98MSV-14

Experimental Study of Open-Wheel Race-Car Front Wings

William J. Jasinski and Michael S. Selig
University of Illinois at Urbana-Champaign

ABSTRACT Since it is difficult to completely match wind-tunnel

conditions or numerical boundary conditions to the track
An experimental study was performed at the conditions, tests have been performed to determine the
University of Illinois at Urbana-Champaign Low-Speed proper way to evaluate the effect of wings on a race car.
Wind Tunnel to quantify the performance and flowfield Berchak and Camosy tested an isolated full size rear
effects of two-element open-wheel-race-car front wing wing and a 40% scale model of a vintage Indy car and
configurations. Four distinct configurations were tested compared results [4]. Their study revealed that
in- and out-of-ground effect and at various speeds performance of the wing is enhanced by the presence of
(Reynolds numbers), angles of attack, and flap positions. the car indicating some degree of cross-coupling
A splitter plate was installed in the wind tunnel to act as between the wing and body. This idea is also supported
the ground plane. Data presented include balance force in Refs. [2][3]. Knowing that wing-alone results will
measurements, surface pressure data, and downstream underestimate performance when used on a race car, it
flow measurements using a seven-hole probe. Results is feasible to perform these tests and still be able to
show that these elementary factors in the design of race- estimate the effects on the complete race car.
car front wings have a significant effect on wing
performance and behavior of the downstream flowfield. The studies mentioned above primarily deal with
rear-wing aerodynamics, and there is still much to learn
INTRODUCTION about how the front wing of an open-wheel race car
effects performance. A series of investigations
BACKGROUND - In recent years, aerodynamics has performed by Ranzenbach and Barlow begin to address
become an important factor in the design and the performance of front wings with two-dimensional
performance of open-wheel race cars. The large experimental and numerical methods [5][6][7]. These
amounts of downforce produced by these cars, typically studies focused mostly on the effects of the various
more than two times the car’s weight, allow cornering simulations of the ground plane boundary conditions on
accelerations of well over 4 g's. Despite the gains made the performance of two-dimensional airfoil sections. In
in aerodynamic design, there is still little known about the case of single-element tests of the NACA 0015 [5]
the influence that race car wings have on the production and 4412 airfoils [6], comparison between experimental
of overall downforce, specifically in CART Cars where and computational results show fairly good agreement at
the underbody channels produce a majority of the ground clearances greater than 0.1 chord lengths. At
downforce. lower heights, boundary layers from the airfoil and
ground plane begin to interact, which becomes difficult to
Previous studies of race-car aerodynamics ranging predict accurately with current computational methods.
from full-car wind-tunnel tests, numerical investigations, An additional study of a two-element airfoil (NACA 632-
and combined experimental and numerical studies have 215 Mod B) with a 30% slotted flap shows very little
shown that the effect of small changes in parameters on agreement of absolute values of lift coefficient, but very
a race car can have significant effects on aerodynamic similar trends [7]. The authors state that this
performance [1][2][3]. Hurst shows that a one-degree discrepancy was due to problems with the experimental
change in wheel camber, a seemingly unimportant setup.
aerodynamic variable, can change downforce by up to 2
percent [1]. Katz demonstrates through the use of a The tests above show that it is possible to perform
panel method that the addition of a front and rear wing to simplified tests and still be able to extract information
an open-wheel race car can change a lifting body to one that is beneficial to the design of a race car despite the
that produces a large amount of downforce [2]. These complexity of the overall problem.
studies show that that in order to use simulations as a
tool to maximize performance, a race car must be MOTIVATION - Motivation for the current study surfaced
modeled properly in the wind tunnel or computations. because of the lack of information available on three-
dimensional downforce wings in ground effect. The
flowfield near the front wing of an open-wheel race car is
very complicated despite being in relatively undisturbed
air, although the close proximity of the wing to the front
wheels tends to have a large effect. The current method
of race-car wing design is based on trial and error where
a part is designed, a model is built and tested in the wind
tunnel. If favorable results occur, the design is improved
until the desired effect is met. The information gained
from this study can help designers improve and shorten
the process by identifying crucial factors in the design of
a front wing and how they affect performance before
testing them in the wind tunnel. Knowledge gained from
this study will be used to further enhance current design
tools and verify various numerical tools that can be used
during the design cycle.

