Abstract
The invention provides an electric circuit device including a conductive element configured to provide an electrical input to an electrical component, the resistivity of the conductive element being variable based at least on the extent of deformation thereof, and a deformable body coupled directly or indirectly to, or integral with, the conductive element such that deformation of the body causes the conductive element to deform, thereby varying the resistance thereof and altering the input to the electrical component. Said devices may be used singularly or in combination as logic elements within electrical circuits, providing drive and/or control functionality. Particular embodiments provide improved electrostatic generators.
Claims
1. An electric circuit device comprising: a deformable dielectric elastomer device comprising compliant electrodes on first and second opposing sides of a dielectric elastomer layer; and a conductive track on the first or second side of the dielectric elastomer layer, the conductive track configured to receive an electrical signal and provide an electrical input to an electrical component, the resistivity of the conductive track being variable based at least on a deformation force applied thereto, the deformation force being normal to a direction from the first side toward the second side of the dielectric elastomer layer; wherein the conductive track is coupled directly or indirectly to, or integral with, the deformable dielectric elastomer device such that deformation of the dielectric elastomer device causes the conductive track to be subjected to the deformation force; and the conductive track is substantially above percolation in a first state and substantially below percolation in a second state, wherein deformation of the dielectric elastomer device causes a transition between the first state and the second state.
2. The electric circuit device of claim 1, wherein variation of the electrical input provides or triggers a predetermined response, and wherein the response includes direct drive of the electrical component by the electrical input.
3. The electric circuit device of claim 1, wherein cooperation of the dielectric elastomer device and the conductive track controls the electrical component, wherein variation of the electrical input provides or triggers a predetermined response, and the response providing a control signal to the electrical component.
4. The electric circuit device of claim 3, wherein the dielectric elastomer device, the conductive track, and the electrical component cooperate to provide closed loop control of the electric circuit device.
5. The electric circuit device of claim 4, wherein the electrical signal is generated at least in part by the electrical component and assists at least in part with the closed loop control.
6. The electric circuit device of claim 1, wherein variation of the electrical input provides or triggers a predetermined response, and wherein the response comprises generating a desired deformation of the conductive track.
7. The electric circuit device of claim 6, wherein the dielectric elastomer device is configured to generate the deformation.
8. The electric circuit device of claim 1, wherein the dielectric elastomer device is deformable in response to application of a voltage thereto or thereacross.
9. The electric circuit device of claim 1, wherein application of a voltage to the compliant electrodes causes the dielectric elastomer device to deform.
10. The electric circuit device of claim 1, further comprising the electrical component.
11. The electric circuit device of claim 10, wherein the electrical component is integrated with the dielectric elastomer device.
12. The electric circuit device of claim 1, further comprising a plurality of dielectric elastomer devices, each of the plurality of deformable dielectric elastomer devices comprising compliant electrodes on first and second sides of a dielectric elastomer.
13. The electric circuit device of claim 12, wherein a first and a second of the plurality of dielectric elastomer devices are formed by at least three compliant electrodes provided on a single dielectric elastomer body, and wherein each of the first and the second of the plurality of dielectric elastomer devices comprises at least a first and a second of the at least three electrodes such that application of a voltage to the electrodes causes the dielectric elastomer device to deform, first and second electrodes of the at least three electrodes operatively coupled to form the first deformable dielectric elastomer device and the first and a third electrodes of the at least three electrodes operatively coupled to form the second deformable dielectric elastomer device.
14. The electric circuit device of claim 1, comprising one or more further conductive tracks and/or functionally distinct portions of the same conductive track, each being configured to provide an electrical input to the electrical component or a further electrical component, the resistivity of each conductive track or portion thereof being variable based at least on the extent of deformation thereof.
15. The electric circuit device of claim 14, wherein first and second ones of the conductive tracks are directly or indirectly electrically and/or mechanically coupled.
16. The electric circuit device of claim 14, wherein each conductive track is configured to be deformed by a respective one or more of the deformable dielectric elastomer devices.
17. The electric circuit device of claim 1, wherein the conductive track is electrically coupled to the dielectric elastomer device.
18. The electric circuit device of claim 1, wherein the dielectric elastomer device is configured to act as a capacitor.
19. The electric circuit device of claim 18, wherein mechanical deformation of the dielectric elastomer device varies the capacitance thereof.
20. The electric circuit device of claim 19, wherein the electric circuit device is configured to act as or form part of a charge pump or a generator.
21. The electric circuit device of claim 20, configured such that: when the dielectric elastomer device deforms or is deformed in a first manner, the resistance of the conductive track changes and results in electrical energy being stored by the dielectric elastomer device as a response; and/or when the dielectric elastomer device deforms or is deformed in a second manner, the resistance of the conductive track changes in a different manner and results in electrical energy being drawn away from the dielectric elastomer device as a response.
22. The electric circuit device of claim 21, comprising a second conductive track and/or a second dielectric elastomer device, the second conductive track and/or dielectric elastomer device configured to cooperate with the first conductive track and/or the first dielectric elastomer device, to perform said storing and/or drawing away electrical energy as said response.
23. The electric circuit device of claim 21, including a second dielectric elastomer device, wherein: the first dielectric elastomer device comprises first and second electroded zones electrically coupled via first and second elements; and the second dielectric elastomer device comprises first and second electroded zones and a third element, the third element being electrically coupled to the first and second electroded zones of the first dielectric elastomer device.
24. The electric circuit device of claim 1, wherein the conductive track is coupled to or configured to couple to an electrical power supply, and wherein the conductive track and the dielectric elastomer device are configured such that deformation of the dielectric elastomer device deforms the conductive track to control the current from said power source as a response.
25. The electric circuit device of claim 24, wherein the conductive track is directly or indirectly electrically coupled to the electrical component, thereby varying the current provided thereto.
26. The electric circuit device of claim 1, configured to act as a logic device.
27. The electric circuit device of claim 1, configured to act as any one or more of an AND gate, a NOT gate, a NAND gate, a voltage clamp, a diode, a summing junction, a neuron or weighted summing junction, a latch, a buffer, an operational amplifier, a frequency multiplier, a Schmitt trigger, an oscillator, and a low current driver, wherein variation of the electrical input provides or triggers a predetermined response, and wherein the response comprises a transformation of one or more input signals to one or more output signals according to a predetermined logic operation.
28. The electric circuit device of claim 1, further comprising means for inhibiting deformation in one or more directions of the dielectric elastomer device and/or the conductive track.
29. The electric circuit device of claim 28, wherein the means for inhibiting comprises one or more ribs, frame elements or areas of increased stiffness.
30. The electric circuit device of claim 1, comprising means for shielding, at least in part, the dielectric elastomer device and/or said conductive track from deformation from at least one direction.
31. The electric circuit device of claim 30, wherein the means for shielding comprises one or more frame elements or regions of increased stiffness.
32. A circuit comprising at least two electric circuit devices of claim 1.
33. The circuit of claim 32, wherein at least first and second ones of the at least two electric circuit devices are electrically and/or mechanically coupled.
34. The circuit of claim 32, configured to provide drive and/or control of an electric motor or a mechanosensitive cilia array.
35. A charge pump and/or a generator comprising at least one electrostatic generator operatively coupled to at least one electric circuit device as claimed in claim 1, whereby in a first state the at least one electric circuit device directs current to a load and/or in a second state the same or another said at least one electric circuit device routes at least a portion of the energy generated to a storage means, the storage means being adapted to provide charge to the or another said electrostatic generator.
36. The electric circuit device of claim 1, wherein the conductive track is on the first side of the dielectric elastomer device.
37. The electric circuit device of claim 1, wherein the conductive track comprises a plurality of terminals for at least one of the terminals configured to connect to the electrical input and a second of the terminals configured to connect to the electrical component.
38. A method of driving and/or controlling an electrical component, the method comprising: providing an electrical signal to a conductive track; and deforming a dielectric elastomer device comprising compliant electrodes on first and second sides of a dielectric elastomer by applying a voltage across the compliant electrodes, the deformable dielectric device having the conductive track associated therewith, said deformation causing the resistivity of the conductive track to change based at least on the extent of deformation thereof; and providing an electrical input to the electrical component, said electrical input being dependent on said electrical signal and said deformation.
