AIAA JOURNAL

Vol. 46, No. 9, September 2008

Control of Flow Separation Using Self-Supplying

Air-Jet Vortex Generators

Andrzej Krzysiak∗
Institute of Aviation, 02-256 Warsaw, Poland
DOI: 10.2514/1.30150
Wind-tunnel experimental investigations were conducted to investigate the use of conventional outsupplying and
proposed self-supplying air-jet vortex generators to delay flow separation over an NACA 0012 airfoil. The
conventional air-jet vortex generators were supplied with the air from an external compressor, and proposed self-
supplying generators got the air from the overpressure region situated in the nose part of the airfoil lower surface.
The lift, drag, and pitching moment for the smooth airfoil equipped with air-jet vortex generators were determined
based on the pressure distribution measurements. The experimental tests were carried out in both the low-speed
wind tunnel (M
0:05–0:1) with the airfoil model of 0.5-m chord length and the high-speed wind tunnel
(M
0:1–0:85) with the airfoil model of 0.18-m chord length. In the high-speed wind tunnel, the influence of using
only the self-supplying air-jet vortex generators was tested. It was found that using both types of air vortex generators
can delay the flow separation on the airfoil, which leads to the increase of both the lift coefficient and the critical angle
of attack.

Nomenclature vortex generators could be successfully replaced by the air-jet vortex

a = length of the nozzle outlet sectional generators for boundary-layer control because of the ease of control
b = width of the nozzle outlet sectional accompanied by a minimal drag penalty. However, the complexity of
CD = drag coefficient the installation of AJVGs in comparison with the simplicity of the
CL = lift coefficient vane vortex generators has limited the practical usage of AJVGs.
CL max = maximum-lift coefficient Furthermore, experimental tests performed by several researchers
Cm = moment coefficient [3–7] showed that identification of the optimum air-jet configuration
Cp = pressure coefficient was not simple and needed careful study, because the effectiveness of
Cx = drag coefficient air-jet vortex generators depended on a number of parameters, such
c = airfoil chord as 1) the pitch and skew angles, 2) the jet mass-flow rate, 3) the ratio
M = Mach number of boundary-layer thickness to the jet diameter, 4) the jet Reynolds
m_ = mass-flow rate number, and 5) the ratio of mean jet velocity to mean freestream
Vj = jet velocity at the air-jet vortex generator exit velocity.
VR = ratio of mean jet velocity to mean freestream velocity The most important conclusions resulting from previous works
V1 = freestream velocity were aimed at identifying the optimum jet parameters (see [8,9]);
x = distance along the chord from the leading edge furthermore, the results of experimental works aimed at explaining

= airfoil angle of attack the physical process of vortex formation were presented in [7]. For

cr = critical airfoil angle of attack optimal results, a strong and persistent vortex should be created at the

= pitch angle jet pitch angle
of about 30 deg and a jet skew angle  of about
 = skew angle 60 deg. Also, the peak vorticity (the value at the center of the core)
rises linearly with an increase of the jet nozzle diameter and the
velocity ratio (VR).
I. Introduction The main aim of this paper is, based on the results of previous
HE control of flow is one of the future areas of interest in fluid works, to suggest a novel simpler solution to AJVGs (namely, self-
T mechanics that is investigated extensively in many research
centers. During the last few years, a lot of emphasis has been placed
supplying air-jet vortex generators) and to prove their usefulness
experimentally. In the first part of the paper, the influence of some
on the development of active control methods. One such method is chosen AJVG parameters on the aerodynamic characteristics of a
based on the use of air-jet vortex generators (AJVGs). In this flow NACA 0012 airfoil equipped with conventional air-jet vortex
control method, the interaction between the air jets and the freestream generators is presented. The methodology and the equipment used in
flow changes the structure of the flow, creating well-organized the experimental measurements and sample results are described
vortical structures that may be used to alleviate boundary-layer next. The results of experimental tests of aerodynamic characteristics
separation. of the NACA 0012 airfoil equipped with the proposed self-supplying
The concept of using AJVGs for boundary-layer control on an air-jet vortex generators are presented in detail.
airfoil has been known for a long time. This idea was originally
proposed in the beginning of the 1960s by Wallis [1] and Wallis and
Stuart [2]. It was found that in many cases, the conventional vane
II. Experimental Setup and Instrumentation
Received 12 February 2007; revision received 29 May 2008; accepted for A. Wind Tunnels
publication 29 May 2008. Copyright © 2008 by the American Institute of The experimental tests were performed both in a low-speed wind
Aeronautics and Astronautics, Inc. All rights reserved. Copies of this paper
tunnel T-1 and a trisonic wind tunnel N-3 at the Institute of Aviation
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
in Warsaw, Poland. In the high-speed wind tunnel N-3, the influence
Drive, Danvers, MA 01923; include the code 0001-1452/08 $10.00 in of using only the self-supplying air-jet vortex generators was
correspondence with the CCC. investigated. The low-speed wind tunnel T-1 is a closed-circuit
∗ continuous-flow wind tunnel with a 1.5-m-diam open test section.
Senior Research Scientist, Department of Aerodynamics, Aleja
Krakowska 110/114; [email protected]. The controlled range of freestream velocity is 15–40 m=s.
2229
2230 KRZYSIAK

