EAP actuator and drive method

Abstract

A field driven electroactive polymer actuator which is actuated using an actuation drive having a profiled portion having a start voltage and an end voltage and a duration of at least 25 ms followed by a steady state drive portion based on a steady state voltage. The profiled portion comprises a voltage curve or a set of voltage points which define a first voltage slope at the beginning of the profiled portion which is steeper than a linear ramp between the start voltage and the end voltage, and a second voltage slope at the end of the profiled portion which is shallower than a linear ramp between the start voltage and the end voltage.

Claims

1. A field driven electroactive polymer actuator, comprising: an electroactive polymer structure; and a driver circuit, wherein the driver circuit is arranged to provide an actuation drive signal to the electroactive polymer structure, wherein the actuation drive signal comprises a drive voltage for charging the electroactive polymer structure from a non-actuated state to an actuated state, wherein the drive voltage comprises a profiled portion having a start voltage and an end voltage, and wherein the profiled portion comprises a voltage curve which defines a first voltage slope at the beginning of the profiled portion which is steeper than a linear ramp between the start voltage and the end voltage, and a second voltage slope at the end of the profiled portion which is shallower than the linear ramp between the start voltage and the end voltage and the profiled portion is applied for a duration of 10 ms to 100 ms followed by a steady state drive portion based on a steady state voltage.

2. The field driven electroactive polymer actuator as claimed in claim 1, wherein the driver circuit is arranged to provide the profiled portion which follows an initial step voltage increase from zero to the start voltage, and wherein the initial step voltage increase is at most 50% of the end voltage.

3. The field driven electroactive polymer actuator as claimed in claim 1, wherein the driver circuit is arranged to provide the profiled portion which comprises a smooth curve with monotonically decreasing gradient.

4. The field driven electroactive polymer actuator as claimed in claim 1, wherein the driver is arranged to provide the profiled portion, and wherein the profiled portion comprises a set of at least 4 voltage steps between voltage points.

5. The field driven electroactive polymer actuator as claimed in claim 1, wherein the driver circuit is arranged to provide the profiled portion of the drive voltage to the electroactive polymer structure, and wherein the profiled portion has a duration of between 10 ms and 200 ms.

6. The field driven electroactive polymer actuator as claimed in claim 1, wherein the driver circuit is arranged to provide the end voltage, and wherein the end voltage is equal to the steady state voltage.

7. The field driven electroactive polymer actuator as claimed in claim 1, wherein the driver circuit is arranged to provide a compensation waveform which is superposed over the profiled portion.

8. The field driven electroactive polymer actuator as claimed in claim 1, wherein the driver circuit is arranged to provide the profiled portion, and wherein the profiled portion comprises: a parabolic curve; or an exponential curve; or a root curve.

9. The field driven electroactive polymer actuator as claimed in claim 1, wherein the profiled portion comprises a set of voltage points which define the first voltage slope at the beginning of the profiled portion which is steeper than the linear ramp between the start voltage and the end voltage, and the second voltage slope at the end of the profiled portion which is shallower than the linear ramp between the start voltage and the end voltage.

10. The field driven electroactive polymer actuator as claimed in claim 1, wherein the driver is arranged to provide the profiled portion, and wherein the profiled portion comprises a set of at least 4 constant slope portions between voltage points.

11. The field driven electroactive polymer actuator as claimed in claim 1, wherein the driver circuit is arranged to provide a step increase from the end voltage to the steady state voltage.

12. The field driven electroactive polymer actuator as claimed in claim 1, wherein the driver circuit is arranged to provide a compensation waveform which is superposed over the steady state voltage.

13. A method of driving a field driven electroactive polymer actuator structure, comprising: providing an actuation drive signal to the electroactive polymer structure, the actuation drive signal comprising: a profiled portion having a start voltage and an end voltage, wherein the profiled portion comprises a voltage curve which defines a first voltage slope at the beginning of the profiled portion which is steeper than a linear ramp between the start voltage and the end voltage, and a second voltage slope at the end of the profiled portion which is shallower than the linear ramp between the start voltage and the end voltage and the profiled portion is applied for a duration of 10 ms to 100 ms followed by a steady state drive portion based on a steady state voltage.

