SAE TECHNICAL

PAPER SERIES

2000-01-3549

Advances in Wind Tunnel Aerodynamics for

Motorsport Testing
Steve Arnette and Bill Martindale
Sverdrup Technology, Inc.

Reprinted From: Proceedings of the 2000 SAE Motorsports

Engineering Conference & Exposition
(P-361)

Motorsports Engineering Conference & Exposition

Dearborn, Michigan
November 13-16, 2000
400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A.

Tel: (724) 776-4841 Fax: (724) 776-5760

The appearance of this ISSN code at the bottom of this page indicates SAEs consent that copies of the
paper may be made for personal or internal use of specific clients. This consent is given on the condition,
however, that the copier pay a $7.00 per article copy fee through the Copyright Clearance Center, Inc.
Operations Center, 222 Rosewood Drive, Danvers, MA 01923 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.
SAE routinely stocks printed papers for a period of three years following date of publication. Direct your
orders to SAE Customer Sales and Satisfaction Department.
Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department.
To request permission to reprint a technical paper or permission to use copyrighted SAE publications in
other works, contact the SAE Publications Group.

All SAE papers, standards, and selected

books are abstracted and indexed in the
Global Mobility Database

No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written
permission of the publisher.
ISSN 0148-7191
Copyright 2000 Society of Automotive Engineers, Inc.
Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely
responsible for the content of the paper. A process is available by which discussions 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 Group.
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 Meetings Board, SAE.

Printed in USA

2000-01-3549

Advances in Wind Tunnel Aerodynamics for

Motorsport Testing
Steve Arnette and Bill Martindale
Sverdrup Technology, Inc.

Copyright 2000 Society of Automotive Engineers, Inc.

ABSTRACT
As the popularity of motorsport continues to surge
throughout the world, so to does the level of competition in
the motorsport community. Participants work to achieve
a performance edge through superior engineering. As an
enabling tool, the wind tunnel has become a focus for
enhancing performance.
This is evidenced by the
increasing interest among motorsport teams in dedicated
wind tunnel facilities, as best exemplified by the Formula
One community. Part of the reason for this increasing
focus on wind tunnels is the availability of breakthrough
technologies that better simulate on-track conditions,
providing new opportunities to enhance performance. Two
areas that are the subject of strong current interest are (i)
test section configurations that eliminate wind tunnel
interference effects to provide the highest possible
aerodynamic simulation fidelity and (ii) high speed rolling
road systems with integrated force measurement systems
that provide high fidelity simulation of underbody effects.
This paper presents an overview of these technologies,
including selected computational and experimental results
that illustrate the simulation advantages obtained with
these new wind tunnel technologies.

INTRODUCTION
The increasing emphasis of wind tunnel usage in the
motorsport community is well exemplified by the Formula
One community. Initially, the wind tunnels employed
were typically of the open-return type designed for testing
of models at 25% scale. Over time, standard practice
evolved to testing at 40% scale. Many of the facilities in
use were converted aeronautical wind tunnels. These
early facilities contained some of the first rolling road
systems to simulate underbody aerodynamics in the wind
tunnel, although the rolling road systems were confined to
low speeds relative to those of actual competition.
A snapshot of the wind tunnels currently being used in
this community shows that the majority of the testing is
occurring at model scales of either 40% or 50%. The

increasing model scale allows better geometry fidelity of

the model to the actual vehicle. The typical configuration
for these wind tunnels is a solid wall test section with a
cross-sectional area of 7-14m2. The maximum wind
speed capability of these facilities is typically 50 m/s, but
ranges as high as 70m/s in some cases. Note that these
speeds represent the limiting capability of traditional
rolling road systems, where concern for the life of the
rolling road belt severely limits high speed operation.
The wind tunnels currently being built or planned for the
Formula One community represent yet another increment
in simulation capability. Testing is planned for a typical
model scale of 60%, with the added requirement to obtain
quality simulation for full-scale testing. The top speed of
these facilities is at least 70m/s, with most being
designed for an 80m/s capability. Most of these facilities
are being designed with non-traditional test section
configurations, driven by the needs for (i) superior
aerodynamic simulation at 60% model scale and (ii) high
quality simulation fidelity for full-scale testing. The rolling
road systems of these facilities represent a dramatic
increment in top speed capability, with systems capable
of routine operation at speeds up to 100m/s if required.
The purpose of this paper is to provide an overview of
emerging technologies that are prominent in the new
generation of wind tunnels dedicated to motorsport
testing.
Chief among these are (i) test sections
employing contoured wall technology to eliminate
boundary interference effects which degrade aerodynamic
simulation quality and (ii) next-generation rolling road
systems which provide a dramatic increase in top speed
capability and the possibility of measuring aerodynamic
forces transmitted to the rolling road through the rotating
tires. A discussion of these technologies is timely given
the increasing emphasis being given to wind tunnel testing
for both stock car and open wheel classifications. The
basis for this invited paper was an invited SAE TOPTEC
presentation given in October 1999. 1

