The Engineering

Resource For
Advancing Mobility

400 COMMONWEALTH DRIVE WARRENDALE, PA 15096

850287

Full-Scale, On-Road Study of the Effect of

Automobile Shape on its Aerodynamic
Characteristics, and Comparison with
Small-Scale Wind Tunnel Results
Ammon Foux

Shimon Fanger-Vexler
and Joseph Katz

Dept. of Biomedical Engineering

Faculty of Mechanical Engineering

Technion, I. I. T., Haifa, Israel

Automotive Program
Technion, I. I.T., Haifa, Israel

International Congress
& Exposition
Detroit, Michigan
February 25 March 1, 1985

The appearance of the code at the bottom of the first page of this paper indicates
SAEs consent that copies of the paper may be made for personal or internal use, or
for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per article copy fee through the
Copyright Clearance Center, Inc., Operations Center, P.O. Box 765, Schenectady,
N.Y. 12301, for copying beyond that permitted by Sections 107 or 108 of the U.S.
Copyright Law. This consent does not extend to other kinds of copying such as
copying for general distribution, for advertising or promotional purposes, for
creating new collective works, or for resale.
Papers published prior to 1978 may also be copied at a per paper fee of $2.50
under the above stated conditions.
SAE routinely stocks printed papers for a period of three years following date of
publication. Direct your orders to SAE Order Department.
To obtain quantity reprint rates, permission to reprint a technical paper or permission to use copyrighted SAE publications in other works, contact the SAE
Publications Division.

ISSN 0148 - 7191

Copyright 1985 Society of Automotive Engineers, Inc.
This paper is subject to revision. Statements and opinions advanced in papers or discussion are the authors and are his
responsibility, not SAE's; however, the paper has been edited
by SAE for uniform styling and format. Discussion will be
printed with the paper if it is published in SAE Transactions.
For permission to publish this paper in full or in part. contact
the SAE Publications Division.

Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300 word abstract of a proposed manuscript to:
Secretary, Engineering Activity Board, SAE.
Printed in U.S.A.

850287

Full-Scale, On-Road Study of the Effect of

Automobile Shape on its Aerodynamic
Characteristics, and Comparison with
Small-Scale Wind Tunnel Results
Shimon Fanger-Vexler
and Joseph Katz
Faculty of Mechanical Engineering

Ammon Foux
Dept. of Biomedical Engineering
Technion, I. I. T., Haifa, lsrael

Automotive Program
Technion, I. I. T., Haifa, Israel

ABSTRACT
The design of passenger vehicles for
improved aerodynamic characteristics will result
in reduced fuel consumption and better road
handling during high-speed driving. In this
research, techniques were developed to measure
the aerodynamic drag and lift forces acting on a
full-scale vehicle under road conditions and
then were compared with results obtained on
reduced-scale models in a wind tunnel. A number
of configurations which characterize common
vehicle forms were investigated for their effect
on aerodynamic efficiency and fuel consumption. Experimental speeds were between 70 and
110 km/h, these being representative of highway
driving conditions.
A typical passenger vehicle of the threebox type was selected for the experiments, and
its exterior form was altered by means of
attaching various configurations to its front,
rear, and underbody portions. These additions
transformed the original vehicle into a fastback and "station wagon," and were used in
combination with underbody alterations, such as
front spoiler, side skirts, and smooth underbody. During road experiments, drag force was
measured by means of a telemetric system receiving data on drive-shaft strains, whereas lift
forces were measured by relative vertical displacements in the front and rear suspensions.
Statistical analyses showed that the different configurations had a significant effect
on the aerodynamic forces. The change in configurations brought about a maximum reduction in
drag coefficient (C ) of 51%, relative to the
D
original vehicle. As a result, fuel consumption
was reduced by 13% (at 110 km/h). Lift forces
dropped by as much a 47%. The most effective
components were a smooth underbody and a "fast-

back" form for drag, and a smooth underbody and

front spoiler for low lift. Results of the road
experiments showed a reasonable correlation with
those obtained using reduced-scale models in a
wind tunnel.
INTRODUCTION
AUTOMOTIVE AERODYNAMICS has an important impact
on the high-speed stability and fuel economy of
passenger cars (1,2,3).* It is therefore important, from the aerodynamic point of view, to
optimize the vehicles shape while it is still
in the preliminary design stages. The most
common, and probably the fastest, method of
obtaining the aerodynamic coefficients for
production automobiles is via full-scale wind
tunnel testing. This is necessary because
analytical prediction methods for the complex
viscous flow about a practical automobile configuration are still limited (4,5,6). In addition, wind tunnel testing has the advantage of
controlled atmospheric conditions, Fixed testing
facility, and (frequently) a fast and reliable
data-acquisition system. One of the few drawbacks of full-scale wind tunnels, apart from
being quite costly to construct, is the absence
of relative motion between the wind tunnel floor
and the tested object. Consequently, the boundary layer on the floor (7) beneath the model is
actually quite different from real road conditions (8,9,10). Moving wind tunnel floor
arrangements to overcome this problem have been
tested in small scale (8) and do provide a
partial remedy for some of the problems. The
use of this method in full-scale tests, however,
*Numbers in parentheses designate references at
end of paper.

