Propulsion of a Microsubmarine Using a Thermally Oscillatory Approach

8/13/2019 Propulsion of a Microsubmarine Using a Thermally Oscillatory Approach

1/12

This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

IP Address: 148.247.1.19This content was downloaded on 01/11/2013 at 03:50

Please note that terms and conditions apply.

Propulsion of a microsubmarine using a thermally oscillatory approach

View the table of contents for this issue , or go to the journal homepage for more

2013 J. Micromech. Microeng. 23 105011

(http://iopscience.iop.org/0960-1317/23/10/105011)

ome Search Collections Journals About Contact us My IOPscience
http://localhost/var/www/apps/conversion/tmp/scratch_2/iopscience.iop.org/page/termshttp://iopscience.iop.org/0960-1317/23/10http://iopscience.iop.org/0960-1317http://iopscience.iop.org/http://iopscience.iop.org/searchhttp://iopscience.iop.org/collectionshttp://iopscience.iop.org/journalshttp://iopscience.iop.org/page/aboutioppublishinghttp://iopscience.iop.org/contacthttp://iopscience.iop.org/myiopsciencehttp://iopscience.iop.org/myiopsciencehttp://iopscience.iop.org/contacthttp://iopscience.iop.org/page/aboutioppublishinghttp://iopscience.iop.org/journalshttp://iopscience.iop.org/collectionshttp://iopscience.iop.org/searchhttp://iopscience.iop.org/http://iopscience.iop.org/0960-1317http://iopscience.iop.org/0960-1317/23/10http://localhost/var/www/apps/conversion/tmp/scratch_2/iopscience.iop.org/page/terms


2/12


3/12

J. Micromech. Microeng. 23 (2013) 105011 L Qiao and C Luo

(a )

(b)

Figure 1. (a ) A large putt-putt boat and its schematic, and ( b) experimentally measured speedtime curve.

actuate large-scale rockets and airplanes. In these approaches,

a thrust is produced by burning fuel to generate a high-speed jet, which provides a force to make the vehicle movealong the opposite direction of the jet. However, since amicrosubmarine has a limited space to store fuel, the adoptionof the thermal approaches would limit the travel distanceof the microsubmarine. Moreover, thermopneumatic [ 22],electrohydrodynamic [ 23] and electrowetting-on-dielectric[24] methods are also applied to actuate small objects inliquids. Nevertheless, it appears that the corresponding thrustsmay not be large, since the total travel distances are 1.7 mm[22] or less [23, 24].

The propulsive approach that is adopted in this work

is a thermally oscillatory method, which is motivated bya propulsive approach employed in a so-called putt-putt orpop toy boat [21, 22]. The adopted propulsive approach onlyrequires a thermal ejector, which includes a resistive heater, areservoir and two nozzles. Due to the ease of fabricating thethermal ejector using UV lithography, the development of amicrosubmarine becomes feasible. Based on this propulsiveapproach, we develop the prototype of a microsubmarine, andalso explore its horizontal motions.

The paper is organized as follows. Preliminary testson putt-putt boats are presented in section 2. Design andfabrication of microsubmarines are described in section 3.Experimental results and discussions are given in section 4.Finally, in section 5, this work is summarized and concluded.

2. Preliminary tests

According to [25], the putt-putt boat is originated in an 1891British patentfor water pulseengines inventedby ThomasPiot.Based on the description and schematic of such a boat [ 25, 26],we manually fabricated a large putt-putt boat (gure 1(a )). Thepropulsive system in this boat includes a candle, a shallowAl chamber, and two pipes that connect to the Al chamberand lead to the rear of the boat. In a water tank, this boatmoved in an approximately periodic manner with the periodof 0.18 s (gure 1(b)). In each period, the speed rst increasedto about 12 cm s 1 and then decreased to 0. We also manuallymanufactureda small putt-puttboatwitha transparent chamberand a transparent pipe for two purposes (gures 2(a ) and ( b)):(i) to gain a deep understanding of the propulsive mechanismdescribed in [26] through in situ observation of the thermallyoscillatory process, and (ii) to examine whether a resistiveheater can replace a candle to generate a water jet, since acandle cannot be used in the microsubmarine.