EXPERIMENTAL APPARATUS

WIND TUNNEL - Experiments using configurations

based on the UIUC700 two-element airfoil were
performed at the University of Illinois at Urbana-
Champaign Low-Speed Wind Tunnel. The wind tunnel Figure 1: Endplate and front wing planforms (in cm).
is a conventional open-return type with a contraction
ratio of 7.5:1 and test section dimensions of 0.853 x while the F1 flap design has shorter tip and root sections
1.219-m. The test section diverges approximately 1.27- with a long, sweeping transition. The 3D flaps used
cm over its 2.43-m length to allow for boundary layer typical foam-core construction with carbon-fiber skins
growth. Flow velocity can be varied up to 71.52-m/s and were not pressure tapped. Flap deflection, as
(160 mph), corresponding to a Reynolds number of measured at the root, was adjustable from 4 to 44 deg in
6
approximately 4.9 x 10 /m. 5 deg increments. Two endplates were designed that
approximate the profile for either an IndyCar or Formula
MODEL - Tests were performed on wing/airfoil One front wing endplate could be combined with the
configurations based on the UIUC700 two-element above wing planforms to form four different front wing
airfoil. Due to structural and space requirements, as well configurations. Front wing and endplate planforms are
as available construction techniques, the UIUC700 airfoil shown in Figure 1. A summary of the possible test
was designed specifically for this experiment using configurations is given in Table 1. Only the results from
methods described in Ref. [8] and has no relation to the front wing configurations will be discussed.
profiles used on actual race cars. The wind-tunnel
model was tested as a semi-span wing with a span of Table 1: Two-element test configurations
46.23-cm, with a main-element chord of 22.86-cm, and
2D flap having a chord of 10.16-cm, resulting in a Designation Description Test Flap Endplate
combined reference chord length of 33.02-cm. Both the UIUC700 Two-element 2D “2D” n/a
main element and 2D flap were constructed using baseline airfoil
molded carbon-fiber skins with steel bar spars. The UIUC700R Indy rear wing 3D “2D” Indy rear
main element was pressure tapped at five spanwise UIUC700I Indy front wing 3D Indy Indy front
locations of 10.16, 20.38, 29.49, 36.83, and 42.42-cm UIUC700F F1 front wing 3D F1 F1 front
from the root of the wing. The 2D flap was pressure UIUC700G Hybrid front wing 3D Indy F1 front
tapped at three locations at 10.16, 29.49, and 42.42-cm. UIUC700H Hybrid front wing 3D F1 Indy front
Taps were aligned parallel to the chord-line. The
UIUC700 two-element airfoil was tested in two-
dimensional flow and, when combined with a large GROUND PLANE - To simulate ground effect, a 183 x
endplate, in a configuration similar to an IndyCar rear 85.41 x 1.905-cm thick clear acrylic sheet with a full-
wing. Flap deflection could be manually set in 5 deg radius leading edge was installed in the wind tunnel.
increments from 6 to 41 deg. The ground plane could be mounted at clearances of
3.302 (0.1c), 6.604 (0.2c), and 9.906-cm (0.3c). The
Two additional flaps were designed such that, when leading edge of the ground plane was placed 66.04-cm
combined with the main element, would approximate the forward of the leading edge of the main element to
planforms of either an IndyCar or Formula One front reduce boundary layer growth and provide for proper
wing planform. For these flaps, the chord lengthens isolation of the wing in the tunnel. Preliminary
from 10.16-cm at the root to 17.78-cm at the tip. The investigations showed that at ground heights at and
IndyCar flap was designed to have a long, constant root below 0.2c and lift coefficients above 2, the trailing
section with a sharp “cutout” that transitions into the tip, vortex system of the wing induced boundary-layer
separation on the ground plane, thereby limiting the
ground clearances that could be tested. Data for the
0
ground clearance of 0.3c will be the only data presented.
Preliminary investigations at this height showed that
wingtip vortex and ground-plane boundary-layer 0.5" x 0.5"
-5
interaction was minimal and acceptable for the scope of
this study.
-10
0.25" x 0.25"
EXPERIMENTAL METHODS - Lift, drag, and pitching

