American Institute of Aeronautics and Astronautics
1
Wind Tunnel Test of Gurney Flaps and T-Strips on an
NACA 23012 Wing
Michael A. Cavanaugh
1
Virginia Tech, Blacksburg, VA, 24061
Paul Robertson
2
Aeronautical Testing Service, Inc., Arlington, WA, 98223
and
William H. Mason
3
Virginia Tech, Blacksburg, VA, 24061
A wind tunnel test was carried out on an aspect ratio 6 wing equipped with Gurney flaps
and trailing edge T-strips. The test was conducted at the University of Washington
Aeronautical Laboratory’s 8 x 12 foot low-speed wind tunnel at Reynolds numbers of
1.95x10
6
, 1.02x10
6
and 0.51x10
6
. The NACA 23012 test wing was unswept and untwisted
with a 90 inch span and a constant chord of 15 inches. Gurney flap heights of 0.21%, 0.52%,
1.04%, 1.46%, 2.08%, 3.33%, 4.00% & 5.00% chord were tested on the model. T-strip
heights of 0.42%, 1.04%, 1.67%, 2.08%, 2.92%, 4.17% & 5.00% chord were also tested.
Results showed that Gurney flaps produced a positive increment in lift coefficient, a negative
shift in the zero-lift angle of attack, and an increase in the wing maximum lift coefficient. T-
strips produced an increase in the slope of the lift curve and an increase in maximum lift
coefficient, but produced no shift in the wing zero-lift angle of attack. Gurney flaps
produced a negative (nose-down) shift in the pitching moment curve and a rearward shift in
the wing aerodynamic center. T-strips also produced a rearward shift in the wing
aerodynamic center, but produced no increment in the pitching moment coefficient near
zero lift. Both devices produced a drag increment that was non-linear with device height,
larger Gurney flaps and T-strips producing a disproportionately larger drag increment.
Nomenclature
C
L
= coefficient of lift
C
Lmax
= maximum lift coefficient
C
Lo
= lift coefficient at zero angle of attack
C
Lα
= lift curve slope, /deg
C
D
= coefficient of drag
C
Do
= coefficient of drag at zero lift coefficient
C
M
= pitching moment coefficient
C
Mo
= pitching moment coefficient at zero lift coefficient
e = Oswald’s Efficiency Factor
Re = Reynolds Number based on chord
x/c
ac
= aerodynamic center location in percent wing chord
x/c
ref
= moment reference center location in percent wing chord (25% chord for this test)
α
OL
= zero-lift angle of attack, deg
Δ = increment in a given coefficient
1
Graduate Student, Department of Aerospace and Ocean Engineering, Senior Member AIAA, mcavanau@vt.edu
2
President, Aeronautical Testing Service, Inc., 18820 59
th
Drive NE, Arlington, WA 98223, Senior Member AIAA.
3
Professor, Department of Aerospace and Ocean Engineering, Associate Fellow AIAA.
American Institute of Aeronautics and Astronautics
2
I. Introduction
s part of a study into the use of micro trailing edge devices in aircraft preliminary design, a wind tunnel test
was carried out on an aspect ratio 6 wing equipped with Gurney flaps and trailing edge T-strips. A Gurney flap
is simply a flat plate attached perpendicularly to the lower (pressure) surface of an airfoil or wing trailing edge. A
T-strip is attached to both the upper and lower surfaces. Gurney flaps and T-strips are used to modify the lifting
characteristics of the baseline airfoil or wing. The simplicity of these micro trailing edge devices is illustrated below
in Figure 1.
Gurney flaps are used to produce a lift increment (ΔC
L
) and to increase the maximum lift coefficient of two-
dimensional airfoils and three-dimensional wings. Typical applications studied have been on the downforce wings
of racing automobiles
1
, in the cove of multi-element airfoils
2
and on wind turbine blades
3
. In more recent studies,
articulated Gurney flap-like devices have been used to tailor the spanwise loading of wings
4
and rotor blades
5
.
