AIAA JOURNAL

Vol. 51, No. 7, July 2013

Suppression of Tonal Trailing-Edge Noise From an Airfoil

Using a Plasma Actuator
Ayumu Inasawa, Chiho Ninomiya, and Masahito Asai
Tokyo Metropolitan University, Tokyo 191, Japan

Downloaded by VIKRAM SARABHAI SPACE CENTRE on October 4, 2013 | http://arc.aiaa.org | DOI: 10.2514/1.J052203

DOI: 10.2514/1.J052203
Suppression control of tonal noise generation at an airfoil trailing edge was conducted using a plasma actuator for a
NACA0012 airfoil at a 2-deg angle of attack and at a Reynolds number 2.2 105 where the acoustic feedback
responsible for the trailing-edge noise generation occurs in the pressure-side boundary layer. To minimize the
possible interference of electrode installation in the boundary-layer stability a specially designed actuator with flushmounted electrode configuration was employed. It was found that when the plasma actuator was installed at 5560%
chord location on the pressure surface a surface flow induced by the actuator stabilizes the downstream boundarylayer most significantly to suppress the strong growth of instability waves, which is responsible for the occurrence of
the acoustic feedback. Consequently, suppression of the tonal trailing-edge noise was successfully achieved.

During the last decade most of the flow-control efforts using PAs
focused on the suppression of boundary-layer separation on the
airfoil [2,3] and the control of vortex shedding in the bluff-body wake
[4,5]. Post and Corke [2] used a SDBD plasma actuator to suppress
the leading-edge stall of an airfoil at high angles of attack, and they
showed that the actuator-induced steady flow very close to the wall
suppressed the flow separation. Huang et al. [3] studied separation
control of low-pressure turbine blades using a PA with steady and
unsteady (time-periodic) operations at low Reynolds numbers, and
they showed that unsteady actuation, which excited vertical
disturbances in succession in the separated shear layer, was more
efficient than steady actuation. Thomas et al. [4] applied both steady
and unsteady operations of a PA to control the vortex shedding in the
circular cylinder wake and the associated sound radiation, and they
demonstrated that the unsteady PA suppressed the vortex shedding
very effectively, although it generated additional sound due to the
unsteady body force.
The present flow-control experiment using a plasma actuator
focuses on the tonal trailing-edge noise of an airfoil that usually
occurs at low and moderate Reynolds numbers up to 106 . A number
of experimental and numerical studies have been conducted to
understand the mechanism of noise generation since Paterson et al.
[6] who first conducted a detailed experiment on the noise radiation
from a NACA0012 airfoil. Arbey and Bataille [7] experimentally
examined the underlying mechanism of discrete tones observed by
Paterson et al., and they proposed a feedback-loop model between the
boundary-layer instability (the growth of Tollmien-Schlichting)
waves, assumed to be excited at the maximum velocity point on the
wing surface, and the acoustic waves generated at the trailing-edge
originally suggested by Tam [8] and Fink [9]. McAlpine et al. [10]
performed local stability analyses of the boundary layer on a
NACA0012 airfoil and pointed out that the frequency of the
boundary-layer instability wave with the highest total amplification
(the growth rate integrated along the chordwise direction) was very
close to the tone frequency. The feedback-loop mechanism was also
supported by the results of experiments by Nash et al. [11] and
Makiya et al. [12] and those of direct numerical simulation by
Desquesnes et al. [13]. Here it is noted that in the case of a symmetric
airfoil with zero angle of attack the generation of trailing-edge noise
may be caused not by acoustic feedback but by the flow instability in
the near wake [14]. According to these studies the feedback-loop
mechanism is responsible for the generation of tonal airfoil trailingedge noise at moderate or transitional Reynolds numbers (except the
case of zero angle of attack). Then, we may expect that if the
development of instability waves (i.e., Tollmien-Schlichting waves in
the airfoil boundary layer) is prevented, the acoustic feedback-loop
mechanism leading to the generation of discrete trailing-edge noise
does not work. The most conventional way to suppress the acoustic
feedback generating the tonal trailing-edge noise is to promote the

Nomenclature
c
EPA
f
fT
IPA
Re
s
sPA
SPL
SPLT
U
u
Ue
U
v
x, y, z
xPA

=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=

wing chord, m
applied voltage to the electrodes, V
frequency, Hz
frequency of the trailing-edge noise, Hz
electric current flow through the actuator, A
Reynolds number based on the wing chord
spanwise length of wing, m
spanwise length of the actuator, m
sound-pressure level, dB
sound-pressure level of trailing-edge noise, dB
mean velocity in the x direction, ms
velocity fluctuation in the x direction, ms
mean velocity at the edge of the boundary layer, ms
freestream velocity, ms
velocity fluctuation in the y direction, ms
coordinate system, m
chordwise location of actuator installation, m
angle of attack, deg
kinematic viscosity, m2 s
spanwise vorticity component, s1

I.