14. The method as claimed in claim 13, further comprising: providing the profiled portion which follows an initial step voltage increase from zero to the start voltage, wherein the initial step voltage increase is at most 50% of the end voltage.

15. The method as claimed in claim 13, further comprising: providing a compensation waveform which is superposed over the profiled portion.

16. The method as claimed in claim 13, further comprising: providing the profiled portion which comprises: a smooth curve with a monotonically decreasing gradient or a set of at least 4 voltage steps between voltage points or a set of at least 4 constant slope portions between voltage points.

17. The method as claimed in claim 13, further comprising providing the profiled portion of the actuation drive signal to the electroactive polymer structure, wherein the profiled portion has a duration of between 10 ms and 200 ms.

18. The method as claimed in claim 13, further comprising: providing the end voltage which is equal to the steady state voltage; or providing a step increase from the end voltage to the steady state voltage.

19. The method as claimed in claim 13, wherein the profiled portion comprises a set of voltage points which define the first voltage slope at the beginning of the profiled portion which is steeper than the linear ramp between the start voltage and the end voltage, and the second voltage slope at the end of the profiled portion which is shallower than the linear ramp between the start voltage and the end voltage.

20. The method as claimed in claim 13, further comprising: providing a compensation waveform which is superposed over the steady state voltage.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Examples of the invention will now be described in detail with reference to the accompanying drawings, in which:

(2) FIG. 1 shows a known EAP actuator, which is unconstrained and thus expands in plane;

(3) FIG. 2 shows a known EAP actuator, which is constrained and thus deforms output of plane;

(4) FIG. 3 shows the waveform of a conventional drive scheme and the response of an actuator to the drive waveform;

(5) FIG. 4 shows a variety of different functions which be used to form a modified drive waveform;

(6) FIG. 5 shows the functions of FIG. 4 with one group replaced by a single representative function for the purposes of further evaluation;

(7) FIGS. 6 to 9 show the effect of a varied initial bias voltage when applying a ramp function as the drive waveform;

(8) FIGS. 10 to 13 show the effect of a varied initial bias voltage when applying a parabolic function as the drive waveform;

(9) FIGS. 14 to 17 show the effect of a varied initial bias voltage when applying an exponential function as the drive waveform;

(10) FIGS. 18 to 21 show the effect of a varied initial bias voltage when applying a root function as the drive waveform;

(11) FIGS. 22 to 25 show the effect of a varied exponent when applying an exponential function as the drive waveform for two different initial bias values;

(12) FIG. 26 shows normalized voltage derivatives for some of the functions used above;

(13) FIG. 27 shows how the waveform may be stepped or approximated rather than smooth;

(14) FIG. 28 shows that there may be a step after the voltage profile;

(15) FIGS. 29 to 30 show the effect of applying a compensation voltage; and

(16) FIG. 31 shows an EAP actuator system.

DETAILED DESCRIPTION OF THE EMBODIMENTS

(17) The invention provides a field driven electroactive polymer actuator which is actuated using an actuation drive having a profiled portion having a start voltage and an end voltage and a duration of at least 25 ms followed by a steady state drive portion based on a steady state voltage. The profiled portion comprises a voltage curve or a set of voltage points which define a first voltage slope at the beginning of the profiled portion which is steeper than a linear ramp between the start voltage and the end voltage, and a second voltage slope at the end of the profiled portion which is shallower than a linear ramp between the start voltage and the end voltage.

(18) This means the voltage has a high initial gradient, and this reduced over time. The resulting curve gives a rapid actuation but with reduced oscillation.

(19) Typically an EAP is controlled by a driving circuit which provides the required electrical voltage to operate the component. If the EAP is being activated, the driver normally generates a step dc voltage with a certain amplitude to bring the EAP into the desired position. Since electronic driving circuits are not ideal (e.g. they always have internal resistances), the actuation response is not only a function of the EAP itself but also of the driving circuit. In order to reduce the impact of the driver, the operating voltage for the EAP is usually stored in a capacitor, parallel to the EAP, and in terms of actuation this stored voltage is fed by an electronic switch (e.g. transistor, MOSFET) to the EAP.

