Sensors and Actuators A 108 (2003) 168–174

A piezoelectrically actuated micro synthetic jet for active flow controlChester Lee, Guang Hong∗, Q.P. Ha, S.G. Mallinson

Faculty of Engineering, University of Technology, Levd. 6 Building 2, Broadway NSW 2007, Sydney, Australia

Received 29 July 2002; received in revised form 18 April 2003; accepted 12 May 2003

Abstract

The synthetic jet actuator (SJA) is a low power, highly compact microfluidic device which has potential application in boundary layerflow control. In recent work we have shown how synthetic jets work without cross flow and how effectively they modify the flow structure inthe boundary layer under an adverse pressure gradient. This paper describes the piezoelectric synthetic jet actuator used in our experiments.The experimental set-up for flow control using this type of actuator is detailed. The results obtained show a significant enhancement of thejet effectiveness by forcing the boundary layer flow at the natural instability frequency. The actuators must have sufficient velocity outputto produce strong enough vortices if they are to be effective for flow control. The forcing effect can occur at a frequency lower than thedriving frequency of the actuator when used without cross flow. The forcing frequency appears to be an important parameter in syntheticjet boundary layer flow control.© 2003 Elsevier B.V. All rights reserved.

Keywords: Synthetic jet; Piezoelectric actuation; Active flow control

1. Introduction

Flow control involves passive or active devices that causea beneficial change in boundary layer or free shear flows,with useful end results including drag reduction, lift en-hancement, mixing augmentation and flow-induced noisesuppression. Active flow control is a multidisciplinary re-search area combining sensing, actuation, flow physics andcontrol with the objective of modifying flow field character-istics to achieve a desired aerodynamic performance. Unlikepassive techniques, such as geometric shaping to adjust thegradient pressure or placement of longitudinal grooves or ri-blets on a surface to reduce drag, active control manipulatesa flow field by using a time dependent forcing system, typ-ically to leverage a natural instability of the flow and thusto amplify the control effectiveness. Advantages of activeflow control include the ability to attain a large effect usinga small, localized energy input, and to control complex dy-namical processes; for example, the reduction of skin frictionand hence viscous drag[1,21] in turbulent boundary layers.

A synthetic jet actuator (SJA) is a jet generator that re-quires zero mass input yet produces non-zero momentumoutput. Developed in recent years, the synthetic jet actuatorfalls within the area of micro-electro-mechanical systems(MEMS) if the characteristic dimension or the diameter of

∗ Corresponding author. Tel.:+61-2-9514-2677; fax:+61-2-9514-2655.E-mail address: guang.hong@uts.edu.au (G. Hong).

the orifice, through which the jets are generated, is less than1.0 mm[1]. Advantages of using SJA include simple com-pact structure, low cost and ease of operation. The basiccomponents of a SJA are a cavity and an oscillating mate-rial. A jet is synthesized by oscillatory flow in and out ofthe cavity via an orifice in one side of the cavity. The flowis induced by a vibrating membrane located on one wallof the cavity. There are many types of actuator that can beused in active flow control, such as thermal, acoustic, piezo-electric, electromagnetic and shape memory alloys. Here, apiezoelectric material is chosen to drive the oscillating di-aphragm because of such desirable characteristics as suchlow power consumption, fast response, reliability, and lowcost[9]. Flow enters and exits the cavity through the orificeby suction and blowing. On the intake stroke, fluid is drawninto the cavity from the area surrounding the orifice. Dur-ing one cycle of oscillation, this fluid is expelled out of thecavity through the orifice as the membrane moves upwards.Due to flow separation, a shear layer is formed between theexpelled fluid and the surrounding fluid. This layer of vor-ticity rolls up to form a vortex ring under its own momen-tum. By the time the diaphragm begins to move away fromthe orifice to pull fluid back into the cavity, the vortex ringis sufficiently distant from the orifice that it is virtually un-affected by the entrainment of fluid into the cavity. Thus,over a single period of oscillation of the diaphragm, whilstthere is zero net mass flux into or out of the cavity, there isalso a non-zero mean momentum flux. This momentum flux

0924-4247/$ – see front matter © 2003 Elsevier B.V. All rights reserved.doi:10.1016/S0924-4247(03)00267-X

C. Lee et al. / Sensors and Actuators A 108 (2003) 168–174 169

is, effectively, a turbulent-like jet that has been synthesizedfrom the coalescence of a train of vortex rings, or vortexpairs, of the ambient fluid[10,18,23]. Flow control can beachieved using traditional devices such as steady[19] andpulsed[20] jets. The obvious benefit of employing SJAs asa flow control device is that they require no air supply andso there is no need for piping, connections, and compressorsassociated with steady jets.