This small boat was xed, and water inside its chamberwas heated up by a commercial ceramic heater (resistance is10 ) at an applied dc voltage of 5 V. It was clearly observedthat, as described in [ 26], a propulsive cycle consisted of twoprocesses: exhaust andsuction. At thebeginning of theexhaustprocess, the shallow chamber and pipe were lled with water(gures 2(a 1) and ( c1)). Shortly after the resistive heater wasturned on, steam formed in the shallow chamber. As morevapors were produced, the pressure increased. Consequently,

2


4/12


(a 1)

(a 2)

(a 3)

(b1)

(b2)

(b3)

(c1) (c2) (c3)

Figure 2. In situ observed exhaust process at ( a 1) t = 0 s, ( a 2) t = 5.9 s, ( a 3) t = 7.5 s. In situ observed suction process at ( b1) t = 8.5 s, ( b2)t = 8.6 s, ( b3) t = 8.7 s. Illustration of a thermally oscillatory process: ( c1) stationary state (heater is switched off), ( c2) exhaust process, and(c3) suction process.

the steam drove part of water in the chamber and pipe outof the boat (gure 2(a 2)). This part of water exited the rearof the boat in the form of a jet, completing the exhaustprocess (gures 2(a 3) and ( c2)). After part of water in the

chamber and the pipes was driven out by the heating, somepart of vapor transported to the section with low temperature.It condensed back into water. Accordingly, pressure wasreduced. When it was lower than the outside (atmospheric)pressure, water surrounding the boat was drawn into thepipes and chamber (gures 2(b) and ( c3)). In the exhaustprocess, the water was jetted out of the pipe in the form of an approximately straight column (gure 2(c2)) while in thesuction process, water was drawn into the pipe through all thedirections (gure 2(c3)). Thus, the jetted water had a largerlongitudinal component of linear momentum than the drawnwater. According to conservation law of linear momentum,

during a single propulsive cycle, the boat was drivenforward.

The testing results of these two putt-putt boats indicatethree points: (i) the thermally oscillatory approach maybe applied to propel a submarine under liquid, since thepropulsion does not require that the propelled vehicle shouldstay on a liquid surface; (ii) a resistive heater, instead of acandle, can be applied to generate a jet at a low voltage(it was 5 V in the test); (iii) compared with other thermalapproaches of propulsion, this thermally oscillatory methoddoes not need fuel. Since it is simple to fabricate the requiredchamber, nozzles and resistive heater using microfabricationtechniques, the thermally oscillatory method is adopted hereto propel microsubmarines.

(a )

(b)

Figure 3. Schematics of ( a ) top and bottom portions, and(b) assembled microsubmarine.

3. Design and fabrication of microsubmarines

Based on the understanding gained in the preliminary tests ontheputt-puttboats, themicrosubmarine is designed to have twoportions (gure 3). The top portion includes two SU-8 layers(gure 3(a )). The second SU-8 layer has a reservoir and twonozzles, and the rst SU-8 layer seals the tops of the reservoirand nozzles. The bottom portion includes a serpentine-shaped

3


5/12


(a 2)

(a 3)

(b2)

Third SU-8 layerS1813

Si waferAu layer

P atterned A u heater Au electrode

(b1)

(b3)

Cr layer

First SU-8 layer

Second SU-8 layer

(d )

(c1)

(c2) Cu or A u wire

Silver epoxy

Chamber

Teflon

Electrodes

Goldheater

(a1)

Figure 4. Procedures to fabricate the microsubmarine (not to scale): generate ( a ) the bottom and ( b) top portions, ( c1) and ( c2) bond top andbottom portions together, and ( d ) bond two interconnects (two Au wires) on this submarine.

Au heater and two contact pads, which are located on top of an SU-8 layer (gure 3(b)). This SU-8 layer also seals thebottoms of the reservoir and nozzles. The Au heater, reservoirandtwo nozzles form a desired thermal ejector. Once theheateris turned on, the thermal ejector produces a jet, providing athrust to drive the submarine along the direction opposite tothat of the jet.

Isopropyl alcohol (IPA) is chosen as the tested liquid,since it has a low mass density (0.76 103 kg m 3) andlow boiling temperature (82.5 C). SU-8 is selected to be the

structural material for microsubmarines for two reasons. First,the density of SU-8 is 1.16 103 kg m 3 (according to ourmeasurement), larger than that of IPA. This difference in thetwo densities enables the SU-8 to stay under IPA. Second,SU-8 is a negative photoresist. It can be patterned todesired shapes using UV lithography, and its thickness canbe controlled for producing SU-8 submarines with properthicknesses. SU-8 has been widely applied in the MEMS areafor various high-aspect-ratio patterning purposes [2729]. Itcanalso function asa good structural material inmicrosystems,such as in bio microuidic device [30, 31], boats [ 58] andotillas [9, 10]. The density of PDMS is 0.97 103 kgm 3 [32]. It is not a photoresist. It is normally patterned by amolding process [33], which may not be good at three-layerpatterning as required in generating a microsubmarine. Since

it is relatively simple to directly fabricate microsubmarinesout of SU-8 using UV lithography, this material is adopted inthis work accordingly. Metals and silicon have much higherdensities than IPA, and might not be good structural materialsfor microsubmarines.