(20) The EAP actuator then starts to deform as a function of its charge, which again depends on the applied voltage amplitude. If the EAP is being deactivated, the applied voltage is being disconnected and accordingly the EAP will slowly discharge via its internal parallel resistance and finally will go back to its initial position.

(21) A problem is that the step increase in voltage at the start of actuation in particular can give rise to mechanical oscillation. In particular, it has been observed by the applicant during experimental investigations that voltage (field) driven EAPs, operated by a rectangular voltage show strong oscillations during the activation phase, whereas a sinusoidal operation does not result in these oscillations. The oscillations during the activation phase are found to be caused by a sudden change of the control (activation) voltage. Mathematically this can be described as the slope or voltage derivative dv/dt.

(22) If this slope exceeds a threshold, oscillations have been found to appear. If the slope is limited by providing a drive voltage which follows one of a set of suitable mathematical functions, these oscillations can be reduced, or even fully or almost fully eliminated. A further measure to assist in providing a non-oscillating activation of an EAP is a smooth transition between the activation phase and the final steady state. Again, a large step in operating voltage may cause oscillations. The change in the voltage amplitude at this transition should thus also be limited. For example, at the point of transition between activation phase and the final steady state position, the slope or voltage derivative dv/dt may be small or zero.

(23) In order to determine the most suitable mathematical functions for the activation voltage, various measurements have been made using a standard voltage-driven EAP actuator.

(24) The table below show the possible voltage actuation shapes.

(25) TABLE-US-00001 TABLE 1 Rectangular v.sub.rect (t) := 0 t < t.sub.0 Equation 1 v.sub.rect (t) := v.sub.drive t.sub.0 ≤ t ≤ t.sub.0 + t.sub.on v.sub.rect (t) := 0 t > t.sub.0 + t.sub.on Ramp $v_{\mathrm{rmp}}\left(t\right):=v_{\mathrm{bias}}+\frac{\left(v_{\mathrm{drive}}-v_{\mathrm{bias}}\right)}{t_{\mathrm{smooth}}}.Math.t$ Equation 2 Parabolic Parameters $v_{\mathrm{para}}\left(t\right):=a+b.Math.t-c.Math.t^2$$a:=v_{\mathrm{bias}}$$c:=\frac{\left(v_{\mathrm{drive}}-v_{\mathrm{bias}}\right)}{t_{\mathrm{smooth}}^2}$$b:=\frac{\left(v_{\mathrm{drive}}-v_{\mathrm{bias}}\right)+c.Math.t_{\mathrm{smooth}}^2}{t_{\mathrm{smooth}}}$ Equation 3 Exponential Parameters $v_{\exp }\left(t\right):=v_{\mathrm{bias}}+\left(v_{\mathrm{drive}}-v_{\mathrm{bias}}\right).Math.\left(1-e^{\left(\frac{-t}{\tau }\right)}\right)$$\tau :=\frac{t_{\mathrm{smooth}}}{\mathrm{nrc}}$ Equation 4 Root Parameters $v_{\mathrm{root}}\left(t\right):=a+b.Math.\sqrt[m]{t}$$a:=v_{\mathrm{bias}}$$b:=\frac{\left(v_{\mathrm{drive}}-v_{\mathrm{bias}}\right)}{\sqrt[m]{t_{\mathrm{smooth}}}}$ Equation 5

(26) The driving schemes form a curved voltage-time profile which will be named a “profiled portion” of the drive waveform, which is followed by the steady state part of the waveform which is intended to hold the actuator in the actuated state.