Introduction

UR ability to control flow improves with each development in

experimental methodology and devices and with our increased
understanding of flow phenomena. Among the flow-control
techniques developed thus far the plasma actuator (PA) is of
particular interest and has been used in several studies of flow control
[1]. The PA consists of simple components. For example, the
commonly used single dielectric barrier discharge (SDBD) plasma
actuator consists of two electrodes and a dielectric material. This
enables easy installation of an actuator on the target-body surface. In
addition, the SDBD actuator is useful for the active control of
unsteady flow phenomena (which require a fast response by the
actuator).

Received 9 July 2012; revision received 30 November 2012; accepted for

publication 25 January 2013; published online 13 May 2013. Copyright
2013 by the American Institute of Aeronautics and Astronautics, Inc. All
rights reserved. Copies of this paper may be made for personal or internal use,
on condition that the copier pay the $10.00 per-copy fee to the Copyright
Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923; include
the code 1533-385X/13 and $10.00 in correspondence with the CCC.
*Assistant Professor, Department of Aerospace Engineering, 6-6
Asahigaoka, Hino, 191-0065.

Graduate Student, Department of Aerospace Engineering, 6-6

Asahigaoka, Hino, 191-0065.

Professor, Department of Aerospace Engineering, 6-6 Asahigaoka, Hino,

191-0065. Senior Member AIAA.
1695

1696

INASAWA, NINOMIYA, AND ASAI

dielectric material; therefore, the induced flow might become weak

compared with that in the usual configuration. However, the
boundary-layer stability is so sensitive to slight changes in the
velocity profile near the wall [25] that we expected that the boundarylayer instability could be controlled by weak induced velocity.

Downloaded by VIKRAM SARABHAI SPACE CENTRE on October 4, 2013 | http://arc.aiaa.org | DOI: 10.2514/1.J052203

II.

Fig. 1 Experimental setup (dimensions in mm).

boundary-layer transition far upstream of the trailing edge [15,16]. In

the present study we applied a plasma actuator to suppress the
acoustic feedback expecting that the plasma-actuator-induced flow
could stabilize the boundary layer and therefore delay the growth of
Tollmien-Schlichting waves.
Although control of the flow involving boundary-layer separation
has been successfully achieved using a PA as described in the
preceding paragraph, the application of PAs to the control of
boundary-layer stability is limited [17]. Here, one should pay close
attention to the configuration of the exposed electrode in controlling
boundary-layer stability using a PA, especially at high velocities,
because the boundary-layer thickness decreases as the velocity
increases. In the experiments reported to date the thickness of the
exposed electrode was between 0.05 and 0.1 mm, which definitely
protrudes from the surface of the dielectric materials. To induce a
large flow velocity near the wall, as in the application of separation
control, the exposed electrode step is important [1821]. Such
electrode protrusion from the wall surface, however, acts as an
isolated roughness element that may promote boundary-layer
transition. It is recognized that the effect of isolated roughness on the
boundary-layer transition is highly dependent on its height. If the
roughness height is much larger than the critical value, for example,
25 in the Blasius boundary layer in terms of the roughness Reynolds
number (defined by the roughness height and the velocity at the
roughness height) [22,23], a roughness element can lead to vortex
shedding immediately behind the roughness element. On the other
hand, a roughness of small-height generally plays a role in the
receptivity process [24] where any departure from surface
smoothness can excite Tollmien-Schlichting waves in cooperation
with freestream disturbances and acoustic noise. Therefore, when a
PA is applied to the control of boundary-layer instability and
transition, its installation on the surface should be done very
carefully.
In the present experimental study of the suppression control of
trailing-edge noise we proposed a novel flush-mounted exposedelectrode arrangement to minimize the possible effects of actuator
installation, and we examined its ability to stabilize the boundarylayer flow. This was done by embedding the exposed electrode in the