OVERVIEW OF TEST SECTION TECHNOLOGY

Of the wind tunnels currently used for motorsport
development, the vast majority possess either solid wall or
semi-open jet test sections. For both types, boundary
interference causes the external aerodynamic simulation
in the wind tunnel to differ from the actual situation on the
road.
SOLID WALL TEST SECTION For the solid wall test
section, boundary interference is fundamentally caused by
the confinement of the flow between the test model (or
vehicle) and the solid test section boundary.
This
unavoidable effect results in over-acceleration of the flow
over the model, as compared to the actual situation on the
road. To minimize these effects, it is advisable to hold the
vehicle blockage ratio (ratio of model frontal area to cross
sectional area of test section) to no more than about 5%.
This is why solid wall wind tunnels devoted to full-scale
automotive development have large test sections, e.g. the
General Motors Aerodynamic Laboratory Wind Tunnel2
has a test section area of approximately 610 ft 2 (57m2).
Because of the bulk of work done in solid wall wind
tunnels, spanning both automotive and aerospace product
development, the topic of solid wall boundary interference
has been investigated extensively.
Seven different wall boundary correction methods were
compared in an SAE-sponsored exercise to assess the
accuracy of correction methods.3 Of the seven, three were
judged to be acceptable. Even the acceptable methods
failed to provide an adequate correction at blockage ratios
of 10%. Of the three acceptable methods, the pressure
signature correction method represents the most rigorous
aerodynamic development. Its use, however, has tended
to be limited by the requirement for wall pressure
signature measurements and lengthy calculations to
complete the correction procedure. Even if the one of the
preferred correction methods is applied, the results
indicate that testing must be carried out at blockage
ratios much less than 10% to produce accurate test
data. 3
It is important to note that sources of degradation other
than boundary interference are possible in the solid wall
test section. For example, the combination of large
model blockage and model location in close proximity to
the contraction section can result in model influence on
the wind tunnels dynamic pressure measurement. If
present, the result will be a facility-indicated dynamic
pressure different than the actual dynamic pressure
incident on the vehicle. Another potential source of
degradation is static pressure influence from the
downstream diffuser. As shown by Garry et al.,4 if the
base of the model is too close to the diffuser at the
downstream end of the test section, the interaction of the
diffuser and the model wake will cause the static pressure
at the base of the model to be artificially elevated, thereby
altering the pressure forces on the vehicle. Note that both