The emergence of micro-electro-mechanical systemstechnology, which employs the methods developed for thefabrication of silicon chips to construct very small-scalemechanical devices, provides a means of batch-fabricatingmechanical parts integrated with electronics and control cir-cuitry. The feasibility of MEMS-based active flow controlhas been demonstrated in recent research[2]. Among themany micro-active devices that have been a focal point ofrecent research into flow control, SJAs have demonstrated apotential in modification of aerodynamic lift and drag[3],forebody flow-asymmetry management[4], jet vectoringand enhanced entrainment[5], acetone mixing control[6],flow separation control[7], and aircraft maneuverability[8].

The properties of synthetic jets in the absence of crossflow have been investigated experimentally and numerically[10,11,24]. For synthetic jets in external cross flow, it wasdemonstrated in[3] that significant lift could be generatedon a two-dimensional cylinder using synthetic jets. The ef-fect of oscillatory blowing on increasing post-stall lift for asymmetric airfoil and the dependency on the location andthe operation of the jet were investigated experimentally, asgiven in[12]. A simulation study on the jet dynamics and itsinteraction with a flat plate boundary layer under zero pres-sure gradient has been systematically carried out using anincompressible Navier–Stokes solver[13]. Recent researchin the field has emphasized on the need of understanding theSJA’s effect on the flow and in particular its operation in aboundary layer under an adverse pressure gradient.

Based on the well-known fact that a turbulent boundarylayer is more resistant to flow separation than a correspond-ing laminar boundary layer[7,9,22], we have focused onhow and at what forcing voltages and frequencies the jetscan delay separation in a boundary layer and particularly un-der an adverse pressure gradient. In our previous work, themean velocity and the local momentum along the jet cen-terline were used to characterize the actuator performance[14]. We have experimentally demonstrated a significant en-hancement of the jet effectiveness by forcing the boundarylayer flow at its natural instability frequency[16]. It wasfound that the forcing effect, in the condition with cross flowcould occur at a frequency lower than resonant frequency ofits membrane material, which is believed to be the effectivedriving frequency of the actuator in the condition withoutcross flow. It was also noticed that the forcing frequency ap-peared to be an important parameter in synthetic jet bound-ary layer flow control, of particular interest the effective-ness at lower frequencies enhanced by Tollmien–Schlichtingwaves. An experimental investigation on how the synthetic

jet and the boundary layer flow interact, coupled with theinstability of the boundary layer flow, is reported in[16].This paper describes the piezoelectric synthetic jet actuatorused in our experiments and examines the effect of forcingparameters on the interaction with an adverse pressure gra-dient boundary layer.

2. Piezoelectrically actuated synthetic jet actuator

Different geometries have been proposed for SJA design.As summarized in[17], the more popular synthetic geome-tries consist of an enclosed cavity with a small orifice on oneface. The shapes of the cavity can be square, rectangular orround. The actuator used in our experiments consists of amembrane located at the bottom of a small cavity which hasan orifice in the face opposite the membrane.Fig. 1 showsthe schematic of the synthetic jet actuator for present studyinstalled underneath a flat plate to be fixed in the work-ing section of a wind tunnel. The actuator membrane is athin circular brass disc, 0.25 mm in thickness, held firmlyat its perimeter. A piezoceramic disc is bonded to the out-side face of the membrane. The lowest resonant frequencyof the membrane is 900 Hz and its lump sum capacitance isapproximately 140 nF. In operation, the piezoceramic discis driven by a sine wave from a standard electrical signalgenerator.