0148-7191/85/0225-0287$02.50
Copyright 1985 Society of Automotive Engineers, Inc.

850287
is both complex and expensive and therefore not
utilized.
Full-scale road testing, on the other hand,
might be very promising. B u t a s a r e s u l t o f t h e
aforementioned difficulties (dependence on
atmospheric conditions and the requirement for
mobility of equipment), their primary importance
is in providing input for the development of
corrections, relative to actual road conditions,
for wind tunnel floor-boundary-layer and blockage effects. Most of the road tests reported
upon in the literature for determining the
aerodynamic coefficients, were based on the
deceleration method for obtaining the drag force
(11,12). The method is such that the test
vehicle is brought to a certain initial speed
and then the drag coefficient is determined from
records of both the rate of deceleration and the
distance traveled by the coasting-down vehicle.
Although the experiment in principle is simple
to perform, it requires the evaluation of inert i a l e f f e c t s . A s a r e s u l t , it is more sensitive
to external (atmospheric) disturbances than
would be expected from a constant-speed test,
and a large number of repetitions for each data
point are required (12). It is also subject to
changes in pitch attitude, Reynolds number,
airspeed and direction, and road noise during a
test. Furthermore, for the extraction of drag
from the total resistance force, the tirerolling resistance needs to be known a priori.
The primary scope of the present research
was, therefore, to develop a method for the
constant-speed aerodynamic testing of a fullscale automobile. Details on this technique are
given in the following two sections (Experimental Setup and Vehicle Configurations).
The second objective of this investigation
was to utilize this method for exploring the
effects of some common geometrical changes in
model configuration on the aerodynamic coefficients; these results are presented in the
Results section.
The third objective was to compare the
results obtained by this method with the results
of small-scale wind tunnel testing and fullscale tow tests. The outcome of this comparison
for six representative automobile shapes is
presented in the Comparison section.
EXPERIMENTAL SETUP
For the experiments, a typical "three-box
sedan" automobile was selected (Opel Record
1700, model 1973, with an automatic three-speed
transmission) having an overall length of
4.567 m, a width of 1.718 m, a height of
1.405 m, and a frontal area of 1.813 m2 (for the
baseline car). The measured variables included

the aerodynamic drag, the lift of the front and

rear axles, and fuel consumption. Drag force
was measured with full-bridge strain gauges
located on the vehicle's drive shaft (Fig. 1).
Front lift was measured by two sets of strain
gauges that were glued to the coil springs of
the front suspension at a half-bridge configuration oriented at 45 to the coil axis. The rear
lift was measured by a leaf spring device which
reacted to the relative displacement between the
vehicle and its rear axle (Fig. 1). In this
case also, two sets of strain gauges, one at
each side of the leaf spring, were used to
measure the rear body displacement (which was
directly calibrated to measure the aerodynamic
lift at the rear axle). The results obtained
for the lift force do not include the lift of
the wheels. The differences in the lift coefficients caused by the various vehicle geometries
must be considered, therefore, only in a relative sense.
The measurement of the drag force was
transmitted from the vehicles drive shaft by
the system shown in Fig. 1 (items 4, 5, and 6).
Both the incoming electromagnetic current and
the outgoing strain gauge information were
transmitted by the fixed antenna (Fig. 1,
item 4) to the amplifying processor (Acurex,
Torsion Measurement System, Model 1206). The
total resisting force (drag and wheels rolling
resistance) was directly obtained by measuring
the strain (due to torque) on the drive shaft.
The drive-shaft strain was calibrated to the
total force, acting on the vehicle body through
its center of gravity, as shown in Fig. 2. The
calibration line in Fig. 2 was obtained by
repetitive attempts to drive the vehicle against
an anchor post (without wheel slippage) at
different engine torque levels, while recording
drive-shaft strain and load-cell tension simultaneously. The results of this calibration test
indicated a highly linear regression correlation
between the total force and drive-shaft strain.
The static calibration of the lift forces
was obtained by loading and unloading the front
and rear axles with external weights (in the
range of l00 kg, front and rear) and then
recording the corresponding strains. To minimize the hysteresis due to the shock absorbers
of the suspension, small-amplitude vertical
oscillations were forced on the vehicles body
at each data point. In this case of the lift
force too, highly linear calibration curves were
obtained. More complicated effects due to
drive-train torque, such as chassis roll and
attitude changes during the road tests, were not
investigated here (buth these data will be
needed for future comparison with full-scale
wind tunnel experiments).