Trailing edge T-strips have been used to improve the performance of aircraft vertical tails. An example is the
use of T-strips on the vertical tail/rudder of the Sino Swearingen SJ30-2 business jet to increase dutch roll damping
6
.
Flight tests showed that a 1% chord T-strip on the lower third of the rudder increased dutch roll damping for the
flaps up, 180 KIAS flight condition from essentially zero to 0.12, more than double the FAR 23 minimum.
A number of references exist reporting the results of wind tunnel tests of two-dimensional airfoils
7,8
and three-
dimensional wings
9,10
equipped with Gurney flaps. Drawing general conclusions on the effects of Gurney flaps is
difficult because each of these tests was conducted on a model with a different airfoil section and different size
Gurney flaps, and each test was run at a different Reynolds number. What is required is a more comprehensive test
for a full range of Gurney flap heights at several different Reynolds numbers. The test described in this paper was
for flap heights ranging from as small as 0.21% chord up to 5.0% chord, with the smaller Gurney flaps (0.21%,
0.52%, 1.04% & 1.46% chord) being tested at Reynolds numbers of 0.51x10
6
, 1.02x10
6
and 1.95x10
6
. The trends
seen in this test were similar to that found in previous Gurney flap wind tunnel tests. The contribution here is a
more complete data set collected for a range of Gurney flap heights for Reynolds numbers from as low as 0.51x10
6
on up to 1.95x10
6
.
Despite being used on a number of aircraft vertical tails
6,11
, very little wind tunnel test data exists in the literature
on the effect of trailing edge T-strips. The only wind tunnel test data found was from Roesch & Vuillet
12
. This data
was for an NACA 5414 wing with a 5% chord T-strip tested at a Reynolds number of 0.75x10
6
. A 5% chord T-strip
would be too large for use on the vertical tail of an aircraft due to its high drag. This paper presents results for T-
strips ranging from 0.42% up to 5.0% chord, with the smaller flaps (0.42%, 1.04% & 1.67% chord) being tested at
Reynolds numbers of 0.51x10
6
, 1.02x10
6
and 1.95x10
6
.
II. Experimental Setup
The test was conducted at the University of Washington Aeronautical Laboratory’s (UWAL) Kirsten Wind
Tunnel
13
located in Seattle, Washington. The Kirsten Wind Tunnel is a closed circuit, double return type tunnel
with an 8 x 12 foot rectangular test section vented to the atmosphere. Model force and moment data is measured by
an external balance located beneath the floor of the test section. For this test, the model wing was mounted atop the
external balance using a yaw fork. The two forward arms of the yaw fork were attached to the model wing at the
quarter chord location. This 25% chord location was also the model moment reference center. A single pitch arm,
Gurney flap
T-strip
Figure 1. Gurney Flap and Trailing Edge T-Strip Detail
American Institute of Aeronautics and Astronautics
3
attached to the rear of the model, was used to adjust angle of attack. Model angle of attack was referenced to the
airfoil chordline. The test wing is shown in Figure 2 mounted atop the external balance.
The baseline model was an unswept, untwisted, aspect ratio 6 wing built by Aeronautical Testing Service, Inc.
Total span was 90 inches with a constant 15 inch chord. The airfoil section was an NACA 23012. The model center
section was machined from aluminum and spanned 86.4 inches. High density polyurethane foam wingtips of 1.8
inch span each were attached to the ends of the wing. The NACA 23012 airfoil shape was machined into the first
inch of each wingtip. The outboard 0.8 inches closed out the wingtip with a half circular shape. The wingtip detail
is shown in Figure 3. The Figure also shows a 5% chord Gurney flap attached to the model wing trailing edge.
Model dimensional tolerances were quoted at ±0.005 inch.