In the experiments without cross flow, the instantaneousvelocity at the centerline of the jet was measured using aDantec hot-wire anemometry system. The sample rates wereset at 3–10 kHz and the sample size is 4096 (212) to 16384(214) for each realization. The sample rates and sample sizewere adjusted in the above range in terms of the forcing fre-quency, which was chose between 0 and 3600 Hz. The wirediameter and length are 5.0�m and 1.25 mm. Consideringthat the length of the hot wire is greater than the orificediameter of the actuator, the instantaneous velocity sensed

Fig. 1. Dimensions of the synthetic jet actuator (enlarged fromFig. 1).Orifice diameterdo = 0.5 mm. Step orifice diameterd1 = 1.42 mm.Membrane diameterD = 40 mm. Orifice heightho = 5 mm. Step orificeheighth1 = 20 mm. Cavity heighthc = 1.25 mm.

170 C. Lee et al. / Sensors and Actuators A 108 (2003) 168–174

0

2

4

6

8

0 10 20 30 40y/d

Vel

oci

ty(m

/s)

U mean m/s

U rms m/s

1/ y

Fig. 2. Variation of jet centreline velocity with the distance from the orifice (no cross flow). Forcing frequencyf = 1506 Hz, forcing amplitude=±7.5 V.

by the hot wire is actually a spatial average of the velocityover the length of the hot wire[10,14]. Measurements madeat distances greater than approximately 20 orifice diametersfrom the orifice were noticeably susceptible to small am-plitude motions caused by air fluctuations in the laboratory[10]. Each measurement was repeated at least five times andaverage values were obtained from these.

In our previous work, the flow generated by a syntheticjet actuator with a circular orifice in the condition withoutcross flow was investigated not only experimentally[10,14]but also computationally[10]. The experimental results andcomputational predictions are in fair to good agreement witheach other and with the theory for a steady turbulent jet. It isfound, however, that the synthetic jet establishes itself muchmore rapidly than the steady jet, primarily because of tur-bulent dissipation. The centerline distributions of the mean,Uc, and fluctuating,u′

c, velocity are presented inFig. 2 forthe present actuator, which differs from previous designs byhaving a sub-cavity connecting the orifice and main cavities.The sub-cavity is present due to difficulties in obtaining ac-

0

2

4

6

8

0 600 1200 1800 2400 3000 3600

Frequency (Hz)

Vel

oci

ty (

m/s

)

U rms

U mean

Fig. 3. Mean and rms centerline velocity varying with the forcing frequency, no cross flow.

cess to the underside of the flat plate for machining. BothUcandu′

c are seen to fall rapidly with distance from the orifice,with the fluctuating component decreasing more rapidly fory/do > 5, wherey is the distance from the orifice normalto the wall,do is the diameter of the orifice. The velocityappears to vary as approximately 1/y for y/do > 10, indi-cating self-similar flow. Thus, the present actuator seems tobehave in a similar manner to our previous designs, sug-gesting that the sub-cavity has only a minor effect on theoutflow properties.

Fig. 3 shows how the mean and rms velocities, on thecenterline of the jet aty = 1.2 mm, vary with the forcingfrequency while the forcing amplitude was kept constant at±10 V. The maximum mean and rms velocity at the center-line of the jet are 6.4 and 7.4 ms−1, respectively, at 1.5 kHz.In the low frequency range, a secondary peak velocity ap-pears at around 100–300 Hz with a much smaller magni-tude. From our ongoing project on the use of synthetic jetsto modify boundary layer flows, we have noticed that thesynthetic jets are more effective when the forcing frequency

C. Lee et al. / Sensors and Actuators A 108 (2003) 168–174 171

is in the range of T-S frequencies[14]. Further, the soundgenerated by the synthetic jet actuators at higher frequencies(>1 kHz) is audible and it is desirable to avoid such noise.