A three-step procedure given below is applied to fabricatethe microsubmarines(gure 4): (i) generate thebottomportion(gure 4(a ), (ii) produce the top portion (gure 4(b), and (iii)bond the two portions together, followed by bonding two Cuor Au wires (interconnects) to their contact pads (gures 4(c)

and ( d ).In the rst step (gure 4(a ), an SU-8 layer is patterned on

top of an S1813-coated Si wafer using UV lithography. S1318is a positive photoresist, and is employed here as the sacricialmaterial to release SU-8 structures from the Si wafer as willbe seen in the second and third fabrication steps. A thin Crlm is then sputter-coated on the SU-8 layer, followed by thesputter-coatingof a Au layer. TheCr lmserves as an adhesionlayer between the Au and SU-8 layers. Subsequently, the Auand Cr layers are patterned in the form of heaters and contactpads using UV lithography.

The second step includes three sub-steps (gure 4(b): (i)spin-coat the rst SU-8 layer on a S1813-coated Si wafer,and pattern it to form the top layer of a submarine using theUVlithography; (ii) spin-coat thesecondSU-8 layer on therst

4


6/12


(a ) (b)

(c)

Figure 5. (a ) Fabricated and ( b) dimensions of the microsubmarine, and ( c) dimensions of the heater (Unit: mm).

one, and pattern it to form the middle layer of the submarineusing the UV lithography; and (iii) remove these twoSU-8 layers (i.e., the top portion of the submarine) from theSi wafer by etching S1813 using its developer.

The third fabrication step includes four sub-steps

(gures 4(c) and ( d )). In the rst sub-step, the SU-8 layerin the top portion is rst coated with a thin SU-8 coating, thetop portion is then ipped over and placed on the top of theSU-8 layer of the bottom portion, and nally the whole sampleis heated at 95 C for 30 min to cure the thin SU-8 coating. Inthe second sub-step, the whole sample is released from the Siwafer by etching S1813 using its developer. In the third sub-step, the top surface of microsubmarine is coated by a layerof Teon. Finally, in the fourth step, two identical Cu or Auwires are connected to the two contact pads of the submarine,respectively, using conductive silver epoxy, completing thefabrication of microsubmarines.

Figures 5(a ) and ( b) show a representative submarinegenerated and its dimensions. This prototype is 8 mmlong, 6 mm wide and 0.96 mm thick. The top, middle

and bottom SU-8 layers are 200, 500 and 260 m thick,respectively. The reservoir and nozzle had the dimensions of 5 4 0.5 mm 3 and 1 1 0.5 mm 3, respectively. Thetotal mass of the microsubmarine is 39 mg, and the volumeof the reservoir is 12 mm 3. The electrical resistance of the Auheater is 450 . The Au heater has a serpentine-shape line,which is 100 m wide and 48.6 mm long, separated by aspacing of 100 m (gure 5(c)).

4. Experimental results and discussions

Three types of tests are done to explore the performance of the microsubmarine, which will be presented in the followingthree sub-sections, respectively. The rst type of tests is toexamine the thermally oscillatory process. The second type isto optimize applied voltage, pulse frequency and duty for thepurpose of maximizing the propulsive force generated duringa thermally oscillatory process. The third type of tests is toinvestigate the horizontal motions of the submarine. In all the

5


7/12


8/12


(a )

(b1) (b2)

Figure 7. (a ) The behavior of the generated bubble when the applied voltage is 13 V, pulse frequency is 1 Hz and duty is 50%. Flow patternbehind the microsubmarine when bubble was ( b1) growing or ( b2) shrinking. The arrows show the moving direction and speeds of particles.

from the nozzles in a jet form (gure 7(b1)), indicating thatan IPA jet was ejected from the reservoir. In the suctionstage, as the bubble gradually shrank, the particles behindthe submarine were moving back towards the nozzles fromalmost all directions (gure 7(b2)), implying that IPA wasdrawn into the reservoir through almost all the directions.Accordingly, due to the difference between the exhaust andsuction processes in the ow directions, a thrust should beproduced to drive the submarine forward.

