the DEA material and configuration. Advance measuring techniques would be required to measure these parameters. This study approximated the total energy attained by DEA2 to be equal to the electrical energy it had before a significant amount was used to deform the DEA.
[13] S. Rosset, B.M. O’Brien, T. Gisby, D. Xu, H.R. Shea, I.A. Anderson, Self-sensing dielectric elastomer actuators in closed-loop operation, Smart Mater. Struct. 22 (2013) 104018. [14] T. Hoffstadt, C. Graf, J. Maas, Optimization of the energy harvesting control for dielectric elastomer generators, Smart Mater. Struct. 22 (2013) 94028.
Biographies
4. Conclusion This work introduces a useful technique for improving the efficiency of DEs and shows the performance that can be achieved. The circuit in our experiment was able to transfer 51% of the 2.4 kV electrical energy stored in one DEA to another almost identical DEA. A total of 75% of the electrical energy was retained in the two DEAs and could be reused. However, the mechanical energy stored in the DEA was not recovered. Experiments have also shown that large series resistance can reduce the efficiency of the circuit to a point where the inductive element in the circuit becomes redundant. Reducing the series resistance of the DEA electrodes improved the circuit’s performance significantly. With careful implementation, high voltage energy transfer can greatly improve the efficiency of DE actuators and generators. Acknowledgements The work was supported by the Auckland Bioengineering Institute and the University of Auckland Doctoral Scholarship.
Ho Cheong Lo received a BE(Hons) in Electrical and Electronics Engineering from the University of Auckland, New Zealand in 2009. He is currently pursuing a PhD degree in Engineering with the Biomimetics Laboratory at the University of Auckland. His current research is focused on getting useful power from dielectric elastomer generators.
Todd Gisby was awarded his PhD from the University of Auckland in 2011 for his thesis on simultaneous sensing and actuation of dielectric elastomer artificial muscles, conducted under the supervision of Associate Professor Iain Anderson of the Biomimetics Laboratory at the Auckland Bioengineering Institute. He continued his work at the Biomimetics Laboratory in the role of Business Development Engineer, and in December 2012 co-founded and became CTO of StretrchSense Limited, a company that has commercialized wireless soft sensor technology for measuring human body motion based on the system integration and soft sensing expertise he and his colleagues developed at the Biomimetics Laboratory.
References [1] I.A. Anderson, T.A. Gisby, T.G. McKay, B.M. O’Brien, E.P. Calius, Multi-functional dielectric elastomer artificial muscles for soft and smart machines, J. Appl. Phys. 112 (2012) 041101. [2] F. Carpi, S. Bauer, D. De Rossi, Stretching dielectric elastomer performance, Science 330 (2010) 1759. [3] J. Maas, C. Graf, L. Eitzen, Control concepts for dielectric elastomer actuators, in: Proc. SPIE, 2011, pp. 79761H–79761H-12. [4] L. Eitzen, C. Graf, J. Maas, Bidirectional HV DC–DC converters for energy harvesting with dielectric elastomer generators, in: Energy Convers. Congr. Expo. (ECCE), 2011 IEEE, 2011, pp. 897–901. [5] T.G. McKay, B. O’Brien, E. Calius, I. Anderson, Self-priming dielectric elastomer generators, Smart Mater. Struct. 19 (2010) 55025. [6] R. Kornbluh, R. Pelrine, J. Eckerle, J. Joseph, Electrostrictive polymer artificial muscle actuators, in: IEEE Int. Conf. Robot. Autom., vol. 3, 1998, pp. 2147–2154. [7] R. Pelrine, R. Kornbluh, J. Eckerle, P. Jeuck, S. Oh, Q. Pei, S. Stanford, Dielectric elastomers: generator mode fundamentals and applications, in: Proc. SPIE, 2001, pp. 148–156. [8] H.C. Lo, T. Gisby, T. McKay, E. Calius, I. Anderson, Transferring electrical energy between dielectric elastomer actuators, in: Proc. SPIE, San Diego, CA, USA, 2012. [9] R. Kaltseis, C. Keplinger, R. Baumgartner, M. Kaltenbrunner, T. Li, P. MäcHler, R. Schwdiauer, Z. Suo, S. Bauer, Method for measuring energy generation and efficiency of dielectric elastomer generators, Appl. Phys. Lett. 99 (2011) 162904. [10] H.C. Lo, E. Calius, I. Anderson, Understanding efficiency limits of dielectric elastomer driver circuitry, in: Proc. SPIE, 2013, pp. 86870E–86870E-6. [11] M. Kaltenbrunner, T. Sekitani, J. Reeder, T. Yokota, K. Kuribara, T. Tokuhara, M. Drack, R. Schwodiauer, I. Graz, S. Bauer-Gogonea, S. Bauer, T. Someya, An ultralightweight design for imperceptible plastic electronics, Nature 499 (2013) 458–463. [12] G. Kettlgruber, M. Kaltenbrunner, C.M. Siket, R. Moser, I.M. Graz, R. Schwodiauer, S. Bauer, Intrinsically stretchable and rechargeable batteries for self-powered stretchable electronics, J. Mater. Chem. A 1 (2013) 5505–5508.