of these potential sources of simulation error can be

minimized through proper facility design.
Boundary
interference, however, can be minimized only through
increasing the test section area, which translates directly
to increased capital cost of the wind tunnel facility.
Although correction procedures such as the pressure
signature method help close the gap between the wind
tunnel and actual on-road aerodynamic performance, it
would obviously be preferable to measure actual on-road
performance in the wind tunnel. This is increasingly true
as motorsport becomes more competitive. With racecar
developers looking for aerodynamic advantage with the
design of essentially every element on the vehicle,
situations will arise where the performance increments
between various design options are much smaller than the
correction increment needed to translate the wind tunnel
measurements to on-road performance. Along a similar
line of thought, the need for correction arises
fundamentally from improper flow simulation over the
model. As advanced diagnostic tools such as Pressure
Sensitive Paint,5 Particle Image Velocimetry,6 and Planar
Doppler Velocimetry 7,8 which provide detailed, spatiallydistributed mappings of local flow structure become
more commonplace in the wind tunnel, proper flow
simulation at localized positions on the vehicle is
becoming increasingly important.
Although post-test
corrections are suitable for force and moment coefficients,
they are not capable of improving flow simulation in the
wind tunnel.
OPEN JET TEST SECTION Many wind tunnels used
for motorsport development possess an open jet test
section. For this configuration, the test model is placed in
a large plenum chamber and the nozzle flow entering the
plenum occupies only a modest portion of the plenums
cross-section. After flowing over the model, the flow is
then collected at the rear of the plenum. Historically, this
type of test section has been preferred in Europe, based
on the idea that aerodynamic simulation fidelity to the
open road is less degraded by increasing model blockage
than it is for a solid wall test section. For the open jet
test section, model blockage ratio is defined as the ratio
of the model frontal area to the cross sectional area of the
nozzle. In the semi-open jet test section (where the model
blockage is typically 10% or more), the main source of
degraded aerodynamic simulation is the finite dimension
of the jet exiting the nozzle. On the road, the vehicle is
immersed in a semi-infinite flow.
Blockage corrections for the semi-open jet test section
are less mature than those for the solid wall test section. 9
The work of Mercker and Wiedemann10 is the most
comprehensive work to date. As a result, it is difficult to
comment definitively on the success of open jet correction
methods. It is noted, however, that the recent
experimental investigation of Hoffman et al.11 indicates
that the aerodynamic simulation obtained from open jet
test sections is superior to that obtained from slotted wall
test sections for model blockages ranging from 7% to

25%. The lack of knowledge regarding quantitative

performance is the driving reason that essentially all new
wind tunnels planned for the near future that will be
dedicated to motorsport testing will not have an open jet
test section.
Open jet test sections are universally beset by an adverse
static pressure gradient in the downstream portion of the
test section. 9,12 The common result is that test vehicles
installed in the test section are subjected to an elevated
static pressure in the base region. This horizontal
buoyancy effect is exactly analogous to that encountered
in a solid wall test section if the test vehicle is located too
close to the downstream diffuser. Although the obvious
answer is to increase the length of the open jet, Arnette et
al.12 have shown that open jet test sections exhibit an
increasing tendency towards low-frequency unsteadiness
as the length of the open jet increases. The result is a
trade-off situation for the wind tunnel designer. Also
similar to the solid wall test section, locating the test
vehicle or model too close to the nozzle can lead to model
influence on the facility measurement of dynamic
pressure, with similar negative effects.
CONTOURED WALL TEST SECTION Because
contoured wall technology has its origin in adaptive wall
technology, a review of adaptive wall test sections is
presented prior to discussing the details of contoured wall
test sections.
Adaptive Wall Test Section The idea of an adaptive wall
test section is simply to shape the side and top walls of
the test section such that they correspond to external
streamlines that would be present over the vehicle on the
open road. More formally stated, if the flow angularity
distribution on a control surface surrounding a body
coincides with the distribution that would be present at
that location for the body located in an infinite domain, the
body will experience no interference effects.13 For the
adaptive wall wind tunnel, the side and top walls of the
test section become the control surface. Shaping these
walls to correspond to streamlines on the open road
means that there will necessarily be no interference
effects to degrade the external aerodynamic simulation.
This implies that no correction of force coefficients is
needed and that the local aerodynamics on the vehicle are
properly simulated.
The concept of an adaptive wall test section actually
originated in aeronautical testing circles, and was
developed for automotive applications by Sverdrup
Technology in the 1980s, resulting in a patent.14 The
theoretical background, principle of operation, and results
from validation experiments for the adaptive wall test
section are presented elsewhere,1.13,14 and are not
repeated here. The result of this internal development
work, which spanned more than a decade, is the ability to
achieve interference-free external aerodynamic simulation
for model blockages of at least 30%. Over the past

decade, the technology has been successfully applied in

the motorsport community.
A cross-sectional schematic of the adaptive wall
configuration employed in Sverdrups sub-scale wind
tunnel laboratory is presented in Figure 1. The slats on
the top and side walls run the length of the test section
and are shaped by actuators spaced evenly along the
length of each slat. For the adaptive wall concept, the
actuators are part of a closed loop system that uses:

The streamwise distribution of static pressures

measured at the test section walls (which captures
the influence of the model)
A potential flow algorithm (which is independent of the
model
geometry)
that
uses
the
pressure
measurements and a convergence rate parameter to
predict new wall positions
A control system that translates the algorithm output
to commands for the wall actuators
A tolerance that defines convergence, which is
typically based on the convergence of the wall
pressure measurements for the final wall shape.