4.2. Optimiz a tion of p a r a meters

In order to optimize applied voltage, pulse frequency and dutyfor the purpose of maximizing the thrust, we rst estimate theaverage thrust generated during a cycle of thermal oscillation.Set x to be the coordinate locating the exit plane of the nozzle.During a single exhaust-suction cycle, the boat is operating inan oscillatory manner. The transient nature of the parameters,such as speeds, can be eliminated by averaging each parameterover one complete cycle. Let [[ G]] = 1T

T

0 G dt , where Grepresents a parameter, T the period of a cycle, and [[ G]] theaverage value of G over a period. Set [[ P ]] = 1T

T e0 P dt , where

P also represents a parameter, and T e the time duration of theexhaust process in a cyclic period. Accordingly, it was derivedin [26] that

[[F ]] = [U 2e ]a , (1)

where F denotes propulsive force along in the x direction, athe total cross-sectional area of nozzles, and U e the exit speedof the jet during the exhaust process, and that

[[m x ]] + [[ D( x )]] = U 2e a , (2)

where m is the mass of the microsubmarine and loaded IPAat time t , x and x are, respectively, speed and accelerationof the microsubmarine, and D( x ) is the drag force on themicrosubmarine. It has been demonstrated in [ 26] that theintake speed during the suction stage of a cyclic period hasnegligible effect on F . In other words, F is negligible duringthe suction stage. Thus, as seen from the right-hand side of equation ( 1), F is not related to the intake speed. It can alsobe seen from equations ( 1) and ( 2) that the thrust, speed andacceleration of the putt-putt boat are directly related to the exitspeed of the jet during the exhaust process. Hence, to controlthe motion of such a boat, it is important to know how tocontrol this speed during the exhaust process. Equations (1)

7


9/12


(a ) (b)

(c) (d )

Figure 8. Volumetime relationship in acycle when frequency is ( a ) 1 Hz. Volumetime relationship and the corresponding curve ttingduring the exhaust process of a cycle when the frequency is ( b) 1, (c) 10 or ( d ) 50 Hz. The applied voltage is 13 V and duty is 50% in allthese tests.

and (2) are also applicable to the microsubmarine, since thesame propulsive approach is used.

Let V and V b denote thevolumes of the reservoir andvaporbubble, respectively. V is constant, while V b is a function of time ( t ) ina period. T is theperiodwhile T e is thepulse durationin a cycle. We have

U 2e =1

a 2T T e

0

d(V V b )dt

2

dt =1

a 2T T e

0

dV bdt

2

dt .

(3)

According to equation ( 3), in order to determine [ U 2e ], itis necessary to know dV bdt , which is subsequently determinedthrough experiments. At a time instant, V b is calculated asthe multiplication of the lateral bubble area with the reservoirheight. As shown in gure 8, the volume-time relations wereexperimentally determined, respectively, at frequencies of 1,10, 25 and 50 Hz when applied voltage and duty were xedto be 13 V and 50%. For each frequency, as well as in thetests that will be subsequently presented, the volume of thebubble rst linearly increased during the expansion process

and then decreased with time during the shrinking process.Let k = dV bdt denote the slope of the volumetime curve at theexpansion process. Accordingly, k is a constant for a givenpulse frequency, and equation (3) becomes

U 2e

=k 2T ea 2T

. (4)

Let D denote the applied duty. Then, we have

D =T eT

. (5)

Subsequently, with the aid of equations (4) and ( 5),equation ( 1) can be re-written as

[[F ]] = k 2 D

a. (6)

This equation implies that the thrust increases withthe increase in both D and k . In addition to duty, appliedvoltage and pulse frequency also affected the thrust. Thus,to generate a large thrust, we determined the optimal valuesof these three parameters in the following order: (i) xed theduty and pulse frequency and determined an optimal value

8


10/12


Figure 9. Thrust- and k-voltage relationships when frequency is1 Hz and duty is 50%.

of applied voltage, (ii) xed this optimal value of appliedvoltage, together with the duty, and found an optimal value of pulse frequency, and (ii) determined an optimal value of theduty when the optimal values of another two parameters werexed. During this process of nding the optimal values, thecorresponding trust was determined using equation ( 6).

