-
400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel:
(724) 776-4841 Fax: (724) 776-5760
SAE TECHNICALPAPER SERIES 2000-01-3549
Advances in Wind Tunnel Aerodynamics forMotorsport Testing
Steve Arnette and Bill MartindaleSverdrup Technology, Inc.
Reprinted From: Proceedings of the 2000 SAE
MotorsportsEngineering Conference & Exposition
(P-361)
Motorsports Engineering Conference & ExpositionDearborn,
Michigan
November 13-16, 2000
-
The appearance of this ISSN code at the bottom of this page
indicates SAEs consent that copies of thepaper 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 Sec-tions 107 or 108 of the U.S. Copyright Law.
This consent does not extend to other kinds of copying such
ascopying 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 yourorders 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 inother works, contact the SAE
Publications Group.
No part of this publication may be reproduced in any form, in an
electronic retrieval system or otherwise, without the prior
writtenpermission of the publisher.
ISSN 0148-7191Copyright 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
solelyresponsible for the content of the paper. A process is
available by which discussions will be printed with the paper if it
is published inSAE 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 300word abstract of a proposed manuscript to: Secretary,
Engineering Meetings Board, SAE.
Printed in USA
All SAE papers, standards, and selectedbooks are abstracted and
indexed in theGlobal Mobility Database
-
2000-01-3549
Advances in Wind Tunnel Aerodynamics for Motorsport Testing
Steve Arnette and Bill MartindaleSverdrup Technology, Inc.
Copyright 2000 Society of Automotive Engineers, Inc.
ABSTRACT
As the popularity of motorsport continues to surgethroughout the
world, so to does the level of competition inthe motorsport
community. Participants work to achievea performance edge through
superior engineering. As anenabling tool, the wind tunnel has
become a focus forenhancing performance. This is evidenced by
theincreasing interest among motorsport teams in dedicatedwind
tunnel facilities, as best exemplified by the FormulaOne community.
Part of the reason for this increasingfocus on wind tunnels is the
availability of breakthroughtechnologies that better simulate
on-track conditions,providing new opportunities to enhance
performance. Twoareas that are the subject of strong current
interest are (i)test section configurations that eliminate wind
tunnelinterference effects to provide the highest
possibleaerodynamic simulation fidelity and (ii) high speed
rollingroad systems with integrated force measurement systemsthat
provide high fidelity simulation of underbody effects.This paper
presents an overview of these technologies,including selected
computational and experimental resultsthat illustrate the
simulation advantages obtained withthese new wind tunnel
technologies.
INTRODUCTION
The increasing emphasis of wind tunnel usage in themotorsport
community is well exemplified by the FormulaOne community.
Initially, the wind tunnels employedwere typically of the
open-return type designed for testingof models at 25% scale. Over
time, standard practiceevolved to testing at 40% scale. Many of the
facilities inuse were converted aeronautical wind tunnels.
Theseearly facilities contained some of the first rolling
roadsystems to simulate underbody aerodynamics in the windtunnel,
although the rolling road systems were confined tolow speeds
relative to those of actual competition.
A snapshot of the wind tunnels currently being used inthis
community shows that the majority of the testing isoccurring at
model scales of either 40% or 50%. The
increasing model scale allows better geometry fidelity ofthe
model to the actual vehicle. The typical configurationfor these
wind tunnels is a solid wall test section with across-sectional
area of 7-14m2. The maximum windspeed capability of these
facilities is typically 50 m/s, butranges as high as 70m/s in some
cases. Note that thesespeeds represent the limiting capability of
traditionalrolling road systems, where concern for the life of
therolling road belt severely limits high speed operation.