Experimental Setup and Procedure

The experiment was conducted in an open-jet-type wind tunnel

with an exit cross section of 600
height 300 mm (span). The
turbulence intensity at the tunnel exit was less than 0.1% of the
freestream velocity in terms of the root-mean-square value of the
streamwise velocity fluctuation. Two Plexiglas sidewalls maintained the two-dimensionality of the main stream in the test section,
although the upper and lower areas were opened. A NACA0012 wing
model with sharp trailing edge made of aluminum whose chord c
and span s were 150 and 298 mm, respectively, was set between the
side walls (Fig. 1). To measure the trailing-edge noise a microphone
(RION-NL22) was installed 900 mm just above the midspan of
the wing trailing edge. The output signal of the microphone was
acquired by a PC through a 16-bit analog-to-digital conversion.
To ensure the insulation of the model the wing surface was covered
with insulating sheet with a thickness of 160 m. The insulating sheet
modified the airfoil contour only by 1.8% of the maximum thickness
(18 mm) and made the otherwise sharp trailing edge blunt with almost
the same thickness as the formal NACA0012 profile. For easy access
to the pressure-side boundary layer the wing was set in such a way
that the angle of attack was measured toward the pitch-down
direction. The coordinates x and y represent streamwise and vertical
directions, respectively, measured from the spanwise z center of the
wing leading edge. For flow control a plasma actuator of SDBD type
[1] was used. The PA consisted of copper electrodes and polyimid
(Kapton) film. For the typical configuration of the PAs used to date an
exposed electrode is simply put on the dielectric materials forming a
protrusion from the wall. Such a configuration can produce a stronger
volumetric flow near the wall and is suitable for flow control such as
the suppression of separated flow, which requires larger momentum
addition near the wall. On the other hand, the protrusion can influence
the boundary-layer transition as an isolated roughness element even
though its height is so small that vortex shedding does not occur from
the electrode roughness. That is, a roughness of small height can
excite Tollmien-Schlichting waves in cooperation with freestream
disturbances and acoustic noise as mentioned in the Introduction.
Therefore, the plasma actuator used in the present experiment had a
flush-mounted electrode configuration where the electrode was
embedded in the dielectric material (Fig. 2). The chordwise location
of the actuator installation is denoted by xPA, which is the distance
between the leading edge and the chordwise center of the actuator
OPA (see Fig. 2). The actuator spans a lateral distance sPA 
270 mm. To maintain the smoothness of the wing surface the wing
surface (other than the actuator) was covered with a 250-m-thick
polyvinyl choroid film.
The actuator was driven with a continuous sinusoidal signal of
15 kHz that was amplified by an audio power amplifier (YAMAHA
P5000S) and boosted in voltage by a transformer with an input-tooutput voltage ratio of 11000. The input voltage and current were
monitored using a high-voltage probe (Tektronix P6015A) and a

Fig. 2 Schematics of the plasma actuator with the flush-mounted electrode configuration.

1697

INASAWA, NINOMIYA, AND ASAI

35

110

30

100

SPL (dB)

IPA (mA)

Background

90

25
20
15

80
70
60

10

50

40

0
0

0.5

1.5

2.5

3.5

EPA (kV)
Fig. 3 Input voltage EPA vs current IPA of the present actuator
(sPA
270 mm).

current probe (Tektronix TCP300, TCP312), respectively. Figure 3

displays the relationship between the input voltage and electric
current flow in the present actuator configuration showing that the
current increases almost linearly with the input voltage at EPA >
2 kV, beyond which the discharge occurred uniformly in the
spanwise direction. In the present study the flow control was
accomplished mainly with an input voltage of EPA  3 kV
(IPA  20 mA).
A particle-image-velocimetry (PIV) system (Dantec) consisting of
a double-pulsed Nd:Yag laser and a Charge-Coupled Device camera
of 1280 1024 pixels was used to obtain instantaneous velocity and
vorticity fields. The vorticity field was obtained by taking spatial
derivatives of the velocity field with the second-order finitedifference method. The spatial resolution was 0.25 mm in the x
direction and 0.12 mm in the y direction. A laser Doppler velocimetry
(LDV) system (Dantec) was also used to examine the blowing effect
of the actuator in detail.
The velocity of the oncoming uniform flow was fixed at U 
21 ms and the chord Reynolds number was 2.2 105 . The angle of
attack examined was  2 deg. According to the past experiments
of the trailing-edge noise for a NACA0012 airfoil the generation of
the trailing-edge noise is expected to be governed by the instability
of the pressure-side boundary-layer and the associated acoustic
feedback mechanism at the Reynolds number 7 104 Re
8 105 for  2 deg [26].