39. A logic element comprising: at least one dielectric elastomer device comprising compliant electrodes on first and second opposing sides of a dielectric elastomer layer and at least one conductive track coupled on the first or second side of the dielectric elastomer layer, wherein the at least one conductive track is configured such that application of a deformation force thereto affects the electrical resistance thereof, the deformation force being normal to a direction from the first side toward the second side of the dielectric elastomer layer; and the at least one conductive track is substantially above percolation in a first state and substantially below percolation in a second state; and the deformation force causes a transition between the first state and the second state.
40. The logic element of claim 39, wherein said at least one dielectric elastomer device and said at least one conductive track cooperate to transform one or more inputs into one or more outputs, said outputs being predetermined but dependent on the one or more inputs.
41. A computing device comprising one or more logic elements of claim 39.
42. A dielectric elastomer switch comprising: a deformable body comprising a dielectric elastomer, wherein the dielectric elastomer body is formed from a dielectric elastomer provided with at least first and second electrodes coupled or integral thereto, such that application of a voltage to the electrodes causes the body to deform; a path comprising a plurality of conductive particles, the path integral with the deformable body such that deformation of the deformable body causes deformation of the conductive path; and the path having a first state wherein the plurality of conductive particles are substantially above the percolation threshold and a second state wherein the plurality of conductive particles are substantially below the percolation threshold; and wherein the deformation of the path, caused by the deformation of the deformable body, causes a transition between the states.
43. The dielectric elastomer switch of claim 42, wherein the dielectric elastomer body is non-planar.
44. The dielectric elastomer switch of claim 42, wherein the path is formed in or through the dielectric elastomer body.
45. The dielectric elastomer switch of claim 42, wherein the dielectric elastomer body comprises a boundary member to shield the dielectric elastomer body from deformation from one or more directions.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
(1) Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings in which:
(2) FIGS. 1A-B are side views of an electrical transformer;
(3) FIGS. 2A-B are schematic and circuit diagrams of a DED arrangement for a sensor;
(4) FIGS. 2C-D are schematic and circuit diagrams of a DED arrangement providing sensing and driver functionality;
(5) FIGS. 3A-B are schematic and circuit diagrams of a DED arrangement providing logic functionality;
(6) FIG. 3C is a schematic diagram of the interaction between DESs and DEAs;
(7) FIG. 4A shows an experimental setup used to assess variation of resistance of a resistive element undergoing deformation;
(8) FIG. 4B illustrates results obtained using the setup of FIG. 4A;
(9) FIGS. 5-7 are perspective, schematic and circuit diagrams of one embodiment;
(10) FIGS. 8A-B are perspective and circuit diagrams of an alternative arrangement to that of FIGS. 5-7;
(11) FIG. 9 is a chart of nominal Maxwell pressure versus displacement;
(12) FIG. 10 is a diagram illustrating advantages that may be conferred by the invention;
(13) FIG. 11 illustrates an example layout of a DES incorporated within a motor;
(14) FIGS. 12-13 illustrate embodiments configured as NOT gates;
(15) FIG. 14 illustrates an embodiment configured as an AND gate;
(16) FIG. 15-16 respectively illustrate an embodiment configured as a voltage clamp and an example current-voltage curve derived therefrom;
(17) FIG. 17A-B illustrate alternative diode embodiments;
(18) FIG. 18A-C illustrate a DEG arrangement and self priming circuits therefor;
(19) FIG. 19 illustrates an embodiment configured as a summing junction;
(20) FIG. 20 illustrates an embodiment configured as a neuron;
(21) FIG. 21 illustrates an embodiment for use in a mechano-sensitive cilia array;
(22) FIG. 22 illustrates an embodiment configured as a low voltage current driver;
(23) FIG. 23A illustrates an embodiment configured as a latch;
(24) FIGS. 23B-C illustrate an alternative embodiment of a latch;
(25) FIG. 23D shows the circuit of FIG. 23C provided with a buffer;
(26) FIGS. 23E-F show results obtained using the arrangement of FIGS. 23B-C;
(27) FIG. 24 illustrates an embodiment configured as a buffer;
(28) FIG. 25 illustrates an embodiment configured as an opamp;
(29) FIG. 26A-B respectively illustrate an embodiment configured as a frequency multiplier and a possible voltage curve derived therefrom;
(30) FIG. 27A illustrates an alternative embodiment of a frequency multiplier;
(31) FIGS. 27B-C provide schematic and circuit diagrams of another embodiment of a frequency multiplier;
(32) FIGS. 28-32B illustrate further embodiments of the invention;
(33) FIGS. 33A-C illustrate a setup, circuit and results for an inverting Schmitt trigger;
(34) FIGS. 34A-B show a ring oscillator;
(35) FIGS. 34C-D show results obtained from the arrangement of FIGS. 34A-B;
(36) FIG. 34E shows an embodiment configured to function as a NAND gate;
(37) FIGS. 35A-C are alternative views of an embodiment of the invention;
(38) FIG. 36 is a chart of the output voltage obtained using the arrangement of FIG. 35;
(39) FIG. 37A shows a mechanical switch-based self-priming arrangement;
(40) FIG. 37B shows an equivalent arrangement to FIG. 37A of another embodiment;
(41) FIG. 38A shows an electrostatic generator arrangement;
(42) FIG. 38B is an equivalent arrangement to FIG. 38A of another embodiment;
(43) FIG. 39 shows arrangements used to test the working of particular embodiments;
(44) FIG. 40A schematically shows an alternative to the embodiment of FIG. 35B; and
(45) FIGS. 40B-D schematically show alternative self-priming circuits.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
(46) As previously discussed, FIGS. 2C and 2D illustrate a simple example arrangement embodying the invention, whereby piezoresistivity is used by a DES to control a small dielectric elastomer actuator holding a mass. Such an arrangement may be used, for example, as a remote actuation device. A user pulls on the ring stretching the piezoresistive element, labelled Rs, causing the resistance to climb rapidly from, say 20 M to 500 M. Electrically, the switching resistor Rs is the discharge part of a voltage divider driving the actuator. When a voltage of 3000V is applied to the high voltage terminal (HV), the voltage applied to the actuator will transition from approximately 1140V to 2730V as the user pulls the ring from the non-stretched to the fully stretched state. The result of this is that the weight moves up and down in synchronization with the motion imparted by the user.
(47) The piezoresistive element Rs may be formed as a conductive track on a dielectric elastomer. There are a number of demands on the design of the resistive element. Namely, it must be configured to allow the correct levels of current therethrough (typically less than 1 mA); be tolerant of high voltages (e.g. greater than 4 kV); vary in resistance due to geometric changes and/or percolation effects; preferably vary in a repeatable manner over time; be compliant up to large strains (e.g. hundreds of percent); be easy to tailor in density; and be easily applied or configured in any desired pattern. It is to be appreciated that these are design considerations and that some level of compromise between the factors will need to be arrived at for any particular implementation.
(48) The inventors have developed and used a carbon grease mix that at least in part meets these needs. One presently preferred constitution includes a mix of Molykote 43G grease and Cabot Vulcan XC72 carbon black, preferably at a ratio of approximately 5:1 (grease:carbon) by weight. The mix essentially forms a paste that can be readily handpainted on to a substrate. It will be appreciated that other techniques could be used, such as deposition, ion implantation, quantum tunnelling composites and carbon nanotubes. Other techniques will also be envisaged.
(49) In the embodiment shown in FIG. 4A, a substantially planar body 20 formed from VHB 4905 (a common, commercially available acrylic material from 3M) was pre-stretched to three times its original size in all directions within the plane of the body 20 and affixed to a 3 mm thick plastic frame 22 with an 80 mm diameter window 24, the frame 22 in turn being supported by frame 25. A 50 mm diameter, 3 mm thick plastic disk 26 was fixed to the centre of the body 20.
