Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Gurney Flap Effects and Scaling

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

GURNEY

FLAPEFFECTS
AND SCALING
FOR LOW-SPEEDAIRFOILS
P. Gigukre*, J. Lemayt and G. Dumas1
Laboratoire de MCcanique des Fluides, DCpartement de gCnie mCcanique
UniversitC Laval, QuCbec, Canada, G1K 7P4

Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

Abstract
T h e effects of Gurney flaps were tested on two
airfoil sections, the LA203A and the Gottingen
797, in a low speed wind tunnel. Lift and drag
forces, wall-pressure distributions and boundarylayer thickness measurements have been carried
out,. Different Gurney flap heights ranging from
0.5% t o 5.0% chord have been tested in order to
identify some optimal configuration of this simple,
yet promising device. T h e results presented are for
a chord Reynolds number of 250 000 and reveal, in
accordance with higher Reynolds number studies,
t h a t Gurney flaps may indeed provide a significant
increase in lift a t very little cost in drag. It is further
shown that through the proper use of Gurney flaps,
the aerodynamic performance of a simple design,
easy-t,o-build airfoil can be made practically as good
as those of a modern, high performance, complex
design. T h e present study suggests that both the
critical and the optimal height of Gurney flaps (in
t,erms of the l / d ratio) scales with the thickness
of the boundary-layer a t the trailing-edge of the
baseline airfoil on the pressure side. Based on the
results of this study and previous investigations, this
conclusion appears supported over a wide range of
Reynolds numbers. A preliminary discussion on the
physical mechanism behind the Gurney's effects and
its scaling is also proposed.

Nomenclature
c

chord length

Cd
CI
CP

drag coefficient
lift coefficient
pressure coefficient

'Graduate student. Present address:


Department of Aeronautical and Astronautical Engineering,
University of Illinois a t Urbana-Champaign.
t~rofessor,Member AI.4.4.
t ~ s s i s t a n tProfessor.
Copyright @ 1995 by Philippe Gigui.re, Jean Lemay and
Guy Dumas. Published by the American Institute of
Aeronautics and Astronautics, Inc., with permission.

h
zld
Re

Gurney flap height

lift-to-drag ratio (airfoil efficiency)


Reynolds number based on chord length
CY
angle of attack
ACd/Cdb drag coefficient increment normalized
by the baseline airfoil drag coefficient
A C l / C l b lift coefficient increment normalized
by the baseline airfoil lift coefficient
6
boundary-layer thickness

Introduction
The use of efficient high-lift low speed airfoils
is required for small, low speed unmanned aerial
vehicles (UAVs). The purpose of such aircraft
demands that they carry heavy payloads for a long
period of time.', 2 , To date, only a few airfoils meet
these requirements. Leading the list is the quite
famous Wortmann FX 63-137. More recently, Selig
and ~ u ~ l i e l mhave
o ~ reported promising results
with the S1210 and S1223 airfoils. Unfortunately,
these airfoil sections and other efficient high-lift
low speed airfoils are distinguished by a rather
complicated trailing-edge geometry.
Since the
performance of these airfoils is sensitive t o their
contour accuracy, and since the wings of these UAVs
usually have a small chord, it becomes difficult t o
build them accurately. Consequently, the achieved
airfoil performance can be substantially less than the
predicted performance.
In addition, the requirement t o minimize the
weight of small UAVs implies the use of light
construction materials. When simplicity and low
cost in construction are design criteria, minimizing
weight further complicates the construction process
of these airfoils, especially in reproducing the
concavity of the lower surface and obtaining a cusp
trailing-edge. In this respect, a study performed
by Stollery and ~
~ shows
e t rh a t ~a decrease in
the concavity of the lower surface on the FX 63137 produces a significant decrease in the airfoil
T h e combined effect of any
efficiency ( l / d ) .