The N-3 wind tunnel is a blowdown type, with partial recirculation

of the flow, equipped with the closed test section (0:6
0:6 m). It
can operate in subsonic, transonic, and supersonic flow regimes at
Mach numbers M  0:2  1:2, 1.5, and 2.3.

B. Airfoil Models
The model of the NACA 0012 airfoil used in the experimental tests
in the low-speed wind tunnel T-1 had a chord of 0.5 m and a span of
1 m. It was made of composite materials. The model was mounted
vertically in the working section between two stationary end plates
(Fig. 1).
The steel model of the NACA 0012 airfoil used in the investigation
in the high-speed wind tunnel N-3 had a chord of 0.18 m and a span of
0.6 m. The model was mounted horizontally between walls of the
wind-tunnel test section (Fig. 2).
Both models were equipped with removable skin panels fastened
to the model’s upper airfoil surface. On the surface of the each skin
panel, 10 air-jet vortex generator nozzles were glued. In accordance
with the recommendations presented in [6,7], the nozzles were
located at the same distance from the leading edge, x=c  0:12, and
spaced at the intervals of 0:11c along the airfoil span (see Fig. 3). The
nozzles’ outlet had a rectangular geometric shape with rounded Fig. 3 Location of air-jet vortex generator nozzles on the airfoil upper
corners. The conventional air-jet vortex generators were supplied surface.
with the air from two compressors of a total flow rate of
800 liter= min.
distributions. The wake rakes made possible the measurement of the
Both models were equipped with measuring orifices situated along
pressure distributions in the wake of the airfoil models.
the chord (in the middle of theirs span) to measure the pressure

III. Idea of Self-Supplying Air-Jet Vortex Generators

The idea of using conventional air-jet vortex generators for
boundary-layer control on airfoil is usually based on the principle of
supplying air from the turbine engine compressor or from any
auxiliary compressors. The complexity of such installation has
limited the practical usage of AJVGs. In this paper, it is proposed to
use the airfoil overpressure regions as a source of the air for the
AJVGs. The rest of the air-jet vortex generator system (i.e., the shape
and location of the nozzles on the upper airfoil surface) remain
unchanged, as in the conventional one. This idea has been called self-
supplying air-jet vortex generators and is presented in Fig. 4.
The self-supplying air-jet vortex generators consist of a number of
nozzles with outlets situated on the upper surface of the airfoil and the
pneumatic pipes supply the nozzles with air. The inlets of these
pneumatic pipes are situated in the nose part of the lower airfoil
surface. The inlet axes cross the outline of the lower airfoil surface as
close as possible to the flow stagnation points for circumcritical
angles of attack and are positioned parallel to the airstream’s
direction in these points. At the higher angles of attack in the nose
part of the lower airfoil surface, the overpressure region is formed.
Fig. 1 Schematic drawing of the low-speed wind-tunnel test stand.
The difference of the pressure between the lower and upper airfoil
surfaces forces the airflow through the pneumatic pipe connecting
the inlet to the nozzle (in the presented investigation, the pipes had
constant diameter). The location of the inlet axis close to the flow
stagnation points makes the self-supplying air-jet vortex generators
more effective. The air jets flowing out from a number of nozzles
with certain velocity Vj, similar to the conventional air-jet vortex
generators, interact with the freestream flow, forming vortices on the
upper airfoil surface. The self-supplying air-jet vortex generators are
characterized by the fact that they remain inactive at low angles of
attack and only become active at the higher angles of attack, close to

Fig. 2 Model of the airfoil in the high-speed wind tunnel N-3. Fig. 4 Concept of self-supplying air-jet vortex generators.
 KRZYSIAK 2231