Emilio Calius is a senior researcher who specializes in the development and application of advanced material and structural concepts at Callaghan Innovation, a New Zealand Crown Agency. He received a PhD in Aeronautics and Astronautics from Stanford University, and has previously worked on lightweight components for spacecraft, bio-inspired mechanisms for flexible machinery, and smart structures that sense and adapt to changing loads. His current research is focused on novel ways of manipulating and controlling static and dynamic mechanical behaviour using metamaterials and other architectured materials and surfaces. Iain Anderson is Group Leader for the Biomimetics Laboratory of the Auckland Bioengineering Institute (www.abi.auckland.ac.nz/biomimetics) and an Associate Professor with the Department of Engineering Science at the University of Auckland. He completed his PhD (Engineering Science, University of Auckland) in 1996. He has worked as a whiteware product designer (Fisher and Paykel Ind.) and a vibrations consulting engineer (NZ Department of Scientific and Industrial Research, Industrial Research Ltd., New Zealand). In 2000 he joined the University of Auckland. His interest in electro-active polymer “artificial muscle” technology led to the formation of the Biomimetics Laboratory in 2005. His Lab’s research has focused on the control and self-sensing of artificial muscles for soft robotics and artificial muscle energy harvesting. The lab fosters an entrepreneurial spirit: students and engineers are encouraged to develop useful devices that are sold to other researchers and industries (www.biomimeticslab.com). In late 2012, with two of his former students, he created the company: Stretchsense Ltd. (www.stretchsense.com), a manufacturer of artificial muscle sensors.
Contents lists available at ScienceDirect
Sensors and Actuators A: Physical journal homepage: www.elsevier.com/locate/sna
Transferring electrical energy between two dielectric elastomer actuators Ho Cheong Lo a,∗ , Todd A. Gisby a , Emilio P. Calius b , Iain A. Anderson a a b
Biomimetics Laboratory, Auckland Bioengineering Institute, University of Auckland, 70 Symonds Street, 1010 Auckland, New Zealand Callaghan Innovation, Brooke House, 24 Balfour Road, 1052 Auckland, New Zealand
a r t i c l e
i n f o
Article history: Received 5 August 2013 Received in revised form 10 March 2014 Accepted 14 March 2014 Available online 24 March 2014 Keywords: Dielectric elastomer Electronics Efficiency Electrode resistance
a b s t r a c t Soft dielectric elastomer (DE) devices used for actuation and power generation are typically operated at kilovolt potential. Effective recovery of the electrical energy inherently stored during operation will greatly improve the efficiency of a DE device. One method of energy recovery involves the use of DC–DC step-down circuits to convert the high voltage energy into a low voltage form. The energy would then go to a storage component and eventually be reconverted back to high voltage for recharging the DE. However, the voltage conversion process incurs significant energy loss. Simply transferring the energy to a capacitor or another DE without large changes in the voltage can achieve better efficiencies. In our experiments, we achieved 51% energy transfer from one DE to another DE of similar size. Improved energy transfer is expected with lower electrode resistance. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Dielectric elastomer (DE) transducers can be used for actuation, electricity generation, strain sensing, and switching [1,2]. They are stretchable capacitors which typically operate at a few kilovolts. Special electronics are required to power these devices [3–5]. This involves electrically charging and discharging the DE at the correct time [6,7]. To achieve good efficiency, discharge of the DE must involve recovery of the stored electrical energy. Step-down circuits can be used to perform the energy extraction and convert the energy into a low voltage form for storage and use [3,4]. However, repeatedly converting electrical energy between low voltage and high voltage forms is not always the best solution due to significant energy loss and the need for complex electronics. It would be more efficient to simply transfer the energy elsewhere while retaining its high voltage form [8]. The energy could be transferred to a capacitor for temporary storage or to another DE for actuation or power generation. One method has been used for DE generators (DEGs) which pumps charge from a low voltage reservoir to a high voltage reservoir [9]. When the voltage across the DEG increases to that of the high voltage reservoir, a connection is made between them. Charge is subsequently pumped into this reservoir as the DEG decreases in
∗ Corresponding author. Tel.: +64 9 9238363. E-mail address: [email protected] (H.C. Lo). http://dx.doi.org/10.1016/j.sna.2014.03.020 0924-4247/© 2014 Elsevier B.V. All rights reserved.
capacitance. However, the connection must be made when their voltages are equal. Failing to do so will cause a large peak current and significant losses through the small resistance of the connecting path, equal to I2 R, where I is the current and R is the resistance of the path which the current flows through. The inclusion of an inductor to the circuit reduces the peak current without adding significant losses, where the current I is related to the inductance L and the integral of the voltage V across it, I = (1/L) V dt. The circuit shown in Fig. 1a would allow efficient transfer of electrical energy from a capacitor at a high voltage to a capacitor at a lower voltage. Furthermore, it is possible for a small capacitor receiving the charge to attain a higher voltage than the larger capacitor being discharged [8]. In this paper, we will briefly assess the practical performance of this circuit when used with DE actuators (DEAs). This will reveal the impact of high series resistance of DEs on the performance of power electronics.
2. Experiment Energy is to be transferred from one charged DE actuator, DEA1, to another uncharged DE actuator of similar size, DEA2. A large inductor of 1 H was used to overcome the detrimental effect of the large electrode resistance [10]. The inductor had the dimensions of 25 mm × 25 mm × 20 mm, had a mass of 24 g, and constitutes approximately 50% of the total circuit mass. Miniature and entirely soft DE systems would enable extremely portable and unobtrusive
124
H.C. Lo et al. / Sensors and Actuators A 212 (2014) 123–126
Fig. 1. (a) Circuit used to transfer electrical energy from one charged capacitor C1 to another uncharged capacitor C2 [8]. (b) Circuit diagram of the experimental setup. Initially, DEA1 was charged to approximately 2.5 kV and DEA2 was discharged to 0 kV. The ‘Transfer Relay’ was turned on to transfer electrical energy from DEA1 to DEA2. The Trek P0865 voltmeter measured the voltages of DEA1 and DEA2 which were recorded using a LabVIEW program.
devices. Many types of electronic components can be made very small, flexible, or soft [11,12]. However, producing physically small or soft inductors with large inductances is a major challenge. A schematic of the experimental setup is shown in Fig. 1b. The electrical efficiency of this process was measured. Equipment and components with low leakage currents were used to minimize energy loss and improve measurement accuracy. Two DAT70510 reed relays were used to establish the initial voltage across DEA1 and DEA2. Another ‘Transfer Relay’ was used in the main circuit to begin the energy transfer.