(50) A stepper motor 27 was used to displace the centre of the disk 26 via dowel 28. Four approximately 30 mm long M-shaped tracks 14 were painted on to the body 20 substantially within the area bound by window 24 and disk 26. The resistance of each track was monitored as it went through a single 0-15 mm sinusoidal displacement at 0.75 Hz.
(51) The results for this taken one day after fabrication are displayed in FIG. 4B. The change in resistance is clearly considerable, and over the 15 mm displacement, sufficient to enable switching of a small DED of a few nanofarads. The variation between samples (possibly due to the handmade process) and the hysteresis across the full displacement range, is negligible when compared to the overall variation and not of concern where it is merely necessary to determine between an on or off state.
(52) FIG. 5 illustrates an arrangement designed and used by the inventors to demonstrate the capabilities of the invention. The setup includes a circular track defined by rails 40 that are mounted on support 41. The support 41 is mounted on DEDs 42a, 42b and 42c via disks 26 (see FIG. 6) and pillars 44. The DEDs 42 were supported by frame 45. The arrangement of FIG. 5 was used to move the ball 46 along the track by appropriate actuation of DEAs incorporated within the DEDs 42. The control to effect this motion was achieved by augmenting the DEDs 42 with DES functionality such that they inherently provided the sensor feedback, logic and driver circuitry.
(53) FIG. 6 is a schematic representation of the active portion of the apparatus of FIG. 5. The DEDs 42 of FIG. 6 include DEAs (dielectric elastomer actuators) formed from actuator sections C.sub.1-3 and DESs R.sub.1a-3a and R.sub.1b-3b including tracks 14. The electrical connections therebetween are also shown. FIG. 7 provides equivalent circuit diagrams.
(54) Each DED 42 was configured as shown in FIG. 6, with the tracks 14 formed in a similar manner to those of FIG. 4A and the body 20 subjected to a similar level of pre-stretching. The actuator sections C.sub.1-3 were formed using Nyogel 756G carbon grease. While only one side of the apparatus is shown in FIG. 6, each actuator section C.sub.1-3 preferably comprises a pair of electrode layers spaced apart by the dielectric elastomer body 20. It will be appreciated that alternative electrode designs could be employed.
(55) The actuator sections C.sub.1-3 had a capacitance of approximately 2 nF and are shown split into two halves, with the tracks 14 positioned therebetween (see FIG. 6), such that each device included two switching sections. The first switching section of each device is marked R.sub.1a-3a and controls the charging current of the next actuator in an anticlockwise direction. The second switching section in each device is marked as R.sub.1b-3b and controls the discharge current of the next actuator in the clockwise direction.
(56) While the operation will be described in more detail, essentially, the downwards displacement of any one actuator will activate the actuator in the clockwise direction and deactivate the actuator in the anticlockwise direction. More particularly, on application of 4 kV to the high voltage terminal Hv and positioning of the support 41 and rails 40 on the DEDs 42, all three DEDs displace downwards approximately 8 mm under the weight being supported. Referring to FIG. 4B, the resistance of the tracks 14 will be approximately 50 M. From FIG. 7, it can be determined that the voltage applied to each actuator C.sub.1-3 will be approximately 2 kV. The application of the voltage causes the actuators to actuate such that the disks 26 are urged downwards. In the example arrangement, the centres sink to a 9 mm displacement, resulting in the resistances increasing to approximately 80 M, but the applied voltage remaining substantially stable at 2 kV. Placement of the 35 g ball 46 on the rails directly above actuator C.sub.1 displaces the disk 26 of that DED to 11 mm. The deformation of the other two disks 26 changed to approximately 8 mm. Table 1 presents the instantaneous effect of these displacements resulting from placement of the ball above actuator C.sub.1, the values for which can be determined using FIGS. 4B and 7.
(57) TABLE-US-00001 TABLE 1 Designator Approximate Value R.sub.1a, R.sub.1b 700 Mohm R.sub.2a, R.sub.2b, R.sub.3a, R.sub.3b 50 Mohm VC.sub.1 2000 V VC.sub.2 266 V VC.sub.3 3733 V
(58) The increased voltage on C.sub.3 causes it to actuate to 12 mm displacement and the decreased displacement on C.sub.2 causes it to relax back to 7 mm displacement. The rails 40 therefore tilt and the ball rolls towards C.sub.3. As the ball rolls around the rails, the tilt of the rails will follow it, accelerating until it reaches a maximum speed. This analysis is illustrative only. In reality, the system is highly dynamic and its states will change continuously over time. The example apparatus operated at 0.7 Hz, corresponding to a maximum ball velocity of approximately 0.5 ms.sup.1. Changing the supply voltage provided an effective mechanism of controlling the speed of the array which means that either the array is rate limited by the charging and discharging currents through the switches, or that increasing the voltage shifted the switches to a more sensitive average displacement.
(59) According to an alternative setup, each of the three DEDs of FIG. 6 was reconfigured to include only a single switch, with the actuators C.sub.1-3 being provided substantially completely around the disks 26 with only one void for the respective switch. This arrangement is illustrated in FIG. 8A, in which terminal A connects to the DES, terminal B connects to the electrode provided on the upper surface of the DEA and the DES, and terminal C connects to the electrode on the lower surface of the DEA.
(60) The equivalent circuit for the three DEDs connected together is provided in FIG. 8B. The switching elements are arranged as shown in FIG. 8 so that each forms a voltage divider with a 50 M fixed resistance. As the central coupling point is displaced out of the plane, the resistance of the switching unit increases substantially due to percolation and geometric effects, this has the effect of raising the output voltage. Terminal A of each of the three phases is connected to terminal C of the next phase in the array so that as each actuator displaces down under the weight of the ball it activates the next, tilting the platform and rolling the ball forward. In this way the ball will continue to roll around the track as long as power is applied.
(61) The nominal Maxwell pressure for each phase using an array of devices according to the embodiment of FIG. 8A with a 4 kV supply and undergoing a sinusoidal 0-15 mm, 0.75 Hz displacement is provided in FIG. 9. The Maxwell pressure is the actuation pressure of the actuator and can be given by P.sub.max=.sub.o.sub.rE.sup.2, where .sub.o and .sub.r, are the absolute permittivity of free space and relative permittivity of the material respectively and E is the electric field strength. A nominal thickness of 55 micron was chosen for the dielectric elastomer membrane and a relative permittivity of 4. Note that improved repeatability and/or accuracy and/or reliability can be achieved by appropriate design of at least the conductive tracks and the results provided in FIG. 9 are only intended to be illustrative and in not limiting of the capabilities of the invention.
(62) The arrangements of FIGS. 5-8B do not merely operate in a predetermined sequence. The arrangements detect the position of the ball and control the tilting of the rails accordingly. This is generally not possible without external control circuitry. Furthermore, the response by the actuators is dependent on the location of the ball on the platform. For smooth operation, the actuation signal is strongest when the ball is over the particular element. Sufficient current is required to drive the actuators to respond quickly.
(63) A more conventional arrangement would possibly use an external sensor placed on each device to detect the location of the ball. A microcontroller could have been used to calibrate the sensor data and to implement some control law. High voltage driver circuits could have been used to charge and discharge the actuators. While this arrangement may achieve the same end result, it adds cost, volume, mass and complexity to the system. These limitations may be tolerable when a small number of actuators are used or where precision in the system is paramount, but they are especially problematic when a large array of actuators are required, or when cost is a bigger factor than precision.
(64) A further problem with conventional approaches is that it is very difficult to create sensors flexible enough to take full advantage of the large strokes that dielectric elastomers are capable of. The addition of self-sensing functionality, whereby the capacitance or resistance of the actuator are monitored in real time and related to actuator displacement, goes some way to solving this but current approaches still require additional circuitry, such as microcontrollers to implement the control law, and driver circuits. Such circuitry may be eliminated from arrangements according to the present invention. The extent to which component parts can be reduced is schematically represented in FIG. 10.
(65) Thus, embodiments of the invention provide for the integration of analogue and digital or Boolean circuitry into the dielectric elastomer actuator, generator, sensor, damper or hybrid device. This integration means that the switching circuitry scales well with the device size or number of devices.