Ground Plane
moment data were acquired with a three-component

Y (in)
external floor balance manufactured by Aerotech ATE -15 1.5" x 1.5"
Limited using methods as described in Refs.[9] and [10].
Data were acquired at both increasing and decreasing
angles of attack. Data for both directions will be shown -20
in the results. The model was mounted to the balance
so that pitching moment measurements were taken 0.5" x 0.5"
about the quarter-chord of the main element. Error -25
analysis of the balance setup indicated that errors in lift,
drag, and moment coefficient were typically less than
-30
2%. At low-speed and low-drag conditions, error in drag
coefficient peaked at approximately 6%. 28 24 20 16 12 8 4 0 -4
Z (in)
Downstream flowfield measurements were acquired Figure 2: Seven-hole probe scan position and
at selected conditions using an Aeroprobe Corporation resolution (in inches).
3.175-mm diameter seven-hole probe. The probe was
mounted to a two-axis traverse manufactured by Lintech, Formula One flap (UIUC700F and H), the maximum
Inc. and was positioned approximately 33-cm, or one Reynolds number was lower than cases with the IndyCar
reference chord length, behind the trailing edge of the flap (UIUC700G and I). This speed was the maximum
wing. Scans included approximately 3600 data points that was possible without exceeding balance load
with the extents and resolution that is shown in Figure 2. ranges.
Scan resolution was chosen to capture as much detail
as possible in areas where large gradients in flow
Table 2: Front Wing Test Matrix
direction were expected while allowing for reasonable
run times of approximately 2 1/2 hours. Three Configuration α (deg) Re δf (deg)
components of flow velocity were determined by using a UIUC700I -3 thru 18 0.7 x 10
6
14
combination of the seven pressures from the probe 6
UIUC700I -3 thru 18 1.1 x 10 4, 14, 24
through methods outlined by Rediniotis et al [11], as 6
UIUC700I -3 thru 18 1.3 x 10 14
originally reported by Zilliac [12]. Measurement errors in 6
dynamic pressure and flow angle from the seven-hole UIUC700F -3 thru 18 0.7 x 10 14
6
probe were typically 0.5 degrees and 1%, respectively, UIUC700F -3 thru 18 1.0 x 10 4, 14, 24
6
as quoted by the manufacturer. The manufacturer UIUC700F -3 thru 18 1.2 x 10 14
6
provided a 1600-point calibration map, which included UIUC700G -3 thru 18 0.7 x 10 14
6
data for cone angles up to 70 degrees. Seven-hole UIUC700G -3 thru 18 1.1 x 10 4, 14, 24
6
probe and surface pressure measurements were UIUC700G -3 thru 18 1.3 x 10 14
6
acquired using a Pressure Systems, Inc. 8400 pressure UIUC700H -3 thru 18 0.7 x 10 14
6
system with ±7 kPa and ±35 kPa electronic pressure UIUC700H -3 thru 18 1.0 x 10 4, 14, 24
6
scanners. Pressure scanner accuracy was quoted as UIUC700H -3 thru 18 1.2 x 10 14
0.05% of full-scale.