We rst xed duty to be 50% and frequency to be 1 Hz,determined the relationship of k and voltage, and furthercalculated the relationship of the trust and voltage. Four pointswere found (gure 9). First, when the applied voltage from 1to 16 V, the thrust also increased, had a peak value at 16 V, andstarted to drop at 19 V. Second, at 22 V, only the second stagewas observed, implying that the bubble grew and oscillatedsimultaneously. After about 20 s of growth at this appliedvoltage, the IPA vapor occupied the whole chamber, no liquidIPA was ejected out of the reservoir, and also no liquid IPA wassucked into the reservoir. Thus, there was no thrustduring sucha cycle. When the IPA bubble further expanded, part of thisbubble got out of the chamber, and was broken away from thepart of thebubble still inside thereservoir. Subsequently, liquidIPA was able to be sucked back in the chamber, re-starting thethermally oscillatory process. Third, when the applied voltagewas 22 V or above, the heater would be broken after runningfor 2 min or longer. Since most of time the reservoir wasoccupied by IPA vapor only, the temperature of the heater kept

increasing, making this heater easy to break. Thus, the largestvoltage tested in our experiments was 22 V. Fourth, if thevoltage was lower than 10 V, the thrust would be small. At 1and4 V, theheat generatedby heaterwas noteven largeenoughto nucleate a visible bubble. Only after applied voltage beyond7 V, the bubble was generated and its size slightly increasedwith the increase of voltage.

We then xed the voltage and duty to be 16 V and 50%,respectively, and found the relationship between the thrust andpulse frequency, which varied from 1 to 500 Hz (gure 10).We found that the thrust has a maximum value at 100 Hz(gure 10). The value of the thrust rst increased with increaseof frequency and then decrease with it. The largest thrust wasobtained at the frequency of 100 Hz. If the frequency wastoo high (such as 300 Hz), the bubble behaved as in the case

Figure 10. Thrust- and k-frequency relationships when voltage is16 V and duty is 50%.

Figure 11. Thrustduty relationship at 16 V and 100 Hz.

that voltage was continuously applied. The accumulation of heat made the corresponding bubble not vary much during acycle, and consequently not much liquid IPA was sucked intothe reservoir, resulting in a small thrust. In case the frequencywas too low (such as 1 Hz), the accumulated heat was notlarge enough for the bubble to grow quickly, also resulting ina small thrust. As observed from gure 10 , 100 Hz was anoptimal frequency, yielding the largest k and thrust.

Finally, we xed the voltage as 16 V and frequency as100 Hz to study the relationship between the thrust and duty.As seen from gure 11, the duty was varied from 20% to 80%with an increment of 10%. The thrust initially increased whenthe duty increased from 20% to 50%, and then decreased whenthe duty increased from 50% to 80%. Hence, the largest thrustappeared at duty of 50%, which was 67.6 nN.

In summary, in the second type of tests, the optimalvalues of voltage, frequency and duty were, respectively, 16 V,100 Hz, and 50%, resulting a thrust of 67.6 nN.

4.3. Motions

In order to examine the horizontal motions of themicrosubmarine, we performed a third type of tests, and

9


11/12


(a )

(b)

Figure 12. Experimental results: ( a ) sequential snapshots of amotion when pulse frequency is 100 Hz, voltage is 16 V and duty is50% (a red circle represents the geometric center of the submarine ata time instant), and ( b) the speedtime curve.

connected it with power-supply devices through two Au wireswith identical diameters of 25 m. Due to the constraintof the Au wires, the moving path of the submarine wasslightly curved. When the applied voltage was 16 V, thepulse frequency was 100 Hz and duty was 50%, it took thesubmarine about 4 s to start moving, reached the highest

speed of 1.8 mm s 1

at 9.5 s, and stopped at 17 s due tothe balance of the thrust with the forces provided by the twoAu wires (gure 12). The total travel distance was 12.6 mm,which was about 7.4 times of the one generated using thethermopneumatic approach [22].

The submarine mainly suffers IPA resistance and the wireconstraint. The IPA resistance slows down the submarine,while the Au wires limit its travel distance. Accordingly,we estimate the effect of these two wires on the travel distance.In the experimental setup, the wires are suspended in air toavoid their contact with IPA, since we found that such contactincreased their resistance to the movement of the submarine.One end of the two wires is xed on the power-supply device,and the other end is attached to the submarine and moveswith it. Let us denote the two ends by a and b , respectively.

We set up a simple model to understand the constraint of these two wires, which are combined together and modeledas a cantilever beam. The end a is considered to be xed.However, the end b is open, but suffers a concentration force,which equals the thrust of the submarine at the end of themotion when this submarine stops. Let P and L denote the

concentration force and the average length of the two wires.Thus, we have [ 34]

=PL 3

3 EI , (7)

where denotes the deection at b, E is Youngs modulus of Au wires, which was measured to be 58 GPa, and I is thearea moment of inertia of the two wires, whose expressionis I = d