(66) The oscillators of FIGS. 5-8B are analogue-based, meaning they require careful tuning and are intolerant to degradation or imperfections in fabrication. Further, they are based on a mass-spring system, which may not be suitable for miniaturization or integration into an arbitrary circuit,
(67) FIGS. 34A-B show a ring oscillator that goes at least some way to addressing these problems, firstly by switching to being digital-based and secondly by removing or at least reducing reliance on mass-spring arrangements. The advantages of the arrangement of FIGS. 34A-B over previous arrangements are summarized below: 1) As a digital element, it should be more resistant to fabrication imperfections/degradation; 2) As an electric, rather than mechanical time constant-based device it should be easier to scale/tune to a particular application and be more tolerant of background mechanical vibrations; 3) It provides a good tool for assessing the life expectancy of high voltage DES devices as well as testing the maximum speed of its constituent inverters; 4) It is simple and requires a small number of components to implement.
(68) The oscillator was formed out of 6 subunits, one of which is shown in FIG. 34A. Each subunit consisted of a DES and a two-part DEA formed on 3.5 equibiaxially prestretched VHB 4905 adhered to a 3 mm thick, 75 mm inner diameter Perspex ring. The DEA electrodes were Nyogel 756G. The DES electrodes were a 5:1 mix of Molykote 44 Medium grease and Cabot Vulcan XC72 carbon black applied during stretching so that they were 35% stretched, and highly non-conductive at rest. The two halves of the DEA were connected electrically and when actuated, compressed the central DES, causing a substantial reduction in resistance.
(69) FIG. 34B shows how the subunits were paired to form three inverters that made up the ring oscillator. An equivalent circuit of one subunit is boxed off in the top left hand corner of FIG. 34B. One subunit in each inverter was referenced against a 2800V supply, and the other subunit against ground. The inputs of each subunit were tied together so that when the input was high, the ground-referenced subunit was active, dragging the inverter output low. When the input was low, the supply-referenced subunit was active, dragging the output high.
(70) The oscillator was powered from a Biomimetics Lab 4 channel EAP control unit. To provide enough current a WYE configuration of 15 kOhm resistors was used to tie three 120 A channels together. To understand the operation of the oscillator, consider an initial condition of all inverters being at half the voltage of the supply. Some small imperfection will cause the output of the 1.sup.st inverter to be a bit higher than half. This will cause the output of the second to be lower than half, then the output of the 3.sup.rd to be higher than half, then the output of the 1.sup.st to be lower than half, and so on (note that other odd numbers of inverters can be used to achieve this functionality). It can be seen here that the gain of the gates must be greater than ()1, or else oscillations will decay to zero, rather than grow. The oscillations build up until the inverters are saturated and the ring oscillator has stabilized.
(71) In testing the supply was switched on and the output connected to a 5 GOhm input impedance voltage probe. Data was acquired using a national instruments DAQ card and Labview 8.6. Table 2 summarises the performance of the oscillator with FIG. 34C showing supply and output voltages against time for the first 50 s of operation. The amplitude and frequency of oscillation built over time until settling into a stable, saturated pattern at around 40 s. FIG. 34D shows the ring stably oscillating at a little over 1 Hz, 100 s into operation with a substantially square wave, demonstrating that the loop gain was greater than one and that the switching delay of each inverter stage was approximately 333 ms.
(72) TABLE-US-00002 TABLE 2 Cycles until failure ~140 Frequency of operation ~1 Hz Switching delay per inverter ~333 ms Amplitude 100% of supply (2800 V) Waveform ~Square
(73) The ring oscillator is particularly well suited to driver applications such as multiphase motors or pumps as each inverter stage can act as an out of phase output. For acting as a clock for artificial muscle digital circuitry, the oscillator does not require a spring mass system and should exhibit lower susceptibility to environmental vibrations and be easier to miniaturise.
(74) An example application is illustrated in FIG. 11, which illustrates a configuration which may be used to generate rotary motion, for example, using a suitably configured crank mechanism coupled substantially centrally in the illustrated device. The arrangement includes three DEAs 12 each coupled to a frame 45. Appropriate actuation/deactivation of the DEAs 12 enables the drive element 110 to move in a desired manner. Between each of the DEAs 12 is provided a DES 10 for switching the adjacent element when any particular element is stretched. Consequently, this arrangement can be used to provide a substantially planar actuator and associated control, which may, for example, be used as a rotary motor.
(75) Further example devices that may be realised using the DES concept of the invention are described hereinafter. Generally, these realizations fall into one of four categories: 1) Bounded membrane where a stretched dielectric elastomer membrane is held in a frame with fixed or moving boundary conditions. These devices are relatively simple to make. A bounded membrane may be included within a larger membrane or a portion of a membrane may be bounded off from other portions thereof. The larger membrane may or may not be bounded. An example of such an arrangement is provided in FIG. 32A where a bounded portion 321 included with a larger dielectric elastomer body 322 is at least partially shielded from the influence of strains due to deformations imparted by elements outside of the bounded portion 321. In such an arrangement, the bounded portion 321 may contain logic circuitry as described hereinbelow. FIG. 32B shows an alternative arrangement in which strains imparted by elements external to the bounded portion 321 are selectively permitted by selective absence of portions of the frame. Note that the concept of using frames is not limited to two dimensions and may be provided between vertically stacked DEDs. It is also not limited to a bounded portion being located within a larger dielectric elastomer body. For example, discrete bodies may be positioned side by side or in any other arrangement where deformation of one element can impact on another element. 2) Infinite membrane where a very large stretched membrane is used compared to the size of the DES elements. This is a useful layout for complex systems of many switches. Designs can be substantially different to the bounded membrane case. 3) Multilayer membrane where multiple stretched dielectric elastomer membranes are stacked on top of each other. This covers the bounded or infinite membrane case. Multilayer solutions can provide improved transmission of switching action between elements as the strain field is the same on two closely bonded membranes. One difficulty inherent in this design is the creation of vias between membranes. 4) General Case where there exists some 3 dimensional piece of dielectric material with a network of dielectric elastomer devices and switches arranged throughout the bulk of the material. Such an arrangement can be viewed as a soft computer or Smart Blob.
(76) Unless otherwise stated or apparent, the underside of the membrane or deformable body in the arrangements described below is provided with a ground electrode.
(77) NOT Gate
(78) A NOT gate produces the negation of its input. If a high voltage signal is connected to the input then the output will drop low. If nothing is connected to the input the output will be high. An example of a bounded membrane type configuration providing NOT gate functionality is provided in FIG. 12. The NOT gate consists of a DEA element adjacent to two below percolation tracks. When the DEA is inactive, the output will be at some value between HV (high voltage) and ground as determined by the voltage divider formed by the top track 121 and the track 14. When the DEA is activated by applying a high voltage to the In terminal, the track 14 will compress substantially such that the two vertical arrays close together, causing the resistivity to drop. This will decrease the output, as required to perform as a NOT gate.
(79) FIG. 13 shows an example of a multilayered NOT gate. Two dielectric elastomer membrane layers 131 are laminated between three electrode layers 132. The lower membrane has a DEA element across it that will expand with the application of a high voltage to the In terminal. The upper membrane will expand in the same place as the lower one since they are laminated together. In this expanding region, an above percolation track is applied. As the area stretches, the track will stretch too, reducing in conductivity markedly and cutting of the high voltage terminal HV from the output. A pull down resistor track is applied away from the local stretch field of the actuator element.
(80) AND Gate
(81) An AND gate has 5 terminals, two inputs, an output, a power supply and ground. An example bounded membrane type configuration for an AND gate is provided in FIG. 14. As is common in the art, the membrane is coupled to the frame such that it is under strain. The input terminals (In1 and In2) are connected to respective DEAs so that when the inputs are high, the DEAs expand. Adjacent to each DEA is a section of below percolation (non-conducting) track. The expansion of the DEA causes this track to compress and begin to conduct. When both DEAs are active the entire track will conduct and the HV supply will be connected to the output. A pull down resistor 141 is also shown.
(82) It will be appreciated that further DEA bodies may be provided to receive a corresponding input, each required to be active to generate a high output from the track.
