American Institute of Aeronautics and Astronautics1 Effect of Longitudinal Ridges on the AerodynamicCharacteristics of an Airfoil

7/29/2019 American Institute of Aeronautics and Astronautics1 Effect of Longitudinal Ridges on the AerodynamicCharacteristic

1/10

American Institute of Aeronautics and Astronautics

1

Effect of Longitudinal Ridges on the Aerodynamic

Characteristics of an Airfoil

Tufan Kumar Guha1, Erik Fernandez

2and Rajan Kumar

3

Florida Center for Advanced Aero-Propulsion

FAMU-FSU College of Engineering, Florida State University, Tallahassee, FL - 32310

A pair of longitudinal ridges is employed to enhance the aerodynamic efficiency of an

USA-35B airfoil used in a remotely controlled aircraft wing. Measurements include surface

static pressure distributions, 2-dimensional and stereo particle image velocimetry. In the

present study, the height and inter-ridge spacing were varied to find optimal control

configuration for maximum effectiveness. The results show that the baseline airfoil exhibits a

massive flow separation on the suction side at an angle of incidence of 12 and beyond. With

the addition of optimal longitudinal ridges, the stall angle was significantly delayed and the

flow was completely attached up to an angle of incidence of 16 . These results clearly show

the benefit of using longitudinal ridges as an effective flow control technique for improving

aerodynamic characteristics of an aircraft wing.

NomenclatureRec = Reynolds number based on airfoil chord length

Cp = Pressure coefficient (= (ps - p)/q)

PS = Local static pressure on the airfoil surface

P = Freestream static pressure

P0 = Freestream total pressure

q = Freestream dynamic pressure

D = Diameter of the longitudinal ridges

S = Inter-ridge spacing

U = Ensemble-averaged (mean) steamwise velocity

Uin = Freestream velocity (=20m/s)

I. IntroductionVER the years, many different methods for enhancing the aerodynamic performance of a wing have been

proposed, verified by wind-tunnel and flight tests and eventually incorporated in the flight vehicle. An airfoil

section, an essential part of a wing, has its primary task as a lift generator, therefore, an optimized design of the

airfoil is the prerequisite to the satisfactory performance of the lifting surface. It has always been a challenge of

optimizing and enhancing its lift without seriously increasing the drag. Also, it is now well known that the flow

separation over an airfoil at high angle of attacks severely degrades its performance, particularly reduction in lift and

increase in drag. To counteract this problem a wide variety of active1-5

and passive6-8

flow control devices have been

developed and researched upon and undoubtedly they help to mitigate the problem to a certain extent. But a lot of

these techniques can be termed as empirical, since there is a lack of complete understanding of the underlying

physics of interaction between the device and the flow. One of the passive control techniques that has been studied

in depth is the use of ribblets for skin friction drag reduction. Riblets9-10

, which are micro-grooves on the surface and

aligned to the freestream direction have shown a viscous drag reduction in the range of 48% on a variety of two-dimensional flows with zero or mild pressure gradients at subsonic speeds. However, the performance of these

devices in adverse pressure gradients and at supersonic speeds is questionable. An experimental study conducted by

Zverkov et al.14

have shown that a wavy wing surface affect the location of the transition point in the boundary layer

and reduces the size of separation bubble. Motivated by the techniques of drag reduction observed in aquatic

animals, Fish et al.11

studied the use of leading edge tubercles (typically found in humpback whale flipper) for lift

1Graduate Research Assistant, Department of Mechanical Engineering, Student Member AIAA.

2Graduate Research Assistant, Department of Mechanical Engineering, Student Member, AIAA

3Assistant Professor, Department of Mechanical Engineering, Senior Member AIAA.

O


2/10


2

enhancement and drag reduction. The leading edge tubercles may function to generate streamwise vortices by

excitation of flow to maintain lift and prevent stall at high angles of attack. Another approach to increase the lift at

moderate angles of attack has been to somehow trap the vortex on the suction side of the wing. Rossow12

at NASA

Ames through his 2D inviscid, incompressible calculations proposed the use of two spanwise wing fences to trap the

vortex.