3. Flow control experimental set-up

The experiments on synthetic jets operating with a crossflow were performed in the Aerodynamics Laboratory atUTS using an open circuit, suction type wind tunnel. Airenters the working section via a bell mouth inlet, a hon-eycomb section, and a short settling chamber fitted withscreens for turbulence reduction. The maximum free streamvelocity achievable is 40 ms−1 and the minimum turbulenceis 0.3%. All of the experiments reported in this paper em-ployed a wind tunnel free stream velocity ofu∞ = 8 ms−1.Fig. 4 shows the arrangement of the working section in thewind tunnel (x is the streamwise distance downstream of theplate leading edge). An octagonal test section, 608 mm×608 mm, was employed and a slot channel was made throughthe center of the top of the test section to allow hot-wireprobes to be placed in the flow. A polished aluminum plate,1500 mm× 608 mm× 25 mm, was mounted horizontallyin the test section. The actuator, stepped orifice instead ofstraight orifice, shown inFig. 1was attached under the platewith the orifice centered atx = 307 mm on the plate center-

Fig. 4. Experimental setting in the wind tunnel.

U inf. = 8 m/s

-0.8

-0.6

-0.4

-0.2

0.0

0.08 0.18 0.28 0.38 0.48 0.58 0.68 0.78

X (m)

C p

Fig. 5. Pressure distribution in the streamwise direction.Xo = 0 at the leading edge of the flat plate.cp is pressure coefficient.

line. For the measurements in the adverse pressure gradientboundary layer, a fairing was set above the flat plate with itsangle adjustable for establishing the desired pressure gradi-ent. Static pressure taps are located every 25 mm along theflat plate and the pressure distribution was measured usinga multi-tube manometer.Fig. 5 shows the pressure distri-bution, represented by pressure coefficient along the centerstreamline. The pressure distribution exhibits an inflectionbetweenx = 390 and 450 mm from the leading edge. De-noting the distance downstream of the jet orifice center asXj ,this indicates the separation point is found betweenXj = 40and 140 mm (note thatXj = 0 whenx = 307 mm)[15,16].

In the condition with cross flow, the hot-wire anemometerwas used to measure the instantaneous velocity in the bound-ary layer in the wind tunnel. The sample rate was 6 kHz, andthe sample size of each realization was 4096 (212). The dis-tance between the hot-wire probe and the wall, for the firsty position at eachx station, was checked using a telescope.Measurements were made atXj = 40–160 mm downstreamof the actuator orifice at 20 mm intervals. At each of thexstations, measurements were made at positions fromy =0.2 mm to the edge of the boundary layer. At each measure-ment position, the streamwise velocity was recorded in bothconditions of jet on and off. The forcing voltage for the syn-thetic jets were±7.5 and±10 V, and forcing frequencieswere 100, 140 and 180 Hz.

4. Results and discussions

The mean and rms velocity profiles, aty = 0.4 mm,Xj =120 mm, with the jet off and on, are shown inFig. 6 versusforcing frequency. The synthetic jet actuator is driven at fre-quencies between 0 and 1.6 kHz, with two different forcingamplitudes of±7.5 and±10 V. In Fig. 6a, it can be seen thatthe synthetic jet is most effective with a driving frequencyof 50–200 Hz when the forcing amplitude is±7.5 V, and40–350 Hz when the forcing amplitude is±10 V. The meanvelocity was 1.9 ms−1 with the jet off, and was amplified up

172 C. Lee et al. / Sensors and Actuators A 108 (2003) 168–174

1

3

5

7

9

0 400 800 1200 1600Frequency (Hz)

U m

ean

(m

/s)

jet on, +/-10V

jet on, +/-7 .5V

jet off

1

1.7

2.4

0 400 800 1200 1600Frequency (Hz)

RM

S(m

/s)

jet on, +/-10V

jet on, +/-7.5V

jet off

(a) (b)

Fig. 6. Mean and rms of the flow velocity vs. frequency aty = 0.4 mm, Xj = 120 mm: (a) mean velocity; (b) rms.

to four times with the jet on. InFig. 6b, rms velocities werechanged with jet on. The effective frequency ranges weresimilar to those for the mean velocity, whilst the rms veloc-ity at lower frequencies seemed to be greater for lower forc-ing amplitude. There are other two peaks, shown inFig. 6a,at 1.3 and 1.5 kHz. As noted earlier, we do not use forcingfrequencies higher than 300 Hz, as we have noticed that thesynthetic jets were most effective when the frequency is inthe range of that associated with flow instability. The re-sults ofFigs. 3 and 6indicate that the effectiveness of thesynthetic jet actuator in boundary layer control may not bepredicted solely on the basis of the output of the actuatoroperating in a condition without cross flow.