The wind tunnels currently being built or planned for theFormula
One community represent yet another incrementin simulation
capability. Testing is planned for a typicalmodel scale of 60%,
with the added requirement to obtainquality simulation for
full-scale testing. The top speed ofthese facilities is at least
70m/s, with most beingdesigned for an 80m/s capability. Most of
these facilitiesare being designed with non-traditional test
sectionconfigurations, driven by the needs for (i)
superioraerodynamic simulation at 60% model scale and (ii)
highquality simulation fidelity for full-scale testing. The
rollingroad systems of these facilities represent a
dramaticincrement in top speed capability, with systems capableof
routine operation at speeds up to 100m/s if required.
The purpose of this paper is to provide an overview ofemerging
technologies that are prominent in the newgeneration of wind
tunnels dedicated to motorsporttesting. Chief among these are (i)
test sectionsemploying contoured wall technology to
eliminateboundary interference effects which degrade
aerodynamicsimulation quality and (ii) next-generation rolling
roadsystems which provide a dramatic increase in top
speedcapability and the possibility of measuring aerodynamicforces
transmitted to the rolling road through the rotatingtires. A
discussion of these technologies is timely giventhe increasing
emphasis being given to wind tunnel testingfor both stock car and
open wheel classifications. Thebasis for this invited paper was an
invited SAE TOPTECpresentation given in October 1999.1
-
OVERVIEW OF TEST SECTION TECHNOLOGY
Of the wind tunnels currently used for motorsportdevelopment,
the vast majority possess either solid wall orsemi-open jet test
sections. For both types, boundaryinterference causes the external
aerodynamic simulationin the wind tunnel to differ from the actual
situation on theroad.
SOLID WALL TEST SECTION For the solid wall testsection, boundary
interference is fundamentally caused bythe confinement of the flow
between the test model (orvehicle) and the solid test section
boundary. Thisunavoidable effect results in over-acceleration of
the flowover the model, as compared to the actual situation on
theroad. To minimize these effects, it is advisable to hold
thevehicle blockage ratio (ratio of model frontal area to
crosssectional area of test section) to no more than about 5%.This
is why solid wall wind tunnels devoted to full-scaleautomotive
development have large test sections, e.g. theGeneral Motors
Aerodynamic Laboratory Wind Tunnel2
has a test section area of approximately 610 ft2 (57m2).Because
of the bulk of work done in solid wall windtunnels, spanning both
automotive and aerospace productdevelopment, the topic of solid
wall boundary interferencehas been investigated extensively.
Seven different wall boundary correction methods werecompared in
an SAE-sponsored exercise to assess theaccuracy of correction
methods.3 Of the seven, three werejudged to be acceptable. Even the
acceptable methodsfailed to provide an adequate correction at
blockage ratiosof 10%. Of the three acceptable methods, the
pressuresignature correction method represents the most
rigorousaerodynamic development. Its use, however, has tendedto be
limited by the requirement for wall pressuresignature measurements
and lengthy calculations tocomplete the correction procedure. Even
if the one of thepreferred correction methods is applied, the
resultsindicate that testing must be carried out at blockageratios
much less than 10% to produce accurate testdata.3
It is important to note that sources of degradation otherthan
boundary interference are possible in the solid walltest section.
For example, the combination of largemodel blockage and model
location in close proximity tothe contraction section can result in
model influence onthe wind tunnels dynamic pressure measurement.
Ifpresent, the result will be a facility-indicated dynamicpressure
different than the actual dynamic pressureincident on the vehicle.
Another potential source ofdegradation is static pressure influence
from thedownstream diffuser. As shown by Garry et al.,4 if thebase
of the model is too close to the diffuser at thedownstream end of
the test section, the interaction of thediffuser and the model wake
will cause the static pressureat the base of the model to be
artificially elevated, therebyaltering the pressure forces on the
vehicle. Note that both
of these potential sources of simulation error can beminimized
through proper facility design. Boundaryinterference, however, can
be minimized only throughincreasing the test section area, which
translates directlyto increased capital cost of the wind tunnel
facility.