850287

Fig. 1. Vehicle drag and lift measuring system: (1) vehicle body, (2) leaf spring for measuring the
rear lift, (3) attachment of leaf spring to differential housing (longitudinally is free to slide),
(4) fixed antenna, (5) rotating transmitter of drive shaft strains, (6) signal receiver and
preprocessor, (7) strain gauges for front lift measurements (on both sides)

Fig. 2. Calibration method and graph for drag force: (1) anchorage, (2) cable (pointing at
vehicles center of gravity), (3) load cell, (4) tow points

4
Throughout the test all data were recorded
by a cassette data recorder (Kyowa,
model RTP-501A/AL) which was located within the
vehicle.
Tires were inflated to 30 psi in order to
minimize their rolling resistance. The rolling
resistance of each tire was measured, off the
v e h i c l e , on a commercial rig (made by Werk
Nurnberg M.A.N.). T h i s r i g i n c l u d e d a h o r i z o n tally rotating cylinder with a very large diameter and surface roughness similar to the concrete surface of the runway that served for the
r o a d t e s t s . In these measurements, the wheels
were driven by the rotating cylinder, and both
test speeds and tire normal load were kept the
same as they were during the road tests.
Because of numerous discontinuations in both the
rolling resistance and the road test, the load
on the tires was small and no temperaturerelated corrections were made to the tire
resistance.
The results for the drag force obtained by
this method do not include the dynamic friction
losses due to the rotation of the differential
assembly and the wheel bearings. The magnitude
of this resistance was measured in separate
towing tests (its variation through the test was
less than 0.5 kg, which is about 0.7% of the
drag at 100 km/h). This resistance, together
with the tire rolling resistance was later
subtracted in the process of computing the
aerodynamic drag.
The fuel-flow measurements were made by a
magnetic, positive-displacement flowmeter
(Brooks, model LS4150). The vehicle speedometer
was also calibrated by means of a radar speed
gun (C.M.I. Inc., model JFl00).
VEHICLE CONFIGURATIONS AND TEST CONDITIONS
The measuring equipment, described in the
former section, was used to measure the various
parameters of 16 vehicle configurations (all
sharing the same running gear). These configurations were obtained by attaching eight different components (as shown in Figs. 3 and 4) in
various combinations to the front, rear, and
underbody portions:
1. Wedge-shaped front hood (which was made
of polyurethane foam covered by a fiberglass
skin). Modification No. 1 in Fig. 3.
2. Fastback, Mod. No. 2.
3. S t a t i o n w a g o n , M o d . N o . 3 .
4. S m o o t h s h e e t m e t a l f l o o r ( r e a r s e c t i o n
o n l y ) , Mod. No. 4.
5. Smooth sheet metal floor (central
section only), Mod. No. 5.
6. Smooth sheet metal floor (front section
o n l y ) , Mod. No. 6 ( s h o w n a l s o i n F i g . 4 ) .

850287

Fig. 3. Configuration alterations from the

o r i g i n a l v e h i c l e : (1) wedge-shaped hood,
(2) fastback, (3) square back, (4,5,
and 6) smooth underbody panel elements

Fig. 4. Original vehicle with side skirts (l),

front spoilers (2), and smooth underbody (3)
7. S i d e s k i r t s ( F i g . 4 ) .
8. Front spoiler (Fig. 4).
With these modifications the 16 test configurations shown in Fig. 5 were obtained.
The road tests were carried out on a
3000-m-long, planar, smooth airfield runway.
The runway was oriented such that wind direction
c o i n c i d e d w i t h t h e p r e v a i l i n g w i n d . Since wind
effect during such experiments is very important
(13), the data usually were recorded with wind
velocity below 0.5 m/s and never exceeding a
velocity of 1.5 m/s. Each configuration was
tested at a constant speed for both directions
on the runway at speeds of 70-110 km/h at
l0-km/h increments. Data were obtained for each
configuration and speed by a continuous recording for 45-60 s. These data points were then
decoded using an X-Y recorder (Yokogawa 3078)
under laboratory conditions, thus producing
graphs as exemplified by Fig. 6.
The continuous drag and lift results were
converted to point samples for statistical
purposes by determining force at l-s intervals
for each graph. Thus approximately 50 data
points were determined for each configuration
and speed tested. Results were statistically

a na l y z e d i n o r d e r t o o b t a i n t h e m e a n ( x ) , s t a n dard deviation (s), and the coefficient of

variation (CV). The average coefficient of

850287

Fig. 5.