(27) For all driving schemes used for the experiments, the maximum driving voltage amplitude (v.sub.drive) was limited to 200V. Furthermore, the frequency of operation was set to 1 Hz with a duty cycle of 50%. All functions are defined to allow for a d-bias (v.sub.bias) at the beginning (t.sub.0), in the form of a step increase from 0V driving voltage to the d-bias voltage at time t.sub.0. The start of the profiled portion may thus be defined as starting after such a dc bias. The profiled portion is limited in time (t.sub.smooth), so that t.sub.smooth is the time at the end of the profiled portion. As reference, an EAP, driven by a rectangular activation pulse as described by Equation 1, is used. With reference to the maximum driving voltage of 200V, this dc-bias was set to 0, 25% (=50V), 50% (=100V) and 75% (=150V).

(28) For each of these settings the smoothing time (during which the profiled portion i.e. smoothing function is applied) was adjusted to ins, 25 ms, 50 ms, 75 ms and 100 ms.

(29) This smoothing time has an influence on the slope of the smoothing functions as can be seen in the corresponding equations of Table 1, so that the voltage waveform reaches the 200V steady state value at time t.sub.smooth.

(30) Apart from the functions defined above, other functions fulfill the requirement of a moderate slope at the start and a (almost) zero slope at the transition point to the steady state operation (at t.sub.smooth). Other functions include the inverse tangent (a tan) functions, sinusoidal-functions (quarter wavelength, 0 to π/2) and logarithmic functions.

(31) FIG. 4 shows seven example functions plotted from time 0 to 0.01 is, where t.sub.smooth=0.01 s. The functions are shown with no d-bias.

(32) The functions are a ramp function vrmp, a parabolic function vpara, an a tan function vatan, a sinusoidal function vsin, a root function vroot, a log function vlog and an exponential function vexp.

(33) As can be seen, some of these functions (log, sinusoidal, a tan and the parabolic function) have a quite similar shape. Therefore, only four functions are selected for further evaluation as shown in FIG. 5, with the parabolic function representing the similar group. Parameters for the parabolic function are chosen giving a gentle slope at the beginning not greatly steeper than the straight ramp function. An advantage of the parabolic function is that it gives a zero slope at the transition to the steady state operation. An exemplarily function representing a medium slope at the beginning is the exponential function (i.e. with nrc=6), whereas the twelfth root function (i.e. with m=12) represents the class of functions with a steep slope at the beginning.

(34) Of course, all functions have parameters (nrc or m) which may be selected and which influence the shape/slope of the function itself.

(35) FIGS. 6 to 21 show measurement results. Each of these graphs shows the driving voltage (as voltage versus time) on top and the corresponding mechanical deflection of the EAP at the bottom (as deflection versus time). The different plots within each individual figure represent a variation in the smoothing time (t.sub.smooth), which was 10 ms, 25 ms, 50 ms, 75 ms and finally 100 ms. As reference, a rectangular driving shape with the correlated large oscillations is also shown in each case. This may be considered to be a plot for t.sub.smooth=0.

(36) In each case, the mechanical response curves show the greatest and most quickly arriving oscillations for the shortest time period t.sub.smooth. For this reason, the plots are labeled only in FIG. 6. It will be immediately apparent how the plots in the other figures correspond.

(37) A first set of measurements is shown in the graphs of FIGS. 6 to 9 for a ramp voltage. FIG. 6 is for no initial step voltage before the “profiled portion” (i.e. the ramp). FIG. 7 is for a 25% initial step voltage before the “profiled portion”. FIG. 8 is for a 50% initial step voltage before the “profiled portion”. FIG. 9 is for a 75% initial step voltage before the “profiled portion”.

(38) Since a fast activation without notable oscillations would be of benefit, in the deflection graphs a 90% threshold line 60 is added. This threshold line 60 indicates a mechanical deflection of 90% of the maximum deflection and can be used to identify the speed of the activation itself. If this threshold is taken into account for the rectangular driven EAP, the time for such an oscillating EAP to reach its stable state is about 75 ms. However during this time the EAP would already have passed the threshold 4 times corresponding to a change from the off-state to the on-state four times.