30
600

800

1000

1200

1400

f (Hz)
Fig. 4 SPL of radiated sound from a single NACA0012 airfoil
(
2 deg, U
21 ms, Re
22 105 ).

upstream of the trailing edge on pressure (upper) surface (Fig. 6a)

indicating that the instability waves (Tollmien-Schlichting waves) of
the selected discrete (or narrow-band) frequency were excited in the
pressure-side boundary layer by the acoustic feedback mechanism.
These vortices can produce a strong pressure fluctuation, such as the
trailing-edge acoustic dipole by diffraction of vortex-induced crossstream fluctuation at the trailing-edge as shown in Fig. 6b. Note that
the side-wall effect was limited within 10% of the whole span, and the
vortex roll-up occurred almost uniformly over 80% of the whole
span. On the suction (lower) surface, on the other hand, the boundary
layer had already undergone a transition to turbulence far upstream of
the trailing edge so that little periodic vortex formation appeared
around the trailing edge. Therefore, we focused on the pressure-side
boundary layer as a target of flow control. Figure 7 illustrates the SPL
spectra when the actuator was installed at three different locations
(xPA c  0.55, 0.65, and 0.80) without operation. The installation of
the actuator did not change the SPL spectrum as seen from the fact
that tone frequency and its magnitude barely change compared to the
case in the absence of actuator (shown in Fig. 4), though by installing
the actuator and polyvinyl choroid film the airfoil contour was varied
from the original one by 4.6% of the wing thickness, and the trailing
edge became 1.5 times thicker than that of NACA0012. Thus, as we
expected, the present actuator with the flush-mounted electrode
configuration little affects the boundary-layer transition.
B. Effect of Actuator Operation on the Trailing-Edge Noise

A. Trailing-Edge Noise That Radiated From the NACA0012 Airfoil

First, the trailing-edge noise from the wing was examined without
actuator installation. Figure 4 illustrates the power spectra of the
sound pressure level (SPL) with and without the wing model in the
test section at  2 deg at U  21 ms (Re  2.2 105 ). Here,
neither the plasma actuator nor polyvinyl choroid film is installed on
the wing surface. We see a distinct discrete tone with the frequency of
fT  998 Hz whose magnitude is more than 40 dB larger than the
background noise level in the presence of the wing. The frequency
variation of the distinct sound is plotted against the uniform flow
velocity U in Fig. 5: all of the tone-frequency components detected
are displayed, and the filled symbols represent the strongest
(dominant) tone frequency in the figure. We see the so-called ladderlike variation where the slope of each rung is U0.85
, and the global
variation of the dominant frequency is proportional to U1.5
. These are
particular features of trailing-edge noise commonly found in past
studies [6,7]. Figures 6a and 6b illustrate instantaneous vorticity and
cross-stream velocity near the trailing edge, respectively, at U 
21 ms and  2 deg. Here, the double-frame PIV images were
captured and synchronized with the signal of the trailing-edge noise
measured by the microphone. One of 150 snapshots is displayed in
the figure. We see the development of periodic vortices immediately

We next examined suppression control of the trailing-edge noise

by using the plasma actuator with the flush-mounted electrode

f (Hz)

0.85

1000

100

10
U

Results and Discussion

III.

Downloaded by VIKRAM SARABHAI SPACE CENTRE on October 4, 2013 | http://arc.aiaa.org | DOI: 10.2514/1.J052203

With wing

(m/s)

1.5

100

Fig. 5 Frequency variation of discrete tone vs freestream velocity at

2 deg. The tonal frequency component with maximum SPL is
represented by filled symbols.

1698

INASAWA, NINOMIYA, AND ASAI

100

x/c=0.8
x/c=0.65

90

x/c=0.55

SPL (dB)

80

70

60

50

30
800

1000

1200

f (Hz)
Fig. 7 Effect of the electrode installation on the radiated sound. Note
that the actuator is not yet operating.

the adverse pressure gradient begins around xc  0.25 and

increases appreciably beyond xc  0.6 on the pressure side at a
2-deg angle of attack for the NACA0012 airfoil. Figure 10 illustrates
the dependency of the electric current on the control effect. The
trailing-edge noise can be suppressed for I PA 12 mA (EPA
2.1 kV) in the present experimental configurations. All of the results
shown in Fig. 9 were obtained for the electric current above this
threshold value. Figures 11a and 11b demonstrate the instantaneous
vorticity and cross-stream velocity fields, respectively, when the
plasma actuator was installed at xPA c  0.55 and operated with
I PA  20 mA. No vortex roll up was found on the pressure (upper)
side, whereas the development of turbulent boundary-layer remained
unchanged on the suction (lower) side (Fig. 11a). In addition, the
diffraction of the cross-flow velocity fluctuation at the trailing edge
became weak (Fig. 11b). Thus, it is understood that the trailing-edge
noise is suppressed by inhibiting the development of instability
waves into strong vortices near the trailing edge.