Table 1 Test conditions for conventional air-jet vortex generators in the low-speed wind tunnel T-1
Parameter Value Measurement uncertainty
Mach number M 0.05, 0.075, 0.1 0.003
Reynolds number Re 0:6 106 (M  0:05) 1:1 106 (M  0:1) ——
Nozzle dimensions 5:6
1:1 mm 0.05 mm
Airfoil angle of attack
a 2:7 deg j21:7 deg 0.1 deg
Jet pitch angle
20, 30 deg 1 deg
Jet skew angle  30, 60, 75 deg 1 deg
VR 1:2  4:6 0.2
Jet mass-flow rate 104  103 kg=s 1  5 105 kg=s
a
With wind-tunnel corrections.

Table 2 Test conditions for self-supplying air-jet vortex generators in the low and high-speed wind tunnels

Parameter Value Measurement uncertainty

Mach number M 0:05  0:85 0.003
Reynolds number Re 0:6 106 (M  0:05) 2:6 106 (M  0:85)
Nozzles dimensions: low-speed wind tunnel 7:2
1:2 mm 0.05 mm
Nozzles dimensions: high-speed wind tunnel 3
0:6 mm 0.03 mm
Airfoil angle of attack
a 2:7 deg j21:7 deg 0.1 deg
Jet pitch angle
20, 30 deg 1 deg
Jet skew angle  30, 60, 75 deg 1 deg
VRb 1.6 0.1
Jet mass-flow rateb 2:1 104 kg=s 105 kg=s
a
With the wind-tunnel corrections.
b
Numerical calculations at M  0:05 and
 14 deg.

critical values, as a result of the greater pressure difference between through the nozzles was realized by electromagnetic valves. In Fig. 6,
the upper and the lower airfoil surfaces in the nose region. the lift coefficient versus angle of attack for different jet velocities Vj
and Mach numbers M  0:05 (V1  18; 3 m=s), 0.075
(V1  25:8 m=s), and 0.1 (V1  34:4 m=s) are shown.
IV. Experimental Uncertainties From the analysis of these results, it can be concluded that use of
The following are estimates of the experimental uncertainties. The the air-jet vortex generators on the NACA 0012 airfoil causes an
angle of attack of the airfoil model is determined to within 0:1 deg increase of the maximum-lift coefficient and an increase of the value
in low-speed wind tunnel T-1 and 0:02 deg in high-speed wind of critical angle of attack. For the preceding test conditions (i.e.,
tunnel N-3. The pressure is accurate to within 0:01 kPa in the range M  0:05  0:1,   60 deg, and
 30 deg) the highest increase
of Mach numbers M  0:05  0:1 and 0:1 kPa in the range of of these values was achieved for the range of velocity ratio
Mach numbers M  0:2. The experimental uncertainties of VR  2:7  3:9. Among the Mach numbers tested in the low-speed
aerodynamic coefficients are the following: lift 0:01, drag wind tunnel T-1, the highest rise of the lift coefficient
0:002, and pitching moment 0:001. (CL max  0:2), due to use of the air-jet vortex generators, occurred
at M  0:05 (VR  2:7). Simultaneously, the increase of critical
angle of attack at about 
cr  2 deg was achieved.
V. Results and Discussion In Figs. 7 and 8 the influence of skew angle and pitch angle on the
A. Conditions of the Tests lift coefficient at Mach number M  0:075 (V1  25:8 m=s) is
In Tables 1 and 2 the test conditions for the experimental presented. The measurements were performed at the mass-flow rate
investigation of control of flow separation on the airfoil with m_  6
104 kg=s (corresponding to Vj  71 m=s) ,at skew
conventional and self-supplying air-jet vortex generators are given. angles   30, 60, and 75 deg, and at pitch angles
 20 and
30 deg.
Test results suggest that a strong and persistent vortex is created at
B. Sample Test Results of Conventional Air-Jet Vortex Generators
the jet pitch angle
of about 30 deg and the jet skew angle  of about
The tests of the NACA 0012 airfoil with a conventional air-jet 60 deg, which has an effect on the increase of the lift coefficient and
vortex generator were performed in the low-speed wind tunnel T-1 in the increase of the critical angle of attack.
the range of Mach numbers M  0:05  0:1 (which correspond to
V1  18:3  34:4 m=s). The lift CL and the moment Cm
coefficients were calculated by integrating pressure distributions
along the airfoil chord, and the CD coefficient was calculated by
integrating pressure distribution along the wake rake. For pressure
integration, the trapezoidal rule was used. The sample result of the
pressure distribution along the airfoil chord is presented in Fig. 5.
To estimate the influence of the air-jet velocity Vj on the lift
coefficient CL , the measurements of the NACA 0012 airfoil
characteristics for various mass-flow rates through the nozzles were
carried out. During these tests, the mass-flow rate was controlled by a
valve in the range m _  4:6
104 –11:7
104 kg=s for each
nozzle (corresponding to Vj  57–112 m=s). To achieve a mass-
flow rate m_ > 8
104 , it was necessary to stop the flow through two
external nozzles (eight nozzles were used), and for mass-flow rate
m_ > 11
104 , it was necessary to stop the flow through two Fig. 5 Influence of the self-supplying air-jet vortex generators on
additional nozzles (only six nozzles were used). Stopping the flow pressure distribution along the airfoil chord.
2232 KRZYSIAK