4

32 with d as the wire diameter. It is observed fromequation ( 7) that increases with the increase in both P and Lwhile decreases with the increase in d . In our case, the wireswith diameters as small as 25 m are used, the maximumlength of the two wires is 10 cm (the longer ones are easy to

bend and get contact with IPA), and P

equals the optimizedthrust of 67.6 nN.Subsequently, by equation ( 7), is estimatedto be 10.1 mm. This result implies that, for the given setup andprototype, the maximum travel distance should be in the orderof 10 mm. This is actually our case, in which the maximumtravel distance measured is 12.6 mm. Although the furtheroptimization of the setup and the submarine, such as reductionof the wire diameters, increase of their effective lengths andimprovement of the thrust, may increase the travel distance,it is expected that the total travel distance is still limited dueto the constraint of the wires. Hence, in the near future, weplan to develop wirelessly powered submarines, whicheliminate the need of wire connection, for the purpose toincrease their travel distances.

In addition to testing the prototype using the optimizedparameters obtained in the second type of tests, we also xedthe duty to be 50%, and examined its motions by varyingapplied voltage and pulse frequency: applied voltage waschanged from 1 to 22 V with an increment of 3 V, and thepulse frequency of the voltage varied from 1 to 500 Hz. Inall cases, it took 15 s to form a visible, oscillating bubblein the reservoir. Three situations were observed regarding themobility of the submarine. First, when the applied voltagewas 7 V or below, the submarine did not move for any testedfrequency, indicating that the generated thrust was not large

enough to overcome the constraint of the Au wires. Second,the submarine began to move when applied voltage was 13 Vfor all the tested frequencies. Third, at an applied voltage of 22 V, the submarine moved rst. However, after t = 10 s, itstopped for 12s and restartedagain. The four points observedin the second type of tests actually explained the occurrenceof these three situations.

5. Summary and conclusions

In this work, through experimental and theoretical investiga-tions, we developed the prototype of a microsubmarine basedon a thermally oscillatory approach of propulsion. We investi-gated the design, fabrication, actuation and horizontal motionsof this prototype. The prototype had a thermally oscillatory

10


12/12


process similar to that of the putt-putt toy boat. The propulsiveparameters were optimized. At an applied voltage of 16 V,pulse frequency of 100 Hz, and duty of 50%, the submarinewas found to have the highest speed of 1.8 mm s 1 and thelongest travel distance of 12.6 mm. The corresponding aver-age thrust per cycle was calculated to be 67.6 nN. Also, it was

considered that the constraint of the wire connections limitedthe travel distance of the submarine. Thus, in the near future,we plan to develop wirelessly powered submarines to elimi-nate this constraint. Furthermore, in this paper, we focused onthe horizontal motions of the submarine. In the near future, wealso plan to explore its vertical motions (i.e., rising and sink-ing under liquid), as well as and corresponding design, basedon our recent work, which used trapped bubbles to increasebuoyancy for the purpose of raising a small object from underliquid [ 35].

References

[1] Tang W C, Nguyen T-C H and Howe R T 1989 Laterallydriven polysilicon resonant microstructures Sensors Actu a tors 20 2532

[2] Fan L-S, Tai Y-C and Muller R S 1989 IC-processedelectrostatic micromotors Sensors Actu a tors20 4147

[3] Yan J, Avadhanula S, Birch J, Dickinson M, Sitti M, Su Tand Fearing R 2001 Wing transmission for amicromechanical ying insect J. Micromech. 1 22137

[4] Mitsumata T, Ikeda K, Gong J P and Osada Y 2000 Controlledmotion of solvent-driven gel motor and its application as agenerator La ngmuir 16 30712

[5] Luo C, Li H and Liu X 2008 Propulsion of microboats usingisopropyl alcohol as a propellant J. Micromech. Microeng.

18 067002[6] Luo C, Qiao L and Li H 2010 Dramatic squat and trimphenomena of mm-scaled SU-8 boats induced byMarangoni effect Microuid. N a nouid. 9 5737

[7] Liu X, Li H, Qiao L and Luo C 2011 Driving mechanisms of CM-scaled PDMS boats of respective close and openreservoirs Microsyst. Technol. 17 87589

[8] Qiao L, Xiao D, Lu F K and Luo C 2012 Control of the radialmotion of a self-propelled microboat through a side rudderSensors Actu a tors A 188 35966

[9] Li H, Qiao L, Liu X and Luo C 2012 Fabrication and testing of a self-propelled, miniaturized PDMS otilla Microsyst.Technol. 18 143144

[10] Li H and Luo C 2011 Development of a self-propelledmicrootilla Microsyst. Technol. 17 77786

[11] Barnaby K C 1967 Ba sic N a va l Architecture 5th edn (London:Hutchinson)[12] Crane J 1984 Subm a rine (London: British Broadcasting)[13] Satchell D 1987 Silicon microengineering for accelerometers