Definition of configurations investigated and their numerical label

Fig. 6 Example of time-dependent lift and drag recordings

850287

variation was low (5%) for the drag force,

whereas for the lift it was larger (up to 60%)
because of the high noise-to-signal ratio caused
by suspension oscillations. Although the oscillations for lift force were large, they were
centered about a constant horizontal line;
therefore the differences in these mean values
as a result of configuration and speed showed
good comparative results. In order to determine
the statistical significance of each configurat i o n a l m o d i f i c a t i o n , a number of analyses of
variance were performed. Each modification was
analyzed in comparison to the basic car and
fastback models. Separation of the means was
determined by use of a Duncans multiple range
t e s t ( 1 4 ) . This information is given in more
detail in Ref. 15.
RESULTS
A typical set of experimental resistive
f o r c e r e s u l t s , versus the vehicles speed is
shown in Fig. 7. T h e t o t a l r e s i s t a n c e , a s
i n d i c a t e d , includes both the aerodynamic drag
a n d t h e r o l l i n g r e s i s t a n c e o f t h e t i r e s . The
tires' rolling resistance (corrected for actual
road conditions; e.g., bearing friction, lift,
etc.) shows only a small increment with increasing speed.
The drag force results of Fig. 7 clearly
indicate the parabolic dependence of the aerodyn a m i c d r a g o n t h e o n s e t a i r s p e e d . The sedan,
with its four windows open (configuration 1*)

Fig. 7.

had the highest drag, while the fastback shape

(configuration 2) had far less air resistance.
This was so since the hatch inclination angle of
about 20 allowed an attached flow regime along
the rear window (observed by means of tufts flow
visualization) and thereby reduced drag was
measured (16,17,18). Furthermore, the complex
unsteady separation pattern behind the sedan
configuration introduced higher aerodynamic
noise into the resistance force data, compared
to configurations 2 and 3. This was the reason
that the least-square curve fitting, shown in
F i g . 7 , improved with the improvements in the
vehicle aerodynamics.
The drag of the station wagon (configuration 3) in this test was found to be lower than
the drag of the baseline car (configuration l),
but it was higher than the drag of the fastback
shape. This is consistent with the findings
reported by other researchers (e.g., Ref. 19).
The drag coefficient (CD = D/0.5V2A), based on
the frontal area A and windspeed V, for each
vehicle's configuration was obtained after
subtracting the tires' rolling resistance from
t h e t o t a l r e s i s t a n c e , which is shown in Fig. 7.
The collective results for the drag coefficients
(based on the average of all speed tests) are
represented by the bar diagrams of Fig. 8. I n a
similar manner the lift variations with the
vehicles speed were measured, and as an example
the test results for configuration 1 are presented in Fig. 9. Here Ltot represents the
total lift of both the front and rear axles,

Variation of total drag force with vehicle speed, as obtained by direct road tests

850287

Fig. 8.

Diagram of comparative drag coefficients

Fig. 9. The effect of vehicle speed on lift

(Lr - rear axle, Lf - front axle, Ltot - front
and rear axles)

is the rearw h i l e L f i s t h e f r o n t - a n d Lr ,
a x l e l i f t c o m p o n e n t , r e s p e c t i v e l y . Most of the
measured lift is generated at the vehicles
leading edge, and therefore the lift at the
front axle (without correction due to drivetorque-induced chassis pitchup) is considerably
higher than the lift at the rear ax1e. This is
consistent with the high suction values shown by
the pressure-distribution data provided in
R e f s . 19 and 20. T h i s s i t u a t i o n , t h a t m o s t o f
the lift was generated at the front, was consistent for all the configurations tested.
In these experiments the lift was not
measured for all the configurations, but rather
only for those modifications with a major effect
on aerodynamic lift forces (judged on grounds of
previous small-scale wind tunnel tests (21)).
The lift-coefficient results for these six tests
are shown graphically by the bar diagrams of
Fig. 10.
Based on the data presented in Figs. 7-10,
several conclusions can be drawn about the
influence of this vehicle's external geometry on
its aerodynamic drag and lift:

850287

CONFIGURATION NUMBERS
Fig. 10. Diagram of comparative lift
coefficients
1. T h e e f f e c t o f t h e m o d i f i e d f r o n t ( w e d g e
shape) can be estimated by comparing the drag
coefficients of configurations 1 and 3 with the
drag of configurations 4 and 6, respectively
(which are essentially the same configurations,
but with a modified wedge-shaped front). I t i s
evident that the modified front reduced the
drag. Consequently, the most aerodynamic conFiguration among models 1 to 6, was model 5,
having the wedge-shaped front and fastback
modifications.
The reduction in the aerodynamic
drag coefficient due to these modifications is
39%, as shown by Fig. 8.
2. The effect of a smooth underbody fairing is demonstrated when comparing the drag of
the original sedan car (configuration 1) with
the drag of configuration 7. Here the addition
of the underbody fairing, between the cooling
intake and the front axle (see details in
Fig. 3), resulted in a drag reduction of 15%.
The extension of the fairing up to the rear axle
(configuration 8) further reduced the drag,
whereas the complete fairing (configuration 16)
did not result in any additional drag reduct i o n . This is probably due to the separated
flow, indicated by the tufts motion in that
region. (Buchheim et al. in Ref. 1 did find
that a longer diffuser section at this region
reduced the vehicle's drag.)