Fig. 8 Lift vs angle of attack for different pitch angles at M
0:075

(V1
25:8 m=s).

Fig. 9 Influence of the air-jet vortex generators on drag coefficient vs

angle of attack at M
0:075 (V1
25:8 m=s).

Fig. 6 Lift vs angle of attack for different jet velocities at M
0:05

(V1
18:3 m=s), M
0:075 (V1
25:8 m=s) and M
0:1
(V1
34:4 m=s).

The forming of the vortices on the upper airfoil surface due to the
use of the air-jet vortex generators does not generally change the
value of the drag coefficient in the range of low angles of attack Fig. 10 The influence of the air-jet vortex generators on pitch moment
(Fig. 9). The delay of the flow separation connected with the coefficient vs angle of attack at M
0:075 (V1
25:8 m=s).

Fig. 7 Lift vs angle of attack for different skew angles at M
0:075 Fig. 11 Influence of the self-supplying air-jet vortex generators on lift
(V1
25:8 m=s). coefficient vs angle of attack at M
0:075 (low-speed wind tunnel).
 KRZYSIAK 2233

phenomenon of the increase of the critical angle of attack causes the generators on the lift coefficient vs angle of attack at M  0:075, 0.2,
decrease in the drag coefficient in this region of the critical angles of 0.3, and 0.4 is presented.
attack in comparison with the smooth airfoil. From the analysis of the preceding results, it can be concluded that
In Fig. 10, the influence of air-jet vortex generators on the pitch the use of the proposed self-supplying air-jet vortex generators on the
moment coefficient of the NACA 0012 airfoil at M  0:75 is NACA 0012 airfoil may be as useful as the conventional air-jet
presented. vortex generators and is technically significantly simpler. The self-
Generally, it can be said that in the less-than-critical range of the supplying air-jet vortex generators provide an increase of the
airfoil angles of attack, the vortices formed on the upper airfoil maximum of lift coefficient and a increase of the critical angle of
surface move the pressure center toward airfoil trailing edge. It is a attack at the wide range of Mach numbers (up to M  0:4 at the test
result of the certain pressure decrease in the middle of the upper condition presented in this work). For the higher Mach numbers (i.e.,
airfoil surface. In consequence, it causes a slight diminishing of the M  0:5  0:85), the influence of these vortex generators on the
Cm =
derivative (e.g., from Cm =
 0:06 to 0.025 for the aerodynamic characteristics of the NACA 0012 airfoil deteriorates
NACA 0012 airfoil at M  0:075). significantly. Although the effectiveness of the self-supplying air-jet
vortex generators (measured as a increase of CL max and
cr ) is less
C. Sample Test Results of Self-Supplying Air-Jet Vortex Generators than conventional ones, the advantages of their use are clear.
The tests of the NACA 0012 airfoil with the self-supplying air-jet At the lower Mach numbers (up to M  0:3) and low angles of
vortex generator were performed in both the low-speed wind tunnel attack, the self-supplying air-jet vortex generators cause a minimal
T-1, in the range of Mach numbers M  0:05  0:1 (which increase of drag coefficient (Cd  0:001  0:0015) (Fig. 13). At
correspond to V1  18:3  34:4 m=s), and the high-speed wind the higher Mach numbers, this drag coefficient increase is a little
tunnel N-3, in the range of Mach numbers M  0:2  0:85. In greater (e.g., Cd  0:003 for M  0:7).
Figs. 11 and 12, the influence of the self-supplying air-jet vortex In Fig. 14, the influence of self-supplying air-jet vortex generators
on the pitch moment coefficient of the NACA 0012 airfoil at M 
0:075 is presented. The forming of the vortices on the upper airfoil
surface causes the slight diminishing of the CL =
derivative,
similar to the case of the conventional air-jet vortex generators.