(39) Some clear trends can be observed. Firstly, short smoothing times of 25 ms and below still have significant oscillation. Secondly, a small d-bias (offset) of about 25% (of the maximum driving amplitude) at the beginning does not cause larger oscillations as seen in FIG. 7, whereas larger d-bias increases the oscillations as seen in FIG. 9. Thirdly, a fast activation with (almost) no oscillations seem contradictory aims when using a ramp waveform, since the 90% deflection target is not reached within a reasonable time. Ideally the steady state position would be reached within a time which is close to the normal activation time, if the EAP is driven by a rectangular pulse.

(40) The performance of the EAP when operated by a parabolic function is shown in FIGS. 10 to 13. FIG. 10 is for no initial step voltage before the “profiled portion”. FIG. 11 is for a 25% initial step voltage before the “profiled portion”. FIG. 12 is for a 50% initial step voltage before the “profiled portion”. FIG. 13 is for a 75% initial step voltage before the “profiled portion”.

(41) A clear improvement to the performance of a rectangular driven as well as the ramp driven EAP can be observed. The oscillations can be completely damped, even at reasonable activation times around 100 ms. Smoothing times between 25 ms and 100 ms all show a good performance, whereas a too short smoothing time of 10 ms still results in oscillations. A dc-bias at the beginning is implementable as well. However a too high d-bias of 75% again shows strong oscillations. By way of example, a d-bias of 50% (i.e. FIG. 12) at smoothing times between 25 ms and 100 ms is for example acceptable for some applications. This excludes the two most oscillating traces in FIG. 12 (for 0 ms and 10 ms).

(42) The performance of the EAP when operated by an exponential function (with nrc=6) is shown in FIGS. 14 to 17. FIG. 14 is for no initial step voltage before the “profiled portion”. FIG. 15 is for a 25% initial step voltage before the “profiled portion”. FIG. 16 is for a 50% initial step voltage before the “profiled portion”. FIG. 17 is for a 75% initial step voltage before the “profiled portion”.

(43) The exponentially driven EAP actuator also shows a clear improvement compared to the rectangular and ramp driven EAPs. In comparison to the parabolic driven EAP of FIGS. 10 to 13, the exponentially driven EAP has a bigger slope at the beginning. Accordingly, the performance for a smoothing time of 25 ms shows slightly larger oscillations. However, the response is basically similar to the parabolic driven EAP, or marginally better in that oscillations are as damped as observed for the parabolic driven EAP, but the activation is a slightly faster. Specifically, an overall activation time of 75 ms (the same as for the rectangular driven EAP) can be seen for a smoothing time of 50 ms for both 0V d-bias (i.e. FIG. 14) as well as 25% d-bias (i.e. FIG. 15).

(44) The performance of the EAP when operated by a root function (with m=12) is shown in FIGS. 18 to 21. FIG. 18 is for no initial step voltage before the “profiled portion”. FIG. 19 is for a 25% initial step voltage before the “profiled portion”. FIG. 20 is for a 50% initial step voltage before the “profiled portion”. FIG. 21 is for a 75% initial step voltage before the “profiled portion”.

(45) A high order root (m=12) is intentionally chosen in order to generate a very steep slope at the beginning of the activation phase. According to this large slope at the beginning the oscillations hardly are damped. Even for a long smoothing time and no d-bias still only slightly damped oscillations are recognizable.

(46) Thus, best performance in terms of reduced oscillations while activating the EAP as quickly as possible is observed, when the device has been operated by a parabolic- and/or an exponential function.

(47) Since the exponential function is more flexible in choosing the exponent (corresponding to the slope of this function) the performance of an EAP when operated by an exponential function with different slopes (by variation of parameter nrc) is shown in FIG. 22.

(48) In FIG. 22, there is a set of plots for t.sub.smooth=25 ms, for t.sub.smooth=50 ms and for t.sub.smooth=75 ms. For each value, there are two plots, one for 0% d bias offset and one for 25% dc bias offset.

(49) FIG. 22 is for nrc=6, FIG. 23 is for nrc=10, FIG. 24 is for nrc=14 and FIG. 25 is for nrc=20.

(50) The parameter nrc defines the exponential time constant in relation to the total smoothing time. A larger value of nrc leads to a steeper slope whereas a smaller value of nrc results in a flat response.