It is important to note that no portion of the adaptive wall

algorithm is dependent on the geometry of the model or
vehicle being tested. As a result, the technology is not
one where one must first know the answer to achieve it
in the wind tunnel, which is a common misperception.
As typically implemented, no more than 6 iterations are
usually required to achieve the final wall position. For a
fully-automated system, the entire wall shaping process
occurs while the wind tunnel is running, with only a few
seconds required for each iteration.
Previous results from Sverdrups sub-scale wind tunnel
have demonstrated the ability to achieve interference-free
results for model blockages up to 30%.13 Recent results
obtained for the standard MIRA fastback model3 in the
sub-scale adaptive wall test section are presented in
Figure 2. 11 The drag coefficient indicated at 11% blockage
in Fig. 2 was obtained in the adaptive wall test section
prior to any wall deflection (i.e. straight walls). The
circular points were obtained in independent tests in
other European wind tunnels.11 Note that, for all data
points, the ratio of the boundary layer displacement
thickness to the model underbody clearance is constant.
This implies consistent underbody effects for all data in
the plot. The faired curve through the data was based on
the data points at 1.7%, 2.7%, 4.8%, and 11% blockage.
The curve intersects the drag coefficient axis at
approximately 0.223. The square data point indicated at
0% blockage in the figure (CD = 0.223) is the result
obtained from the adaptive wall test section with a model
blockage of 11%.
Its exact coincidence with the
extrapolation to zero blockage (indicated by the faired
curve) illustrates the simulation quality attainable with
adaptive wall technology. Similar results are attained in
the adaptive wall test section for model blockages
approaching 30%.

Contoured Wall Test Section With an eye toward

motorsport applications, Sverdrup has investigated
different levels of wall adaptation to determine
requirements for different levels of simulation accuracy.
This has taken the form of both experimental and
computational investigations. The result of this work is
the contoured wall test section, in which both side walls
and the top wall of the test section each assume a twodimensional contour (i.e. there is only a single slat in the
side walls and top wall). The wall shapes can still be
modified, but the shape is modified via manual actuators
in an open loop configuration. This concept is similar to
the streamline wall configuration described by Hucho. 15
Both adaptive and contoured wall test sections enjoy a
decided advantage relative to solid wall, semi-open jet, or
slotted wall test sections. Because of their ability to
generate interference-free results at blockage ratios
approaching 30%, the size of the test section (and wind
tunnel) can be much smaller than if a traditional test
section configuration is employed. This represents an
important reduction in both capital cost and operating
cost for the facility, for superior simulation quality.
There are two main advantages of the contoured wall test
section versus the adaptive wall test section. The first is
that the contoured wall test section does not include the
automated, closed loop control system that shapes the
walls of the test section. This plus the reduced number
of wall slats creates a much simpler system, which also
represents an increment in reduced capital cost. The
basic principle of the contoured wall test section is to
obtain most of the simulation benefit of adaptive wall
technology in a simpler configuration.
The absence of an automated control system is most
ideally suited to motorsport applications.
This is
because a given team tests essentially the same model
for an entire season, if not multiple seasons. The same
would hold true for various motorsport teams within a
given classification using the same wind tunnel facility.
Hence the requirement for wall re-shaping is greatly
diminished relative to that which would be present in an
OEM contoured wall wind tunnel devoted to passenger
car development. In summary, the contoured wall test
section is a good solution for testing a single type of
vehicle where infrequent wall reshaping is required.
When an occasional wall redefinition is required, it is
achieved via manual actuation.
Simulation Quality in the Contoured Wall Test Section
Computational simulations have been carried out to
investigate the simulation quality attainable with the
contoured wall test section. A single sedan shape with a
frontal area of 23.0ft 2 (2.14m2) was employed as the test
article for all of the simulations. All simulations were
carried out with the PMARC potential flow code using a
parameter-based wake model to represent the vehicle
wake. The simulations were carried out for the full vehicle
oriented at 0 yaw and 7 yaw, and for the half-vehicle

oriented at 0 yaw, for each of the following three

geometries:

The open-road condition

The model in an 11.6ft x 23.2ft (3.54m x 7.07m) solid
wall wind tunnel (test section area of 269ft 2),
representing a model blockage of 8.5%
The model in a 9.0ft x 14.0ft (2.74m x 4.27m)
contoured wall test section (test section area of
126ft 2), representing a model blockage of 18.3%

Note that the solid wall test section area is 210% larger
than the contoured wall test section area. Figure 3
presents an illustration of the zero yaw simulation results
for the open road condition.
Table 1 presents the cumulative results of the
simulations. No attempt has been made to correct the
force and moment coefficients from the solid wall test
section for blockage effects. Examining Table 1, the
correspondence of the contoured wall test section to the
open road test section is very good, with deviations
ranging from 0.0% to approximately 3.5%. Comparing
the solid wall test section to the open road, the deviations
range from approximately 15% to more than 100%. The
contoured wall test section clearly provides superior
aerodynamic simulation quality to the solid wall test
sectioneven though the latter is more than twice as
large!
Just as important, it should be noted that the contoured
wall simulations for the vehicle yawed at 7 were run with
the test section wall shape that was optimized for 0
model yaw. Despite not having optimized the wall
shapes for the 7 yaw condition, Table 1 shows that there
is essentially zero degradation of the lift and drag
coefficients, and only very minimal degradation for the
pitching moment coefficient. This illustrates that wall
reshaping is not required for the range of yaw angles
expected for motorsport testing. Further, it directly
supports the fact that, for a given model geometry, wall
reshaping is required only infrequently for the contoured
wall test section.
In the final analysis, it is useful to compare the quality of
simulation attainable from adaptive wall wind tunnels and
contoured wall wind tunnels. Our internal computational
and experimental work, including the results presented in
Table 1, shows that the adaptive wall test section offers
the best possible aerodynamic simulation. The adaptive
wall test section has been proven capable of providing
force and moment coefficients with errors of no more than
0.5% to 1.5%, with no data correction required. These
results can be achieved for model blockages of at least
30%. The contoured wall test section provides force and
moment coefficients with errors of 1% to 5%, with no
data correction required. This level of performance can
be achieved for model blockages in excess of 20%with
20% blockage suggested as a comfortable design point.
As illustrated by the computational exercise presented

here, these simulation accuracies are far superior to that

which can be attained with traditional test section
configurations.
The ability to achieve quality aerodynamic simulation for
model blockages in excess of 20% with no need for data
correction has opened new possibilities for motorsport
development. For instance, this type of facility makes it
possible to do both model-scale and full-scale testing in
the same moderately-sized wind tunnel. The ability to do
meaningful full-scale testing in the wind tunnel, even if it
is possible only intermittently (e.g. during the racing offseason), could represent a major enhancement to a
teams ability to achieve performance on the track.

OVERVIEW OF MODERN ROLLING ROAD

TECHNOLOGY
Rolling road systems are a critical system for wind
tunnels dedicated to motorsport testing, and represent a
second area where modern wind tunnels are achieving
dramatic gains in simulation capability. As discussed in
the introduction, traditional systems were limited to a top
speed of 40-50m/s. This is fundamentally a limitation of
elastomer belt systems, which typically experience
severely degraded belt life for even intermittent operation
at high speeds.
The advanced rolling road systems of today have broken
through this limitation, offering a top speed capability of up
to 100m/s if required. This has been achieved by using a
stainless steel belt in place of the traditional elastomer
belt, a technology pioneered by MTS Systems through
their experience with tire testing. For the stainless steel
belt, routine operation at high speeds does not translate
to degraded belt life. A typical belt lifetime of 2000 testing
hours per unit is achieved with these high-speed systems.
The other primary breakthrough of modern rolling road
systems is the ability to measure forces transmitted
through the belt by the rotating tires. This evolution has
been driven by motorsport applications.
Two primary configurations for these high-speed rolling
road systems have emerged: a single, wide belt system
and a five belt system including narrow center belt.
Single, wide belt system Figure 4 presents a
photograph of an MTS Flat-Trac rolling road system.
The belt is fabricated of stainless steel. Belt widths can
range up to 10.5ft (3.2m). The flat length of the rolling
road can range up to 33.0ft (10.0m). The systems are
capable of top speeds up to 100m/s.
Figure 5 presents a photograph of an MTS Flat-Trac
downforce measurement system. These modules are
located under the belt directly beneath a rotating tire on
the model. The presence of four downforce measurement
systems at the locations of the rotating tires on the model