Gurney flap heights of 0.21%, 0.52%, 1.04%, 1.46%, 2.08%, 3.33%, 4.00% & 5.00% chord were tested on the
model wing. Metal angles of standard sizes were mounted to the lower surface trailing edge of the wing to simulate
Gurney flaps. The mounted angles spanned the 86.4 inch center section of the wing. They did not extend to the
wing tips. The percentage of wing span covered by the angles (Gurney flaps) was 96%. The angles were attached
to the underside of the wing using double-sided tape. The thickness of the double-sided tape was measured to be
approximately 0.003 inches. However, the double-sided tape was soft and appeared to lose thickness when
compressed. Aluminum tape was used as an additional means of securing the angles to the wing. The Gurney flap
heights quoted above are the height of the angle alone. The thickness of the wing trailing edge (0.050”) and the
height of the double-sided tape were not included. The aluminum tape did not change the height of the Gurney flap.
Trailing edge T-strip heights of 0.42%, 1.04%, 1.67%, 2.08%, 2.92%, 4.17% & 5.00% chord were also tested on
the model. The same metal angles were mounted to the wing upper and lower surface trailing edges to simulate T-
strips. The T-strip angles spanned the 86.4 inch center section of the model and did not extend into the wingtips.
Double-sided tape was used to attach the angles to the wing trailing edge. Aluminum tape was used as an additional
means of securing the angles. The T-strip heights quoted above are based on the height of the upper and lower
angles alone. The model wing’s 0.050” trailing edge thickness and the height of the double-sided tape were not
included. There was also a small loss in T-strip height because the wing upper and lower surfaces are not parallel at
the trailing edge. The NACA 23012 trailing edge closure angle is approximately 15.5°. This trailing edge angle
caused the upper and lower angles to tilt aft, causing a small loss in total T-strip height. This lost height was also
not included in the T-strip chords (heights) reported above.
A single row of trip dots (0.0114” height, 0.150” spacing, 0.075” diameter) were applied along the wing 10%
chordline on the upper and lower surfaces to fix the position of boundary layer transition from laminar to turbulent
flow. The test was conducted at nominal dynamic pressures of 75, 20 & 5 lbs/ft
2
; corresponding to Reynolds
numbers based on wing chord of 1.95x10
6
, 1.02x10
6
and 0.51x10
6
, respectively. Trip sizing studies showed that
boundary layer transition was likely forward of the trip for the 75 and 20 lbs/ft
2
runs. However, the trip was too
small and too far forward to fix the position of boundary layer transition for the 5 lbs/ft
2
runs.
Figure 3. Wing Tip Detail
Figure 2
.
Wing
Mounted in Test Section
American Institute of Aeronautics and Astronautics
4
III. Experimental Results
Force and moment data was collected for the baseline wing, the baseline wing equipped with Gurney flaps and
the baseline wing equipped with trailing edge T-strips. Each configuration was run at Reynolds numbers of
1.95x10
6
, 1.02x10
6
and 0.51x10
6
. The data presented in this paper has been corrected for weight tares, for the tare &
interference of the mounting fork, for flow angularity in the test section, for blockage and for the effect of the tunnel
walls on the model’s aerodynamic characteristics.
The weight tare was applied to account for the offset of the model wing’s center of gravity from the balance
moment center. That weight tare was taken with the wind-off during Run 1. A UWAL provided tare was applied to
lift, drag and pitching moment to account for the presence of the mounting fork on the data. That generic fork tare
was based on a UWAL model inversion test of a similar sized rectangular wing with an NACA 0015 airfoil section.
The blockage correction was made to dynamic pressure using Shindo’s Simplified Tunnel Correction Method
14
to
account for the increased velocity in the test section due to the presence of the model. A small correction to the drag
coefficient was made to account for the rotation of the lift vector due to the -0.12° of upflow in the test section.
Wall corrections were applied to angle of attack and drag coefficient to account for the proximity of the tunnel walls
to the model. The constraint of the flow field due to the tunnel walls makes the model wing appear to have a higher
aspect ratio, and thus too little induced angle of attack and too little induced drag. Nominal turbulence intensity in
the test section is quoted in the tunnel Technical Guide
13
as 0.72%.