A similar trend was found when these lower

panel modifications were adapted to the already
a e r o d y n a m i c a l l y e f f i c i e n t c o n f i g u r a t i o n 5 . The
addition of the smooth underbody panel, extended
to the rear axle (configuration 9) resulted in a
drag reduction of DC
= 0.056, which was not
D
further improved by full extension of the lower
fairing (configuration 13). Actually these two
configurations (9 and 13) were the most aerodynamic throughout the experiment, with a drag
coefficient of C
= 0.231, which is 51% lower
D
than the aerodynamic drag of the baseline
vehicle.
3. T h e e f f e c t o f t h e s e m o d i f i c a t i o n s o n
the cooling airflow was not thoroughly investig a t e d i n t h i s s t u d y , and the only concern in
that regard was that engine overheating should
be avoided (during the road tests, with ambient
temperatures of 18-22C). I t i s w o r t h w h i l e t o
point out that cooling drag losses can be
reduced considerably by careful aerodynamic
design as it was shown for light aircraft engine
installations (22,23). Therefore, a large
portion of the drag-reduction potential (if not
all) of the modifications tested here can fully
be exploited.
One of the problems that can be introduced
by a continuous smooth underbody panel (apart
from additional weight) is the blockage of the
engine-cooling airflow. The simplest solution
that was tested is to duct the cooling air
beneath the vehicle (as in configurations 11
and 12) and eject it into the wake flow. This
was accomplished by covering the lower surface
of the car with the fairing (modification 5 in
F i g . 3 ) , and by a small rearward displacement of
It was
the rear bumper, as shown in Fig. 11.
found that sufficient airflow (based on static

Fig. 11. Geometry of the airflow ducting

beneath the vehicle:
(1) rear section of duct,
(2) smooth underbody panel, (3) outcoming
airflow

850287
pressure difference) was obtained in this manner
so that engine cooling could be resolved with a
complete underbody smooth panel. The comparison
of the drag of configurations 11 and 12 (having
a smooth underbody) with the drag of configurations 5 and 2, respectively, show that the dragreduction effect of the smooth underbody was
partially lost. Improved cooling-air inlet and
e x i t , and smoother ducting (22,23) probably will
increase the aerodynamic benefits due to smooth
underbody fairings.
4. The effect of the underbody fairing can
be obtained by comparing the results of the
sedan (configuration 1) and the same shape with
the smooth underbody panel (configuration 16).
The data of Fig. 10 show a 12% reduction of the
total lift due to this modification. The combination of the lowest drag shape with the smooth
underbody fairing (configuration 13) has further
reduced the lift to a value which is lower by
37% than the lift of the baseline car (see
Fig. 10).
5. The effects of some lift-reducing
modifications on vehicle drag can be obtained by
comparing the results of configurations 10, 14,
and 15. The addition of a front spoiler (see
Fig. 4) to configuration 11 (which is a direct
modification to configuration 10) resulted in an
increase of the drag by CD = 0.054. A similar
change in drag (C D = 0.048) was measured when
the spoiler was added to the baseline shape with
the smooth underbody panel (configuration 16) to
obtain configuration 15. The lift-reducing
effect of the front spoiler was effective, as
can be deduced from Fig. 10. Actually, the
spoiler and the smooth underbody in configuration 10 resulted in a 47% lift reduction relative to the baseline configuration. The net
effect of the front spoiler on the lift can also
be estimated by comparing the lift of configurations 15 and 16, which showed a lift difference
of C
= 0.051. Similar lift-reducing effect,
L
for the front spoiler, were reported in Refs. 16
and 20. By further optimization of the spoiler
geometry, drag reduction was also obtained.
The effect of the side skirts (see Fig. 4)
on drag is negligible. This can be observed
when comparing the results of configuration 14
in Fig. 8 with the results of configuration 16.
The reduction of the lift obtained by this
device is about C = 0.032, as shown in
L
Fig. 10.
In summary, the best configurations, with
respect to drag, were configurations 9 and 13,
both with a drag coefficient of 0.231. This
vehicle still had one side mirror (the effect of
similar external modifications, such as mirrors,
are given in Ref. 24) and a front-pillar-mounted
external antenna. The improvements resulted in

an overall drag coefficient reduction of 51%,

together with a 37% reduction of the lift coefficient, compared to the baseline shape. Further reduction of the lift coefficient was
obtained with the addition of the front spoiler
(configuration 10). However, this modification,
with the tested vehicle geometry, did increase
the drag force.
The gain in fuel economy owing to the drag
difference between the baseline and the lowestdrag vehicle is demonstrated in Fig. 12. The