is illustrated in Figure 6a. Using load cell technology,

each downforce measurement system measures the full
vertical load at its respective tire, including both the weight
of the model and the aerodynamic downforce. As shown
in the figure, the force measurement modules and rolling
road system are typically integrated into a turntable to
allow the complete system to be yawed. Measurement
of side and axial forces is achieved through independent
means.
The measurement systems for these force
components are integrated into the vehicle restraint
system.
As shown in Figure 6b, modern rolling road systems are
now commonly being designed to handle both scale
model testing as well as full-scale testing. Vertical force
measurement through the belt is possible for both.
Five Belt System The other primary configuration for
modern rolling road systems is the five belt system.
Figure 7 presents a schematic of the MTS Flat-Trac five
belt system. As shown in the figure, this system is ideal
for use with a traditional external force balance, which
makes it an attractive option for upgrading existing wind
tunnels. The system consists of four mini-belts, one
beneath each tire, plus a narrow central belt that runs
from upstream of the vehicle to some downstream of the
vehicle. As suggested by its integration with an external
balance, these systems have so far been intended mainly
for full-scale vehicle testing.
As shown in Figure 8, the vehicle is actually supported by
four rocker panel support struts with integrated actuation.
These struts are connected directly to the vehicle
chassis, and provide automatic control of vehicle ride
height during testing. Both the vehicle support struts and
the mini-belts are supported by an intermediate support
frame which is connected to the external balance. The
net result is that all forces are transmitted to the external
balance. Thus, unlike the single belt system, the five belt
system does not alter the method with which forces are
measured.
The narrow center belts typically range in width from 1.0m
to 1.4m. The mini-belts beneath each tire typically range
in width from approximately 240mm to 410mm. Similar to
the single belt system, the five belt system is capable of
top speeds of up to 100m/s with excellent belt life
characteristics.
Single belt versus five belt Organizations developing
new motorsport wind tunnels have to this point generally
preferred the single belt system. For example, this is
especially true in the Formula One community. However,
the five belt system is ideal for upgrading a wind tunnel
that has an existing external balance. For example, a
similar upgrade solution was successfully implemented at
the Pininfarina Wind Tunnel.16 Even for a new wind tunnel,
the traditional force measurement configuration may
represent an advantage.

It is interesting to compare these advanced rolling road

systems to the common underbody simulation of a
boundary layer treatment system with blowing or
distributed suction, but no rolling roadthe conditions
under which a large portion of the wind tunnel testing for
stock car racing in North America occurs. Compared to
this, the five belt system provides the overwhelming
majority of the underbody simulation benefit to be gained
from a single-belt system. The five belt configuration is
also attractive for passenger car testing, providing dual
use possibilities that can be important for facilities that
execute both OEM testing and motorsport testing. Again,
this type of situation is most logical for upgrading a wind
tunnel with an existing external balance.

CONCLUSION
As the popularity of motorsport continues to grow, so to
does the effort spent on wind tunnel testing to gain
competitive advantage. The purpose of this paper is to
provide an overview of two areas of technology that are
having a major impact on wind tunnel testing dedicated to
motorsport.
Contoured wall test section technology has come to be
recognized as a cost-effective means of achieving quality
simulation. The technique is grounded in adaptive wall
technology, and provides the bulk of the advantages
associated with adaptive wall technology in a simpler,
less costly system. By enabling accurate aerodynamic
simulation at large model blockage, the technology allows
test objectives to be met in a much smaller facility than
would be required for traditional test section
configurations. These emerging test objectives include
full-scale testing in a moderately-sized facility, which can
be achieved. The results presented here demonstrate the
simulation advantage to be gained from a contoured wall
test section. Because the test section area drives both
the size and power consumption of a wind tunnel,
contoured wall technology also provides a gain in facility
cost effectiveness (capital and operational).
The other major impact area regarding wind tunnel testing
is rolling road systems. Single belt and five belt systems
are now available that substantially enhance the
underbody simulation capability of wind tunnel facilities,
including routine operation at speeds up to 100m/s. For
the single belt system, vertical forces can be measured
directly through the rolling road, for both model scale and
full scale testing. The five belt system can be integrated
directly into a traditional external balance, providing an
enhanced ability to simulate underbody effects while
maintaining the traditional force measurement system.
Both of these test section-focused technologies directly
enhance wind tunnel simulation quality, and therefore data
quality. As a result, the contoured wall test section
configuration and high-speed rolling road systems with