966
American Institute of Aeronautics and Astronautics

Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

alterations of the airfoil shape is usually an overall


reduction in performance. In this context, the
use of the most efficient high-lift low speed airfoils
is questionable for small wing chords when using
simple and inexpensive construction techniques.
In connection t o this problem, the present
study demonstrates that it is possible to modify
a low speed airfoil which has a simple geometry
(finite trailing-edge thickness and flat bottom lower
surface) so that its performance is enhanced to a
level comparable to an efficient high-lift low speed
airfoil design. The airfoil performance enhancement
is achieved by the addition of a Gurney flap (figure
1) which consists of a small flat plate positioned
at the trailing-edge, perpendicular to the pressure
side of the airfoil.6 The Gurney flap was named
after Dan S. Gurney who first used this type of
flap on Indy Car wings to increase the down force,
improving the lateral traction in high speed turns.
It seems however that a fairly similar device has
been used previously by free flight enthusiasts in an
effort t o improve the glide characteristics of model
airplanes.7

brought about a new hypothesis on the efficiency


of Gurney flaps which is presented in this paper
as a new scaling for this device. Neuhart and
Pendergraft also mentioned that the increase in
lift is most likely due to the effective increase in
trailing-edge camber provided by the Gurney flap
and that the potential benefits in terms of drag
reduction may be limited to airfoils with rehtively
large trailing-edge closure angle and/or operating
at high lift coefficients. Another beneficial Gurney
flap in the study of Neuhart and Pendergraft was a
serrated version of the larger flap. Further testing
on this particular configuration of the Gurney flap
was performed by Vijgen et al.' They tested the
effects of a plain and serrated Gurney flap on the
aerodynamic characteristics of two tripped wings at
Reynolds numbers 1.1x lo6 and 3.7 x lo6. Serrations
was found to reduce both the increase in lift and
drag. The overall improvement in the wing lift-todrag ratio was greater in the case of the serrated
flap. Vijgen et al. also noticed that the height of
the serrated flap was less than the thickness of the
turbulent boundary-layer near the lower surface at
the trailing-edge.

e
Chord line

~h

Gurney flap/

'-

Figure 1: Airfoil equipped with a Gurney flap.


The benefits of this sort of flap were quantified
by the work of ~ i e b e c kon
~ a Newman airfoil for
Reynolds numbers between 1 - 2 x lo6. Liebeck
found that the addition of a 1.25% chord (1.25% c )
Gurney flap increased the lift and reduced the
drag for high lift coefficients and that in order
t o maximize the benefits of this device, its height
should be kept between 1% c and 2% c. In this
paper, we refer to this conclusion as the "GF
criterion". Liebeck found that beyond 2% c, the
drag increases noticeably. Liebeck also proposed
a hypothetical flow field at the trailing-edge of
an airfoil with a Gurney flap.6 This flow field
was later qualitatively confirmed by Neuhart and
Pendergraft8 who conducted a water tunnel study
of Gurney flaps, testing different configurations on
a NACA 0012 symmetrical airfoil at a Reynolds
number of 8 500. These tests revealed that the
most beneficial Gurney flap height was 4 . 2 % ~ .This
Gurney flap had the most favorable effects on the
upper surface flow separation at an angle of attack
up to 3.5'.
It is interesting to note that this
result does not agree with the G F criterion. This

Experimental data on the Gurney flap at low


speeds or low Reynolds numbers are scarce. Selig
et al.1 have reported experimental results on the
addition of a Gurney flap on a low speed airfoil.
They tested three Gurney flap heights (0.6, 1.2 and
2.6% c) on the S2091 airfoil at Reynolds numbers
between 60 000 t o 300 000. For all Reynolds
numbers, the increase in lift is significant while a
decrease in drag is only observed for moderate to
high lift coefficients. The efficiency of the S2091
airfoil was improved for all angles of attack before
stall.
While in the past, the main focus of interest
has been on the Gurney flaps in car racing
applications1l1 more recently, interest in the aviation community has been growing due t o promisin
results published in the last four years. 12, 13, 14, l f
Henne and Gregg12 have reported a new airfoil
concept which is similar to a Gurney flap. This
concept is known as the divergent trailing-edge
which is somewhat like a faired Gurney flap. It
has been used on the new wing design of a
commercial airliner. Jang et a1.13 have performed
a computational evaluation of an airfoil with a
Gurney flap. This computational study provided
insights on the trailing-edge flow structure around
the Gurney flap. Jang et al. mentioned that the
flow mechanism that makes the Gurney flap work so

967
American Institute of Aeronautics and Astronautics

effectively is its effect on the K u t t a condition. This


study was followed by some wind tunnel experiments
conducted by Storms and an^'^, verifying the
computational results. In these two studies, a
NACA 4412 was tested a t a Reynolds number
of approximately 2 x l o 6 . Storms and
have reported that t o avoid a drag penalty during
cruise, a split-flap configuration can be used with
a hinge line forward to the trailing-edge, provlding
perfornlancp comparable to the standard Gurney
flap. They also provided pressure measurements
w h ~ c hshow t h a t the addition of a Gurney flap
(between 0.5% c and 1% c ) considerably increased
the aft loading of the airfoil. It is also noted
t h a t much of the lift mcrernent is derived from a
general increase in loading and a higher suction
peak. Storms and Jang also quantified the effect
of the Gurney flap on the pitching moment. This
effect becomes more important as the Gurney flap
height is increased. Also, Jang el a1.13, as well as
Storms and
were in general agreement with
tlhe results of previous studies regarding the effect of
the Gurney flap on aerodynamic forces.

Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

an^'^

speed airfoils, namely t,he LA203A and the


Gottingen 797 (both shown in figure 2).
Demonstrate t,hat the addition of a Gurney
flap can enhance the performance of a low
speed airfoil having a sirnple geometry t o a
level comparable t o an efficient high-lift low
speed airfoil when built with a small chord using
simple and inexpensive construction techniques.
Propose a new flow-based scaling for optimum
height of the Gurney flap.

an^'^

Storms and Ross15 recently performed an


experimental study of lift-enhancing tabs, which are
similar t o Gurney flaps, on a two-elcment NACA
632-215 Mod B airfoil a t a chord Reynolds number of
approximately 3.7 x l o 6 . These lift-enhancing tabs
were positioned near the trailing-edge of the main
element and directly a t the trailing-edge of the flap.
T h e t a b located near the trailing-edge of the main
elenicrit produced a n increase in maximum lift only
for large t o moderate flap deflection, while the t a b
a t the trailing-edge of t,hc fiap produced a significant
lift increment for all flap defle~t~ions.
Recent interest in the Gurney flap is also
present in wind t,urbine applications. Kentfield
has presented several studies on the Gurney
flap in the context of its use on wind turbine
blades.16,
18, l 9 According t o these studies, the
Gurney flap improves the l/d characteristics of the
blades, consequently increasing the local torque and
therefore energy productmion.
From the above review, there clearly appears t o be
a need for further Gurney data on low speed airfoils
a s well as on the fundamental physics involved with
the mechanism of this device. T h e main objectives
of this study are therefore the following:
1. Provide experimental d a t a on the effect of
Gurney flaps of different heights on two low

Figure 2: The airfoils used in this study


a ) LA203A; b) Gottingen 797.
T h e LA203A was selected because of Liebeck's
results which show that it has even better overall
performance than the well known FX 63-137~'.

Ex~erimentalProcedure
All tests were performed in a low speed, closedloop wind tunnel a t Universitg Laval. T h e wind
tunnel used has a 2 meter long test section with
a cross section of approximately one meter in
diameter. The experiments were performed a t three
chord Reynolds numbers: 150 000, 250 000 and
375 000. Using hot-wire anemometry, the freestream
turbulence level was measured as 0.20% and 0.18%
a t the velocities corresponding t o a Reynolds
number of 150 000 and 375 000 respectively.
T h e models have been machined out of aluminum
using a CNC milling machine and have a span of just
under one meter. Their chord was chosen according
t o recommendations given by Rae and popez1 t o
minimize tunnel interference, as well as according t o
the Reynolds number range of the tests. This led
t o a chord length of 229 m m (9 in). Unfortunately,
it was not possible t o machine the last 7 m m (0.25
in) of the chord of the LA203A airfoil adequately so
that the final chord length was set a t 222 m m (8.75
in) for both airfoils. As a result, the suffix "-UL"
(standing for Universit4 Laval) was added t o the

968
American Institute of Aeronautics and Astronautics

Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

respective name of the airfoils. This demonstrates


the difficulties of reproducing the exact trailingedge shape of the most efficient high-lift low speed
airfoils at a small scale even if the material used is
aluminum. At the mid-span section of the Gottingen
model, 24 pressure taps were machined around the
airfoils contour (figure 3). The models are mounted
vertically in the test section with a 2 mm gap
between the model tip and the ceiling of the test
section. A CNC milling machine was also used to
machine a total of 7 Gurney flaps (0.5, 1.0, 1.5, 2.0,
2.5, 3.0 and 5.0% c) from aluminum angles. The
thickness of the Gurney flaps is 0.5 mm (0.20 in).
The flap was added to the models using a doublesided tape of thickness 0.05 mm (0.002 in), a simple
and yet effective method.