(51) From this parameter variation, the best-suited exponent (i.e. the corresponding value of nrc) can be identified. The largest value of nrc=20 (FIG. 25) results in a fast activation while still having small oscillations. Similar but not as fast responses of about 70 ms have been observed for nrc=6 (with a smoothing time t.sub.smooth=25 ms) and nrc=14 (with a smoothing time t.sub.smooth=50 ms).

(52) Slightly faster activation was reached with an exponential parameter of nrc=10 (t.sub.smooth=25 ms and 50 ms), For all these settings, a dc bias of 0 or 25% seems to be feasible.

(53) The invention provides an approach by which the EAP actuator is actuated fast as possible while minimizing mechanical oscillations. At the beginning of the profiled portion of the activation waveform, the voltage derivative dv/dt (slope) is higher than that of an ordinary ramp but smaller than ‘infinite’ (i.e. an ideal rectangular pulse). At the end of the activation time (t.sub.smooth) the voltage derivative is as small as possible in order to provide a smooth transition to the steady state driving with an ordinary constant or pulsed dc-voltage (or ac voltage).

(54) The normalized voltage derivatives of the smoothing functions discussed above are shown in FIG. 26 for t.sub.smooth=100 ms, and labeled using same notation as in FIG. 5 (note that the y-axis values are normalized and therefore arbitrary). In comparison to the constant slope of a ramp-function vrmp better suited smoothing functions such as the parabolic function vpara and exponential function vexp give larger slope during the start of the activation and smaller slopes during the transition to the steady state operation mode (at t>100 ms). The higher order root vroot is less satisfactory.

(55) The parabolic function has a constant slope, whereas the time-dependent behavior of an exponential or root function can be adapted by their parameters as defined in Table 1 above. These parameters may be fine-tuned according to the requirements of the envisaged application, especially the activation (smoothing) time, maximum driving amplitude and tolerable oscillation amplitudes, but also adapted to the EAP actuator itself.

(56) Accordingly, the numbers presented above are purely by way of example. In addition, lower order root functions with a small root exponent (e.g. m=3 or 4) also have much smaller voltage derivatives, and they may also be used, and their performance can be optimized.

(57) Instead of using only one smoothing function, the overall profiled portion may be implemented by means of a combination of several smoothing functions (also including step and ramp function portions). For example, within a certain time frame Δt.sub.m of the whole smoothing time, smoothing function 1 may be used. For another sub-frame Δt.sub.m+1 function 2 may be used and even within another time frame Δt.sub.m+n maybe again function 1 (with a different or the same parameter setting) may be used.

(58) The mathematical function may be realized by a gradual approximation. For example, a stepwise function may be used as shown in FIG. 27.

(59) In this case, the function has a set of points 70 which define the decreasing gradient as explained above. There may be step increases between these points, as shown by steps 72 or there may be linear interpolations of intermediate values. Approximated or interpolated values may be exactly on the corresponding points of the smoothing function but also might be nearby. The voltage steps or interpolations need to be chosen in order not to generate any large oscillations. As a rule of thumb, the variation in voltage amplitude should not be larger than 25% of the maximum driving voltage amplitude.

(60) FIG. 28 shows that there may be a step portion 80 between the profiled portion 82 and the steady state portion 84 of the drive waveform. Although it may be preferred to provide a zero slope at the point of transition from the activation phase (the “profiled portion”) to the steady state operation mode, a jump in the driving voltage as shown in FIG. 28 may be applicable. Depending on the tolerable level of oscillation (i.e. the amplitudes) and the threshold limits, the end point of the profiled portion may be below the required maximum operation voltage and a final step, or any other linear or non-linear increase, may be included to reach the end voltage. There may be such steps at any point in time during the activation phase.

(61) For actuators with known oscillations, a specific driving signal may be generated. This driving signal may for example include an initial dc bias and a function dependent on the time period t.sub.smooth as indicated above. However, a third component may also be introduced in the form of an alternating signal superimposed on the function formed by the two first parameters. The alternating signal for example has a certain amplitude a, characteristic frequency f, an initial time delay t.sub.1 and a phase delay t.sub.2.