were approximately 25 k for the configuration in Fig. 2a and 5 k for the configuration in Fig. 2b. 2.3. Method Using relays, DEA1 was charged to approximately 2.5 kV using a Biomimetics Laboratory EAP Controller while DEA2 was fully discharged to ensure it holds no energy initially. The relays were then turned off, disconnecting both the voltage source from DEA1 and the discharge path from DEA2 so no more energy would enter the system. The energy transfer would then occur by switching on the
2.1. DE capacitances DEA1 and DEA2 were similar dot actuators of 30 mm diameter, fabricated using 3 M VHB 4905 with 3-by-3 equi-biaxial pre-stretch and Nyogel 756G carbon grease electrode. The capacitances of the charged DEA1 and the uncharged DEA2 were determined to be approximately 0.96 nF and 0.73 nF respectively. These capacitance values were assumed to have remained unchanged when the voltages were measured immediately after the energy transfer. This assumption can be made because the mechanical time constant of the DEA was several orders of magnitude greater than the electrical time constant of the circuit. The time to transfer the electrical energy is governed by the damped natural frequency of the series RCL system as shown in Eq. (1). Given that the series resistance R was measured to be approximately 25 k, C1 and C2 were 0.97 nF and 0.73 nF, and L was 1 H, the electrical energy would be transferred within 74 s. The much slower mechanical response time of VHB is in the order of 1 s [13]: t=
2
(1)
(((C1 + C2 )/(LC1 C2 )) − (R/L) )
2.2. Electrical connections Two different methods of establishing an electrical connection to the DEAs were used to investigate how the length of the carbon grease connection affects the performance of the circuit. A shorter electrode path would have lower series resistance. Fig. 2 illustrates the two configurations. The measured equivalent series resistance
Fig. 2. Cross-sectional diagrams of the DEA configurations used in the experiment. (a) High resistance: a thin carbon grease track establishes an electrical connection to the DEA’s active area. (b) Low resistance: an aluminium foil track establishes an electrical connection to the centre of the DEA’s active area.
H.C. Lo et al. / Sensors and Actuators A 212 (2014) 123–126
125
Fig. 3. Images showing the changes in DEA deformation before and after energy transfer for the low resistance configuration shown in Fig. 2b. The voltage was measured before deformation while the images show the DEA after deformation. Only a small amount of deformation occurred in DEA2 because not all the energy was transferred. Regarding the results for ‘After Transfer’, the voltages were measured before deformation occurred, while the images show the DEAs after deformation occurred.
‘Transfer Relay’. The voltages of DEA1 and DEA2 immediately before and after switching on the ‘Transfer Relay’ were recorded using a high-impedance Trek P0865 electrostatic voltmeter and a LabVIEW program. The voltages were measured before considerable changes in capacitance arising from voltage-induced deformation had occurred. A relatively low voltage was used so that the deformations were small yet observable, minimizing the changes in DEA capacitances. 3. Results and discussion 3.1. Voltage measurements Experimental results show that highly resistive electrical connections greatly reduced the amount of energy transferred. When resistance was large (Fig. 2a), charge was distributed almost equally between the two DEAs, with each DEA reaching approximately 1.4 kV. When resistance was small (Fig. 2b), DEA1 discharged to 1.2 kV and DEA2 charged to 2.0 kV. The equal distribution of charge when the resistance was large implies the resistance was so large that the inductor in the circuit had negligible effect on the amount of energy transferred. Reducing the resistance increased the energy transfer. These measurements were consistent between consecutive tests, with a variability of <1% in a series of three tests. However, performance was poorer a week later where DEA2 was only charged to 1.8 kV (down from 2.0 kV) using the configuration shown in Fig. 2b. This was likely due to the degradation of the VHB dielectric and carbon grease electrode.