VI. Conclusions
The results of the experimental measurements of aerodynamic
characteristics of the NACA 0012 airfoil with the conventional air-jet
vortex generators and proposed idea of the self-supplying air-jet
vortex generators are presented. These results show that air-jet vortex
generators are able to delay the flow separation, which, in
consequence, leads to the increase of the critical angle of attack and

Fig. 13 Influence of the self-supplying air-jet vortex generators on drag

coefficient vs angle of attack at M
0:3.

Fig. 12 Influence of the self-supplying air-jet vortex generators on lift

coefficient vs angle of attack at M
0:2, 0.3 and 0.4 (high-speed wind Fig. 14 Influence of the self-supplying air-jet vortex generators on
tunnel). pitch moment coefficient vs angle of attack at M
0:075.
2234 KRZYSIAK

lift coefficient. Furthermore, experimental tests performed for the [2] Wallis, R. A., and Stuart, C. M., “On the Control of Shock-Induced
different values of the air-jet vortex generators parameters suggest Boundary Layer Separation with Discrete Air Jets,” Aeronautical
that the highest effectiveness (i.e., increase of lift coefficient and Research Council, Rept. CP 595, London, 1962.
[3] Johnston, J., and Nishi, M., “Vortex Generator Jets—A Means for
increase of critical angle of attack) of air-jet vortex generators should
Passive and Active Control of Boundary Layer Separation,” Aerospace
be achieved at the jet skew angle  of about 60 deg and the jet pitch Sciences Meeting, Reno, NV, AIAA Paper 89-0564, 1989.
angle
of about 30 deg, as per previous references. [4] Selby, G., “Experimental Parametric Study of Jet Vortex Generators for
The wind-tunnel investigation showed that the proposed self- Flow Separation Control,” NASA CR 187836, Dec. 1990.
supplying air-jet vortex generators may be a viable alternative to [5] Compton, D. A., and Johanston, J. P., “Streamwise Vortex Production
conventional ones, because they may also cause the delay of flow by Pitched and Skewed Jets in a Turbulent Boundary Layer ,”
separation. In this concept, the air-jet vortex generators are supplied Aerospace Sciences Meeting, Reno, NV, AIAA Paper 91-0038, 1991.
with air from the overpressure region situated in the nose part of the [6] Pearcey, H. H., “Shock Induced Separation and Its Prevention by
airfoil lower surface. Although the effectiveness of the self- Design and Boundary Layer Control,” Boundary Layer and Flow
supplying air-jet vortex generators is less than the conventional ones, Control, edited by Lachmann, G., Pergamon, Oxford, 1961, pp. 1166–
1134.
the advantages of their use seems to be interesting, especially [7] Pearcey, H. H., Rao, K., and Sykes, D. M., “Inclined Air Jets Used as
because of the simplicity of their manufacture. Vortex Generators to Suppress Shock Induced Boundary Layer
Separation ,” Computational and Experimental Assessment of Jets in
Acknowledgment Cross Flow, AGARD-CP-534, AGARD, Neuilly-sur-Seine, France,
1993, Paper 40.
The experimental tests were done under grant from the Polish [8] Barberopoulos, A. A., and Garry, K. P., “The Effect of Skewing on the
Ministry of Scientific Research and Higher Education “Experimental Vorticity Produced by an Airjet Vortex Generator,” The Aeronautical
and Numerical Investigation of Active Control of Flow by Usage of Journal, Vol. 102, No. 1013, Mar. 1998, pp. 171–177.
Air-Jet Vortex Generators” program (grant 4T12C 008 27). [9] Bray, T. P., and Garry, K. P., “Optimisation of Air-Jet Vortex
Generators with Respect to System Design Parameters,” The
Aeronautical Journal, Vol. 102, No. 1013, Oct. 1999, pp. 475–479.
References
[1] Wallis, R. A., “A Preliminary Note on a Modified Type of Air Jet for
Boundary Layer Control,” Aeronautical Research Council, J. Gore
Rept. CP 513, London, 1960. Associate Editor