Fig. 6 Contours of a) cU and b) vU without actuator. The

contour intervals are 12 in a) and 0.04 in b), and the positive values are
shown by broken lines.

configuration. Figure 8 compares the effect of actuator operation at

two different chordwise locations of xPA c  0.8 and 0.55. In both
cases the input voltage of EPA  2 kV (I PA  20 mA) was supplied
to the electrodes. The sound-pressure level of the trailing-edge noise
at f  fT , SPLT , decreased down to the background noise level
when the actuator was operated at xPA c  0.55, whereas no
appreciable reduction in the sound level occurred in the xPA c  0.8
case indicating that the control effect strongly depends on the location
of the actuator operation. Figure 9 illustrates the dependency of the
noise reduction on the actuator location (xPA c) showing that the SPL
of the trailing-edge noise decreased rapidly as the actuator location
moved upstream and could be reduced below the background noise
level when the actuator approached xPA c  0.6. Thus, the
suppression control of the trailing-edge noise can be achieved most
effectively when the actuator is operated at and around xPA c 
0.550.6 at this angle of attack. Here it is important to mention that

C. Mechanism of Tonal Noise Suppression

To clarify how the plasma actuator can suppress the growth of

boundary-layer instability waves leading to the occurrence of the
acoustic feedback loop we examined the near-wall flow developing

100
90

100

PA off
PA on (IPA=20mA)

80

70

70

60

60

50

50

40

40

1000

f (Hz)

PA off
PA on (IPA=20mA)

80

30
800

a)
Fig. 8

90

SPL (dB)

SPL (dB)

Downloaded by VIKRAM SARABHAI SPACE CENTRE on October 4, 2013 | http://arc.aiaa.org | DOI: 10.2514/1.J052203

40

1200

30
800

1000

1200

f (Hz)

b)
SPL vs frequency with and without actuator operation at a) xPA c
0.8 and b) xPA c
0.55 (
2 deg).

1699

INASAWA, NINOMIYA, AND ASAI

110

PA on (IPA=20mA)

100

PA off

SPL (dB)

90
80
70
60
50

Background noise level

40
0.6

0.7

0.8

xPA/c
Fig. 9

Chordwise variation of SPL of the tonal trailing-edge noise.

90
80
70

SPLT (dB)

Fig. 11 Contours of a) cU and b) vU with actuator operation.

The contour intervals are 12 in a) and 0.04 in b), and the positive values
are shown by broken lines.

60
50

Background noise level

40
30
0

10

15

20

25

30

I PA (mA)
Fig. 10 SPL of the trailing-edge noise vs actuator operating current IPA .

downstream of the actuator in detail by means of LDV. Figure 12a

compares the normal-to-wall y distribution of time-averaged
streamwise velocity U at xc  0.6 immediately downstream of the
actuator installed at xPA c  0.55 with and without the plasma
actuator operation. Here, the data were stored over 5 s during which
about 2 104 Doppler signal was detected. We observed a slight
increase in the streamwise velocity U inside the boundary layer
(y 1 mm). The near-wall flow directly induced by the plasma

actuator is shown in terms of the increment velocity (or the difference

between the streamwise velocities with and without the actuator
operation) U in Fig. 12b. The present actuator can induce the nearwall flow (like a wall-jet blowing) with a maximum velocity of about
0.04U (0.84 ms) despite the flush-mounted electrode configuration. Such weak near-wall flow can markedly alter the boundarylayer velocity profiles at the downstream locations up to the trailing
edge as shown in Fig. 13, which compares the velocity profiles at
xc  0.9 with and without the actuator operation. The boundarylayer profile under the actuator operation is not inflectional near the
trailing edge.
Figures 14a and 14b display the waveform of the streamwise
velocity fluctuation and its power spectrum measured at a maximum
shear y-position at xc  0.9, respectively, for the case in which the
plasma actuator was not operated. In estimating the power spectrum
the velocity signal detected with various time intervals was resampled
and converted into a sequence of data with equal time intervals by

1.2

0.05

0.04

0.8

U/U

0.03
U/U

Downloaded by VIKRAM SARABHAI SPACE CENTRE on October 4, 2013 | http://arc.aiaa.org | DOI: 10.2514/1.J052203

30
0.5

0.6

0.02

0.4
PA off

0.2

0.01

PA on (IPA=20mA)
0
0

0.5

y (mm)

1.5

0.5

y (mm)

1.5

b)
a)
Fig. 12 The Y-distribution of mean velocity at xc
0.6 in a) and the increment of the streamwise velocity in b) The solid and broken lines in a) represent
the Falkner-Skan boundary-layer profile.