integrated force measurement are both gaining wide

acceptance in the motorsport community

ACKNOWLEDGMENTS
The authors would like to acknowledge the substantial
contributions of David Meier of MTS Systems
(David.Meier@MTS.com) regarding advanced rolling road
systems.

REFERENCES
1.

Arnette, S.A., Martindale, W.R., Meredith, W.S., and

Hoffman, J.H., Advances in Wind Tunnel
Aerodynamics for Motorsport Testing, SAE TOPTEC,
October 1999.
2. Kelly, K. B., Provencher, L. G., and Schenkel, F. K.,
"The General Motors Aerodynamic Laboratory - A
Full-Scale Automotive Wind Tunnel," SAE Paper
820371, 1982.
3. SAE Standards Committee on Closed Test Section
Wind Tunnel Boundary Corrections, "Closed-TestSection Wind Tunnel Blockage Corrections for Road
Vehicles," SAE SP-1176, 1996.
4. Garry, K. P., Cooper, K. R., Fediw, A., Wallis, S. B.,
and Wilsden, D. J., "The Effect on Aerodynamic Drag
of the Longitudinal Position of a Road Vehicle Model
in a Wind Tunnel Test Section," SAE Paper 940414,
1994.
5. Duell, E., Everstine, D., Mehta, R., Bell, J., and
Perry, M., "Pressure Sensitive Paint Technology
Applied to Low Speed Automotive Testing," to be
presented at the 2001 SAE Congress, March 2001.
6. Hoffman, J.M., Arnette, S.A., Porter, C.B., Sung, B.,
and Arik, B.E., Application of Particle Image
Velocimetry in the Korea Aerospace Research
Institute Low Speed Wind Tunnel, AIAA-2000-0411,
2000.
7. Elliott, G.S., and Beutner, T.J., "Molecular Filter
Based Planar Doppler Velocimetry," Progress in
Aerospace Sciences, 35(1999), pp. 799-845.
8. Arnette, S.A., Elliott, G.S., Mosedale, A.D., and
Carter, C.D., "A Two-Color Approach to Planar
Doppler Velicimetry," AIAA-98-0507, 1998.
9. SAE Standards Committee on Open Jet Wind Tunnel
Adjustments, Aerodynamic Testing of Road Vehicles
in Open Jet Wind Tunnels, SP-1465, 1999.
10. Mercker, E., and Wiedemann, J., "On Interference
Effects in Open Jet Wind Tunnels," SAE Paper
960671, 1996.
11. Hoffman, J., Martindale, W., Arnette, S., Williams, J.,
and Wallis, S., "Effect of Test Section Configuration
on Aerodynamic Drag Measurements," to be
presented at the 2001 SAE Congress, March 2001.
12. Arnette, S.A., Buchanan, T.D., and Zabat, M., "On
Low-Frequency Pressure Pulsations and static
Pressure Distribution in Open Jet Automotive Wind
Tunnels," SAE 1999-01-0813, 1999.

13. Whitfield, J.D., Jacocks, J.L., Dietz, W.E., and Pate,

S.R., "Demonstration of the Adaptive-Wall Concept
Applied to an Automotive Wind Tunnel," AIAA 820584, 1982.
14. Whitfield, J.D., Jacocks, J.L., Dietz, W.E., and Pate,
S.R., "Demonstration of the Adaptive-Wall Concept
Applied to an Automotive Wind Tunnel," SAE Paper
820373, 1982.
15. Hucho, W.-H., ed. Aerodynamics of Road Vehicles,
4th edition, SAE, 1998, p. 641.