F i g . 1 2 . Fuel consumption improvement by drag

coefficient reduction
fuel flow, at constant speed, was noticeably
reduced, and at a speed of 110 km/h a 12.7%
lower fuel consumption was measured. T h i s f u e l
consumption could further be improved, in view
of the above drag reduction, by mechanical
modifications such as higher gear ratio or a
standard clutch instead of the torque converter.
Mechanical modifications of this type, however,
affect the overall performance of the vehicle
( e . g . , low-speed acceleration) and their

10

850287

application requires some additional

considerations apart from the aerodynamic
consideration.
COMPARISON WITH OTHER TESTING METHODS
The experimental method which was presented
throughout this paper has the advantage of being
a direct method of measuring actual road resistance. On the other hand, wind tunnel experiments with full- or small-scale models are far
more simple to conduct and are less dependent on
external atmospheric conditions. However, they
do have problems such as boundary layer interference at the tunnel floor and wall, effects of
blockage and of nonrotating wheels, etc. For
this reason it is essential to compare wind
tunnel data with other experimental methods, so
that either of these techniques could be used
when such engineering data are required.
The comparison was made on six basic automotive shapes (configurations 1 to 6), and the
results for the drag coefficient are given in
Fig. 13. In addition to the present measurements, measurements were made by the towing
method and by small-scale (1:17) wind tunnel
testing (results of these experiments were
reported in Ref. 21). The towing experiments
were actually performed on the same vehicle with
a 40-m-long towing cable. The method of testing
was identical to the testing reported here, but
the drag was measured at constant speed by
measuring the cable tension at the tow-car
end. Since the tow force also included the
force required for rotating the drive shaft and
part of the transmission, the drag was defined
accurately only relative to the baseline shape.
Once the drag of the baseline car (configuration 1) was determined by the current method,
t h e p l o t t i n g o f F i g . 1 3 w a s m a d e p o s s i b l e . More
details on these tests are reported in Ref. 25.
In the small-scale wind tunnel experiments,
the models were mounted on a vertical sting
balance in a 1- by l-m (test section) wind
tunnel (Fig. 14). An elevated ground plane
served to divide the wind tunnel boundary layer
from the models (for more details refer to
Ref. 21). The average Reynolds number for these
t e s t s w a s a b o u t 2 . 5 x 1 05 ( b a s e d o n m o d e l
length) compared with the value of 3.5 x 106 for
the full-scale tests. The models had underbody
details (gas tank, exhaust system, etc.) that
were carved up to 3 mm deep into the model, and
upper body details such as headlights, separate
wheels, hoods, etc., but not cooling air passages. Special care was taken in sharpening the
rear end and window details in order to achieve
a separation-line location similar to those of
the full-scale model. Blockage corrections were

Fig. 13. Diagram of the drag coefficients--a

comparison of different testing methods
not applied in the process of data reduction,
since the ratio between model cross section to
wind tunnel test section area was less than 1%.
The results of the towing tests, as shown
by Fig. 13, compared well with the results of
the current road tests. The result of the
small-scale wind tunnel experiments showed the
same order of drag decrease as the full-scale
tests, but the increments were smaller. The
larger values of these drag increments in the
full-scale tests are probably a result of the
larger Reynolds number (more turbulent boundary

11

85

Fig. 14. Small-scale wind tunnel model and its

mounting in the test section
layer), so the tendency of the flow to stay
attached, because of geometrical modifications,
is higher. Flow visualizations by tufts ensured
t h i s , and the flow was fully attached on the
rear hatch (configuration 5) in full-scale
tests, while in the small-scale tests it was
partially separated. This can provide an
explanation to why the full-scale drag reduction
effect of the fastback configuration was larger.
A comparison of the lift coefficient is not
reported here, but the results for the baseline
c o n f i g u r a t i o n ( c o n f i g u r a t i o n 1 ) w e r e CL = 0.361
in full scale and 0.35 in small scale (Ref. 21).
Similarly, the lift of the most aerodynamic
shape (configuration 13) was 0.227 in full
scale, whereas in the small scale the value of
0.12 was obtained.
In this case the larger
deviation is again due to the flow being fully
attached in the rear hatch region in full scale.
This can be explained by the larger suction
peaks that are present on the upper surface of
the vehicle when the attached streamlines bend
downward, as it was in the larger Reynolds
number experiments.
CONCLUDING REMARKS
The road-testing technique for the aerodynamic lift and drag measurements was found to be
very practical, and configuration changes were
fast and simple to perform. This method is
suitable for research-oriented experiments,
since the development of an instrumented
r e s e a r c h v e h i c l e i s r e q u i r e d . The adoption of
this technique for industrially efficient measurements of the aerodynamic forces on numerous
production vehicles therefore seems not to be
economical.
The geometrical changes that were investigated had significant effects on the aerodynamic