1700

INASAWA, NINOMIYA, AND ASAI

1.2

0.1
0.05

u/U

-0.05

0.8

-0.1
0

0.6

U/U

0.005

0.01

0.015

0.02

t (s)
a)

0.4

-5

10

PA off

0.2

PA on (IPA=20mA)
1

1.5

2.5

Power spectrum

0.5

y (mm)
Fig. 13 The Y-distribution of mean velocity at xc
0.9. The solid and
broken lines in a) represent the Falkner-Skan boundary-layer profile.
0.1

-6

10

u/U

0.05
0
-0.05

-7

10

10

-0.1

0.005

0.01

0.015

0.02

t (s)
0.0001

10

10

-5

-6

10

-7

10

100

1000

f (Hz)
b)
Fig. 14 a) The streamwise velocity fluctuation and b) its power
spectrum in the absence of actuator operation.

using the sample-and-hold technique to perform the fast-Fouriertransformation. A nearly periodic velocity fluctuation corresponding
to the frequency of the trailing-edge noise (998 Hz) was dominant in
the boundary layer without the actuator operation, which is no doubt
due to the streamwise growth of instability waves (TS waves) excited
by the acoustic feedback. On the other hand, Figs. 15a and 15b
display the corresponding waveform and power spectrum,
respectively, in the flow controlled by the actuator. The evolution
of the broadband spectrum over 3001200 Hz was seen, whereas the
line spectrum (at 998 Hz) observed when there was no actuator
operation disappeared. Thus, the weak blowing by the plasma
actuator can suppress the acoustic feedback mechanism successfully.
To investigate the noise suppression from the instability viewpoint,
we conducted the linear stability analysis for the boundary-layer
profiles on the pressure side. The Falkner-Skan velocity profile was
used to model the boundary layer profiles under nonzero pressure
gradients on the pressure surface. The approximated velocity profiles
are shown by the solid and broken lines in Figs. 12a and 13. Then, the
0.08

PAoff
PAon

0.06

0.04

0.02

Spatial growthrate (mm- 1)

Spatial growthrate (mm- 1)

1000
f (Hz)

0.08

0
200

100

b)
Fig. 15 a) The streamwise velocity fluctuation and b) its power
spectrum with actuator operation xPA c
0.55.

a)

Power spectrum

Downloaded by VIKRAM SARABHAI SPACE CENTRE on October 4, 2013 | http://arc.aiaa.org | DOI: 10.2514/1.J052203

0
0

PAoff
PAon

0.06

fT

0.04

0.02

0
400

600

800 1000 1200 1400 1600

f (Hz)

200

400

600

800 1000 1200 1400

f (Hz)

a)
b)
Fig. 16 Spatial growth rate of linear disturbances predicted by the Orr-Sommerfeld equation (xPA c
0.55). a) xc
0.6 and b) xc
0.9. Chain lines
represent the frequency of the trailing-edge noise.

INASAWA, NINOMIYA, AND ASAI

linear stability of these velocity profiles was analyzed by solving the

Orr-Sommerfeld equation with the Chebyshev collocation method.
The spatial growth rates calculated for the approximated profiles at
xc  0.6 and 0.9 are plotted in Fig. 16 in which the frequency of
trailing-edge noise (998 Hz) is depicted by the chain lines. In the case
of no actuator operation, we see that the unstable frequency range
both at xc  0.6 and xc  0.9 includes the frequency of the
trailing edge. When the actuator was operated, on the other hand, the
growth rate decreased drastically down to about one-third of that
without actuator operation at xc  0.6, and the unstable frequency
range covers the broadband spectra developing up to xc  0.9
shown in Fig. 15b, that is, 3001200 Hz.

Downloaded by VIKRAM SARABHAI SPACE CENTRE on October 4, 2013 | http://arc.aiaa.org | DOI: 10.2514/1.J052203

IV.