The flow separation suggested by the pressure distribu-tion, as shown inFig. 5, can be confirmed by examining theboundary layer velocity profiles.Fig. 7 demonstrates thatthe velocity profile stays almost constant fory < 0.8 mmwhen the jet is off, which is characteristic of boundary layerflow in the separation region[15]. As a result of the back-flow close to the wall, a strong thickening of the boundarylayer takes place and with this, boundary layer mass is trans-ported away into the outer flow. At the point of separation,the streamlines leave the wall at a certain angle. The posi-tion of separation is given by the condition that the velocity

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.2 0.4 0.6 0.8 1.0u / U

y /

del

ta

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.2 0.4 0.6 0.8 1.0u / U

y /

del

ta

jet off

100 Hz, +/-7.5V

180 Hz, +/-7.5V

jet off

100 Hz, +/-7.5V

180 Hz, +/-7.5V

(a) (b)

Fig. 7. Mean velocity profiles with jet on and off atXj = 120 mm,U∞ = 8 ms−1. (a) Effect of forcing amplitude; (b) effect of forcing frequency.

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.2 0.4 0.6 0.8 1.0u / U

y /

del

ta

jet off

180 Hz, +/-5V

180 Hz, +/-7.5V

180 Hz, +/-10V

Fig. 8. Critical forcing amplitude for boundary layer flow control atXj = 120 mm,U∞ = 8 ms−1.

gradient perpendicular to the wall vanishes at the wall[22].With the jet on, the profile is fuller, indicating that bound-ary layer separation was suppressed. The mean velocity pro-files remain almost unchanged when the forcing amplitudechanges, whilst varying the frequency appears to have a moresignificant effect. For example, in the regiony < 0.3 mm,

C. Lee et al. / Sensors and Actuators A 108 (2003) 168–174 173

Xj=40mm 60mm0

0.2

0.4

0.6

0.8

1

u / U

y/

del

tajet off 40mm, jet on

100mm, jet on

Xj=40mm 60mm 80mm 100mm 120mm 140mm 160mm0

0.2

0.4

0.6

0.8

1

Turbulence level (m/s)

y /

del

ta

jet off 40mm, jet on 60mm, jet on 80mm, jet on

100mm, jet on 120mm, jet on 140mm, jet on 160mm, jet on

120mm, jet on 120mm, jet on

60mm, jet on 80mm, jet on

120mm, jet on

80mm 100mm 120mm 140mm 160mm

Fig. 9. Mean velocity profiles and turbulence profiles in the separation region with the synthetic jet on and off. Freestreamvelocity= 8 m/s,forcingfrequency= 100 Hz, forcingamplitude= ±7.5 V.

the mean velocity is amplified more than 20 times with thejet on. These results, together with those above, suggest thatforcing frequency may be a more important parameter foractuation effectiveness than forcing amplitude.

The piezoelectric synthetic jet actuators must have suffi-cient velocity output to produce strong longitudinal vorticesif they are to be effective for flow control[11]. Fig. 8showsthe effect of forcing amplitude at a forcing frequency of180 Hz. There were three different forcing amplitudes,±5,±7.5 and±10 V, and 180 Hz forcing frequency applied withjet on. The mean velocity profile was hardly changed whenthe forcing amplitude was±5 V. When forcing amplitude isjust over±5 V, the effectiveness of the flow control seemsto be stable. Thus, it would appear that a critical forcing am-plitude exists, below which the control effect of the presentactuator is negligible.