The aim of our research is to focus on the understanding of the mechanisms associated with the stall

characteristics, particularly flow separation at high angles of incidence and develop a flow control technique to

enhance the aerodynamic performance of an airfoil over a wide range of Reynolds number. In the present study we

have used USA-35B airfoil (Fig. 2) and characterized its performance over a range of test conditions. We have

introduced a flow control method by attaching longitudinal (chordwise) thin ridges, near the mid-section of the

airfoil. We anticipate that the placement of chordwise ridges along the airfoil section will have multiple flow control

effects in terms of restricting the spanwise vortex movement, generation of streamwise vorticity and energizing the

flowfield due to entrainment near the surface. Measurements include surface static pressure distributions and particle

image velocimetry. The initial results of this control technique are very encouraging in terms of increase in

maximum lift coefficient, reduction in size or even elimination of separation bubble and a delay in stall angle.

II. Test Facility and MeasurementsA. Low Speed Wind Tunnel

Measurements were made in the low speed wind tunnel at the Florida Center for Advanced Aero propulsions

(FCAAP) at the Florida State University. The facility is a closed loop wind tunnel with continuous operation over

extended period. The test section is manufactured with acrylic walls and has dimensions of24 x 24 x 60, which

allows for flow visualization from all the four sides. It is driven by a 240 HP fan which is controlled with a variable

frequency drive (VFD; Toshiba H9425KAA) and a precision regulator. The precision regulator is used to

manipulate the fan pitch angle while maintaining a particular fan RPM with the help of the VFD. The flow travels

through a heat exchanger and a flow straightener section which allows for constant flow temperature and minimum

flow fluctuations. Accurate velocities can be obtained between a range of 2 90 m/s with a test section freestream

turbulence intensity of 0.5% at 20m/sec. The present tests were carried out at a freestream velocity of 20m/sec and

corresponding Reynolds number of 5.1 x 105

based on a chord length of 14.45 inches. Wind tunnel velocity is

measured by a pitot-static probe mounted 0.75 m upstream of the test section and is monitored using a 0 - 0.1 in H2O

Omega differential pressure transducer with 0.5% full scale accuracy.

B. Test ModelUSA-35b AirfoilThe airfoil studied is a section of a USA-35b wing used in a RC plane. The span of the airfoil section is 24, so

that it is flushed with the walls of the test section and its chord length is 14.45 (Fig. 2). The airfoil is mounted at a

distance of 8 downstream of the beginning of the test section and 12 from the floor of the wind tunnel , as shown in

in Fig. 1.

a) Subsonic windtunnel facility b) Mounting of Airfoil

Fig. 1. Experimental Setup


3/10


3

To measure the surface static pressure distributions, 31 pressure ports of 0.02 diameter are drilled at the mid -

section of the airfoil. The airfoil can be rotated with respect to the freestream and fixed at a chosen angle of attack in

the range of -4 to 20. The initial tests were carried out on the baseline airfoil and later two ridges on the two sides

of pressure generator were introduced to study the effect of control (Fig. 2a). The separation distance (S) between

these two ridges was varied between 1 8 (Fig. 2b). The ridges covered the airfoil surface up to 90 % of chord

length on both the suction and the pressure side. This was done so that the airfoil maintains the characteristic of a

sharp trailing edge. The longitudinal ridges were made of two flexible circular tubes and were mounted on the airfoil

surface along its contour (Fig. 1). The diameter of the circular tubes was also varied from 0.05 0.2 in theseexperiments (Fig. 2b). These tubes were covered with a smooth tape to achieve a nearly Gaussian surface profile and

to avoid non-uniformity near the surface. The non dimensionalized height (D/C) of the ridges with respect to the

chord length is 0.0035 to 0.014. The incoming boundary layer thickness () at quarter-chord was estimated to be

approximately 0.15-inch and the corresponding non-dimensionalized /C is equal to 0.01. Table 1 shows the set of

experimental test conditions.