Effects of Gurney flaps on lift and drag

As expected, figure 4 confirms that the addition


of all Gurney flaps considered here significantly
increases the lift, while basicaly preserving the stall
characteristics. With the longer flap (5% c ) , the
lift coefficient increases from 1.55 to 1.95 for the
LA203A-UL and from 1.50 to 2.00 for the Gottingen
797-UL. However, the additional lift gained is not a
linear function of the height of a Gurney
- flap- used;
the lift increment slows down with the flap size.

Figure 3: Pressure taps on the Gottingen 797-UL.


The lift and drag were measured using an electromechanical balance for angle of attacks from -4.5'
to 15' with increments of 1.5' (0.75' into stall).
The uncertainty for these measurements is on
the order of 1.7% on the lift and 12.1% on the
drag. All the measured data have been corrected
for tunnel interference using the two-dimensional
tunnel corrections outlined by Rae and
The
static pressures around the Gottingen model were
measured and compared to freestream conditions
using a pressure transducer. The boundary-layer
thicknesses at the trailing-edge on the pressure side
of the airfoils were obtained from the velocity profiles
measured with a flattened Pitot tube mounted on a
traversing mechanism. Boundary layer thicknesses
were so measured for 4 angles of attack (-4.5'
, 0' , 4.5' and 9'). The uncertainty for these
measurement,^ increases with the angle of attack (the
boundary-layer gets thinner) and on the average, is
on the order of 10%.

Experimental Results
The results presented in this section are for a
chord Reynolds number of 250 000 which was the
primary Reynolds number for this wind tunnel study
of Gurney flaps. Due to space limitations, the results
for the tests performed at Reynolds numbers of
150 000 and 375 000 have been omitted here but
can be found in a separate document. 22

Figure 4: Variation of the lift coefficient with


angle of attack and Gurney flap size
a) LA203A-UL; b) Go 797-UL.
At first sight, it may seem that the addition of
the flap causes the stalling angle of attack to
decrease. However, it is important to note that
this is essentially due to the chord line convention
used here (see figure 1) which is defined in all cases
969

American Institute of Aeronautics and Astronautics

with respect t o the baseline airfoil chord without


considering the displacement of the actual trailing
edge toward the end of the Gurney flap. A chord line
defined, for example, from the leading-edge of the
airfoil to the end of the Gurney flap, would increase
the angles of attack in figure 4 in such a proportion
t,hat the stalling angles of attack for all flap heights
would basically appear the same as for the baseline
case.

Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

a)

b)

2.0

2.0

1.

smallest one. Similar results are shown for the


LA203A-UL for lift coefficients above 0.5 and 1.4
respectively. For both airfoils, it thus appears
however that only the smallest flap tested here allows
to avoid a detectable increase in drag for most
practical lift coefficients.

NoGurney

cd
NoGurney

Figure 6: Variation in the airfoil efficiency with


angle of attack and flap size
a) LA203A-UL; b) Go 797-UL.

Figure 5: Variation in the drag polars with


Gurney flap size
a) LA203A-UL; b) Go 797-UL.
The drag polars for both airfoils are presented
in figure 5 which shows, in agreement with some
previous reports, that a drag reduction for a given
lift is indeed observed. For the Gottingen 797-UL,
one observes a drag reduction for all lift coefficients
when using our smallest flap, and a drag reduction
for lift coefficients above 0.8 when testing our second

The most interesting beneficial effects of the


Gurney flap are perhaps best illustrated by
considering the l / d ratios (figure 6). With the
smallest flap, the efficiency of both airfoils are
enhanced compared to the baseline cases. One also
notes that the global increase in airfoil efficiency
is more significant for the simple Gottingen 797UL than for the more optimized LA203A-UL.
Furthermore, the results presented in figure 6
show that for angles of attack between -4.5' and
approximately 2O, and for both airfoils, Gurney
flap heights up to 2% c yield an increase in airfoil
efficiency.
970

American Institute of Aeronautics and Astronautics

point where the two curves intersect on figure 7.


This critical Gurney flap height is plotted in figure 8
as a function of the angle of attack and for both
airfoils tested. This figure clearly indicates that
the critical height decreases with increasing angle
of attack.

The effectiveness of Gurney f l a p s

-0.5Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

b)

2.0

h i t i c a l flap height
'

h/c (%)

CI

c*

A comparison of the efficiency of the LA203A-UL


without a Gurney flap, and the Gottingen 797-UL
equiped with the two smallest flaps is presented in
figure 9. For lift coefficient below 1.2, the Gottingen
797-UL with the smallest flap has a comparable or
better efficiency than the LA203A-UL. T h e use of a
longer flap (1.0% c or more) on the Gottingen 797UL does not allow t o enhance its l / d efficiency t o
the level of the LA203A-UL.