TEST MATRIX – The test conditions that were run were Downstream flowfield measurements are
chosen to cover the effects of changing major variables summarized in Table 3. The aim of these
of flap planform, endplate planform, angle of attack, flap measurements was to gain an understanding about the
deflection, and Reynolds number at a ground clearance general flow structure behind the wing, including vortex
of 9.906-cm, or 30% of the reference chord. Table 2 development, and changes that would occur with
summarizes the test matrix that was performed. The changes in the aforementioned parameters. The
four possible flap and endplate combinations formed the baseline UIUC700I configuration was tested extensively,
UIUC700I-H base configurations as described in Table 1 while a limited number of scans were performed with
above. Due to maximum balance loads, especially at other configurations. The changes in the flowfield due to
higher speeds, the number of possible flap deflections parameter changes such as angle of attack, Reynolds
that could be tested was limited. Reynolds number was number, and flap deflection that occurred for the
set by varying tunnel RPM and was based on the UIUC700I can be used to infer results for the other
reference chord of 33.02-cm. In the cases with the configurations.
 Table 3: Seven-hole probe scan test matrix
6
Re = 0.7 x 10
Re = 1.1 x 106
Configuration α (deg) Re δf (deg)
6
Re = 1.3 x 106 UIUC700I 6 0.7 x 10 14
3 6
UIUC700I 6 1.1 x 10 4, 14, 24
6
UIUC700I 3, 6, 9 1.1 x 10 14
6
2.5
UIUC700I 6 1.3 x 10 14
6
UIUC700F 6 0.7 x 10 14
6
UIUC700G 6 1.0 x 10 14
6
2 UIUC700H 6 1.2 x 10 14

RESULTS
CL

1.5

REYNOLDS NUMBER – In a typical road or street

1 course, race-car speeds may vary from as low as 50
mph to upwards of 200 mph. With this extreme change
in conditions, performance of a wing can vary greatly.
0.5 Therefore, it becomes important to quantify the effects
that speed, or Reynolds number, has on performance of
a race-car wing. Figure 3 shows the effect of Reynolds
0 number on lift curves and drag polars for the UIUC700I
-5 0 5 10 15 20
α (deg) configuration. As Reynolds number increases, lift
3 coefficient increases and drag coefficient decreases.
Typical increases in lift coefficient for an increase in
6 6
Reynolds number from 0.7 x 10 to 1.1 x 10 averaged
2.5 2.5%, while an increase in Reynolds number from 1.1 x
6 6
10 to 1.3 x 10 caused an average increase in lift of
1.9%. Decreases in total drag for the same changes in
2 Reynolds number were, on average, 2.3% and 1.2%,
respectively. In the sense of an open-wheel race car, a
desired effect would be a decrease in lift as speed (Re)
increases, since high speeds typically occur on
CL

1.5
straightaways where downforce is not needed. From an
aerodynamic standpoint, as the present data show, this
1 is not possible with the current fixed wing configurations
allowed in open-wheel racing. An expected result is an
increase in induced drag with higher speed, since the
0.5 wing is operating at a higher lift coefficient. But the
results show a reduction in overall drag. For this wing
design the results indicate that the decrease in profile
0
0 0.25 0.5 0.75 1
drag due to Reynolds number effects is larger than the
CD associated increase in induced drag.

Figure 3: CL vs. α and CL vs. CD for UIUC700I at δf =

14 deg and h/c=0.3
 Figure 4 (included at end for draft) presents the
downstream flowfield measurements for the UIUC700I δf = 4 deg
δf = 14 deg
configuration at α = 6 deg, δf = 14 deg, and Re = 1.1 x δf = 24 deg
6
10 . Note the formation of two large trailing vortices 3
rolling up from the endplate edges, as well as a smaller
vortex forming off of the “cutout.” This smaller vortex is
a result of the large change in lift that occurs between 2.5
the smaller chord root region and longer chord tip and
does not appear in measurements taken with the
UIUC700H, which had the Formula One flap (Figure 5 – 2
included at end for draft). The measurements taken in
the region between the ground plane and wing near the

CL
root of the wing caused concern. It is likely that this 1.5
feature, which appeared in all flowfield measurements
taken with varying strength, is a result of the wing/wind-
tunnel floor/ground plane juncture. It is difficult to 1
ascertain what effect this region had on the rest of the
6
flowfield. Data taken at Re = 0.7 x 10 and Re = 1.3 x
6 0.5
10 of the UIUC700I show no significant movement of
the trailing vortex system with changes in Reynolds
number.
0
-5 0 5 10 15 20
FLAP DEFLECTION – During a race, particularly in the α (deg)
CART series, the ability to change the front-wing flap 3
angle is the only way to adjust performance of the car
barring a complete change of the front nose cone.
Therefore, the ability to quantify the effects of changing 2.5
flap deflection on wing performance become very
important. As described in Table 2, each configuration
was tested at flap deflections of 4, 14, and 24 deg. 2
Figure 6 shows the lift curves and drag polars for the
6
UIUC700I configuration at Re = 1.1 x 10 for these flap
CL

deflections. The average, minimum, and maximum 1.5

increases in lift coefficient for each step change in flap
deflection are summarized in Table 4. These results
show that as flap deflection is increased by 10 degrees, 1
at a constant angle of attack, an average increase of 0.5
in lift coefficient occurs. Figure 6 also shows that
although there is a large shift in the lift curves, there is 0.5
no appreciable change in overall drag, when compared
at a constant CL. Because the values of overall drag are
essentially constant at similar lift coefficients and 0
0 0.25 0.5 0.75 1
constant Reynolds number, it is inferred that the CD
changes in profile drag are small with respect to the
overall drag, and that performance of a front wing is Figure 6: CL vs. α and CL vs. CD for UIUC700I at Re =
dominated by induced drag. 1.1 x 106 and h/c = 0.3
wing at the endplate moves closer to the root by
Table 4: Increase in lift with changes in flap approximately 1.2-cm, as well as 1.2-cm further from the
deflection of UIUC700I ground. The movement of the vortex with increasing flap
deflection has two competing effects. First, the
Change in δf (deg) Avg ∆CL Min ∆CL Max ∆CL increased lift at high flap deflections will cause more
4 to 14 0.5540 0.4492 0.6842 induced drag, resulting in a loss of power from the
engine, which decreases top speed. Additionally, more
14 to 24 0.4472 0.3685 0.5360
air might be forced through the underbody because of
the vortex being moved closer to the root of the car. An
increase in flow to the underbody channels of a race car,
Figure 4 and Figure 7 (included at end of draft) plot particularly in the CART series, will manifest itself as an
the flowfield measurements for the UIUC700I at α = 6 increase in downforce, and higher cornering speeds can
deg with flap deflections of 14 and 24 deg, respectively. be attained. The wing/floor/ground juncture region also
Lift coefficients for these cases are 1.95 for δf = 14 deg appears to grow with increasing flap deflection. This is
and 2.37 for δf = 24 deg. Comparison of these plots likely due to the higher speeds over the suction side of
show that the trailing vortex from the suction side of the the wing that result from the increased loading provided
 -4

UIUC700I
UIUC700H
-3
3

UIUC700I, 2y/b=0.22
-2
UIUC700I, 2y/b=0.64
2.5 UIUC700I, 2y/b=0.92

Cp
UIUC700H, 2y/b=0.22
δf = 24° -1 UIUC700H, 2y/b=0.64
UIUC700H, 2y/b=0.92
2

δf = 14° 0
CL

1.5
1
0 0.25 0.5 0.75 1
δf = 4°
x/cm
1
Figure 9: Main element pressure distributions for
UIUC700I and UIUC700H at α = 6 deg, δf = 14 deg, Re
0.5 = 1.1 x 106, and h/c = 0.3
the UIUC700F and G comparison are not shown, but are
0 consistent with the UIUC700I and H configurations. As a
-5 0 5 10 15 20 result of increasing the wing planform area by 4.5%, the
α (deg) average increase in CL is 0.15. This effect, however, is
3
not explained by the increase in area. At first glance, it
might be expected that an increase in area without an
increase in span would result in higher induced drag and
2.5
lower lift, which are results of increased aspect ratio.
But Figure 8 shows that increased lift occurs. This is
largely in part due to the increase in local chord that
2
occurs over much of the flap. The main element can be
loaded more because of this, and as a result lift is
increased. Examination of the pressure distributions of
CL

1.5
Figure 9 confirms this result. The drag polar of Figure 8
also shows that there is no significant effect on drag due
to a change in wing planform, again suggesting that
1
profile drag is insignificant compared to induced drag.

0.5
The downstream flowfield measurements for the
UIUC700H are shown in Figure 5. Comparing these
measurements with those for the UIUC700I (see Figure
0 4), it is seen that the suction side vortex is stronger,
0 0.25 0.5 0.75 1 moves closer to the root and further from the ground with
CD
the Formula One-style flap (UIUC700H). An increase in
Figure 8: CL vs. α and CL vs. CD for UIUC700I and induced drag would be the expected result; however, no
UIUC700H at Re = 1.1 x 106 and h/c = 0.3 significant increase is shown.
by the flap deflection increase. Since the complete area ENDPLATE PLANFORM – The overall effect that
of this flow was not captured in the seven-hole probe endplate planform area has on performance of the front
scans, it is not known how this affects the rest of the wing is, perhaps, the most interesting. Although racing
flowfield. rules in both CART and Formula One are very restrictive
in terms of endplate size and position, it is still essential
FLAP PLANFORM – In addition, the effect of flap that a car manufacturer maximizes performance. Figure
planform was compared. Here, the two different flap 10 compares the UIUC700F with the UIUC700H, both
planforms were tested with the same endplate. having the Formula One flap and Formula One and Indy
Comparisons were made between the UIUC700I and front-wing endplates, respectively. When the Indy
UIUC700H, both having the Indy-style front endplate, endplate was used, lift coefficient increased by and
and the UIUC700F and UIUC700G, both having the average of 0.0958 at constant α, while drag at constant
Formula One-style front endplate. Lift and drag data for CL decreased by an average of 13.7%. Hence, adding
the I and H configurations are shown in Figure 8, while endplate area produces a more efficient wing. This
selected pressure distributions over the main element for effect can also be partially explained by the positioning
both configurations are shown in Figure 9. Results for of the wing within the endplate, and it is difficult to
 drag of typically 1-2% for 50% changes in Reynolds
UIUC700H number. Changes in flap deflection of 10 degrees, on
UIUC700F average caused an increase in CL of 0.5, while drag was
largely unaffected indicating that induced drag is a
3
dominant factor in race-car wing design. A study of
changes in flap planform show significant increases in
lift, largely due to the increased loading that occurs on
2.5
the main element. Examination of the results from
changing endplate planform reveal that endplate design
is quite significant in the performance of a race-car front
2
wing. When endplate area is increased, overall lift
increases while there is a significant reduction in drag.
CL

1.5
ACKNOWLEDGMENTS

1
The support of Ford Motor Company, AVT
Motorsports Technology Group with technical monitors
John LaFond, Paul Carrannanto, Fredrico Hsu is
0.5 gratefully acknowledged. Also, the early input of Frank
Hsu leading to the initiation of this work is greatly
appreciated.
0
-5 0 5 10 15 20 REFERENCES
α (deg)
3
1. Hurst, D.W., “Modern Wind Tunnel Testing of Indycars,”
SAE Paper 942497, 1994 Motorsports Engineering
Conference Proceedings, Vol. 1, pp. 151-159, December
2.5 1994.
2. Katz, J., “Aerodynamic Model for Wing-Generated Down
Force on Open-Wheel-Racing-Car Configurations,” SAE
2 Paper 860218, February 1986. SAE Transactions, Vol.
95, pp. 129-137, 1986.
3. Coiro, D.P., F. Nicolosi, A. Amendola, D. Barbagallo, L.
Paparone, S. Beccio, P. Castelli, and S. Limone,
CL

1.5 “Experiments and Numerical Investigation on a Multi-

Component Airfoil Employed in a Racing Car Wing,” SAE
Paper 970411, Topics in Vehicle Aerodynamics, pp. 221-
1 231, 1997.
4. Berchak, M.J. and Camosy, M.W., “Comparison of Full-
Scale Wing Wind Tunnel Test to Scale-Model Test for
Open Wheel Race Cars,” SAE Paper 942495, 1994
0.5 Motorsports Engineering Conference Proceedings, Vol. 1,
pp. 141-149, December 1994.
5. Ranzenbach, R. and Barlow, J.B., “Two-Dimensional
0 Airfoil in Ground Effect, An Experimental and
0 0.25 0.5 0.75 1
CD Computational Study,” SAE Paper 942509, 1994
Motorsports Engineering Conference Proceedings, Vol. 1,
Figure 10: CL vs. α and CL vs. CD for UIUC700F and pp. 241-249, December 1994.
6. Ranzenbach, R. and Barlow, J.B., “Cambered Airfoil in
UIUC700H at δf = 14 deg, Re = 1.1 x 106, and h/c = 0.3 Ground Effect - An Experimental and Computational
separate the two without further study. As shown in Study,” SAE Paper 960909, 1996 Motorsports Engineering
Conference Proceedings, Vol. 1, pp. 269-276, December
Figure 1, the positioning of the wing is closer to the edge
1996.
of the Formula One endplate than with the Indy front 7. Ranzenbach, R., Barlow, J.B., and Diaz, R.H., “Multi-
endplate. The lack of endplate area between the wing’s Element Airfoil in Ground Effect - An Experimental and
suction surface and the freestream allows the trailing Computational Study,” AIAA Paper 97-2238, June 1997.
vortex to roll-up earlier, which can result in more induced 8. Gopalarathnam, A., Selig, M.S. and Hsu, F., “Design of
drag and less lift. High-Lift Airfoils for Low Aspect Ratio Wings with
Endplates,” AIAA Paper 97-2232, June 1997.
9. Noe, S.C., “Force Balance Measurements of Wind-Turbine
CONCLUSION Airfoil Performance with Simulated Leading-Edge Ice
Accretions,” Master’s Thesis, University of Illinois at
The results from this study show that many Urbana-Champaign, August 1996.
simplified factors in the design and setup of open-wheel 10. Jasinski, W.J., Noe, S.C., Selig, M.S. and Bragg, M.B.,
race-car front wings have a significant effect on “Wind Turbine Performance Under Icing Conditions,”
performance. It was shown that Reynolds number Journal of Solar Energy Engineering, Vol. 120, No. 1,
February 1998, pp. 60-65.
effects were least significant, with changes in lift and
11. Rediniotis, O.K., Hoang, N.T. and Telionis, D.P., “The
Seven-Hole Probe: Its Calibration and Use,” Forum on
Instructional Fluid Dynamics Experiments, Vol. 152, pp.
21-26, June 1993.
12. Zilliac, G.C., “Calibration of Seven-Hole Pressure Probe
For Use in Fluid Flows with Large Angularity,” NASA TM
102200, Decemeber, 1989.
 UIUC700I
α = 6 deg, δf = 14 deg, Re = 1.1 x 106
0

UX (ft/sec)
300
-5 290
280
270
260
-10 250
240
230
Ground Plane

220
Y (in)

210
-15 200
190
180
170
-20 160
150
140
130
-25 120
110
100

-30
25 20 15 10 5 0
Z (in)

Figure 4: Downstream flow measurements (Ux, Uy, Uz) of UIUC700I configuration as viewed from upstream.
 UIUC700H
α = 6 deg, δf = 14 deg, Re = 1.1 x 106
0

Ux (ft/sec)
300
-5 290
280
270
260
-10 250
240
230
Ground Plane

220
Y (in)

210
-15 200
190
180
170
-20 160
150
140
130
-25 120
110
100

-30
25 20 15 10 5 0
Z (in)
Figure 5: Downstream flow measurements (Ux, Uy, Uz) of UIUC700H configuration as viewed from upstream.
 UIUC700I
α = 6 deg, δf = 24 deg, Re = 1.1 x 106
0

UX (ft/sec)
300
-5 290
280
270
260
-10 250
240
230
Ground Plane

220
Y (in)

210
-15 200
190
180
170
-20 160
150
140
130
-25 120
110
100

-30
25 20 15 10 5 0
Z (in)

Figure 7: Downstream flow measurements (Ux, Uy, Uz) of UIUC700I configuration as viewed from upstream.