Parameter Symbol Parametric Values

Angle of Attack 0, 12, 14, 16

Diameter of RidgeD 0, 0.05, 0.1, 0.2

D/C 0, 0.0035, 0.007, 0.014

Inter-ridge Spacing

S 0, 1, 2, 4, 6, 8

S/C0.07, 0.14, 0.28, 0.42, 0.56

D

x/C = 0.9C = 14.45

S

b = 24

Fig. 2. Airfoil with longitudinal ridges

a) Experimental wing section b) Schematics of test model

Table 1. Experimental test conditions


4/10


4

a)2D PIV measurement plane

b) Stereo PIV measurement plane

Fig. 3. PIV measurement locations on the airfoil section

C. MeasurementsMeasurements include surface static pressure distributions and particle image velocimetry at the mid chord of the

airfoil. The Reynolds number based on chord length at a freestream velocity of 20 m/s is 5.1 x 105. The angle of

incidence was varied from 0 - 20. The static pressures were measured with a Omega 0-0.3 psid pressure

transducer (Model PX138) with a full scale error of 1%. The pressure data was sampled (1000 samples) at each port

at 100 Hz for 10 secs to ensure flow stabilization and minimize statistical errors.

Particle Image Velocimetry (PIV) was used to obtain quantitative measurements of the flow field of interest.

Two dimensional planar PIV and three dimensional stereoscopic PIV were employed in this study. For all planar 2D

PIV cases presented, measurements were carried out at the airfoil mid-span (x-y plane) and the flow was examined

over the aft 60% of the airfoil to capture the

separation effects. Stereo PIV cross planes (y-

z plane) at selected chordwise locations

covered a center span of approximately 0.52

z/C. Figure 3 shows the PIV measurement

locations and extent of laser sheet on the

airfoil section. The flow was illuminated by a

pulsed Quantel

Nd:YAG laser triggered at a

specified time interval. The beam is focused

using a single spherical lens of appropriate

focal length and the laser sheet is created when

the beam passes through two cylindrical lenses(two lenses were used in order to obtain a

sheet wide enough to cover the area of

interest).

The air flow is seeded using a ROSCO

fog machine and introduced into the wind

tunnel upstream of the flow straighteners. The

seed particles are approximately 2-5m in

size. The time interval between laser pulses

was approximately 30s for the presented

cases. Image pairs were acquired at 15Hz with

a resolution of 2560 x 2160 pixels yielding a

spatial resolution of approximately 12.4pixels/mm. Each PIV test case shown here consists of 500 ensemble averaged instantaneous image pairs in order to

estimate the mean statistics of interest. Images were acquired and processed using LaVision Davis 8 software with a

5.5 megapixel sCMOS camera equipped with a 55mm focal length lens. Velocity correlations were made using a

final adaptive interrogation window size of 48 x 48 pixels with a decreasing size multipass algorithm. All passes

used a 50% window overlap. The measurement uncertainty is estimated to be about 1% in ensemble-averaged

velocity measurements with a 95% confidence level.

III. Results and DiscussionAs briefly mentioned in the introduction, the main objective of the present study is to characterize the airfoil near

stall conditions particularly measure the effect of control on stall angle, the location and the reduction in the extent

of flow separation. Here we present results in the form of surface static pressure distributions, 2D velocity field and

a few stereo PIV measurements at selected conditions.

FLOW

x/C z/b

A 0.5 0.4

B 0.5 0.7

A

B

x

zy


5/10


5

A. Surface Static Pressure DistributionsSurface static pressure distributions are presented in the form of coefficient of pressure Cp= (p - p)/q, (Where

p

is the free-stream static pressure and qis the free-stream dynamic pressure) as a function of non-dimensionalized

chordwise distance, x/C. Figures 4-5 show the effect of longitudinal ridges at 0, 12 and 14 angles of attack,

respectively. As may be seen from these pressure distributions that for the baseline airfoil, the flow is attached at 0

angle of incidence (Fig. 4) and remain attached up

to angle of 11

(not shown here). The flow isseparated at 12 (Fig. 5) with a large separation

bubble as indicated by a plateau region on the

suction side of the airfoil. At higher angles the

size of separation bubble is further increased and

larger area of airfoil is subjected to reverse flow.

With the addition of two longitudinal ridges (D/C

= 0.008, S/C = 0.14) on the surface of airfoil, it is

clearly observed that there is an increase in the

value of suction peak at all angles of incidence

and the effects are more pronounced at higher

angles of incidence, particularly near the stall

angles. The plateau region corresponding to

separation bubble is modified to attached flow

profile. From these pressure distributions it

appears that with the addition of longitudinal

ridges the flow is attached up to 14 angle of

attack. As expected the flow on the pressure side

of the airfoil remain unaffected with ridges at these angles of incidence.