Although correction procedures such as the pressuresignature
method help close the gap between the windtunnel and actual on-road
aerodynamic performance, itwould obviously be preferable to measure
actual on-roadperformance in the wind tunnel. This is increasingly
trueas motorsport becomes more competitive. With racecardevelopers
looking for aerodynamic advantage with thedesign of essentially
every element on the vehicle,situations will arise where the
performance incrementsbetween various design options are much
smaller than thecorrection increment needed to translate the wind
tunnelmeasurements to on-road performance. Along a similarline of
thought, the need for correction arisesfundamentally from improper
flow simulation over themodel. As advanced diagnostic tools such as
PressureSensitive Paint,5 Particle Image Velocimetry,6 and
PlanarDoppler Velocimetry7,8 which provide detailed,
spatially-distributed mappings of local flow structure becomemore
commonplace in the wind tunnel, proper flowsimulation at localized
positions on the vehicle isbecoming increasingly important.
Although post-testcorrections are suitable for force and moment
coefficients,they are not capable of improving flow simulation in
thewind tunnel.
OPEN JET TEST SECTION Many wind tunnels usedfor motorsport
development possess an open jet testsection. For this
configuration, the test model is placed ina large plenum chamber
and the nozzle flow entering theplenum occupies only a modest
portion of the plenumscross-section. After flowing over the model,
the flow isthen collected at the rear of the plenum. Historically,
thistype of test section has been preferred in Europe, basedon the
idea that aerodynamic simulation fidelity to theopen road is less
degraded by increasing model blockagethan it is for a solid wall
test section. For the open jettest section, model blockage ratio is
defined as the ratioof the model frontal area to the cross
sectional area of thenozzle. In the semi-open jet test section
(where the modelblockage is typically 10% or more), the main source
ofdegraded aerodynamic simulation is the finite dimensionof the jet
exiting the nozzle. On the road, the vehicle isimmersed in a
semi-infinite flow.
Blockage corrections for the semi-open jet test sectionare less
mature than those for the solid wall test section.9
The work of Mercker and Wiedemann10 is the mostcomprehensive
work to date. As a result, it is difficult tocomment definitively
on the success of open jet correctionmethods. It is noted, however,
that the recentexperimental investigation of Hoffman et al.11
indicatesthat the aerodynamic simulation obtained from open jettest
sections is superior to that obtained from slotted walltest
sections for model blockages ranging from 7% to
-
25%. The lack of knowledge regarding quantitativeperformance is
the driving reason that essentially all newwind tunnels planned for
the near future that will bededicated to motorsport testing will
not have an open jettest section.
Open jet test sections are universally beset by an adversestatic
pressure gradient in the downstream portion of thetest section.9,12
The common result is that test vehiclesinstalled in the test
section are subjected to an elevatedstatic pressure in the base
region. This horizontalbuoyancy effect is exactly analogous to that
encounteredin a solid wall test section if the test vehicle is
located tooclose to the downstream diffuser. Although the
obviousanswer is to increase the length of the open jet, Arnette
etal.12 have shown that open jet test sections exhibit anincreasing
tendency towards low-frequency unsteadinessas the length of the
open jet increases. The result is atrade-off situation for the wind
tunnel designer. Alsosimilar to the solid wall test section,
locating the testvehicle or model too close to the nozzle can lead
to modelinfluence on the facility measurement of dynamicpressure,
with similar negative effects.
CONTOURED WALL TEST SECTION Becausecontoured wall technology has
its origin in adaptive walltechnology, a review of adaptive wall
test sections ispresented prior to discussing the details of
contoured walltest sections.
Adaptive Wall Test Section The idea of an adaptive walltest
section is simply to shape the side and top walls ofthe test
section such that they correspond to externalstreamlines that would
be present over the vehicle on theopen road. More formally stated,
if the flow angularitydistribution on a control surface surrounding
a bodycoincides with the distribution that would be present atthat
location for the body located in an infinite domain, thebody will
experience no interference effects.13 For theadaptive wall wind
tunnel, the side and top walls of thetest section become the
control surface. Shaping thesewalls to correspond to streamlines on
the open roadmeans that there will necessarily be no
interferenceeffects to degrade the external aerodynamic
simulation.This implies that no correction of force coefficients
isneeded and that the local aerodynamics on the vehicle areproperly
simulated.