(83) It should be noted that a NOT gate and an AND gate are the only primitives required to form a Boolean computer. In principle, it should therefore be possible to form a dielectric elastomer computer. Such computers have the huge advantage that they can be seamlessly integrated with other dielectric elastomer devices such as actuators, generators and sensors.
(84) NAND Gate
(85) It is also possible to build a NAND gate using the repeating unit shown in FIG. 34A. NAND gates can also be used to form any Boolean circuit. A NAND gate design is provided in FIG. 34E and consists of four said units of FIG. 34A arranged in the circuit shown. The operation of the NAND gate can be summarised in the following way: 1. If both input 1 and input 2 are low, units A and B will be off and units C and D will be on. The output will thus be pulled to the High Voltage (HV) supply, and will be high. 2. If either input 1 or input 2 is high (but not both), either unit A or B will be off, and unit C or D will be on. The output will be high. 3. If both input 1 and input 2 are high, both units A and B will be on, and C and D will be off, and the output will be low.
Voltage Clamp
(86) A voltage clamp, as shown in FIG. 15, has three terminals, one input (In) and two grounds (Gnd). The voltage to be clamped is applied to the input. As the voltage rises the DEA begins to actuate. As the DEA actuates it compresses the below percolation track, raising its conductivity. At some critical voltage, the track starts to draw significant charge off the DEA. As the voltage increases, the current required to maintain it rises rapidly. The effect could be very sensitive as the DEA will actuate with voltage squared and percolation is also a highly nonlinear phenomenon. This could be a suitable device for protecting overvoltaging in a small dielectric elastomer generator. FIG. 16 shows an example current-voltage curve for a dielectric elastomer voltage clamp. Above the critical voltage, the amount of current required grows exponentially.
(87) Diode
(88) A diode can be formed using a bounded membrane type device as shown in FIG. 17A. The membrane is segmented into three zones: 1) a forward bias DEA (FWD DEA); 2) a reverse bias DEA (REV DEA); 3) a dead zone with the switching element.
(89) The rear side of the diode is grounded. When the diode is in forward bias (the FWD terminal is at a higher potential than the REV bias terminal), the FWD DEA will expand more than the REV DEA. This will cause the centre dead zone with the percolating track 14 to move towards the REV DEA side and compress the track, lowering its resistance. When the diode is in reverse bias the REV DEA will expand more than the FWD DEA, increasing the resistance of the percolating track as it is stretched.
(90) This diode is particularly useful for dielectric elastomers configured to act as generators. For example, FIG. 18 shows a self-priming circuit for a generator taken from New Zealand patent application no. 578275. The circuit of FIG. 18 includes two capacitors of capacitance C and three diodes arranged as illustrated. When a capacitor bank supplies a priming voltage to a Dielectric Elastomer Generator (DEG), current flows from node B to node A. Because of the diodes, when current flows in this direction, the capacitors are effectively arranged in parallel as far as current flow is concerned, meaning the energy is in a high charge, low voltage form. To transfer energy back from the DEG to the capacitor bank, current flows in the direction of node A to node B. When current flows in this direction, the arrangement of the diodes means the capacitors are effectively in series, meaning the energy is in a high voltage, low charge form. In this arrangement, diodes are used to divert a portion of the energy (i.e., charge) generated in a particular cycle, enabling it to be used to charge or prime the dielectric elastomer for a subsequent cycle. With the new diode design presented herein, a self-priming generator could be formed substantially entirely from DEDs, providing a substantially entirely polymer-based, stretchy, self-priming, high-voltage, shock tolerant, and efficient generator.
(91) FIG. 17B shows an alternative diode design that is largely identical to that of FIG. 17A and only differences are described. The embodiment of FIG. 17B includes ribs 171 that serve to constrain the DES to deform substantially only in the length direction. The ribs may be on the underside of the dielectric elastomer body. While not shown, the ribs 171 may be provided for other embodiments of the invention (e.g. in the arrangements of FIGS. 24-25).
(92) Neuron
(93) An artificial neuron produces an output corresponding to a series of weighted inputs. In a sense, it is a weighted summing junction. Such a device could be created using multiple DESs to control the total current moving into a junction. The first step in the design is a summing junction, which is shown in FIG. 19. Each input to the junction terminates at a DEA element. Increasing the voltage causes the DEA to stretch more. On the top electrode layer there is a just-above percolation track above each input DEA. The stretching of the DEA causes the track to increase substantially in resistivity. The material could be tuned to enable the transition about the percolation point to occur over the full actuation range of the DEA. This increase in resistivity controls the current that each track allows into the large central DEA 191. The discharge resistance is held constant with the pull down resistor and the voltage across the large DEA 191 will be the balance between the different charging and discharging currents. The large DEA 191 then controls the current through a final switch which acts to invert the output, providing a summing junction, not an inverse summing junction.
(94) To make a neuron we need to be able to weight the inputs. This can be achieved by adding a second set of inputs corresponding to the weights. A two input neuron is illustrated in FIG. 20. The voltage on the two DEAs in the centre of the membrane is proportional to the ratio of the voltages applied to the input and weight terminals. The voltages on the two middle DEA 201 are summed and the sum is the output.
(95) Mechanosensitive Cilia Array
(96) A mechanosensitive DEMES (Dielectric Elastomer Minimum Energy Structure) array made using DES is shown in FIG. 21. DEMES arrangements have been described by B. O'Brien, T. Gisby, S. Xie, E. Calius, and I. Anderson. In FEA of dielectric elastomer minimum energy structures as a tool for biomimetic design, Proc. SPIE 7287 (2009); by B. O'Brien, T. McKay, E. Calius, S. Xie, and I. Anderson in Finite element modelling of dielectric elastomer minimum energy structures, Applied Physics A: Materials Science & Processing 94, 507-514 (2009); and by B. O'Brien, E. Calius, S. Xie, and I. Anderson in An experimentally validated model of a dielectric elastomer bending actuator, Proc. Spie 6927 (2008).
(97) Each DEMES has two sides, a high voltage side HV and a ground side GND. The upper left hand portion of FIG. 21 shows the high voltage underside of the nth DEMES in the array, and the upper right hand portion shows the ground side of the n+1th DEMES. Down the centre of each side of the DEMES between the electrodes of the DEDs is an above percolation track that conducts when the DEMES is in its equilibrium state, and becomes non-conducting when the DEMES moves towards the flattened state. The high voltage side of the nth DEMES forms the discharge resistance for the nth DEMES. The ground side of the n+1th DEMES forms the charging resistance for the nth DEMES. The discharge track is shorter than the charging track to ensure that the DEMES is substantially off in the equilibrium state. A side sectional view of the array of DEMES is shown in the lower portion of FIG. 21. The operation of the array is as follows: The tip of the first DEMES is displaced forward by an external load, the discharge path on the DEMES is stretched, raising the voltage on the DEMES and causing it to actuate to hold its displacement. The 1st DEMES pushes on the second DEMES, causing it too to actuate in the same manner as the first. There is an additional effect however; the actuation of the second DEMES stretches the charging resistance of the 1st DEMES, causing the voltage to drop back to low levels, turning the 1st DEMES off. The wave of actuation propagates down the array. If there is a load on the array (such as the ball on rails shown) then the wave speed will be mediated by the load forming a useful conveyor/propulsion system.
Low Current Driver
(98) As DES are essentially variable resistance switches, they can be used to directly drive low voltage devices. An example arrangement embodying this concept is shown in FIG. 22 where the current through an electric motor is controlled with the resistance on the ground side of a DEA/DES. Another application is for the tuning of a variable resistance for other devices. A variable filter could, for example, be created where the voltage on a DEA is used to control the resistance of a filter component.
(99) Latch
(100) DES/DEA pairings in a state of positive feedback can be used to create a memory element. That is, activation of the actuator deforms the switch so that the voltage on the actuator rises further, in turn deforming the switch ever more until the system reaches saturation. In addition to a state retention property, the memory element also requires an ability to set and read the state of the device.