B. Velocity FieldParticle image velocimetry (PIV) was carried out at a few selected conditions to obtain 2D velocity field along

the streamwise direction perpendicular to the airfoil surface. The PIV window was selected from x/C of 0.45 to 1.03

to capture the reverse flow in the separation bubble.

Fig. 5. Effect of ridges (D/C = 0.008 & S/C = 0.14) on surface static pressure distributions

Fig. 4. Surface static pressure distributions at = 0


6/10


6

a)Baseline airfoil, = 0 b) Baseline airfoil, = 12

c)Baseline airfoil, = 14 d) Baseline airfoil, = 16 Fig. 6. Effect of angle of incidence on streamwise velocity distribution

a)Ridge height D/C = 0.007 b) Ridge height D/C = 0.014Fig. 7. Effect of ridge height on streamwise velocity distribution, S/C = 0.14, = 14


7/10


7

a)Ridge Spacing S/C = 0.07 b) Ridge Spacing S/C = 0.014

c)Ridge Spacing S/C = 0.028 d) Ridge Spacing S/C = 0.042

e) Ridge Spacing S/C = 0.056

Fig. 8. Effect of inter-ridge spacing on streamwise velocity distribution, S/C = 0.14, = 14


8/10


8

a) Baseline airfoil b) With longitudinal ridgesFig. 10. Effect of longitudinal ridges (D/C = 0.014 & S/C = 0.42) on streamwise vorticity

i. Baseline FlowfieldFigure 6 shows the effect of angle of attack on the streamwise velocity contours for the baseline airfoil. The

arrowheads in the streamlines indicate the direction of flow.As expected, the flow is completely attached at = 0(Fig. 6a) and the surface streamlines follow the contour of the airfoil. An increase in angle of incidence to = 12

(Fig. 6b), the flow is separated on the airfoil suction side approximately at x/C = 0.65 and the separation bubbleshows an open separation. With further increase in angle of incidence to 14 and then 16 (Figs 6c and 6d), the size

of separation bubble is increased and the separation location is moved upstream leading to massive stall (indicated

by reverse flow) characteristics. Due to the presence of separation bubble, the streamlines in the freestream flow

have been pushed away from the surface.

ii. Airfoil with Longitudinal RidgesTwo chordwise longitudinal ridges were placed on the airfoil near the mid-section on either side of the airfoil

centerline. As mentioned in the experimental section the height of ridges and their inter-spacing was varied in this

study. Figure 7 shows the effect of the ridge height in controlling the separation on the airfoil for a fixed inter-ridge

a) Baseline airfoil b) With longitudinal ridges

Fig. 9. Effect of longitudinal-ridgs on streamwise velocity distribution, D/C = 0.014,

S/C = 0.14, = 16


9/10


9

spacing of S/C = 0.14 at an angle of incidence of 14. The dotted line indicates the airfoil surface and the solid line

represent the new surface due to chordwise ridges. The results clearly show that in comparison to baseline flow at

= 14(Fig. 6c), even a small thickness ridges (D/C = 0.007) reduces the size of separation bubble and pushes the

separation location downstream. With an increase in ridge height to D/C = 0.014, the flow on the entire surface is

completely attached. In respect to the boundary layer thickness estimated at quarter-chord ( = 0.15), the small

thickness ridges are well within the boundary layer whereas a ridge height of D/C =0.014 is slightly outside.