Conclusions

Suppression control of tonal noise generation at an airfoil trailing

edge was conducted by using a plasma actuator for a NACA0012
airfoil at an angle of attack of 2 deg, at a chord Reynolds number
Re  2.2 105 . Under these experimental conditions the boundary
layer on the suction surface of the wing had undergone a transition to
turbulence upstream of the trailing edge so that the generation of tonal
trailing-edge noise was governed by the acoustic feedback-loop
mechanism operating on the pressure surface. To minimize the
possible interference of electrode installation itself in the boundarylayer transition we used a specially designed actuator with a flushmounted electrode configuration in the present flow-control
experiments. Our results confirmed that the installation of the
present actuator itself did not affect the generation of trailing-edge
noise and therefore did not affect the boundary-layer stability at all
without blowing by the actuator operation.
The suppression effect of the plasma actuator was dependent on
the chordwise position of actuator operation, and the effective
suppression of the trailing-edge noise was achieved when the plasma
actuator was operated at 5560% chord location from which the
adverse pressure gradient started to increase appreciably on the
pressure side. When the PA was not operated the boundary-layer
profiles started to be inflectional beyond there. In the acoustic
feedback mechanism responsible for the generation of tonal trailingedge noise the excitation and subsequent growth of boundary-layer
instability plays an important role. The present actuation technique
controlled the development of instability waves very successfully by
changing the otherwise inflectional velocity profiles to more stable
ones. This stabilization of the boundary layer was also confirmed by
the linear stability analyses of velocity profiles downstream of the
actuator. In this effective condition the magnitude of the near-wall
blowing velocity induced by the actuator was at most 4% of the
uniform flow velocity U. We stress that such weak addition of nearwall flow velocity could control the development of boundary-layer
instability.
The present noise-suppression control is applicable for other
experimental conditions whenever the noise generation is governed
by the acoustic feedback mechanism. The actuator location at which
the most effective flow control is achieved of course depends on the
pressure distribution on the wing surface and the actuator should be
installed further downstream at higher angles of attack.

Acknowledgments
This work was partly supported by a Grant-in-Aid for Young
Scientists (B) from the Japan Society for Promotion of Science
number 24760664 and a Grant for Scientific Research from Tokyo
Metropolitan Government.

References
[1] Corke, T. C., Enloe, C. L., and Wilkinson, S. P., Dielectric Barrier
Discharge Plasma Actuators for Flow Control, Annual Review of Fluid
Mechanics, Vol. 42, 2010, pp. 505529.
doi:10.1146/annurev-fluid-121108-145550
[2] Post, M. L., and Corke, T. C., Separation Control on High Angle of
Attack Airfoil Using Plasma Actuators, AIAA Journal, Vol. 42, No. 11,
2004, pp. 21772184.
doi:10.2514/1.2929