(101) It is possible to make a latch using a DES and an example arrangement is shown in FIG. 23A. When the SET DEA is actuated, the track 14a below it will be compressed, raising the voltage on the Q DEA. The Q DEA will in turn actuate and compress the track 14b under it. The compressed track 14b has the effect of a latch, maintaining the high voltage on Q, even if the SET DEA drops low. The actuated Q can be reset by activating the RESET DEA, which will compress the track 14c under it, and discharge the Q DEA. With a NOT gate as described hereinbefore, this SR or SET-RESET latch could easily be made into a D latch.
(102) The basic operation of a Set Reset (SR) flip flop or latch is given in Table 3. The flip flop remembers the last input and only changes if the set or reset line goes high.
(103) TABLE-US-00003 TABLE 3 SR Flip Flop Operation Set Reset Q 0 0 Previous Q 1 0 1 0 1 0 1 1 NA
(104) An alternative, simpler arrangement is shown in FIG. 23B. This arrangement was evaluated experimentally with a membrane of VHB 4905 pre-stretched 3.25:3.25 and affixed to a circular Perspex frame with an 85 mm inner diameter. In the free area of the membrane, two DEA were patterned using Nyogel 756G carbon grease so that there was an approximately 20 mm wide strip between them. Within the strip a DES element was patterned out of a 5:1 mix of Molykote 44 Medium grease and Cabot Vulcan XC72 carbon black. The DES had a total of 60 mm of track aligned horizontally between the DEA and stretched 1.36:1.36 at rest.
(105) The two DEA of FIG. 23B were connected together so that when they were activated, they compressed the central DES track below percolation, substantially lowering its resistance. FIG. 23C shows how the flip-flop was formed with two repeating parts connected together. The output Q is remembered when Set is low and Reset is high. Driving Set high will set Q high, and driving Reset low will set Q low. The state where Set is high and Reset is low is not allowed and will set Q to some middle tri state.
(106) Note that the voltage divider between the Set and Reset line consisted of two active DES. The divider could have been achieved using one DES and a fixed reference resistance however this would have suffered from the problems that if the percolation threshold of the DES wandered, the divider would become substantially high or low, and also that either the discharge or the charge rate would be limited. Using a dual active divider avoided these problems. Further, the element was designed so there could be no short circuit between the Set and Reset inputs as there was always one DES with a high resistance in the on and off configurations. Additionally the DES did not conduct when their partnered DEA were at half voltage.
(107) The flip-flop of FIG. 23B-C is a SR flip-flop with an inverted reset input. Thus for actual implementation as a SR flip-flop, an inverter is required. The inputs of the flip-flop directly drove the outputs, meaning it could not be cascaded without a buffer circuit. FIG. 23D demonstrates the flip-flop with the addition of a buffer stage to enable cascading.
(108) The performance of the device of FIGS. 23B-C was tested using a 4-channel Biomimetics Lab EAP controller and the output measured with a 5 GOhm input-impedance high voltage probe. Data acquisition was conducted with a National Instruments DAQ card and Labview 8.6. The results are presented in FIG. 23E which shows that the flip-flop behaved correctly when connected to a 3200V HV supply with 1 second set and reset pulses. The response to a set or reset pulse had two clear stages as shown in FIG. 23F which is zoomed in on a set pulse. Stage 1 was the instantaneous response where Q immediately steps up to some fraction of S. During stage 2 positive feedback took over and Q continued to climb until it was saturated and locked in place.
(109) Inverting Schmitt Trigger
(110) Schmitt triggers are essentially digital hysteretic components. They track an input, but with hysteresis defined by two thresholds (see Table 4). If the input goes above the upper threshold the output is high. If the input drops below the lower threshold the output is low. If the input is between the thresholds the output does not change. Schmitt triggers thus remember their previous state and are useful for de-bounce, oscillator, and timing circuits.
(111) TABLE-US-00004 TABLE 4 Schmitt Trigger Operation Input Q Input > T.sub.High 1 Input < T.sub.Low 0 T.sub.Low < Input < T.sub.High Previous Q
(112) A DES inverting Schmitt trigger based on a SR flip flop was designed and a single element of the setup is shown in FIG. 33A. The design was based on three VHB 4905 membranes pre-stretched 3.67:3.67 and affixed to frames with 100 mm by 100 mm windows in the middle. A single approximately 6070 mm DEA was painted on each membrane and a 36% stretched DES was applied next to it. Expansion of the DEA resulted in compression of the DES and a drop from a high to low resistance. Each frame had four corner legs which served as electrical interconnects for the four terminals of the device.
(113) Three elements shown in FIG. 33A were connected together to form the device and the resulting circuit layout is shown in FIG. 33B. Note that the R unit had its DES element halved in size to ensure that at rest, its resistance was approximately half that of the S and Q elements. The input of the inverting Schmitt trigger R was connected to S which was inverted by referencing it against a high voltage instead of ground. The input-output waveforms of the fabricated Schmitt trigger are given in FIG. 33C for a high voltage supply of 4 kV, an inverting reference voltage of 3.5 kV for the S DEA, and a 0.35 Hz 0-3.5 kV triangle input signal.
(114) As can be seen in FIG. 33C, switching between high and low outputs occurred when the input fell below about 600 V and rose above about 2900 V, respectively. The behaviour and thresholds were governed by the displacements required to drive the switches past their percolation thresholds as well as the reference voltages. Thresholds could be tuned by changing the reference voltages or the geometry of the components.
(115) Buffer and Opamp (Operational Amplifier)
(116) FIG. 24 shows a design for a DES buffer. The purpose of the buffer is to produce an output voltage equal to the input voltage despite changing current demands. This acts to shield the input line from drawing more current than it is able to supply. The buffer presented here uses two curved percolating tracks that form a voltage divider between four DEA formed on a pre-stretched membrane mounted on a frame.
(117) The back of the DEA is grounded. When the voltage applied to the input DEA is larger than the voltage on the output DEA the centre dead zones are distorted down. This compresses the left and stretches the right percolating tracks 14, increasing the resistance on the right and lowering it on the left, thereby increasing the voltage on the output DEA. If the output DEA has a larger voltage than the input DEA, the reverse happens, lowering the voltage.
(118) As the buffer adjusts the output in proportion to the error, the system can only track a voltage within the tolerances of the materials and fabrication efficiency.
(119) With a small modification, it is possible to arrive at a DES opamp as shown in FIG. 25. In this embodiment, the output voltage is approximately equal to K*((Input+)(Input)). Where K is the gain of the system and input+ and input are the voltages on the input terminal. Note that in reality the relationship will be more complex than this, the equation is merely illustrative. The creation of an OPAMP is a critical piece of analogue circuitry, from which many circuits can be derived such as filters, drivers, feedback, and analogue logic devices. Thus, while many embodiments of the invention have been described using digital components or configured to provide a digital output, the invention is not limited thereto.
(120) Frequency Multiplier
(121) A frequency multiplier is shown in FIG. 26A. The frequency multiplier consists of a DEA and a DES voltage divider formed by two below percolation tracks 14. The tracks are arranged so that when the DEA expands due to the application of a voltage to the In terminal the lower member of the ladder is compressed, whilst the length of the upper member remains largely unchanged. This has the effect of making the ratio of the voltage divider drop with increasing voltage. As both Maxwell pressure and percolation effects are nonlinear the output voltage is approximately 50% of the input voltage for low voltages, and drops back to zero for very high input voltages. Thus if a sine wave of sufficient amplitude is applied to the input terminal the output terminal will output a periodic wave with twice the frequency, as illustrated in FIG. 26B.
(122) FIG. 27A shows a frequency multiplier that uses above percolation tracks 14 and a hybrid DES/DEA for one of the electrodes. The top electrode is made of an above percolation electrode, whereas the bottom electrode is simply grounded with a normal electrode. The input terminal is at one end of the rectangular DEA and the output terminal is in the middle, the other end of the strip is grounded. For low voltages the ratio of output to input voltage is approximately 50%. As the voltage climbs this ratio drops, as the top end of the device will begin to charge up, causing the electrode to expand, and raise its resistance at the top end at a greater rate than the resistance raises at the bottom end. At some point an increase in input voltage will start to produce a decrease in output voltage, with the output voltage tending to zero. This will form a frequency multiplier as before. Successful working of this device is dependent on careful tuning of electrode properties. This arrangement illustrates an example embodiment where the DEA and the DES are more integrally formed.