Figure 7: Normalized lift and drag increment with


Gurney flap size for the LA203A-UL
a)a=OO; b)a=6O.

Figure 9: Airfoil efficiency with lift coefficient

Figure 8: Critical flap height as a function of the


angle of attack for both airfoils tested.
Figure 7 presents the relative lift and drag
increments with respect t o the flap height. Defining
t,he critical Gurney flap height as being the
maximum height still providing a net benefit in the
l / d ratio, one thus finds the critical height a t the

Hence, this confirms that owing to the Gurney


flap, it is indeed possible t o make the simple
Gottingen airfoil reach and even surpass the
performance of the LA203A-UL. It is important to
note however that the LA203A-UL is not quite as
efficient as the original design of Liebeck. This
somewhat under-design performance of the actual
LA203A-UL is a direct result of the practical
difficulty in achieving predicted performances when
constructing a small chord airfoil with inexpensive
and simple construction techniques. When design
simplicity and low cost in construction are desired, it
seems that the use of a Gurney flap on an easily built
airfoil is an especially advantageous and attractive
option.

971
American Institute of Aeronautics and Astronautics

Effects on the pressure distribution


As an additional source of information to help
in understanding the performance enhancement
related to efficient Gurney flaps, pressure measurements were taken on the Gijttingen 797-UL with
and without Gurney flaps for 4 angles of attack (see
figure 10).

Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

a) -1.5

No Gurney

for the shortest Gurney flap, the change in pressure


coefficient is more important on the lower surface
while it becomes more significant on the upper
surface for larger flaps.
c) -3.5
No Gurney

d ) -5.0

xlc
NoGurney

-4.5

b) -2.5

O 0.5%~

No Gurney
O 0.5% c

Figure 10: Pressure distributions for the Go797UL (baseline case and with Gurney
flap size of h / c = 0.5%, 2% and 5%)
a) a = 0'; b) 5'; c) 10'; d) 15'.
The increase in lift provided by all Gurney flaps
tested is seen here as an increase in the area between
the corresponding upper and lower side curves. With
all Gurney flaps, the loading of the airfoil is clearly
seen to increase over the entire body, although
more so a t the suction peak and near the trailing
edge. The results in figure 10 also suggest that,

Boundary-layer thickness
Table 1 presents the results for the boundary-layer
thickness at the trailing-edge on the pressure side of
the baseline LA203A-UL and Go797-UL airfoils. As

9 72
American Institute of Aeronautics and Astronautics

expected, the boundary-layer thickness is found to


decrease significantly with increasing angle of attack.

Range of optimal
Gurney flap size

Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

Table 1: Boundary-layer thickness at the trailingedge on the airfoil pressure side as a


function of the angle of attack.

Range of optimal
Gurney flap size

Discussion
Scaling for optimum Gurney flap heinht
In an effort t o relate the efficiency of the Gurney flap
with respect to its height, a scaling based on actual
flow characteristics is proposed. This scaling relates
the Gurney flap efficiency (in terms of lid) to the
thickness of the boundary-layer on the pressure-side
of the baseline airfoil at the trailing-edge and comes
from the observation that the optimum Gurney flap
size is related t o the angle of attack.
Figure 11: Boundary layer thickness and range
for efficient Gurney flap heights as a
function of angle of attack
a) LA203A-UL; b) Go 797-UL.

Figure
11 presents
the
boundary-layer
measurements for the two airfoils as well as
the experimentally determined range for optimal
Gurney heights, both as functions of the angle of
attack. In the present context, an "optimal Gurney"
means a Gurney flap which yields the maximum
increase in our efficiency parameter, i.e., the l / d
ratio. Due t o the uncertainty in the experimental
data and also to the limited range of Gurney
heights tested, a single formal optimum for each
angle of attack cannot be inferred. Rather, an
approximate, "most-likely" range for the optimal
Gurney is presented in figure 11.
The result quite eloquently suggests that the
height of the most efficient Gurney scales with the
boundary- layer thickness. This observation is valid
for both airfoils tested within the uncertainty of the
data. Unfortunately, it is not yet possible at this
point to refine this new scaling. More data taken
with increased accuracy and covering a wider range
of Gurney's heights would be necessary.