The effect of inter-ridge spacing on the streamwise velocity is shown in Fig. 8. As observed, when the ridges aretoo close to each other (Fig. 8a), the effectiveness in terms of separation reduction is minimal, if any. As the spacing

between the two ridges was increased to S/C = 0.14 (S = 2) and 0.28 (Figs. 8b and 8c), its effectiveness improved

and the separation was eliminated. With further increase in spacing to S/C = 0.42 (Fig. 8d), the results were even

better in terms of completely attached flow, higher velocities near the airfoil surface and reduction in boundary layer

thickness. However, when the spacing was further increased to S/C = 0.56 (Fig. 8e), a small separation bubble was

observed near the trailing edge and the flow velocity near the surface was reduced. These results clearly indicate that

there are optimal values of ridge height and inter-ridge spacing for which the effectiveness is the best. The optimal

case of D/C = 0.014 and S/C = 0.42 was further tested at higher angles of incidence of 16 (Fig. 9) and the results

show attached flow features. These results are consistent with those observed in pressure distributions, confirming

the effectiveness of longitudinal ridges in enhancing the performance of airfoil near stall angles.

iii. Stereo Particle Image VelocimetryIn order to better understand the mechanisms responsible for the effectiveness of longitudinal ridges in

eliminating separation near stall angles, the flowfield was further investigated using stereo particle image

velocimetry at a few selected spanwise planes (Fig. 4). Figure 10 shows the contour plots of streamwise vorticity for

the baseline flowfield and with longitudinal ridges (D/C = 0.014 and S/C = 0.42) at = 14. It may be observed that

the longitudinal ridges significantly alter the flowfield close to the surface through the production of streamwise

vorticity and energizing the boundary layer. These results will be further analyzed to understand the flow physics

associated with this control scheme.

IV. Concluding RemarksAn experimental study measuring the effect of longitudinal ridges on the aerodynamic characteristics of an

airfoil is presented. Surface static pressure distributions and particle image velocimetry are used to measure the flow

control effectiveness. The results show that two chordwise ridges of optimal height and inter-ridge spacing can

eliminate the separation on the suction side of this airfoil. This will result in a significant lift enhancement and drag

reduction at these moderate angles of incidence and a delay in its stall angle by about 4 - 5. These results alsoindicate that there may be multiple flow control mechanisms responsible for the effectiveness of this flow control

technique.

References1Gad-el Hak, M., Flow Control, Passive, Active, and Reactive Flow Management, Cambridge University Press, 2000

2Wood, N. and Robert, L., Control of Vortical Lift on Delta Wings by Tangential Leading-Edge Blowing, Journal of

Aircraft, vol. 25, pp. 236-243, 19883McManus, K., and Magill, J., Airfoil Performance Enhancement Using Pulsed Jet Separation Control, AIAA Paper 97-1971, 1997.4Greenblatt, D., and Wygnanski, I. J., The control of flow separation by periodic excitation, Progress in Aerospace

Sciences, 2000; 36(7):487545.5Ron, B., and Faokhi, S., On the Aerodynamics and Performance of Active Vortex Generators, AIAA-93-3447-CP, 376-386, 1993.6

Bushnell, D. B., Turbulent drag reduction for external flows. Aircraft drag prediction and reduction, AGARD-R-72 3,1985, Paper No.5.

7Lin, J. C., Review of research on low-profile vortex generators to control boundary layer separation,Progress inAerospace Sciences 2002, 38, 3894208Mekhanika, Z., Flow on a Swept Wing in the Region of a Fence, Journal of Fluid Dynamics, Vol. 3, No. 6, pp-84-86, July

19689Walsh, M. J. , Riblets,Progress in Astronautics and Aeronautics, 1990. 123:2036110Viswanath, P. R., Aircraft viscous drag reduction using riblets, Progress in Aerospace Sciences, 38 (2002) 57160011Fish, F.E. and Lauder, G.V., Passive and Active Flow Control by Swimming Fishes and Mammals , Annual Review in

Fluid Mechanics, 2006. 38:19322412Rossow, V. J., Aerodynamics of airfoils with vortex trapped by two spanwise fences, Journal of Aircraft, Vol. 31, No. 1,


10/10


10

994, pp. 146-15313Yoon, H.S., Hung, P.A., Jung, J.H., Kim, M.C., Effect of wavy leading edge on hydrodynamic characteristics for flowaround low aspect ratio wing, Computers & Fluids 49 (2011) 276-289

14Zberkov Ilya, Zanin Boris, and Kozlov Viktor , Disturbances Growth in Boundary layers on Classical and Wavy SurfaceWings, AIAA Journal, Vol. 46, No. 12, December 2008

American Institute of Aeronautics and Astronautics1 Effect of Longitudinal Ridges on the AerodynamicCharacteristics of an Airfoil

Documents