forces. Whenever these modifications allowed a

larger transfer of free-stream momentum to the
rear section of the vehicle, a drag reduction
was obtained, since the amount of rear-flow
s e p a r a t i o n s w a s r e d u c e d . This was achieved by
adding the smooth underbody to the baseline,
resulting in less blockage of the flow beneath
the vehicle and in a reduction of the drag
coefficient by about 0.2. Similarly, the
improvement of the upper flow, with the aid of
the wedge-shaped front and the fastback configuration, resulted in a reduction of about 0.19
i n t h e d r a g c o e f f i c i e n t . The combination of all
these modifications, however, did not add up in
a linear manner; but even so a very low drag
c o e f f i c i e n t o f CD = 0 . 2 3 1 o n a p r a c t i c a l r o a d going car was obtained. This result could be
even further improved by detailed body work,
wheel fairings, and the removal of the side
mirror and radio antenna. The net drag-reducing
effect of the wedge-shaped front ranged between
C D = 0 . 6 a n d CD = 0 . 3 , a n d o f t h e f a s t b a c k
m o d i f i c a t i o n w a s a b o u t CD = 0 . 1 3 t o
CD = 0 . 1 6 , w h e n a d d e d t o t h e b a s e l i n e v e h i c l e .
With the addition of the smooth underbody, these
amounts were reduced, but not eliminated. I t i s
concluded, therefore, that the modifications
tested here will result in similar trends when
applied to a different three-box-type automobile, but the linear breakdown of the contribution of each modification is not exact.
The comparison of the towing tests (which
are considered to be less accurate) with the
direct road-test data showed good agreement.
The comparison with the low-Reynolds-number wind
tunnel data showed good qualitative results, but
both drag and lift increments due to geometric
changes were larger in the full-scale tests.
Small-scale tests are therefore useful to provide qualitative information about the influence
of geometrical changes. They could be useful in
developing general configurations or for calib r a t i n g a n a l y t i c a l p r e d i c t i o n m e t h o d s . For
information about fine details of a production
automobile, however, full-scale (wind tunnel)
testing is still the best way to go.
ACKNOWLEDGMENTS
The authors wish to acknowledge their
appreciation to the staff of the Energy Center
and the Aeronautical Center at the Technion,
I.I.T. for providing the resources and the
technical support for this research. We wish to
express our gratitude to Laura Fanger-Vexler for
her assistance in many phases of this research
and to Leah and Ami Kronenberg for their assist a n c e i n d a t a c o l l e c t i o n . The authors also wish

12
to thank Dr. Gino Sovran from the G.M. Research
Laboratories for his many useful comments on
this
manuscript.
REFERENCES
1. B u c h h e i m , R . , D e u t e n b a c h , K . R . , a n d
Luckoff, H. J., "Necessity and Premises for
Reducing the Aerodynamic Drag of Future Passenger Cars," SAE 81-0185, February 1981.
2. J a n s s e n , L . J . , a n d E m m e l m a n n , H . J . ,
"Aerodynamic Improvements--A Great Potential for
Better Fuel Economy," SAE 78-0265, 1978.
3. Hucho, W. H., Janssen, L. J., and
Emmelman, H. J., "The Optimization of Body
Details--A Method for Reducing the Aerodynamic
Drag of Road Vehicles," SAE 76-0185, 1976.
4. L o s i t o , V . , D e N i c o l a , C . , A l b e r t o n i ,
S ., a n d B e r t a , C . , " N u m e r i c a l S o l u t i o n s o f
Potential and Viscous Flows Around Road Vehicles," International Journal of Vehicle Design,
Technological Advances in Vehicle Design Series,
SP3, Impact of Aerodynamics on Vehicle Design,
1983, pp. 429-440.
5 . Chometon, F., " C a l c u l a t i n g T h r e e Dimensional Separated Flow Around Road Vehicles," International Journal of Vehicle Design,
Technological Advances in Vehicle Design Series,
SP3, Impact of Aerodynamics on Vehicle Design,
1983. pp. 374-386.
6. Markatos, N. G., "The Theoretical
Prediction of External Aerodynamics of Road
Vehicles," International Journal of Vehicle
Design, Technological Advances in Vehicle Design
S e r i e s , SP3, Impact of Aerodynamics on Vehicle
Design, 1983, pp. 387-400.
7 . Mason, W. T., Jr., and Sovran, G.,
t'Ground Plane Effects on the Aerodynamic Characteristics of Automobile Models--An Examination
of Wind Tunnel Test Technique," Advances in Road
Vehicle Aerodynamics BHRA Fluid Engineering,
1973, pp. 291-310.
8. Hucho, W. H., Janssen, L. J., and
Schwarz, G., "The Wind Tunnel's Ground Plane
Boundary Layer--Its Interference with the Flow
U n d e r n e a t h t h e C a r s , " SAE 75-0066, February
1975.
9. Gross, D. S., and Sekscienski, W. S.,
"Some Problems Concerning Wind Tunnel Testing of
Automotive Vehicles," SAE 66-0385, June 1966.
10.
Beauvais, F. N., Tignor, S. C., and
T u r n e r , T . R . , "Problems of Ground Simulation in
Automobile Aerodynamics," SAE 68-0121, 1968.
11. R o u s s i l l o n , G . , " C o n t r i b u t i o n t o
Accurate Measurement of Aerodynamic Drag on a
Moving Vehicle From Coast-Down Tests and Determination of Actual Rolling Resistance," Journal
of Wind Engineering and Industrial Aerodynamics,
Vol. 9, 1981, pp. 33-48.