incurred from energy transfer. Partial discharge may also be useful for DE generators (DEGs). In theory, a DEG achieves maximum energy generation when it is fully discharged every oscillation cycle [7,14]. However, in practice the energy consumed to fully discharge and recharge the DEG may exceed the amount of energy gained. 3.3. Energy measurements The distribution of electrical energy was determined using the measured capacitances and voltages. When series resistance was large, DEA1 and DEA2 ended up with 26% and 25% of the energy respectively. In theory, an equal energy distribution of 25% would occur if an inductor was not used. The results are similar to this, hence the inductor had negligible effect when series resistance was large. When series resistance was small, DEA1 and DEA2 ended up with 24% and 51% of the energy respectively. The allocation of electrical energy is depicted in Fig. 4. In this study, only the electrical efficiency of the energy transfer was measured. Since energy was converted between electrical and mechanical forms when deformation occurred, both the electrical and mechanical energy must be measured to determine the
3.2. Observed actuation After transferring the energy, both DEAs became partially charged. However, neither were at a high enough voltage to undergo significant actuation as shown in Fig. 3. Nevertheless, we now require less input energy to actuate DEA2. This would also be true for the partially charged DEA1, requiring less input energy to recharge and actuate. The partial discharge of DEA1 may seem undesirable, but it can also be an advantage. Since little actuation occurs below 2 kV, it would be more efficient to deactivate the DEA by partially discharging it to 2 kV and avoid some of the losses
Fig. 4. Electrical energy distribution after discharge of DEA1. Simply discharging DEA1 would result in losing 100% of the stored energy. With large series resistance (Config-1), 25% of the energy was transferred to DEA2. With small series resistance (Config-2), 51% of the energy was transferred to DEA2.
126
H.C. Lo et al. / Sensors and Actuators A 212 (2014) 123–126
overall efficiency of the energy transfer. The stored mechanical energy is difficult to measure and will depend on the DEA material and configuration. Advance measuring techniques would be required to measure these parameters. This study approximated the total energy attained by DEA2 to be equal to the electrical energy it had before a significant amount was used to deform the DEA.
[13] S. Rosset, B.M. O’Brien, T. Gisby, D. Xu, H.R. Shea, I.A. Anderson, Self-sensing dielectric elastomer actuators in closed-loop operation, Smart Mater. Struct. 22 (2013) 104018. [14] T. Hoffstadt, C. Graf, J. Maas, Optimization of the energy harvesting control for dielectric elastomer generators, Smart Mater. Struct. 22 (2013) 94028.
Biographies
4. Conclusion This work introduces a useful technique for improving the efficiency of DEs and shows the performance that can be achieved. The circuit in our experiment was able to transfer 51% of the 2.4 kV electrical energy stored in one DEA to another almost identical DEA. A total of 75% of the electrical energy was retained in the two DEAs and could be reused. However, the mechanical energy stored in the DEA was not recovered. Experiments have also shown that large series resistance can reduce the efficiency of the circuit to a point where the inductive element in the circuit becomes redundant. Reducing the series resistance of the DEA electrodes improved the circuit’s performance significantly. With careful implementation, high voltage energy transfer can greatly improve the efficiency of DE actuators and generators. Acknowledgements The work was supported by the Auckland Bioengineering Institute and the University of Auckland Doctoral Scholarship.
Ho Cheong Lo received a BE(Hons) in Electrical and Electronics Engineering from the University of Auckland, New Zealand in 2009. He is currently pursuing a PhD degree in Engineering with the Biomimetics Laboratory at the University of Auckland. His current research is focused on getting useful power from dielectric elastomer generators.
Todd Gisby was awarded his PhD from the University of Auckland in 2011 for his thesis on simultaneous sensing and actuation of dielectric elastomer artificial muscles, conducted under the supervision of Associate Professor Iain Anderson of the Biomimetics Laboratory at the Auckland Bioengineering Institute. He continued his work at the Biomimetics Laboratory in the role of Business Development Engineer, and in December 2012 co-founded and became CTO of StretrchSense Limited, a company that has commercialized wireless soft sensor technology for measuring human body motion based on the system integration and soft sensing expertise he and his colleagues developed at the Biomimetics Laboratory.