A. Effect of Gurney Flaps and T-Strips on Baseline Wing Lift Curve
The effect of Gurney flaps on the baseline wing lift curve is shown below in Figure 4. Figure 5 shows the effect
due to trailing edge T-strips. The coefficient data presented in Figures 4 and 5 was taken at a Reynolds number of
1.95x10
6
. The coefficient data shows that Gurney flaps produced a positive increment in lift coefficient, a negative
shift in the zero-lift angle of attack, and an increase in the wing maximum lift coefficient. Larger Gurney flaps
produced larger lift increments. T-strips produced an increase in the slope of the lift curve and an increase in
maximum lift coefficient. However, T-strips produced no shift in the wing zero-lift angle of attack.
Angle of Attack - deg
-10 -5 0 5 10 15 20 25
C
L
-0.5
0.0
0.5
1.0
1.5
2.0
baseline
0.21% GF
0.52% GF
1.04% GF
1.46% GF
2.08% GF
3.33% GF
4.00% GF
5.00% GF
Re 1.95x10
6
Figure 4. Effect of Gurney Flaps on Wing Lift Curve, Re 1.95x10
6
American Institute of Aeronautics and Astronautics
5
The increments in zero-lift angle of attack due to Gurney flaps and T-strips are plotted versus device height in
Figure 6. Data is plotted for all three Reynolds numbers tested. The larger Gurney flaps and T-strips with chords
above 2% were only tested at a Reynolds number of 1.95x10
6
. As the coefficient data indicated, Gurney flaps
produced a negative shift in the wing zero-lift angle of attack and T-strips produced essentially no shift. The low
Reynolds number (0.51x10
6
) T-strip data does show a small (~0.4°) shift in α
OL
. The negative zero-lift angle of
attack shift due to Gurney flaps was not linear with device height. Smaller Gurney flaps produced a
disproportionately larger shift in the wing zero-lift angle of attack.
The lift coefficient at 0° angle of attack (C
Lo
) increment data in Figure 7 shows a similar trend. Gurney flaps
produced a positive shift in C
Lo
, T-strips did not. Notice that the 1.95x10
6
and 1.02x10
6
Reynolds number Gurney
flap and T-strip data agree quite well for both the α
OL
and C
Lo
shifts. This may indicate a Reynolds number
independence above 1.02x10
6
Re.
The percentage increase in wing lift curve slope due to Gurney flaps and T-strips is shown in Figure 8. Note that
lift curve slopes were taken from the coefficient data (C
L
vs. α) between 0° and 5° angle of attack. The data shows
that trailing edge T-strips produced a larger increase in wing lift curve slope. The increase for both was non-linear
with device height. Smaller Gurney flaps and T-strips produced a disproportionately larger increase in lift curve
slope. Above 2% chord, the lift curve slope increase for both devices was nearly constant at about 10%. Both
trailing edge devices produced smaller increases in wing lift curve slope at a Reynolds number of 0.51x10
6
. Below
0.5% chord, the low Reynolds number data actually showed a drop in lift curve slope when a Gurney flap or T-strip
was added to the wing.
The increase in the model wing maximum lift coefficient due to Gurney flaps and T-strips is plotted in Figure 9.
At each Reynolds number tested, Gurney flaps produced larger increments in wing maximum lift coefficient than
trailing edge T-strips. The C
Lmax
increments for both devices increased with increasing Reynolds number and
increased device height. That increase was non-linear with device height. Notice that the C
Lmax
increment for the
4% chord Gurney flap was quite substantial at 0.60.
Figures 6, 7 & 8 showed little change in the increments in zero-lift angle of attack, lift coefficient at 0° angle of
attack and the percentage increase in wing lift curve slope when Reynolds number was increased from 1.02x10
6
to
1.95x10
6
. This was the case for both Gurney flaps and trailing edge T-strips. This agreement may indicate a
Reynolds number independence for these lift curve characteristics above 1.02x10
6
.
Angle of Attack - deg
-10 -5 0 5 10 15 20 25
C
L
-0.75
-0.50
-0.25
0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
baseline
0.42% T-strip
1.04% T-strip
1.67% T-strip
2.08% T-strip
2.92% T-strip
4.17% T-strip
5.00% T-strip
Re 1.95x10
6
Figure 5. Effect of T-Strips on Wing Lift Curve, Re 1.95x10
6