850287
12. K o r s t , H . H . , a n d W h i t e , R . A . , " E v a l uation of Vehicle Drag Parameters From CoastDown Experiments Conducted Under Nonideal
Environmental Conditions," Journal of Fluids
Engineering, Vol. 103, March 1981, pp. 133-140.
13. S o v r a n , G . , " T h e E f f e c t o f A m b i e n t
Wind on a Road Vehicle's Aerodynamic Work
Requirement and Fuel Consumption," SAE 84-0298,
February 1984.
14. D u n c a n , D . B . , " M u l t i p l e R a n g e a n d
Multiple F Tests," Biometrics, Vol. 11, 1955,
pp. l-42.
15. F a n g e r - V e x l e r , S . , " E x p e r i m e n t a l S t u d y
on the Effect of Vehicle Shape on Aerodynamic
Drag and Lift Forces," M.Sc. Thesis, Faculty of
Mechanical Engineering, Technion, I.I.T., Haifa,
Israel, March 1984.
16. Hucho, W. H., "The Aerodynamic Drag of
Cars, Current Understanding, Unresolved Problems
and Future Prospects," Symp. on Aerodynamic Drag
Mechanisms of Bluff Bodies and Road Vehicles,
General Motors, Warren, Michigan, 1971; Plenum
Press, 1978, pp. 7-40.
17. M o r e l , T . , "Aerodynamic Drag of Bluff
Body Shapes Characteristic of Hatch-Back Cars,"
SAE 78-0267, 1978.
18. A h m e d , S . R . , " I n f l u e n c e o f B a s e S l a n t
on the Wake Structure and Drag of Road Vehicles," Journal of Fluids Engineering, Vol. 105,
December 1983, pp. 429-434.
19. C o g o t t i , A . , B u c h h e i m , R . , G a r r o n e ,
A., and Kuhn, A., "Comparison Tests Between Some
Full Scale European Automotive Wind Tunnels-Pininfarina Reference Car," SAE 80-0139,
February 1980.
20. S c h e n k e l , F . K . , " T h e O r i g i n s o f D r a g
and Lift Reductions on Automobiles With Front
and Rear Spoilers," SAE 77-0389, 1977.
21. K a t z , J . , "Evaluation of Small-Scale
Automotive Wind Tunnel Testings," to be published by Israel J. of Technology, Vol. 21,
1984.
22. C o r s i g l i a , V . R . , K a t z , J . , a n d
Kroeger, R. A., "Full-Scale Study of the Effect
of Nacelle Shape on Cooling Drag," Journal of
Aircraft, Vol. 18, No. 2, 1981, pp. 82-88.
23. K a t z , J . , C o r s i g l i a , V . R . , a n d
Barlow, P. R., "Cooling Air Inlet and Exit
Geometries on Aircraft Engine Installations,"
Journal of Aircraft, Vol. 19, No. 7, 1982,
pp. 525-530.
2 4 . M o r e l l i , A . , "Aerodynamic Basic Bodies
Suitable for Automobile Applications," International Journal of Vehicle Design, Technological
Advances in Vehicle Design Series, SP3, Impact
of Aerodynamic on Vehicle Design, 1983,
pp. 70-98.
25. K a t z , J . , " E f f e c t o f E x t e r n a l G e o m e t r y
on Automobile Aerodynamic Drag," E.E.C. Report
No. 132, Technion, I.I.T., Haifa, Israel, 1982.

This paper is subject to revision. Statements and opinions advanced in papers or discussion are the authors and are his
responsibility. not SAEs; however. the paper has been edited
by SAE for uniform styling and format. Discussion will be
printed with the paper if it is published in SAE Transactions.
For permission to publish this paper in full or in part, contact
the SAE Publications Division.

Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300 word abstract of a proposed manuscript to:
Secretary, Engineering Activity Board, SAE.
Printed in U.S.A.