References [1] I.A. Anderson, T.A. Gisby, T.G. McKay, B.M. O’Brien, E.P. Calius, Multi-functional dielectric elastomer artificial muscles for soft and smart machines, J. Appl. Phys. 112 (2012) 041101. [2] F. Carpi, S. Bauer, D. De Rossi, Stretching dielectric elastomer performance, Science 330 (2010) 1759. [3] J. Maas, C. Graf, L. Eitzen, Control concepts for dielectric elastomer actuators, in: Proc. SPIE, 2011, pp. 79761H–79761H-12. [4] L. Eitzen, C. Graf, J. Maas, Bidirectional HV DC–DC converters for energy harvesting with dielectric elastomer generators, in: Energy Convers. Congr. Expo. (ECCE), 2011 IEEE, 2011, pp. 897–901. [5] T.G. McKay, B. O’Brien, E. Calius, I. Anderson, Self-priming dielectric elastomer generators, Smart Mater. Struct. 19 (2010) 55025. [6] R. Kornbluh, R. Pelrine, J. Eckerle, J. Joseph, Electrostrictive polymer artificial muscle actuators, in: IEEE Int. Conf. Robot. Autom., vol. 3, 1998, pp. 2147–2154. [7] R. Pelrine, R. Kornbluh, J. Eckerle, P. Jeuck, S. Oh, Q. Pei, S. Stanford, Dielectric elastomers: generator mode fundamentals and applications, in: Proc. SPIE, 2001, pp. 148–156. [8] H.C. Lo, T. Gisby, T. McKay, E. Calius, I. Anderson, Transferring electrical energy between dielectric elastomer actuators, in: Proc. SPIE, San Diego, CA, USA, 2012. [9] R. Kaltseis, C. Keplinger, R. Baumgartner, M. Kaltenbrunner, T. Li, P. MäcHler, R. Schwdiauer, Z. Suo, S. Bauer, Method for measuring energy generation and efficiency of dielectric elastomer generators, Appl. Phys. Lett. 99 (2011) 162904. [10] H.C. Lo, E. Calius, I. Anderson, Understanding efficiency limits of dielectric elastomer driver circuitry, in: Proc. SPIE, 2013, pp. 86870E–86870E-6. [11] M. Kaltenbrunner, T. Sekitani, J. Reeder, T. Yokota, K. Kuribara, T. Tokuhara, M. Drack, R. Schwodiauer, I. Graz, S. Bauer-Gogonea, S. Bauer, T. Someya, An ultralightweight design for imperceptible plastic electronics, Nature 499 (2013) 458–463. [12] G. Kettlgruber, M. Kaltenbrunner, C.M. Siket, R. Moser, I.M. Graz, R. Schwodiauer, S. Bauer, Intrinsically stretchable and rechargeable batteries for self-powered stretchable electronics, J. Mater. Chem. A 1 (2013) 5505–5508.
Emilio Calius is a senior researcher who specializes in the development and application of advanced material and structural concepts at Callaghan Innovation, a New Zealand Crown Agency. He received a PhD in Aeronautics and Astronautics from Stanford University, and has previously worked on lightweight components for spacecraft, bio-inspired mechanisms for flexible machinery, and smart structures that sense and adapt to changing loads. His current research is focused on novel ways of manipulating and controlling static and dynamic mechanical behaviour using metamaterials and other architectured materials and surfaces. Iain Anderson is Group Leader for the Biomimetics Laboratory of the Auckland Bioengineering Institute (www.abi.auckland.ac.nz/biomimetics) and an Associate Professor with the Department of Engineering Science at the University of Auckland. He completed his PhD (Engineering Science, University of Auckland) in 1996. He has worked as a whiteware product designer (Fisher and Paykel Ind.) and a vibrations consulting engineer (NZ Department of Scientific and Industrial Research, Industrial Research Ltd., New Zealand). In 2000 he joined the University of Auckland. His interest in electro-active polymer “artificial muscle” technology led to the formation of the Biomimetics Laboratory in 2005. His Lab’s research has focused on the control and self-sensing of artificial muscles for soft robotics and artificial muscle energy harvesting. The lab fosters an entrepreneurial spirit: students and engineers are encouraged to develop useful devices that are sold to other researchers and industries (www.biomimeticslab.com). In late 2012, with two of his former students, he created the company: Stretchsense Ltd. (www.stretchsense.com), a manufacturer of artificial muscle sensors.