Fig. 9shows the mean velocity and turbulence profiles ateach measurement station along the streamwise direction.Examining the mean profiles, it can be seen that in the con-dition with the jet off, the separation happens betweenXj =40 and 60 mm. AtXj = 60, 80, 100 and 120 mm, the meanvelocity profile at they positions close to the wall is nearlyconstant, and the velocity profile has an inflection point. Thisseparation continues to a position betweenXj = 120 and140 mm. When the synthetic jet actuator is switched on, thejet amplifies the turbulence which leads to resistance of sep-

aration[16,22]. The turbulence profiles inFig. 9 are usefulto help understand how the synthetic jet works to preventflow separation.

At Xj = 40 mm, although the mean velocity profile isnot yet modified, significant turbulence has been generatedby the jet aty/δ = 0.04–0.5 (equivalenty = 0.2–2.5 mm),whereδ is the thickness of the boundary layer. The peak tur-bulence level and vertical extent of disturbed flow increasesup to Xj = 120 mm. As the flow proceeded further down-stream, the turbulence decreases. It is interesting to noticethat atXj = 160 mm, the mean velocity profile with the jeton is “fuller” than that jet off, but the corresponding turbu-lence profiles show that the turbulence level with the jet onis less than jet off. This indicates that the synthetic jets mayplay dual roles in both accelerating and decelerating turbu-lence in different applications[16].

5. Conclusions

Previous experiments and simulations have been intro-duced and demonstrated the feasibility of active boundarylayer control in turbulent flows using synthetic jets. Thecurrent work has focused on the effect of synthetic jetswithout cross flow and on boundary layer flows under ad-verse pressure gradients. Significant enhancement of the

174 C. Lee et al. / Sensors and Actuators A 108 (2003) 168–174

jet effectiveness is observed by forcing the boundary layerflow at the natural instability frequency, which is lower thanthe frequency at which the actuator produces the maximumvelocity without cross flow. The piezoelectric synthetic jetactuators must have sufficient velocity output to producestrong longitudinal vortices if they are to be effective for flowcontrol. The effectiveness of synthetic jets seems to dependmore strongly on the forcing frequency than on the forcingvoltage. The variation of the control effectiveness of SJAwith frequency does not correlate exactly with output of thesynthetic jet actuator operating in a condition without crossflow and this is most probably due to an interaction with thenatural instability frequency of the flow. The results obtainedhave shown that the synthetic jet generator is an effectiveand promising device for controlling separation and turbu-lence levels in an adverse pressure gradient boundary layer.

References

[1] M. Gad-el-Hak (Ed.), The MEMS Handbook, CRC Press, BocaRaton, 2001.

[2] C.M. Ho, Y.C. Tai, Review: MEMS and its applications for flowcontrol, J. Fluids Eng. 118 (1996) 437–447.

[3] D.R. Smith, M. Amitay, V. Kibens, D.E. Oarekh, A. Glezer, Modifi-cation of Lifting Body Aerodynamics Using Synthetic Jet Actuators,AIAA 98-0209, 1998.

[4] F.W. Roos, Synthetic Jet Microblowing for Forebody Flow-Asymmetry Management, AIAA 98-0212, 1998.

[5] L.G. Pack, A. Seifert, Periodic Excitation for Jet Vectoring andEnhanced Spreading, AIAA 99-0672, 1999.

[6] B.D. Richie, J.M. Seitzman, Acetone Mixing Control of Fuel JetsUsing Synthetic Technology: Scalar Field Measurements, AIAA99-0448, 1999.

[7] A. Crook, A.M. Sadri, N.J. Wood, The Development and Implemen-tation of Synthetic Jets for the Control of Separated Flow, AIAA99-3176, 1999.

[8] F.-J. Chen, C. Yao, G.B. Beeler, R.G. Bryant, R.L. Fox, Developmentof Synthetic Jet Actuators for Active Flow Control at NASA Langley,AIAA 2000-2405, 2000.

[9] L.D. Kral, Active Flow Control Technology, ASME Fluids Engineer-ing Technical Brief, 2000.

[10] S.G. Mallinson, J.A. Reizes, G. Hong, An experimental and numericalstudy of synthetic jet flow, Aeronaut. J. 105 (2001) 41–49.

[11] L. Jenkins, S. Althoff Gorton, S. Anders, Flow Control DeviceEvaluation for an Internal Flow with an Adverse Pressure Gradient,AIAA 2002-0266, 2002.