(123) FIGS. 27B and 27C show a physical setup and circuit layout, respectively, of an alternative frequency multiplier design. According to this embodiment, DEA bodies are positioned either side of a central resistive track which acts as a DES. Actuation of the DEAs compresses the DES element, lowering its resistance. According to the arrangement shown, the DEA is referenced against potential of half that of the high voltage supply HV. If the input is low, the DEA will be active and result in a low output. If the input is at 50% HV, the DEA will be inactive and the output will be high. If the input switches to high, the DEA will activate again, dropping the output to low. Consequently, with a sine wave input, the output waveform will have twice the frequency.
Further Embodiments
(124) FIG. 28 shows a basic oscillator unit formed by stretching a dielectric elastomer (VHB) membrane over an 80 mm diameter support frame. A ring is attached to the centre of the membrane with a 4 mm hole in its centre. Electrodes are patterned on the membrane forming a variable capacitor (the DEA unit) and a variable resistor. An electrode is also painted on the outer frame to form a fixed resistor. The variable resistor has a rest resistance of approximately 6 M, increasing to approximately 20 M when the unit centre ring is depressed down 30 mm. As the centre ring is depressed further the resistance increases substantially up to 50 M due to percolation effects and the stretching of the electrode. This is shown in FIG. 29. The unit has 5 terminals. One for either side of the DEA section and three for the voltage divider. The ladder is formed by connecting the fixed resistor to the variable resistor. When a voltage is applied to the high voltage terminal and the base of the ladder is connected to ground, the voltage divider will output a voltage between the two. Initially when the variable resistance is low, the output voltage will also be low. As the resistance of the variable resistor increases the output tends towards that of the high voltage supply. Note that the arrangements of FIGS. 28 and 29 were configured for incorporation within a wider setup substantially the same as that of FIGS. 5 and 6.
(125) FIG. 30 shows an alternative voltage clamp arrangement for use with a dielectric elastomer generator (DEG). As discussed previously, the amount of charge stored by a DEG can be increased provided that mechanical energy is added to the system. However, there exists a problem in that the voltage should not climb above some threshold or the device will fail due to dielectric breakdown. It is desirable therefore to incorporate a voltage clamp into the device. The clamp shown in FIG. 30 consists of a DES coupled electrically to the generator, such that there is a variable capacitive element connected to the generator (or incorporated into it) and a variable resistor. The capacitor and resistor are arranged in such a way that as the capacitor expands it compresses the length of the resistor, reducing its resistance. This reduction is dramatic because the resistor is formed below a percolation threshold and its slight contraction will dramatically lower its resistance. The resistor is connected between either electrode of the capacitor. When the capacitor is charged above some level, the resistance will drop and begin to conduct current. Further charging above this point requires much more current to overcome the plummeting resistance of the track. In this way the generator is protected from over volting itself and the dangerous excess energy is dissipated as heat.
(126) FIG. 31 shows an arrangement for controlling a dielectric elastomer actuator to be able to hold a particular position using an applied voltage. Without feedback circuitry the actuator is unable to reject a disturbance and is also electromechanically unstable for some load conditions. This can be remedied using a DES. This arrangement is similar to the arrangements used to transport a ball around a circular track. An additional capacitive element is placed radially out from the variable resistance element Rp. A second change is that the resistance Rp is placed to act as the charging resistor for the main actuator Cb, while a fixed resistor Rd acts to discharge. With the application of a bias load placed to put the central loading point down into the page, the actuator will act to move the load up and down with variation of the supply voltage. The percolating track Rp is held in the sensitive regime by the expansion and contraction of the actuator A, allowing the charging of the main actuator at any supply voltage. When a disturbance is applied to the load, such as dragging of the central ring down, the resistor Rp will increase in value and lower the equilibrium voltage on the main actuator, rejecting the disturbance. If the disturbance drives the ring up the opposite will happen. This feedback functionality could alternatively be achieved using an opamp arrangement of the invention.
(127) Particular embodiments advantageously combine the use of one or more DES and one or more DEG, the DES providing all or a portion of the control circuitry for the DEG. More particularly, the DESs may act as switches that toggle between open and closed when a dielectric elastomer body is deformed. Open and closed may be relative and not require complete opening or closing of a circuit. By integrating or otherwise coupling the DES with the associated DEG such that deformation of the DEG membrane causes the conductive properties of the DES to vary (i.e., by corresponding deformation thereof), the DEG can be configured to cause the DES to switch so that currents within the wider arrangement are appropriately controlled.
(128) Thus, embodiments of the invention provide a charge pump that reduces or eliminates the need for a high voltage power source for a generator, more particularly, a DEG. The charge pump described herein does not require external capacitors to store energy and external diodes to control current flow. Instead, the charge pump is integrated into the DEG by using the capacitance of the DEG for charge storage and switching effected to control current flow achieved using one or more DES, eliminating or reducing the need for rigid parts.
(129) FIG. 35A is a perspective view of an antagonistic pair DEG configuration of one embodiment. The basic configuration is similar to the arrangements described in WO 2011/005123. Two dielectric elastomer membranes 3a, 3b are provided between frame members 2. According to the embodiment shown, the frame members 2 are fixed spatially relative to one another such that movement of central member or hub 4 causes the bodies of DEG1A, DEG1B and DEG2 to deform.
(130) FIGS. 35B and 35C are circuit and schematic diagrams, respectively, of the embodiment of FIG. 35A. As more clearly shown in FIG. 35C, the first membrane 3a is configured to define two DEG, DEG1A and DEG1B (referred to collectively as DEG1). DEG1 may be formed by discrete electrode portions of the same dielectric elastomer body or may be two separate bodies. Membrane 3a is further provided with two switches Q1 and Q2. The second membrane 3b includes a single switch Q3 and DEG2.
(131) When driven 180 out of phase, one of the membranes 3a, 3b is stretched while the other is relaxed. With DEG1 in a stretched state and DEG2 relaxed, the resistance of the switch Q3 was low (in a prototype based on the FIG. 36 arrangement, 5 M) while that of Q1 and Q2 was high (in the same prototype, 10 G). Thus DEG1A and DEG1B were effectively connected in series. When the deformation was reversed, the resistance of the switches toggled between high and low to place DEG1A and DEG1B in parallel. By toggling the configuration of DEG1 between series and parallel, the system converted generated energy into a higher charge form. This ability to boost system charge allows losses to be replenished, avoiding the need for an external charge source (i.e., a battery etc) or the self-priming arrangements described in WO 2011/005123. More particularly, self-priming is achieved without the need for external circuitry, particularly diodes.
(132) The embodiment of FIG. 35A was embodied in a prototype and the output thereof, starting with an initial priming voltage of 100V and cycling at 2 Hz, is given in FIG. 36. In later tests, it was found that an initial priming voltage of 30V was acceptable. As can be seen from FIG. 36, the output voltage rose quickly after application of the initial priming voltage, to as high as nearly 1.4 kV. Further, it is apparent from FIG. 36 that by about 100 seconds, the system was stabilised with the output voltage oscillating between about 700V and 1350V.
(133) The particular configuration shown in FIG. 35A was chosen as it does not require a large voltage swing meaning that it was possible to over design the generator so that the system was tolerant to non-ideal switches. Further, since it only contained three switches, it was relatively straightforward to identify any issues with the setup. However, it is possible to use other self-priming circuit configurations.
(134) Self-priming is advantageous for autonomous power generation as it is not necessary to provide an external high voltage source. However, mechanical switches could be provided in combination with or instead of switches Q1-Q3, said switches being configured to open/close at the appropriate time in the DEG cycle. A circuit diagram of such an arrangement is provided in FIG. 37A. Operation is effected by switch S.sub.1 only being closed when the DEG is stretched and S.sub.2 only being closed when the DEG is relaxed. More particularly, switch S.sub.1 controls priming of the DEG whereas switch S.sub.2 controls energy extraction (i.e., connection to a load).