Table 2: Heights of the most beneficial Gurney


tested by different investigators.

aFlat plate estimate


b ~ o sbeneficial
t
Gurney flap tested
'The G F criterion

It is, however, possible to test, although crudely


at this stage, the new scaling against other available

973
American Institute of Aeronautics and Astronautics

terms of two distinct base cases, i.e., when there is no


local boundary-layer separation on the upper surface
of the baseline airfoil and when there is one. One can
expect that the Gurney flap, although potentially
efficient in both cases, act quite differently in each
case.

Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

results in the literature. Table 2 provides the heights


of the most beneficial Gurney flaps tested by other
investigators against rough but indicative estimates
of the boundary-layer thickness at the trailingedge in each case. These estimates are obtained
either by the simple Blasius boundary-layer result
or by the standard empirical relation for a turbulent
boundary-layer on a flat plate.

It is common knowledge and intuition that a


flat plate normal to the flow experiences significant
pressure drag. How is it then that an efficient
Gurney flap does not appear to exhibit such a
phenomenon? The answer to this may perhaps be
found in the previously discussed scaling of these
flaps as well as in the flow field behind them. An
efficient Gurney flap is very small and within the
boundary-layer thickness. This suggests that one
should not think of it as the usual bluff body
case. The Gurney flap is submerged in a high-shear
layer (vorticity layer) where viscous effects are very
important. This may indeed allow a steady-in-themean bubble to exist right behind the flap as was
. ~can be inferred that if
first suggested by ~ i e b e c kIt
an attached bubble exists, there might be an actual
compression of the wake behind the airfoil. Such
a compression effect would lead t o a thinner wake
and thus a smaller drag. It is interesting t o note
that this situation could provide a drag reduction
not only when considering a given lift coefficient but
also for a given angle of attack.

These above estimates are admittedly quite


crude since they correspond to a boundarylayer evolution under zero pressure gradient, but
they are nonetheless reasonable ones short of
actual measurements. Furthermore, comparing the
measured data of figure 11 and the present estimate
of 6 / c = 1% (flat plate estimate at Re=250000), such
estimates appear quite satisfactory, especially for
moderate angles of attack (4' to 5'). It is important
to note that the data of Table 2 cover a wide range
Reynolds numbers, from a few thousands to a few
millions. In the table, the result representing this
study comes from figure 6 (lid vs a ) where it is
clearly seen that our smallest Gurney flap is the most
efficient we have tested. Although the data of Table
2 are quite approximate and must be interpreted
with care at this early stage of the investigation, a
significant and safe conclusion can be expressed as
follows:
Beneficial Gurney flaps are always found to have a
height not greater than the boundary-layer thickness.

Phvsical mechanism-Preliminarv

discussion

As stated earlier, the Gurney flap is an intriguing


device. The word "intriguing", as used by Jang
e t a1.13, demonstrates that the physical mechanism
associated to this special flap is still an open
question.
This study and all others on Gurney flaps have all
shown that this device can yield important increases
in lift without increasing the drag significantly. The
fundamental question about the physical mechanism
related to the Gurney flap may thus be considered
twofold. On the one hand, there is the question of
how such a small device can increase lift so much,
and on the other hand, how can it do that without
the expected increase in drag.
As far as lift is concerned, it is obvious that such a
flap increases the local camber at the trailing-edge.
It might be more instructive however t o consider
this question in terms of a blunt trailing-edge in
connection with the Kutta condition, and also in

In an attempt to further describe the drag


reduction phenomenon related to the Gurney flap,
it seems appropriate to consider its effects on both
skin-friction and pressure drag. It is likely that the
addition of a Gurney flap decreases skin-friction drag
since t,he flow is forced to separate from the lower
surface, forming a recirculation region in front of the
flap. However, this reduction in skin-friction drag is
modest since an efficient Gurney flap is small.
The Gurney flap has a better potential in reducing
pressure drag. It has been shown8' 9 ' l 3 that the
Gurney flap can delay and even alleviate separation
on the airfoil suction side for a range of angles of
attack. In doing so, a Gurney flap could decrease the
pressure drag of a baseline airfoil for which trailingedge separation occurs. From the point of view
of the base pressure behind the flap, it is likely,
due to the relative size of the bubble, that the two
counter-rotating vortices forming it are fed in so
much energy by the adjacent shear layers that there
is almost a match in pressure on both sides of the
flap. It would thus seem crucial t o the efficiency
of a Gurney flap that an attached bubble exists

974
American Institute of Aeronautics and Astronautics

Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

behind it. This attached bubble can be visualized


as a fluid extension of the airfoil which provides an
off-the-surface pressure recovery. The off-the-surface
pressure recovery and the previously discussed wake
compression effect can then allow a reduction in
pressure drag.