16. Cogotti, A., Ground Effect Simulation for Full-Scale

Cars in the Pininfarina Wind Tunnel, SAE Paper
960996, 1996.

CONTACT
The primary contact for this paper is Stephen A. Arnette,
Ph.D., Vice President, Sverdrup Technology, Inc.
(arnettsa@sverdrup.com).

Table 1. Results of computational investigation of contoured wall simulation quality.

Complete Vehicle 0 Yaw
CL (lift force)

CD (drag force)

Cm (pitching moment)

Open Road

-0.2518

0.3680

0.1001

Contoured Wall Tunnel (18.3% blockage)

-0.2558

0.3654

0.1014

Solid Wall Tunnel (8.5%blockage)

-0.3048

0.2918

0.0852

(Cont Wall Open Road) / Open Road

1.59%

-0.71%

1.30%

(Solid Wall Open Road) / Open Road

21.05%

-20.71%

-14.89%

CY (side force)

Cn (yawing moment)

Cl (rolling moment)

Open Road

0.3852

-0.2896

-0.3371

Contoured Wall Tunnel (9.2% blockage)

0.3854

-0.2976

-0.3369

Solid Wall Tunnel (4.3% blockage)

0.8986

-0.1256

-0.746

(Cont. Wall Open Road) / Open Road

0.05%

2.76%

-0.06%

133.28%

-56.63%

121.30%

CL (lift force)

CD (drag force)

Cm (pitching moment)

Open Road

-0.2536

0.3768

0.1008

Contoured Wall Tunnel (18.3% blockage)

-0.2497

0.3792

0.1043

Solid Wall Tunnel (8.5% blockage)

-0.3088

0.3019

0.0853

(Cont Wall Open Road) / Open Road

-1.54%

0.64%

3.47%

(Solid Wall Open Road) / Open Road

21.77%

-19.88%

-15.38%

Half Vehicle 0 Yaw

(Solid WallOpen Road) / Open Road

Complete Vehicle 7 Yaw

Figure 1. Schematic of Sverdrups sub-scale adaptive wall test section.

Figure 2. Vehicle surface pressure distribution obtained from CFD for

a sedan shape oriented at zero yaw for interference-free conditions.

0.30

0.29

Passenger Car Shape

0.28
Sverdrup Adaptive Wall Tunnel
Straight Wall
Geometric Blockage = 0.11
*/H = 0.15

0.27
Trendline

Cd

0.26

0.25

0.24

0.23

College Of Aeronautics
Cranfield Institute of Technology
Straight Wall Test Section
reported in SAE SP-1176
*/H = 0.16

0.22

0.21

0.20
0.00

Sverdrup Adaptive Wall Tunnel

Adapted Wall
Geometric Blockage = 0.11
*/H = 0.15
0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

0.11

Blockage Ratio

Figure 3. Drag coefficients for the MIRA fastback model from various wind tunnels. The regressed curve
serves as an extrapolation to zero blockage for the four points obtained at various blockages. The solid
point at 11% blockage was obtained in the adaptive wall test section with no wall shaping (i.e. straight
walls). The solid point at 0% blockage was also obtained in the adaptive wall test section at a model
blockage of 11%, but after optimizing the wall contours.

Figure 4. Photograph of an MTS Flat-Trac rolling road system.

0.12

Figure 5. Photograph of an MTS Flat-Trac vertical force measurement system.

Figure 6. Scale model (top) and full scale (bottom) installations on a rolling road system, with vertical
force measurement systems located beneath the rolling road. (illustration courtesy of MTS).

DISTRIBUTED SUCTION
FLOOR PANELS
NARROW
(CENTER) BELT

ADD-ON
AIR BEARING
PANELS
FORCE
BALANCE

ROTATING WHEEL
MINI-BELT

5-BELT CONCEPT
MOVING BELT SYSTEM

Figure 7. Schematic of a car installed on an MTS Flat-Trac five belt rolling road system that has
been integrated with an external force balance (illlustration courtesy of MTS).

TURNTABLE
NARROW
BELT
ROCKER PANEL
SUPPORT ACTUATORS

MINI-BELTS

FORCE
BALANCE

Figure 8. Schematic of the MTS Flat-Trac five belt system integrated with an external force
balance (illustration courtesy of MTS).