[12] A. Seifert, A. Darai, I. Wygnanski, Delay of airfoil stall by periodicexcitation, J. Aircraft 33 (1996) 691–698.

[13] R. Mittal, P. Rampunggoon, H.S. Udaykumar, Interaction of a Syn-thetic Jet with a Flat Plate Boundary Layer, AIAA Paper 2001-2773,2001.

[14] G. Hong, S.G. Mallinson, C. Lee, Q.P. Ha, On centerline distributionsof the mean velocity in synthetic jets, in: Proceedings of the 14thAustralian Fluid Mechanical Conference, Adelaide, Australia, 2001,pp. 737–740.

[15] C. Lee, G. Hong, Q.P. Ha, S.G. Mallinson, Synthetic jet generator—amicro actuator for air flow control, in: Proceedings of the Pacific RimWorkshop on Transducers and Micro/Nano Technologies, Xiamen,China, 2002, pp. 339–342.

[16] G. Hong, C. Lee, Q.P. Ha, A.N.F. Mack, S.G. Mallinson, Effec-tiveness of Synthetic Jets Enhanced by Instability of Tollmien–Schlichting Waves, AIAA Paper 2002-2832, 2002.

[17] J. Kiddy, P. Chen, J. Niemczuk, Active Flow Control Using Mi-croelectromechanical Systems, 41st AIAA/ASME/ASCE/AHS/ASCStructures, Structural Dynamics and Materials Conference and Ex-hibit, AIAA-2000-1561, Atlanta, USA, 2000.

[18] D.R. Smith, A. Glezer, The formulation and evolution of syntheticjets, Phys. Fluids 10 (1998) 2281–2297.

[19] X. Zhang, Co- and contrarotating streamwise vortices in a turbulentboundary layer, J. Aircraft 33 (1995) 1095–1101.

[20] I. Wygnanski, Boundary Layer and Flow Control by Periodic Addi-tion of Momentum, AIAA Paper 97-2117, 1997.

[21] S.M. Kumar, W.C. Reynolds, T.W. Kenny, MEMS based actuators forboundary layer control, microelectromechanical systems (MEMS),DSC 66 (1998) 103–109.

[22] H. Schlichting, K. Gersten, Boundary Layer Theory, Eighth Revisedand Enlarged Edition, Springer, Berlin, 2000.

[23] Q. Gallas, J. Mathew, A. Kaysap, R. Holman, T. Nishida, B.Carroll, M. Sheplak, L. Cattafesta, Lumped Element Modelingof Piezoelectric-Driven Synthetic Jet Actuators, AIAA-2002-0125,2002.

[24] S.G. Mallinson, C.Y. Kwok, J.A. Reizes, Numerical simulation ofmicro-fabricated zero mass-flux jet actuators, Sens. Actuators, A 105(2003) 229–236.

Biographies

Chester Lee received his BE degree in mechanical engineering fromSoutheast University, China. He is currently a PhD student in the Facultyof Engineering at the University of Technology, Sydney. His researchtopic is “The interaction between synthetic jets and boundary layer flowsunder an adverse pressure gradient”.

Guang Hong received her Master of Engineering degree from HuazhongUniversity of Science and Technology (HUST), China, and her PhDfrom Cambridge Univeristy, UK. She is currently an associate professorat the University of Technology, Sydney. Her research interests includeactive flow control and IC engine emission reduction.

Q.P. Ha received his BE degree in electrical engineering from Ho ChiMinh City University of Technology, Vietnam, the PhD degree in en-gineering science from Moscow Power Engineering Institute, Russia,and the PhD degree in electrical engineering from the University ofTasmania, Australia. He is currently a senior lecturer at the Universityof Technology, Sydney, Australia. His research interests include controland automation, robotics, and artificial intelligence applications.

S.G. Mallinson received his Bachelor of Science from the AustralianNational University in 1991, and his PhD in aerospace engineeringfrom the University of New South Wales in 1995. He is currently em-ployed as a MEMS design engineer at Silverbrook Research. His re-search interests include hypersonic aerothermodynamics, synthetic jets andmicrofluidics.