In light of this discussion, it seems that the offthe-surface pressure recovery, which results from the
presence of an attached bubble behind an efficient
Gurney flap, might be a most important factor that
allows a reduction in drag. The Kutta condition is
then satisfied at the end of the attached bubble.

It would thus appear that the question of drag


reduction associated with an efficient Gurney flap is
Concluding Remarks
intimately related to the presence of local separation
on the baseline airfoil (unattached case). When
The present wind tunnel study of Gurney flap
there is no local separation on the baseline airfoil
effects
for the LA203A-UL and Gottingen 797-UL
(attached case), the possibility of having lower
airfoils
at a chord Reynolds number of 250 000 has
drag relies mainly on the aforementioned skinled
to
the
following conclusions:
friction drag reduction and compression effect on
the wake. Therefore, the potential of reducing
1. The effects of the Gurney flap on the lift and
drag in the attached case is possible but is inferior
drag characteristics are in general agreement
to the unattached case. This last statement is
with
the conclusions of other investigators. A
consistent with a previous observation from the
significant
increase in lift is provided with in
literature. More specifically, the potential benefits
some cases little or even no cost in drag. The
of the Gurney flap in terms of drag reduction were
airfoil efficiency and endurance characteristics
said8 t o be limited to airfoils with relatively large
are
therefore usually improved.
trailing-edge closure angle and/or operating at high
lift coefficients. It is exactly in this type of condition
2. An efficient Gurney flap can make a simple
that local separation on the suction side is more
design airfoil reach and even surpass the
predominant. It is also consistent with the reduction
performance of an efficient high-lift low speed
in the lift-hysterisis loop documented by ~ i g u g r e ~ ~ airfoils when actually built with a small chord
since lift-hysterisis is related t o flow separation.
using simple and inexpensive construction
techniques.
The question of the drag reduction associated with
an efficient Gurney flap is also related t o the cases
where one considers either a given angle of attack
or a given lift coefficient when doing the comparison
with the baseline airfoil. A drag reduction when
both the airfoil fitted with an efficient Gurney flap
and the baseline case operate at the same angle of
attack is indeed interesting. However, it is not so
important because the required lift force that an
aircraft must produce to remain aloft corresponds
to a required lift coefficient and not a required angle
of attack. When both the baseline airfoil and the
airfoil with the Gurney flap are operating at the
same lift coefficient, the pressure drag of the baseline
airfoil should increase more rapidly since it must
operate a t a larger angle of attack. The difference
in pressure drag becomes even more important as
the baseline airfoil approaches stall in which case
separation becomes more dominant. Taking the
skin-friction drag as essentially constant, the total
drag of the airfoil equipped with a Gurney flap may
become smaller than the total drag of the baseline
airfoil, especially at moderate to high lift coefficients.
This statement is supported by t.he drag polars
shown in figure 5 as well as by the results of other

3. The proposed flow-based scaling which relates


the efficiency of Gurney flaps with the
boundary-layer thickness at the trailing-edge on
the airfoil's pressure side not only represents
the results of this study, but also agrees with
other investigators' data. In this respect, it
seems important that the Gurney flap remains
submerged in the boundary-layer in order to
increase the airfoil efficiency.
4. The potential of the Gurney flap in reducing
drag appears to be greatest when local
separation is present on the suction side of the
baseline airfoil. This usually corresponds to
moderate to high lift conditions.
Finally, it would appear that an efficient Gurney
flap should not only be considered as a high-lift
device, but also as a performance enhancer, even
for cruise conditions.

Acknowledgments
The authors wish to thank Professor Dinh N.
Nguyen and Mr. Etienne Jean for their valuable

975
American Institute of Aeronautics and Astronautics

contributions to this work. Financial support from


NSERC of Canada and FCAR of Qukbec is also
gratefully acknowledged.

l2

Henne, P.A. and Gregg, R.D. 111, "New Airfoil


Concept," Journal of Aircraft, Vol. 28, No. 5, May
1991, pp. 300-311.

l3

Jang, C.S., Ross, J.C. and Cummings, R.M.,


"Computational Evaluation of an Airfoil with a
Gurney Flap," AIA A Paper 92-.$?708-CP, June
1993.