Int. Conf. on the Mech a nic a l Technology of Inerti a l Devicesvol 1 (Newcastle, England) pp 1913

[14] Venkatesh S and Culshaw B 1985 Optically activatedvibrations in a micromachined silica structure Electron. Lett. 21 3157

[15] Andres M, Tudor M and Foulds K 1987 Analysis of aninterferometric optical bre detection technique applied tosilicon vibrating sensors Electron. Lett. 23 7745

[16] Greenwood J 1984 Etched silicon vibrating sensor J. Phys. E:Sci. Instrum. 17 650

[17] Nathanson H C, Newell W E, Wickstrom R A and Davis J R Jr1967 The resonant gate transistor IEEE Tr a ns. Electron. Devices 14 11733

[18] Petersen K E and Guarnieri C 1979 Youngs modulusmeasurements of thin lms using micromechanics J. Appl.Phys. 50 67616

[19] Ikeda K, Kuwayama H, Kobayashi T, Watanabe T,Nishikawa T, Yoshida T and Harada K 1990 Siliconpressure sensor integrates resonant strain gauge ondiaphragm Sensors Actu a tors A 21 14650

[20] Wang W, Yao Z, Chen J C and Fang J 2004 Composite elasticmagnet lms with hard magnetic feature J. Micromech. Microeng. 14 1321

[21] Dijkink R, Van Der Dennen J, Ohl C and Prosperetti A 2006The acoustic scallop: a bubble-powered actuator J. Micromech. Microeng. 16 1653

[22] Ku C M, Liao H H and Yang Y J 2011 MEMS-based buoyancy

and propelling mechanisms for a micro-submarine NEMS11: IEEE Int. Conf. on N a no / Micro Engineered a nd Molecul a r Systems (T a iwa n, K a ohsiung) pp 7447

[23] Wang J M, Lin T Y and Yang L J 2007 Electrohydrodynamic(EHD) micro-boat NEMS07: IEEE Int. Conf. on N a no / Micro Engineered a nd Molecul a r Systems (Th a ila nd, Ba ngkok) pp 5847

[24] Mita Y et a l 2009 Demonstration of a wireless driven MEMSpond skater that uses EWOD technology Solid-St a te Electron. 53 798802

[25] Bass V 2012 The pop-pop pageswww.nmia.com/ vrbass/pop-pop/

[26] Finnie I and Curl R L 1963 Physics in a toy boat Am. J. Phys.31 28993

[27] Lorenz H, Despont M, Vettiger P and Renaud P 1998

Fabrication of photoplastic high-aspect ratio microparts andmicromolds using SU-8 UV resist Microsyst. Technol.4 1436

[28] Dellmann L et a l 1998 Two steps micromoulding andphotopolymer high-aspect ratio structuring for applicationsin piezoelectric motor components Microsyst. Technol.4 14750

[29] Luo C et a l 2006 Thermal ablation of PMMA for water releaseusing a microheater J. Micromech. Microeng. 16 580

[30] Nijdam A et a l 2005 Fluidic encapsulation inSU-8 -reservoirs with -uidic through-chip channelsSensors Actu a tors A 120 17283

[31] Gadre A et a l 2004 Fabrication of a uid encapsulated dermalpatch using multilayered SU-8 Sensors Actu a torsA 114 47885

[32] Mark J 1999 Polymer D a t a H a ndbook (New York: OxfordUniversity Press)[33] Xia Y and Whitesides G M 1998 Soft lithography Annu. Rev.

M a ter. Sci. 28 15384[34] Gere J and Timoshenko S 1990 Mech a nics of M a teri a ls

(Boston, MA: PWS Publishing)[35] Qiao L, Xiang M and Luo C 2012 Increase buoyancy of a solid