(62) The superimposed alternating signal may be a sine wave in which case the signal will be defined by

(63) $f\left(t\right)\{\begin{array}{cc}0.& \mathrm{if}t\le t_1\\ a\left(t\right)*\sin \left(2\pi f*\left(t-t_2\right)\right).& \mathrm{if}t>t_1\end{array}$

(64) The superimposed signal can have a length of several periods, n. The number of periods, n, may however be smaller than 1 and may be constant or include a damping factor so that the sine wave will slowly decay, as determined by the function a(t).

(65) Preferably, the alternating voltage component is driven out of phase with the oscillations of the EAP (from a step input) in order to counteract the oscillations from the step voltage input. The frequency of the superimposed anti-oscillation voltage is then substantially the same as the resonance frequency of the actuated device or at least similar to the damped resonance frequency of the step voltage actuated device. The resonance frequency can be determined for a clamped actuator after manufacturing or can be calculated based on the geometry and material properties. The time constants t.sub.1 and t.sub.2 can be determined during a calibration procedure or via calculation.

(66) The procedure for calculating the suitable signal to superimpose will be different for low dc bias and high d bias actuation voltages. By way of example, FIG. 29 shows the response (plot 90) of an actuator in response to a clean ramp signal (plot 92). FIG. 30 show the response (plot 94) as a function of a ramp input voltage (plot 96) with a superimposed anti-oscillation voltage. This reduces the oscillations in the EAP actuator by counteracting the displacement.

(67) In the above example the sinusoidal compensation is added to the already constant operation voltage, i.e. during the steady state voltage. Depending on the chosen slope of the applied driving voltage the compensation may be part of the profiled portion of the drive waveform.

(68) In order to generate the required driving scheme several implementations may be used. Analogue circuits may be used or else a digital-microcontroller may calculate the required analogue data points which may be amplified by any conventional (power) amplifier solution. A look-up table approach may instead be implemented; where all required data points are already pre-calculated and saved in a memory of any controlling device. As a function of time, the digital or analogue data points are read out and may be amplified.

(69) In the above examples, functions with specific response times have been considered for a specific actuator. The response times of the functions may be scaled with the resonance response of a random actuator. In particular, the time constants are substantially linearly related to the resonance frequency of the actuated device or at least similar to the damped resonance frequency of the step voltage actuated device. The resonance frequency can be determined for a clamped actuator after manufacturing or can be calculated based on geometry and material properties. The time constants t.sub.1 and t.sub.2 for the compensation signal as explained above can be determined during a calibration procedure or via calculation.

(70) FIG. 31 shows that a driver 100 is used to apply the drive voltage to the EAP actuator 102. It also shows an optional feedback path 104 (mechanical, optical or electrical).

(71) The output of the driver 100 is the voltage profile as discussed above, namely before the load and any resistances connecting to the load. The voltage at the capacitive load will differ as a result of the series resistances between the driver and the load (or forming part of the load) and thus giving rise to RC charging time constants. Thus, the drive voltage is the direct output from the driver, and is the profile which the driver is designed to deliver as its output, for example to an open circuit output terminal.

(72) This invention relates in particular to actuation of EAP actuators. However, it can be used in applications where an EAP device is performing both a sensing and an actuation function.

(73) Materials suitable for the EAP layer are known. Electro-active polymers include, but are not limited to, the sub-classes: piezoelectric polymers, electromechanical polymers, relaxor ferroelectric polymers, electrostrictive polymers, dielectric elastomers, liquid crystal elastomers, conjugated polymers, Ionic Polymer Metal Composites, ionic gels and polymer gels.

(74) The sub-class electrostrictive polymers includes, but is not limited to:

(75) Polyvinylidene fluoride (PVDF), Polyvinylidene fluoride-trifluoroethylene (PVDF-TrFE), Polyvinylidene fluoride-trifluoroethylene-chlorofluoroethylene (PVDF-TrFE-CFE), Polyvinylidene fluoride-trifluoroethylene-chlorotrifluoroethylene) (PVDF-TrFE-CTFE), Polyvinylidene fluoride-hexafluoropropylene (PVDF—HFP), polyurethanes or blends thereof.