l4

Storms, B.L. and Jang, C.S., "Lift Enhancement


of an Airfoil Using a Gurney Flap and Vortex
Generators," AIAA Paper 93-064 7-CP, Jan.
1993.

l5

Storms, B.L. and Ross, J.C., "An Experimental


Study of Lift,-Enhancing Tabs on a Two-Element
Airfoil," AIA.4 Paper 94-1868-CP, 1994.

l6

Kentfield, J.A.C., "The Potential of the


Gurney Flap for Improving the Performance of
Wind-Turbine Blading," Proceedings of the
WINDPO WER '91 conference, Palm Springs,
California, Sept. 1991.

l7

Kentfield, J.A.C. and Clavelle, E.J., "The Flow


Physics of Gurney Flaps, Devices for Improving
Turbine Blade Performance," Proceedings of the
ASME Wind Energy Conference, Houston, Texas,
Jan. 1993.

References

Downloaded by UNIVERSITY OF CALIFORNIA - DAVIS on May 7, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.1995-1881

Foch, J.R. and Ailinger, K.G., "Low Reynolds


Number, Long Endurance Aircraft Design,"
AIAA Paper 92-1263, Feb. 1992.
Foch, J.R. and Toot, P.L., "Flight Testing
of Navy Low Reynolds Number Unmanned
Aircraft," Lecture Notes in Engineering: Low
Reynolds Number Aerodynamics, T.J. Mueller
(ed.), Vol. 54, Springer-Verlag, New-York, June,
1989.
Mueller, T.J., "Low Reynolds Number Vehicles,"
AGARDograph No. 288, Feb. 1985.
Selig, M S . and Guglielmo, J . J . , "High-Lift Low
Reynolds Number Airfoil Design," AIAA Paper
94-1866, June 1994
Stollery, J.L. and Dyer, D.J., "Wing-Section
Effects on the Flight Performances of a Remotely
Piloted Vehicle," Journal of Aircraft, Vol. 26, No.
10, Oct. 1989, pp. 932-938.

IsKentfield,
J.A.C.,
"Theoretically
and
Experimentally Obtained Performances of
Gurney Flaps Equipped Wind Turbines,"
Proceedings of the ASME Wind Energy
Conference, New Orleans, Louisiana, Jan. 1994.

Liebeck, R.H., "Design of Subsonic Airfoils for


High Lift," Journal of Aircraft, Vol. 15, No. 9,
Sept. 1978, pp. 547-561.
Gieskieng, B. J r . , "Indoor H .L. Glider Section,"
Model Aeronautic Year Book, Frank Zaic (ed.),
1964-65.

l9

Kentfield, J.A.C., "The Flow Over Gurney


Flaps, Devices for Improving Wind Turbine
Performance," Proceedings of the WINDPO WER
'94 Conference, Minneapolis, Minnesota, May
1994.

20

Liebeck, R.H. and Camacho, P.P., "Airfoil Design


at Low Reynolds Number with Constrained
Pitching Moment," Proceedings of the Conference
on Low Reynolds Number Airfoil Aerodynamics,
University of Notre Dame, U.S.A., June 1985.

Neuhart, D.H. and Pendergraft, O.C. Jr., "A


Water Study of Gurney Flaps," NASA TM 4071,
Nov. 1988.
Vijgen, P.M.H.W., van Dam, C.P., Holmes, B.J.
and Howard, F.G., "Wind-Tunnel Investigations
of Wings with Serrated Sharp Trailing Edges,"
Proceedings of the Conference on Low Reynolds
Number Airfoil Aerodynamics, University of
Notre Dame, U.S.A., 1989.
Selig, M.S., Donovan, J.F. and Fraser, D.B.,
"Airfoils at Low Speeds," SoarTech 8, SoarTech
Publications, 1504 N . Horseshoe Circle, Virginia
Beach, VA, 23451, 1989.

Rae, W.H. Jr. and Pope, A. J r . , "Low-Speed


Wind Tunnel Testing," John Wiley St Sons (2nd
ed.), U.S.A., 1984.
22

Katz, J . and Dykstra, L., "Study of an OpenWheel Racing Car's Rear-Wing Aerodynamics,"
SAE Paper 890600, 1989.

Gigukre, P., "Effets d'un Volet Gurney sur des


Profils d'Ailes b Faibles Nombres de Reynolds,"
Master's Thesis (#13207), Universite Laval,
1994.

976
American Institute of Aeronautics and Astronautics

You might also like