1701

[3] Huang, J. H., Corke, T. C., and Thomas, F. O., Unsteady Plasma
Actuators for Separation Control of Low-Pressure Turbine Blades,
AIAA Journal, Vol. 44, No. 7, 2006, pp. 14771486.
doi:10.2514/1.19243
[4] Thomas, F. O., Kozlov, A., and Corke, T. C., Plasma Actuators for
Cylinder Flow Control and Noise Reduction, AIAA Journal, Vol. 46,
No. 8, 2008, pp. 19211931.
doi:10.2514/1.27821
[5] Kozlov, A. V., and Thomas, F. O., Plasma Flow Control of Cylinders in
a Tandem Configuration, AIAA Journal, Vol. 49, No. 10, 2011,
pp. 21832193.
doi:10.2514/1.J050976
[6] Paterson, R. W., Vogt, P. G., Fink, M. R., and Munch, C. L., Vortex
Noise of Isolated Airfoils, Journal of Aircraft, Vol. 10, No. 5, 1973,
pp. 296302.
doi:10.2514/3.60229
[7] Arbey, H., and Bataille, J., Noise Generated by Airfoil Profiles Placed
in a Uniform Laminar Flow, Journal of Fluid Mechanics, Vol. 134,
Sept. 1983, pp. 3347.
doi:10.1017/S0022112083003201
[8] Tam, C. K. W., Discrete Tones of Isolated Airfoils, Journal of the
Acoustic Society of America, Vol. 55, No. 6, 1974, pp. 11731177.
doi:10.1121/1.1914682
[9] Fink, M. R., Prediction of Airfoil Tone Frequencies, Journal of
Aircraft, Vol. 12, No. 2, 1975, pp. 118120.
doi:10.2514/3.44421
[10] McAlpine, A., Nash, E. C., and Lowson, M. V., On the Generation of
Discrete Frequency Tones by the Flow Around an Aerofoil, Journal of
Sound and Vibration, Vol. 222, No. 5, 1999, pp. 753779.
doi:10.1006/jsvi.1998.2085
[11] Nash, E. C., Lowson, M. V., and McAlpine, A., Boundary-Layer
Instability Noise on Aerofoils, Journal of Fluid Mechanics, Vol. 382,
March 1999, pp. 2761.
doi:10.1017/S002211209800367X
[12] Makiya, S., Inasawa, A., and Asai, M., Vortex Shedding and Noise
Radiation from a Slat Trailing Edge, AIAA Journal, Vol. 48, No. 2,
2010, pp. 502509.
doi:10.2514/1.45777
[13] Desquesnes, M., Terracol, M., and Sagaut, P., Numerical Investigation
of the Tone Noise Mechanism over Laminar Airfoils, Journal of Fluid
Mechanics, Vol. 591, Nov. 2007, pp. 155182.
doi:10.1017/S0022112007007896
[14] Tam, C. K. W., and Ju, H., Aerofoil Tones at Moderate Reynolds
Number, Journal of Fluid Mechanics, Vol. 690, Jan. 2012, pp. 536
570.
doi:10.1017/jfm.2011.465
[15] Longhouse, R. E., Vortex Shedding Noise of Low Tip Speed Axial
Flow Fans, Journal of Sound and Vibration, Vol. 53, No. 1, 1977,
pp. 2546.
doi:10.1016/0022-460X(77)90092-X
[16] Akishita, S., Tone-Like Noise from an Isolated Two-Dimensional
Airfoil, AIAA Paper 86-1947, 1986.
[17] Grundmann, S., and Tropea, C., Experimental Damping of BoundaryLayer Oscillations Using DBD Plasma Actuators, International
Journal of Heat and Fluid Flow, Vol. 30, No. 3, 2009, pp. 394402.
doi:10.1016/j.ijheatfluidflow.2009.03.004
[18] Enloe, C. L., McLaughlin, T. E., VanDyken, R. D., Kachner, K. D.,
Jumper, E. J., Corke, T. C., Post, M., and Haddad, O., Mechanisms and
Responses of a Single Dielectric Barrier Plasma Actuator: Geometric
Effects, AIAA Journal, Vol. 42, No. 3, 2004, pp. 595604.
doi:10.2514/1.3884
[19] Jukes, T. N., Choi, K. S., Johnson, G. A., and Scott, S. J.,
Characterization of Surface Plasma-Induced Wall Flows Through
Velocity and Temperature Measurements, AIAA Journal, Vol. 44,
No. 4, 2006, pp. 764771.
doi:10.2514/1.17321
[20] Thomas, C. L. E., McLaughlin, E., VanDyken, R. D., Kachner, K. D.,
Jumper, E. J., and Corke, T. C., Mechanisms and Responses of a Single
Dielectric Barrier Plasma Actuator: Plasma Morphology, AIAA
Journal, Vol. 42, No. 3, 2004, pp. 589594.
doi:10.2514/1.2305
[21] Thomas, F. O., Corke, T. C., Iqbal, M., Kozlov, A., and Schatzman, D.,
Optimization of Dielectric Barrier Discharge Plasma Actuators for
Active Aerodynamic Flow Control, AIAA Journal, Vol. 47, No. 9,
2009, pp. 21692178.
doi:10.2514/1.41588
[22] Morkovin, M. V., On Roughness-Induced Transition: Facts, Views,
and Speculations, Instability and Transition, Vol. 1, 1990, pp. 281
295.

1702

INASAWA, NINOMIYA, AND ASAI

Downloaded by VIKRAM SARABHAI SPACE CENTRE on October 4, 2013 | http://arc.aiaa.org | DOI: 10.2514/1.J052203

[23] Reshotko, E., Disturbances in a Laminar Boundary Layer Due to

Distributed Surface Roughness, Turbulence and Chaotic Phenomena
in Fluids: Proceedings of the International Symposium, North-Holland,
Amsterdam, 1983, pp. 3946.
[24] Saric, W., Reed, H. L., and Kerschen, E. J., Boundary Layer
Receptivity to Freestream Disturbances, Annual Review of Fluid
Mechanics, Vol. 34, Jan. 2002, pp. 291319.
doi:10.1146/annurev.fluid.34.082701.161921
[25] Inasawa, A., Asai, M., and Floryan, J. M., Certain Aspect of Instability
of Flow in a Channel with Expansion/Contraction, Proceedings

of the Seventh International Union of Theoretical and Applied

Mechanics Symposium on Laminar-Turbulent Transition, edited by
Schlatter, P., and Henningson, D. S., SpringerVerlag, New York, 2009,
pp. 501504.
[26] Lowson, M. V., Fiddes, S. P., and Nash, E. C., Laminar Boundary Layer
Aero-Acoustic Instabilities, AIAA Paper 94-0358, 1994.

A. Naguib
Associate Editor