(135) A design for a soft DEG system that behaves similarly is provided in FIG. 37B. In FIG. 37B, the two DEGs are an antagonistic pair (similar to the arrangement shown in FIG. 35A) that forms two circuits represented in a similar fashion to FIG. 35C. Soft switches Q1A and Q1B of FIG. 37B correspond to mechanical switch S.sub.1 in FIG. 37A whereas switches Q2A and Q2B correspond to switch S.sub.2. Note that in FIG. 37B, all terminals with the same net label are electrically connected.
(136) An electrostatic generator circuit is shown in FIG. 38A. In this arrangement, a portion of the generated energy is used to charge batteries used to prime the system, increasing the length of time before they need to be replaced or more actively charged. FIG. 38B shows the soft equivalent schematic arrangement that uses an antagonistic DEG/DES pair. Again the basic setup is similar to that shown in FIG. 35A.
(137) Rather than the more conventional componentry of FIG. 38A which includes diodes D.sub.1 and D.sub.2 for controlling balancing of priming of the DEG with extraction of energy to a load, soft switches Q1A and Q1B replace diode D.sub.1 and soft switches Q2A and Q2B replace diode D.sub.2. In FIG. 38B, all terminals with the same net label are electrically connected and a battery is required, as shown.
(138) FIG. 39 shows cast silicone membranes in the foreground with an assembled device in the background. DEGs may be formed using Nusil silicone and Perspex dies formed using laser engraving. For DEGs, high stiffness is not generally of particular significance, at least when compared with DEA although this may provide some improved level of control.
(139) FIG. 40A schatically shows an alternative configuration to the circuit of FIG. 35B in which both DEG1 and DEG2 include self-priming functionality. While FIG. 40A shows more conventional diodes, these are preferably in the form of DES.
(140) FIGS. 40B-D show further alternative, novel self-priming circuits. Again, while more conventional diodes notation has been used, preferably they are in the form of DES.
(141) While example embodiments have been described using dielectric membranes provided between fixed, substantially rigid frame elements, the invention is not limited thereto and the invention could alternatively be embodied in a flexible arrangement that forms a soft, self supporting structure. Alternatively, a soft support structure could be cast or adhered onto a Danfoss DEG. Further, RS silicone may for example be used to stitch Danfoss membranes together to arrive at a desired configuration.
(142) Antagonistic pairs are used in the examples because the shown arrangements require current to flow along different paths when the DEG is stretched and relaxed. An alternative to an antagonistic pair would be to utilize structures that deform so that as one segment is stretched another region compresses. For instance, if a band is uniaxially stretched it will neck-in and it is possible to use the compression in the transverse direction to activate a switch that turns off or on as the band is stretched. The band could alternatively be a tube or inflated tube with a band around the centre which relaxes when the tube is axially stretched.
(143) Another application of the invention is as a bending mode generator. When bent, a relatively thick structure is compressed on one side as the other is stretched. This could be solid, inflated, or a soft spring roll.
(144) Finite element analysis applications such as ABAQUS can be used to evaluate the strain fields in potential structures to assess likely deformations and the stresses involved.
(145) While embodiments of the generator of the invention have been described with particular DES configurations, others are included within the scope of the invention. For example, above or below percolation tracks may be used.
(146) Particular embodiments of the invention provide DEGs of low mass and complexity, and high flexibility to be maintained at the system level, rather than just comprising a flexible dielectric elastomer body. The system, including the DEG and power electronics described may essentially simply consist of an acrylic membrane and carbon grease used to formed the electrodes and elements.
(147) Thus the production of lighter, more flexible, cheaper and less complex DEG systems is enabled than current state of the art technologies including the DEG described in WO 2011/005123, electromagnetic, piezoelectric, and traditional electrostatic generators. Broadly speaking, these advances may be achieved by integrating soft electronics (for example but not limited to deformable conductive tracks) onto the DEG membrane, so that external circuitry is no longer required. Further, the system is potentially environmentally tolerant, autonomous, scalable and suitable for use in distributed arrays. Further, the system does not require a high voltage charge source, so low cost electronic components can be used, as required. Key advantages of the generator of the invention are set out below.
(148) Environmental Tolerance:
(149) DEGs according to the invention can be fabricated with corrosion resistant and non-magnetic components making them suitable for systems operating in corrosive environments or an environment with low tolerance to magnetic noise (e.g. marine or MRI environments).
(150) Autonomy:
(151) By integrating self-priming electronics onto the DEG membrane, the system can operate without being constantly connected to a power source because self-priming systems boost the charge of the energy generated by DEG so that losses can be replenished. The soft DEG presented here can generate energy (or transform mechanical energy to electrical energy) from intermittent sources and has been demonstrated to be self-primed from 30V, meaning it was able to generate power at 30V. DEG usually operate in the kV range, meaning expensive DC-DC converters are required. In response to a mechanical stimulus, the system could boost its voltage from 30V up to the kV range without a DC-DC converter. Consequently, costs of producing the system are much reduced.
(152) Scalability:
(153) One advantage of DEGs is that they are inherently scale-invariant. However, the mass and bulk of the external circuitry is not often considered and poses a problem when producing small scale DEG for applications such as wearable devices and self-powered sensors. Also, because of the operation at high voltages and low current, there has been mention in the art that the cost of DEG electronics may be a significant disadvantage for small scale applications. The DEG of embodiments of the invention with integrated electronics is very low cost. Further, the integrated electronics could possess similar scaling laws to that of DEG.
(154) Distributed Control:
(155) An array of a plurality including thousands of kelp-like soft DEG could be provided, each with their own local control electronics integrated onto their dielectric elastomer membrane, all supplying energy to a large centralised storage reservoir and regulator circuitry. The distribution of the control electronics greatly reduces the complexity and enhances the robustness of the array.
(156) The generator of the invention shows considerable promise for harvesting energy from environmental sources such as ocean waves, wind, and human motions because DEG can be directly coupled to large broadband motions without gearing while maintaining a high energy density and having few moving parts. These advantages are greatly enhanced by using the DEG embodiments presented here.
(157) The invention could be adapted to produce a highly flexible self-powered sensor. DEG produce a voltage swing that is proportional to the amount of stretch in a cycle. By probing the voltage we can determine the stretch. Also, since the DEG is a generator, it does not require a power source.
(158) Since DEGs consist of materials that lend themselves to a wide range of manufacturing methods, a large range of configurations and scales are possible, from devices for harvesting Mega-Watts of wind energy to those for harvesting energy from human movement for powering mobile electronics.
(159) Unless the context clearly requires otherwise, throughout the specification, the words comprise, comprising, and the like, are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense, that is to say, in the sense of including, but not limited to.
(160) The entire disclosures of all applications, patents and publications cited above and below, if any, are herein incorporated by reference.
(161) Reference to any prior art in this specification is not, and should not be taken as, an acknowledgement or any form of suggestion that that prior art forms part of the common general knowledge in the field of endeavour in any country in the world.
(162) The invention may also be said broadly to consist in the parts, elements and features referred to or indicated in the specification of the application, individually or collectively, in any or all combinations of two or more of said parts, elements or features.
(163) Wherein the foregoing description reference has been made to integers or components having known equivalents thereof, they are herein incorporated as if individually set forth.
(164) It should be noted that various changes and modifications to the presently preferred embodiments described herein will be apparent to those skilled in the art. Such changes and modifications may be made without departing from the spirit and scope of the invention and without diminishing its attendant advantages. It is therefore intended that such changes and modifications be included within the scope of the invention. Particular reference is made to the component devices described with reference to FIG. 11 onwards. It will be appreciated that such devices may be coupled together or otherwise integrated to form other elements. Also, at least portions of said combinations may be formed from more conventional components and for particular arrangements, at least some components may be removed where there is duplication. Further, while illustrated embodiments have dielectric elastomer bodies in the form of membranes, the invention is not limited thereto and may use any configuration for the deformable body, including being three dimensional. Deformations may be imparted from any angle, including from out of the plane of substantially planar membrane-type devices. Deformations to a DES or a DEA may be imparted as a result of an electrical or mechanical input, including from user interaction with devices of the invention.