(76) The sub-class dielectric elastomers includes, but is not limited to:

(77) acrylates, polyurethanes, silicones.

(78) The sub-class conjugated polymers includes, but is not limited to:

(79) polypyrrole, poly-3,4-ethylenedioxythiophene, poly(p-phenylene sulfide), polyanilines.

(80) Ionic devices may be based on ionic polymer-metal composites (IPMCs) or conjugated polymers. An ionic polymer-metal composite (IPMC) is a synthetic composite nanomaterial that displays artificial muscle behavior under an applied voltage or electric field.

(81) In more detail, IPMCs are composed of an ionic polymer like Nafion or Flemion whose surfaces are chemically plated or physically coated with conductors such as platinum or gold, or carbon-based electrodes. Under an applied voltage, ion migration and redistribution due to the imposed voltage across a strip of IPMCs result in a bending deformation. The polymer is a solvent swollen ion-exchange polymer membrane. The field causes cations travel to cathode side together with water. This leads to reorganization of hydrophilic clusters and to polymer expansion. Strain in the cathode area leads to stress in rest of the polymer matrix resulting in bending towards the anode. Reversing the applied voltage inverts the bending.

(82) If the plated electrodes are arranged in a non-symmetric configuration, the imposed voltage can induce all kinds of deformations such as twisting, rolling, torsioning, turning, and non-symmetric bending deformation.

(83) In all of these examples, additional passive layers may be provided for influencing the electrical and/or mechanical behavior of the EAP layer in response to an applied electric field.

(84) The EAP layer of each unit may be sandwiched between electrodes. The electrodes may be stretchable so that they follow the deformation of the EAP material layer. Materials suitable for the electrodes are also known, and may for example be selected from the group consisting of thin metal films, such as gold, copper, or aluminum or organic conductors such as carbon black, carbon nanotubes, graphene, poly-aniline (PANI), poly(3,4-ethylenedioxythiophene) (PEDOT), e.g. poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate) (PEDOT:PSS). Metalized polyester films may also be used, such as metalized polyethylene terephthalate (PET), for example using an aluminum coating.

(85) The invention can be applied in many EAP and photoactive polymer applications, including examples where a passive matrix array of actuators is of interest.

(86) In many applications the main function of the product relies on the (local) manipulation of human tissue, or the actuation of tissue contacting interfaces. In such applications EAP actuators for example provide unique benefits mainly because of the small form factor, the flexibility and the high energy density. Hence EAP's and photoresponsive polymers can be easily integrated in soft, 3D-shaped and/or miniature products and interfaces. Examples of such applications are:

(87) Skin cosmetic treatments such as skin actuation devices in the form of a responsive polymer based skin patches which apply a constant or cyclic stretch to the skin in order to tension the skin or to reduce wrinkles;

(88) Respiratory devices with a patient interface mask which has a responsive polymer based active cushion or seal, to provide an alternating normal pressure to the skin which reduces or prevents facial red marks;

(89) Electric shavers with an adaptive shaving head. The height of the skin contacting surfaces can be adjusted using responsive polymer actuators in order to influence the balance between closeness and irritation;

(90) Oral cleaning devices such as an air floss with a dynamic nozzle actuator to improve the reach of the spray, especially in the spaces between the teeth. Alternatively, toothbrushes may be provided with activated tufts;

(91) Consumer electronics devices or touch panels which provide local haptic feedback via an array of responsive polymer transducers which is integrated in or near the user interface;

(92) Catheters with a steerable tip to enable easy navigation in tortuous blood vessels;

(93) Measurements of physiological human body parameters such as heart beat, SpO2 and blood pressure.

(94) Another category of relevant application which benefits from such actuators relates to the modification of light. Optical elements such as lenses, reflective surfaces, gratings etc. can be made adaptive by shape or position adaptation using these actuators. Here one benefit of EAPs for example is a lower power consumption.

(95) Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measured cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope.