1

Aerodynamics of a wing in ground eect in generic

racing car wake ows
M D Soso* and P A Wilson
School of Engineering Sciences, University of Southampton, Southampton, UK
The manuscript was received on 17 March 2005 and was accepted after revision for publication on 7 September 2005.
DOI: 10.1243/095440705X69632

Abstract: In an eort to provide more detailed insight into the aerodynamic factors that may
inuence the creation of overtaking opportunities in modern open-wheeled racing series, a set
of wind tunnel experiments was initiated in the moving ground facilities at the University of
Southampton. To generate data typical of one car following another, a single-element wing in
ground eect was tested downstream of a blu body that incorporated a diuser and rear
wing. The tests included variations in the height and angle of attack of the wing, while data
collection was achieved via force and pressure measurements, ow visualization and oweld
surveys. The results were then compared with baseline data that were obtained without the
presence of the blu body. It was found that, while behind the upstream body, the wing
experienced a decrease in its downforce values, with the amount of downforce lost depending
on its height above the ground. It was also shown that more downforce was lost from sections
closer to the mid-span of the wing than was the case from sections closer to the tips of the wing.
Keywords: wing, ground eect, overtaking, diuser, blu body, aerodynamics, racing

1 INTRODUCTION
The issue of the lack of frequent overtaking
opportunities typical of open-wheeled racing series
such as Formula 1 has received a great deal of
publicity in recent times. It is common to hear racing
commentary suggesting that the aerodynamics of
a following car have been severely aected when
travelling in the ow produced by a leading car of
similar specications. In order further to investigate
this particular situation, and to provide information
on strategies that could possibly be exploited to
develop more robust aerodynamic packages, a series
of experiments were initiated to model the conditions that may be experienced during a typical
race, when one vehicle follows in the wake produced
by another.
Previous studies of this type of vehicle interaction
were carried out as early as the 1970s for the NASCAR
racing series in America. Romberg et al. [1] investigated the aerodynamic forces on wind tunnel models
of stock cars of the time. The experiments were

* Corresponding author: 2 Summerton Place, Chipping Norton,

Oxon OX7 5AZ, UK. email: soso_md@yahoo.com

JAUTO56 IMechE 2006

carried out for a variety of drafting and passing

positions. Tunnel restrictions limited the maximum
distance between the models to two car lengths.
Within this range, the force results that were presented showed that there was as much as a 37 per
cent reduction in the drag of the following car.
Signicant changes in the distribution of the lift
forces at the front and rear axles of both models were
also reported.
Howell [2] preformed wind tunnel investigations
into the resulting lift, drag, and pitching moment
forces occurring when two Can-Am racing car models
were in close proximity to each other. The main
objective was to highlight the fact that the following
car could experience sucient aerodynamic changes
to cause it to overturn. Data were provided to
show that the following car experienced reduced
drag and downforce values, along with increased
pitching moment values, when in the wake of the
car ahead.
For the case of open-wheeled racing cars, Dominy
[3] carried out a detailed investigation of the aerodynamic eects occurring when quarter-scale models
of a 1989 Formula 1 car followed in the wake of an
identical model. Measurements on the following car
were taken with it placed on a moving ground, while
Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

M D Soso and P A Wilson

the leading car was positioned less than a car length

ahead, on the oor of the test section. In the
measurements of lift and drag that were presented,
it was shown that, when fully immersed in the wake
of the leading car, the downforce on the following
car was reduced by 36 per cent and the drag by 23 per
cent. As the lateral oset between the two models
increased, both variables commenced to recover to
their freestream values, with the drag taking much
longer to do so than the downforce. Data were also
provided to show that the centre of pressure changed
from 68 per cent wheelbase to 90 per cent wheelbase
when directly behind the leading car. This result
implied that more downforce was lost from the front
of the car than from the rear.
Experimental tests aimed at investigating the downstream wake of vehicle models and vehicle-like blu
bodies were carried out by a number of authors [48].
The results that were obtained typically highlighted
the existence of a pair of counter-rotating vortices
in the ow eld. The vortices induced an upwash or a
downwash, depending on the manner in which they
rotated, which in turn depended on the particular
conguration.
Published studies of racing car wings in ground
eect were the result of experiments and computations in which the wing was placed in oncoming ow conditions that were undisturbed and
of low turbulence intensity. Ranzenbach and Barlow
[911] carried out a series of two-dimensional, comparative computational and experimental studies
of a symmetric and a cambered airfoil in ground
eect. The work centred on the investigation of the
force coecients as the height of each airfoil was
varied. The authors also investigated the eect of a
moving ground in the computational simulations.
Their results showed that, as the distance between the
airfoil and the ground was reduced, the downforce
experienced by the wing increased to a maximum,
and then decreased with further height reduction.
The drag was shown to rise monotonically for the
case of the cambered airfoil. With regard to the
ground boundary, they showed that the airfoil experienced greater downforce coecients when that
particular surface moved with the freestream velocity,
as opposed to when it was stationary.
Knowles et al. [12] carried out an experimental and
limited computational study of a GA(W)-1 wing with
endplates, in ground eect. The experimental facility
was equipped with a rolling road. They varied the
angle of attack of the wing at dierent heights above
the ground to generate lift curves, while recording
force and pressure data. They obtained results showing that the lift curve slope of the wing increased,
Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

while the angle of stall was reduced, as the ground

was approached. The computational comparisons,
which were carried out with a two-dimensional airfoil in a panel method, were limited to pressure
distribution plots. The plots showed fairly good
agreement at low angles of attack and large ride
heights.
Jasinski and Selig [13] performed an experimental
ground eect study of wing and endplate combinations that were representative of Champ Car
and Formula 1 front wings at the time. A rolling road
was not used. They specically looked at the eect
of Reynolds number, ap deection, ap planform
shape, and endplate shape. The measurements that
were taken included forces, pressures, and oweld
data. They obtained results showing that the drag
coecient at constant C was relatively unaected
L
while increasing ap deection, and that ap planform and endplate shape had a large eect on the
aerodynamics of the wing. Their oweld data clearly
highlighted the presence of two vortices, the larger
emanating from the bottom of the endplate and the
smaller from the top. The Reynolds number was
found to have the least eect on the aerodynamics.
Zerihan and Zhang [1416] performed extensive
experimental investigations of a single element and
then a double-element wing in ground eect in
the moving ground facilities at the University of
Southampton. The height above the ground of both
wings was varied, as was the angle of attack in
some tests. Among the important conclusions to be
drawn from the single-element wing investigation
were that an increase in the angle of attack of the
wing engendered greater maximum downforce, while
reducing its sensitivity to changes in downforce,
and that trailing edge separation increased with
decreasing ride height. Transition xing was shown to
produce less downforce on the wing and to produce
force reduction at a greater ride height.
The current study focuses on assessing the
aerodynamic performance of a wing in ground
eect, when placed in dierent oncoming ow conditions. The dierent ow states were intended to be
representative of racing scenarios in which an openwheeled racing car was travelling in undisturbed
freestream ow, and in the ow generated by a leading racing car. Focus was given to a wing in ground
eect because it was the most forward part of
the vehicle that would experience changes in the
oncoming ow, because the majority of the downforce was typically lost from this component, and
because the remainder of the car operated in the
wake that it generated.
JAUTO56 IMechE 2006

Aerodynamics of a wing in ground eect

2 OUTLINE OF EXPERIMENT
2.1 Facilities
The experimental tests were carried out in the
large-scale moving ground wind tunnel facilities at
the University of Southampton. Flow visualization
images, along with force and pressure measurements were gathered in the 2.11.5 m tunnel, while
laser doppler anemometry (LDA) measurements
were taken in the 3.52.5 m R. J. Mitchell tunnel.
Both tunnels are of a closed-circuit, single-return
design, with boundary layer suction mechanisms
located just ahead of the moving ground belt.
2.2 Models and installation
The race scenario of one car following another was
replicated generically in the wind tunnel by the construction and use of simple experimental models that
were selected to represent the salient characteristics
of each particular vehicle. Starting with simple models
will allow for more complexity to be added in the
future, as a greater understanding of the inherent
interactions develops. The component of the following car that was chosen to commence the study was
the front wing. This device was selected because it
is the most forward part of the vehicle that would

experience changes in the oncoming ow, because

the majority of the downforce is typically lost from
this component, and because the remainder of the
car operates in the wake that it generates.
The front wing was idealized as a single-element
LS(1)-0417 (GA(W)-1) wing, scaled to 40 per cent of
the dimensions of the front wing of a typical F1 car
for the year 2002. The resulting chord and span were
220 and 550 mm respectively. Endplates of 5 mm
thickness were attached to each end of the wing.
The wing was specically designed to pivot at the
quarter-chord point, while allowing the endplates to
remain parallel to the ground at all times. It was
thought logical to establish baseline data with this
conguration before attempting to investigate multielement devices. A second, pressure-tapped model
was also constructed. The pressure taps were located
in the direction of the oncoming ow, at the spanwise
stations of 2z/b=0.09, 0.49, and 0.89. There were
44 taps at each station, 24 on the suction surface and
20 on the pressure surface.
The conguration adopted for the 2.11.5 m wind
tunnel is shown in Fig. 1. The wing was installed
above the moving ground by attaching it to an overhead balance via movable struts. The struts had
machined slots along their top ends that allowed a
degree of freedom in the vertical direction in order

Fig. 1 Experimental conguration adopted in the 2.11.5 m wind tunnel

JAUTO56 IMechE 2006

Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

M D Soso and P A Wilson

to raise and lower the model. Machined blocks of

varying thickness were then placed between the
bottom of the endplates and the top of the rolling
road surface to set specic ride heights.
The representation of the leading car developed
in two stages. Firstly, a wing without endplates was
used to idealize the upper elements of a typical rear
wing. This wing had a chord of 140 mm and a span
of 400 mm, as it was scaled to 40 per cent of the
dimensions of the rear wing of a typical 2002 F1 car.
Following initial tests with this device, a generic
diuser was then incorporated into a blu body
shape so as to improve the model further. The rear
wing was then attached to the diuser blu body via
endplates. The height and lateral position of the blu
body were adjustable, as was the angle of the diuser
ramp. Its width was 400 mm, as it was scaled to be
40 per cent of the actual width of a diuser of a
typical 2002 F1 car. The entire length of the blu
body was 900 mm. The model was mounted to a
ground board that was positioned just ahead of the
rolling road and suction box in the wind tunnel test
section. The distance from the base of the blu body
to the leading edge of the wing was 2160 mm
(approximately 1.5 car lengths at full scale). A sketch
of both models in the 2.11.5 m test section is
provided in Fig. 2.
In the larger 3.52.5 m facility, where the LDA
system was permanently located, both the test wing
and the blu body were mounted using the same
strategy. The only dierence was an increase in
the distance between the two models owing to the
longer moving ground belt system in that tunnel. The
distance from the base of the blu body to the leading edge of the wing was 3100 mm in this tunnel
(approximately 2.2 car lengths at full scale).

2.3 Methodology and uncertainties

The experimental tests involved placing the wing in
dierent oncoming ow conditions produced in the
wind tunnel facilities. The baseline ow condition,
FC1 or clean air, was the uniform freestream that
entered the test sections. The remaining ow conditions were categorized as dirty air, since the bodies
that were placed upstream of the wing disturbed the
oncoming ow.
The dirty air ow condition was developed in a
series of incremental steps. Each step was designed
to add a physical component that would produce a
wake more closely resembling that of a generic racing
car. As a result, the following subdivisions were used:
1. FC2 the wing used to simulate the rear wing of
a typical open-wheeled racing car was placed
upstream of the test wing.
2. FC3 the blu body which incorporated a diuser
and a rear wing was placed upstream of the test
wing.
Owing to limits placed on the time available in the
wind tunnel facility, it was only possible to investigate FC1 and FC3 in an in-depth manner. FC2 was
investigated to a lesser extent.
The physical procedure undertaken was to vary the
height and angle of attack of the test wing when
placed in the dierent oncoming ow conditions.
The height of the endplate of the wing above the
tunnel oor typically ranged from 2 to 169.6 mm.
The angle of attack ranged from 5 to 30, in order
to ensure that the region of stall of the wing was
covered. The investigations were carried out at a
constant wind tunnel dynamic pressure of 25 mm
water, which corresponded to a freestream velocity of

Fig. 2 Partial schematic showing the relative locations of the diuser blu body and the test
wing in the 2.11.5 m wind tunnel
Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

JAUTO56 IMechE 2006

Aerodynamics of a wing in ground eect

20 m/s. Owing to variations in the ambient pressure

and in the temperature of the wind tunnel, the
Reynolds number based on wing chord varied from
300 000 to 309 000.
For some test conditions, the angle of the diuser
ramp was changed in order to generate downstream
ow typical of high-angle and low-angle diusers. For
other test conditions, the height or lateral position
of the diuser was varied, while keeping the ramp
angle constant. In the baseline conguration, the
model was positioned 60 mm above the ground
board and incorporated a 16.7 ramp. Techniques
such as ow visualization, force measurements,
pressure tapping, and LDA were then used to extract
the relevant data from the wing and its oweld.
For the ow visualization studies, a mixture of
invisible blue uorescent pigment, paran, and oleic
acid was used. It was then applied to the model with
a paint roller or paintbrush. When dried, the mixture
formed a white aky coating which highlighted the
ow of air around the object.
Forces were measured on the overhead balances
present in each wind tunnel. Typically, a series of
75 samples was taken at each ride height. An average
was then calculated and output by the controlling
software. Uncertainties were calculated following the
methods outlined by Coleman and Steele [17]. The
maximum uncertainties in the downforce and drag
coecients were estimated as being 0.002 and
0.0004 respectively.
Pressures were measured by connecting the tubes
that emanated from the taps to a ZOC pressure transducer system [18]. The ZOC was located externally

to the tunnel test section in order to minimize the

eect of temperature changes on the transducer.
For each ride height investigated, 22 values were
recorded at each tap. An average was then calculated
to provide the most representative value. Using the
same procedure as that used for the force coecients,
the uncertainty in the pressure measurements was
estimated to be 0.01.
The LDA system at the University was previously
quoted as producing uncertainties in u/U and v/U
2
2
of 0.005 [16].

3 RESULTS AND DISCUSSION

3.1 Diuser oweld
Since FC3 was representative of the conditions that
most closely resembled the ow generated by a leading open-wheeled racing car, an outline of its main
features will be presented for the baseline case. Flow
visualization was performed on the ramp and endplate region of the diuser blu body. The resulting
pattern, which is illustrated in Fig. 3, indicated the
presence of vortex ow. The vortex ow was highlighted by the swirling lines that trailed along the
edge of the ramp, close to the endplate, as was
previously described by Senior and Zhang [19]. The
ow was found to be symmetric about the blu
body centre-line and showed the characteristics of
high-angle diusers as outlined by Rhurmann and
Zhang [20].

Fig. 3 Swirling lines on the diuser ramp, indicating the presence of vortex ow
JAUTO56 IMechE 2006

Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

M D Soso and P A Wilson

Smoke trails, which were released from a portable

wand, showed that the ow in the vicinity of the rear
of the blu body, close to the diuser sideplates, was
sucked in towards a plane that coincided with the
centre-line of the test section. The ow then progressed downstream, seemingly concentrated in the
middle of the test section as it did so. There was also
a signicant increase in the amount of audible noise
associated with this conguration, an indication of
the generation of turbulent ow [21].
Floweld tests were carried out in the R. J. Mitchell
facility, as it was equipped with a three-component
LDA system. The objective was to provide a more
detailed map of the wake of the blu body. The
measurements, which were taken 3l downstream of
the blu body, are presented in Figs 4(a) to 4(d) for
clarity. It was only possible to extract two components
of the velocity owing to a failure of some of the
system hardware. The plots reveal a variation in the
freestream and vertical velocity components while

progressing through a plane perpendicular to the test

section centre-line.
The wake decit highlighted by plots of u/U
2
is seen to have been greater at locations closer to the
centre of the tunnel (2z/b=0) than at locations
closer to where the tips of the wing would have been
positioned (2z/b=1). The plots for v/U highlight
2
an upwash close to the centre of the tunnel. Away
from the centre-line, the upwash gradually decreased,
transitioning to a downwash by the location of
2z/b=0.80. At 2z/b=0 the maximum turbulence
intensity was found to be 11 per cent while at
2z/b=0.87 it was found to be 7.4 per cent.
3.2 Eect of upstream bodies
The aerodynamic changes that the upstream bodies
induced on the downstream wing will be presented
in the form of a comparison with the baseline data
that were obtained from the clean air case. Figure 5

Fig. 4 Velocity proles 3l downstream of the diuser blu body in the 3.52.5 m wind tunnel
Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

JAUTO56 IMechE 2006

Aerodynamics of a wing in ground eect

Fig. 5 Experimental downforce coecients in ground

eect in the three ow conditions in which they
were measured

shows plots of the downforce coecients while varying the height of the wing in all ow conditions. Its
angle of attack was held constant at 5. On initial
observation, it is clear that there were successive
decreases in downforce as the oncoming ow
progressed from FC1 to FC2, and nally to FC3.
It can be deduced that more downforce was lost
at greater ride heights than was the case at lower ride
heights. For example, at h /c=0.833, the downforce
r
value for FC2 was approximately 13 per cent less than
that of FC1, while the downforce value for FC3 was
approximately 33 per cent less than the same FC1
baseline value. At h /c=0.401, the corresponding
r
losses in FC2 and FC3 were 7.6 and 25 per cent
respectively. At h /c=0.204, the corresponding losses
r
in FC2 and FC3 were 4.8 and 18 per cent respectively.
The shape of the curves for FC1 and FC2 is consistent with previously published data. There was the
characteristic increase in downforce to a maximum
value as the ride height was reduced to a certain
point (h /c#0.09). Below this height, the downforce
r
then began to decrease, and continued to do so for
further height reductions. The curve for FC3 followed
a similar trend of increasing downforce between the
ride heights of h /c=0.83 and h /c#0.09. At ride
r
r
heights below this range, however, a new characteristic was revealed. The region of downforce decrease
was halted by an abrupt increase in downforce at
very low ride heights. The second region of downforce increase produced coecients that were higher
than the maximum that was achieved during the rst
region of force increase. Tests that were carried out
at a higher freestream velocity indicated that the
second region of force increase commenced at a
slightly greater ride height.
JAUTO56 IMechE 2006

The variation in the drag coecients with ride

height in the three ow conditions is presented in
Fig. 6. The plots indicate that the ow generated by
the upstream bodies caused an increase in the drag
of the downstream wing. Above h /c=0.4, the highest
r
drag values occurred in FC3, while below this point
FC2 produced the highest. At ride heights below
h /c=0.153 in all ow conditions, uctuations in
r
the values became evident. It is not entirely certain
why this characteristic occurred. Flow visualization
images did, however, highlight the presence of
recirculating regions of ow at the junction of the
wing and endplate. The regions were found to
change position and size with ride height changes.
It is plausible to assume that this movement may
have had some eect on the drag values.
An angle of attack variation was performed in FC1
and FC3 to highlight any changes that may have
occurred. The results are plotted in Fig. 7 for two ride
heights, h /c=0.833 and h /c=0.153. It can be seen
r
r
that, throughout the useful angle of attack range, the
wing generated less downforce in FC3 than it did in
FC1. Furthermore, previous authors reported that the
lift (downforce) curve slope of a wing increased with
decreasing ride height. It has now become evident
that this trend also existed in dirty air conditions.
The results of Fig. 7 also highlighted the fact that
more downforce was lost at greater ride heights than
at lower ride heights.
The dirty air conditions also had the potential to
alter the stall characteristics of the wing. At h /c=
r
0.833 in FC1 there was an abrupt stall at 23. However, in FC3, the stall became more gradual. This result
suggested that the boundary layer characteristics

Fig. 6 Experimental drag coecients in ground eect

in the three ow conditions in which they were
measured
Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

M D Soso and P A Wilson

Fig. 7 Lift curves at h /c=0.153 and 0.833 in ow

r
conditions FC1 and FC3

of the wing could potentially be altered in the dirty

air ow.
A series of measurements was taken while moving
the diuser blu body laterally away (indicated by
2z/b) from its original position in front of the test
wing. The resulting downforce curves at h /c=0.153
r
are plotted in Fig. 8. It can be seen that throughout
the prestall angle of attack range, the least downforce was generated when the blu body was directly
in front of the wing (2z/b=0). As the blu body
was moved laterally away, the downforce gradually
recovered to the freestream values (as was similarly
reported in reference [3]), and in the case of
2z/b=1.5 it surpassed them.
Flow visualization was performed in FC1 and FC3
in order to ascertain the physical eects of change
in the oncoming ow conditions. Figure 9(a) shows
the ow over the suction surface of the wing in FC1.

Fig.8 Downforce curves at h /c=0.153 while moving

r
the blu body laterally away from its original
position in front of the wing
Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

Observation of the pattern shows that there was a

laminar separation bubble across the majority of the
span of the wing (a). At mid-span, it can be deduced
that the bubble trapped some of the ow visualization
uid while it was drying (b), thereby causing an
additional obstruction to the ow, which then went
on to cause premature trailing edge separation (c).
Premature separation was also aided by a blob of
unmixed solution (d). Close to each endplate, at the
trailing edge of the wing, a region of recirculating
bubble ow can be identied (e).
In comparison, Fig. 9(b) shows the ow over
the suction surface of the wing in FC3. It can be
deduced that there is now a dramatic dierence in
the surface ow when compared to FC1. At a considerable distance either side of mid-span the laminar
separation bubble has disappeared (f) as the ow was
turbulent from the outset. The ow also appeared to
have been antisymmetric, as evidenced by the comparable sizes of the regions labelled (g) and (h).
The asymmetry was also reected in the slightly
dierent shape of the separation region at (i) and (j).
The recirculating bubble regions (k, l), which now
have distinctly dierent shapes when compared with
each other, are still observed at either endplate, close
to the trailing edge.
The ow on the pressure surface of the wing in FC1
can be seen in Fig. 10(a). A large region of laminar
ow is visible over the majority of the surface. Also
present was a laminar separation bubble (m), which
formed across the entire span.
In comparison, Fig. 10(b) shows the ow over
the same surface in FC3. There was again another
striking dierence when compared with FC1. The
laminar separation bubble was again eliminated
from a considerable distance either side of the midspan of the wing (n), but was still clearly visible
towards the tips (o). In contrast to the overall ow
on the suction surface, the ow on the pressure
surface appeared to be symmetric.
A comparison of the surface pressure distributions
in FC1 and FC3 can be seen in Fig. 11. Measurements
were taken at the ride heights of h /c=0.833, 0.401,
r
0.153, and 0.077, but only those at h /c=0.153 are
r
shown. The data indicate that changing the ow
from FC1 to FC3 resulted in a decrement in the
pressure distribution at each station investigated.
The decrement appeared to be greater at stations
closer to the centre of the wing than at stations
further away. Also evident was the fact that the
majority of the loss occurred from the suction
surface of the wing, especially at 2z/b=0.09 and
0.49. These results also held for the other ride heights
investigated.
JAUTO56 IMechE 2006

Aerodynamics of a wing in ground eect

Fig. 9 Comparison of the suction surface ow in (a) FC1 and (b) FC3 for h /c=0.153. Leading
r
edge uppermost in image

The pressure distribution data also show the existence and disappearance of the separation bubbles.
A typical location of the bubble can be pinpointed
by a plateau-like region followed by a steep drop,
both of which produce an area that does not appear
to t with the natural curvature of the plot. For
example, the region bounded by x/c#0.45 and
x/c#0.6 at 2z/b=0.09. When the ow was changed
to FC3, the region just described vanished. A similar
scenario existed on the pressure surface between
x/c#0.65 and x/c#0.75.
A more detailed analysis of the same plot can
provide the reader with further insight into the
aerodynamic changes experienced by the wing. The
presence of a separation bubble on the suction
surface in FC1 has just been highlighted. This surface
ow feature implied that the boundary layer prior to
its existence was laminar, while the boundary layer
after the bubble was turbulent. A comparison with
the corresponding curve for FC3 will show that, in
general, the greatest dierence in values between the
two plots occurred between x/c#0.01 and x/c#0.6.
As the bubble had disappeared in this ow condition,
it was evident that the ow prior to that point did
JAUTO56 IMechE 2006

not support its existence. That is, the boundary layer

was now turbulent. It can be inferred that the greatest
loss in pressures on the suction surface occurred
because of the elimination of an extensive region of
laminar ow when progressing from FC1 to FC3. In
contrast, between x/c#0.6 and x/c#0.95 the loss
in pressure was not as great, suggesting that the
turbulent boundary layer was not that sensitive to
the change in the characteristics of the oncoming
ow.
The two-dimensional sectional downforce coefficients were estimated for each station, at each
ride height at which measurements were taken.
The trapezium rule was used to integrate the distributions between x/c=0 and x/c=0.9. Owing to
manufacturing constraints, no pressure tap was
located beyond x/c=0.9 on the wing pressure surface
and beyond x/c=0.95 on the suction surface. The
integration was therefore not carried out for the last
10 per cent of each section. Table 1 summarizes the
changes as the oncoming ow evolved from FC1 to
FC3. It is clear that, at each ride height, most lift was
lost at 2z/b=0.09. This was followed by the section
at 2z/b=0.49, then 2z/b=0.89.
Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

10

M D Soso and P A Wilson

Table 1 Sectional downforce coecient values

for each of the ride heights investigated
2z/b=0.09

2z/b=0.49

2z/b=0.89

C : FC1
l
C : FC3
l
%DC
l

0.779
0.361
53.7

h /c=0.833
r
0.750
0.472
37.1

0.648
0.572
11.7

C : FC1
l
C : FC3
l
%DC
l

0.956
0.626
34.5

h /c=0.401
r
0.916
0.680
25.8

0.766
0.679
11.4

C : FC1
l
C : FC3
l
%DC
l

1.54
1.17
24.0

h /c=0.153
r
1.45
1.17
19.3

1.14
1.05
7.89

C : FC1
l
C : FC3
l
%DC
l

1.44
1.15
20.1

h /c=0.077
r
1.30
1.10
15.4

0.613
0.580
4.89

In FC3, the shape of the load distribution of the

wing changed depending on its ride height. At
h /c=0.833 and 0.401 it can be deduced that the
r
outboard sections of the wing generated more down-

force than the inboard sections. As the ride height

was further reduced, this trend gradually reversed.
At h /c=0.153 and 0.077, the inboard sections genr
erated more downforce than the outboard sections.
From the data presented thus far, a more detailed
understanding has emerged regarding the changes
experienced by the downstream wing. To supplement
these data further, the sketches in Figs. 12 and 13 have
been produced. The rst image shows a plan view of
the experimental conguration in the 2.11.5 m
wind tunnel. The arrows on the image show the main
path taken by the ow immediately surrounding the
blu body, as revealed by the smoke wand ow
visualization tests. The entrainment of the outer uid
by the diuser vortices led to a signicant increase
in turbulent ow along the centre-line of the tunnel.
The concentrated turbulent ow in this region was
responsible for the distinctive patterns that appeared
in the central portions of the wing, as previously
shown in Figs 8(b) and 9(b).
Plane aa is shown in the second image. In this
view, the hypothetical positions of the diuser vortices

Fig. 10 Comparison of the pressure surface ow in FC1 and FC3 for h /c=0.153. Leading edge
r
uppermost in image
Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

JAUTO56 IMechE 2006

Aerodynamics of a wing in ground eect

Fig. 11 Pressure distributions in FC1 and FC3 at the

spanwise stations of 2z/b=0.09 (a), 0.49 (b),
and 0.89 (c), for the ride height of h /c=0.153
r

have been sketched. The gure is used to visualize

the ow structure that was reasoned to have been
occurring at that spatial location, in the vicinity of the
wing. The circular dashed lines are used to represent
the vortices generated by the upstream diuser,
while the arrows signify the vertical component of
velocity in the plane.

11

The vortices are seen to rotate in such a manner

as to induce an upwash along the wing centre-line,
while inducing some degree of downwash close to
the tips. The upward velocity along the centre-line
would have been noticeably greater than the downward velocity closer to the wing tips owing to the
additive eect of the velocity components from both
vortices. This oweld structure is supported by the
LDA measurements that were presented in Fig. 4.
The structure of the oweld in which the wing
was placed can be used to help explain the reduction
in downforce that it experienced when placed downstream of the diuser blu body. In the dirty air, the
downforce reduction followed from the generation
of lower C values, and hence lower total forces.
P
The lower C values were induced by a reduction
P
in the freestream velocity, an upwash of the ow,
and increased turbulence. The reduced freestream
velocity essentially lowered the Reynolds number at
which the wing was operating, while the upwash had
the eect of reducing its eective angle of attack. Both
scenarios would have contributed to lower pressures.
The increased turbulence in the oncoming ow promoted early laminar-to-turbulent boundary layer
transition on the wing. Consequently, the boundary
layer was much thicker from the outset, therefore
having a greater decambering eect on the prole.
That is, the wing eectively lost camber, which would
have also resulted in lower pressures.
Also evident from the results was the fact that more
downforce was lost at greater ride heights than was
the case at lower ride heights. This observation can
be linked to the change in upwash with increasing
distance above the ground, when in the ow produced by the diuser. As the wing ride height was
increased, it operated in a region of greater upwash,
which continued to reduce the eective angle of
attack at which it operated. At greater ride heights,
there was also continued slowing of the oncoming

Fig. 12 Plan view of the experimental conguration in FC3

JAUTO56 IMechE 2006

Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

12

M D Soso and P A Wilson

Fig. 13 View through plane aa of Fig. 12

ow, in addition to increased turbulence levels. The

end result for the wing was therefore lower overall
downforce values.
Comparisons at the spanwise stations at which
pressure distributions were recorded indicated that
more lift was lost from the mid-span of the wing than
was the case from the ends of the wing. The midspan was thought to have experienced a greater loss
in downforce than the tips because the majority of
the disturbed ow appeared to be concentrated in
that region, and because that region experienced an
upwash, while the tips experienced ow with a neutral
to slight downward component of velocity, in addition
to a slightly higher freestream velocity.
The increase in the drag of the wing can be
explained by the resultant interaction of a number
of factors. Firstly, ow visualization in the disturbed
conditions highlighted the fact that the laminar
separation bubble was eliminated from the middle
portion of both the upper and lower surfaces of the
wing. This result in itself should have accounted for
a decrease in the drag, but this decrease may have
been outweighed by drag increments from increased
induced drag [22] and the increased extent of the
turbulent boundary layer. Increased induced drag
resulted from the fact that the wing operated in the
upwash of the upstream diuser blu body. The
increased extent of the turbulent boundary layer
would have caused increased skin friction drag.

2.

3.

4.

5.

6.

wing in ground eect experienced a decrease in

its downforce values and an increase in its drag
values.
When varying the height of the wing in the dirty
air ow, more downforce was lost at greater ride
heights than was the case at lower ride heights.
When in the dirty air ow, more lift was lost from
sections closer to the mid-span of the wing than
was the case for sections closer to the tip of the
wing.
When in the dirty air ow, the shape of the spanwise load distribution altered, depending on ride
height. At greater ride heights, the load distribution was lower at mid-span than it was towards
the wing tips. At lower ride heights, the trend
reversed.
The downstream wing was aected by the upwash
oweld of the upstream diuser blu body. The
presence of an upwash would have resulted in an
increase in the induced drag experienced by the
wing.
The disturbed ow emanating from the upstream
body had the ability signicantly to alter the surface ow patterns on the downstream wing. The
altered characteristics included earlier laminarto-turbulent boundary layer transition, and the
elimination of laminar separation bubbles.

ACKNOWLEDGEMENT
4 CONCLUSIONS
The data and associated analyses that have been presented were aimed at cataloguing the aerodynamic
changes that may be experienced by a single-element
wing when it operated in ground eect, downstream
of a leading racing car. The following conclusions can
be drawn from the research:
1. In the ow produced by bodies used to simulate
a leading racing car, a downstream single-element
Proc. IMechE Vol. 220 Part D: J. Automobile Engineering

The authors would like to recognize the support of

the School of Engineering Sciences at the University
of Southampton, Higheld, Southampton, UK, in
providing a scholarship to author Michael Soso.

REFERENCES
1 Romberg, G. F., Chianese Jr, F., and Lajoie, R. G.
Aerodynamics of race cars in drafting and passing
situations. SAE paper 710213, 1971.
JAUTO56 IMechE 2006

Aerodynamics of a wing in ground eect

2 Howell, J. Catastrophic lift forces on racing cars.

J. Wind Engng and Ind. Aerodynamics, 1981, 9,
145154.
3 Dominy, R. G. The inuence of slipstreaming on the
performance of a Grand Prix racing car. Proc.
IMechE, Part D: J. Automobile Engineering, 1990,
204, 3540.
4 Ahmed, S. R. Wake structure of typical automobile
shapes. Trans. ASME, J. Fluids Engng, March 1981,
103, 162169.
5 Ahmed, S. R. An experimental study of the wake
structures of typical automobile shapes. J. Wind
Engng and Ind. Aerodynamics, 1981, 9, 4962.
6 Hamidy, E. The structure of wakes of 3D blu
bodies in proximity to the ground, PhD Thesis,
Department of Aeronautics, Imperial College of
Science, Technology and Medicine, London, 1991.
7 Bearman, P. W. Near wake ows behind two- and
three-dimensional blu bodies. J. Wind Engng and
Ind. Aerodynamics, 1997, 69, 3354.
8 Lienhard, H., Stoots, C., and Becker, S. Flow and
turbulence structures in the wake of a simplied car
model (Ahmed model), Lehrstuhl fu
r Stromungsmechanik (LSTM), Universitat Erlangen Nu
rnberg,
Erlangen, Germany.
9 Ranzenbach, R. and Barlow, J. B. Two-dimensional
airfoil in ground eect, an experimental and computational study. SAE paper 942509, 1994.
10 Ranzenbach, R. and Barlow, J. B. Cambered airfoil
in ground eect wind tunnel and road conditions.
AIAA-95-1909-CP, 1995.
11 Ranzenbach, R. and Barlow, J. B. Cambered airfoil in ground eect an experimental and computational study. SAE paper 960909, 1996.
12 Knowles, K., Donoghue, D. T., and Finnis, M. V. A
study of wings in ground eect. Loughborough
University Conference on Vehicle Aerodynamics,
1994, pp. 22.122.13.
13 Jasinski, W. J. and Selig, M. S. Experimental study
of open-wheel race-car front wings. SAE paper
983042, 1998.
14 Zerihan, J. D. C. An investigation into the aerodynamics of wings in ground eect, PhD thesis,
University of Southampton, April 2001.
15 Zerihan, J. and Zhang, X. Aerodynamics of a single
element wing in ground eect. J. Aircraft, November
December 2000, 37(6).

JAUTO56 IMechE 2006

13

16 Zhang, X. and Zerihan, J. Aerodynamics of a double

element wing in ground eect. Am. Inst. Aeronaut.
Astronaut. J., June 2003, 41(6).
17 Coleman, H. W. and Steele, W. G. Experimentation
and uncertainty analysis for engineers, 1999 (John
Wiley).
18 ZOC22B and ZOC23B instruction and service manual
(Scanivalve Corporation).
19 Senoir, A. E. and Zhang, X. The force and pressure
of a diuser-equipped blu body in ground eect.
Trans. ASME, J. Fluids Engng., March 2001, 123,
105111.
20 Ruhrmann, A. and Zhang, X. Inuence of diuser
angle on a blu body in ground eect. Trans. ASME,
J. Fluids Engng., March 2003, 125, 332338.
21 Gad-el-Hak, M. Flow control passive, active
and reactive ow management, 2000 (Cambridge
University Press).
22 McCormick Jr, B. W. Aerodynamics of V/STOL ight,
1967 (Dover Publications).

APPENDIX
Notation

h
r
l
U
2
u, v, w
x, y, z

wing span (m)

wing chord (m)
coecient for the horizontal component
of the resultant force on the wing
two-dimensional sectional lift coecient
coecient for the vertical component of
the resultant force on the wing
pressure coecient
clean air ow
ow generated by the upstream wing
ow generated by the upstream diuser
blu body
ride height (m)
length of the diuser blu body (m)
length of the diuser blu body
Cartesian velocity components
Cartesian coordinates

angle of attack (deg)

b
c
C

C
l
C
L
C
P
FC1
FC2
FC3

Proc. IMechE Vol. 220 Part D: J. Automobile Engineering