fragment using micropillars Sensors Actu a tors A182 13645

11
http://dx.doi.org/10.1016/0250-6874(89)87098-2http://dx.doi.org/10.1016/0250-6874(89)87098-2http://dx.doi.org/10.1016/0250-6874(89)87100-8http://dx.doi.org/10.1016/0250-6874(89)87100-8http://dx.doi.org/10.1163/156856301760132123http://dx.doi.org/10.1163/156856301760132123http://dx.doi.org/10.1021/la990483ohttp://dx.doi.org/10.1021/la990483ohttp://dx.doi.org/10.1088/0960-1317/18/6/067002http://dx.doi.org/10.1088/0960-1317/18/6/067002http://dx.doi.org/10.1007/s10404-010-0569-4http://dx.doi.org/10.1007/s10404-010-0569-4http://dx.doi.org/10.1007/s00542-010-1201-yhttp://dx.doi.org/10.1007/s00542-010-1201-yhttp://dx.doi.org/10.1016/j.sna.2012.04.004http://dx.doi.org/10.1016/j.sna.2012.04.004http://dx.doi.org/10.1007/s00542-012-1569-yhttp://dx.doi.org/10.1007/s00542-012-1569-yhttp://dx.doi.org/10.1007/s00542-010-1178-6http://dx.doi.org/10.1007/s00542-010-1178-6http://dx.doi.org/10.1049/el:19850223http://dx.doi.org/10.1049/el:19850223http://dx.doi.org/10.1049/el:19870549http://dx.doi.org/10.1049/el:19870549http://dx.doi.org/10.1088/0022-3735/17/8/007http://dx.doi.org/10.1088/0022-3735/17/8/007http://dx.doi.org/10.1109/T-ED.1967.15912http://dx.doi.org/10.1109/T-ED.1967.15912http://dx.doi.org/10.1063/1.325870http://dx.doi.org/10.1063/1.325870http://dx.doi.org/10.1016/0924-4247(90)85028-3http://dx.doi.org/10.1016/0924-4247(90)85028-3http://dx.doi.org/10.1088/0960-1317/14/10/005http://dx.doi.org/10.1088/0960-1317/14/10/005http://dx.doi.org/10.1088/0960-1317/16/8/029http://dx.doi.org/10.1088/0960-1317/16/8/029http://dx.doi.org/10.1016/j.sse.2009.02.020http://dx.doi.org/10.1016/j.sse.2009.02.020http://www.nmia.com/~vrbass/pop-pop/http://www.nmia.com/~vrbass/pop-pop/http://www.nmia.com/~vrbass/pop-pop/http://dx.doi.org/10.1119/1.1969435http://dx.doi.org/10.1119/1.1969435http://dx.doi.org/10.1007/s005420050118http://dx.doi.org/10.1007/s005420050118http://dx.doi.org/10.1007/s005420050119http://dx.doi.org/10.1007/s005420050119http://dx.doi.org/10.1088/0960-1317/16/3/014http://dx.doi.org/10.1088/0960-1317/16/3/014http://dx.doi.org/10.1016/j.sna.2004.11.004http://dx.doi.org/10.1016/j.sna.2004.11.004http://dx.doi.org/10.1016/j.sna.2004.01.049http://dx.doi.org/10.1016/j.sna.2004.01.049http://dx.doi.org/10.1146/annurev.matsci.28.1.153http://dx.doi.org/10.1146/annurev.matsci.28.1.153http://dx.doi.org/10.1016/j.sna.2012.05.046http://dx.doi.org/10.1016/j.sna.2012.05.046http://dx.doi.org/10.1016/j.sna.2012.05.046http://dx.doi.org/10.1146/annurev.matsci.28.1.153http://dx.doi.org/10.1016/j.sna.2004.01.049http://dx.doi.org/10.1016/j.sna.2004.11.004http://dx.doi.org/10.1088/0960-1317/16/3/014http://dx.doi.org/10.1007/s005420050119http://dx.doi.org/10.1007/s005420050118http://dx.doi.org/10.1119/1.1969435http://www.nmia.com/~vrbass/pop-pop/http://dx.doi.org/10.1016/j.sse.2009.02.020http://dx.doi.org/10.1088/0960-1317/16/8/029http://dx.doi.org/10.1088/0960-1317/14/10/005http://dx.doi.org/10.1016/0924-4247(90)85028-3http://dx.doi.org/10.1063/1.325870http://dx.doi.org/10.1109/T-ED.1967.15912http://dx.doi.org/10.1088/0022-3735/17/8/007http://dx.doi.org/10.1049/el:19870549http://dx.doi.org/10.1049/el:19850223http://dx.doi.org/10.1007/s00542-010-1178-6http://dx.doi.org/10.1007/s00542-012-1569-yhttp://dx.doi.org/10.1016/j.sna.2012.04.004http://dx.doi.org/10.1007/s00542-010-1201-yhttp://dx.doi.org/10.1007/s10404-010-0569-4http://dx.doi.org/10.1088/0960-1317/18/6/067002http://dx.doi.org/10.1021/la990483ohttp://dx.doi.org/10.1163/156856301760132123http://dx.doi.org/10.1016/0250-6874(89)87100-8http://dx.doi.org/10.1016/0250-6874(89)87098-2

Documents

Propulsion of a Microsubmarine Using a Thermally Oscillatory Approach