The concept of an adaptive wall test section actuallyoriginated
in aeronautical testing circles, and wasdeveloped for automotive
applications by SverdrupTechnology in the 1980s, resulting in a
patent.14 Thetheoretical background, principle of operation, and
resultsfrom validation experiments for the adaptive wall
testsection are presented elsewhere,1.13,14 and are notrepeated
here. The result of this internal developmentwork, which spanned
more than a decade, is the ability toachieve interference-free
external aerodynamic simulationfor model blockages of at least 30%.
Over the past
decade, the technology has been successfully applied inthe
motorsport community.
A cross-sectional schematic of the adaptive wallconfiguration
employed in Sverdrups sub-scale windtunnel laboratory is presented
in Figure 1. The slats onthe top and side walls run the length of
the test sectionand are shaped by actuators spaced evenly along
thelength of each slat. For the adaptive wall concept, theactuators
are part of a closed loop system that uses:
The streamwise distribution of static pressuresmeasured at the
test section walls (which capturesthe influence of the model)
A potential flow algorithm (which is independent of themodel
geometry) that uses the pressuremeasurements and a convergence rate
parameter topredict new wall positions
A control system that translates the algorithm outputto commands
for the wall actuators
A tolerance that defines convergence, which istypically based on
the convergence of the wallpressure measurements for the final wall
shape.
It is important to note that no portion of the adaptive
wallalgorithm is dependent on the geometry of the model orvehicle
being tested. As a result, the technology is notone where one must
first know the answer to achieve itin the wind tunnel, which is a
common misperception.As typically implemented, no more than 6
iterations areusually required to achieve the final wall position.
For afully-automated system, the entire wall shaping processoccurs
while the wind tunnel is running, with only a fewseconds required
for each iteration.
Previous results from Sverdrups sub-scale wind tunnelhave
demonstrated the ability to achieve interference-freeresults for
model blockages up to 30%.13 Recent resultsobtained for the
standard MIRA fastback model3 in thesub-scale adaptive wall test
section are presented inFigure 2.11 The drag coefficient indicated
at 11% blockagein Fig. 2 was obtained in the adaptive wall test
sectionprior to any wall deflection (i.e. straight walls).
Thecircular points were obtained in independent tests inother
European wind tunnels.11 Note that, for all datapoints, the ratio
of the boundary layer displacementthickness to the model underbody
clearance is constant.This implies consistent underbody effects for
all data inthe plot. The faired curve through the data was based
onthe data points at 1.7%, 2.7%, 4.8%, and 11% blockage.The curve
intersects the drag coefficient axis atapproximately 0.223. The
square data point indicated at0% blockage in the figure (CD =
0.223) is the resultobtained from the adaptive wall test section
with a modelblockage of 11%. Its exact coincidence with
theextrapolation to zero blockage (indicated by the fairedcurve)
illustrates the simulation quality attainable withadaptive wall
technology. Similar results are attained inthe adaptive wall test
section for model blockagesapproaching 30%.
-
Contoured Wall Test Section With an eye towardmotorsport
applications, Sverdrup has investigateddifferent levels of wall
adaptation to determinerequirements for different levels of
simulation accuracy.This has taken the form of both experimental
andcomputational investigations. The result of this work isthe
contoured wall test section, in which both side wallsand the top
wall of the test section each assume a two-dimensional contour
(i.e. there is only a single slat in theside walls and top wall).
The wall shapes can still bemodified, but the shape is modified via
manual actuatorsin an open loop configuration. This concept is
similar tothe streamline wall configuration described by
Hucho.15
Both adaptive and contoured wall test sections enjoy adecided
advantage relative to solid wall, semi-open jet, orslotted wall
test sections. Because of their ability togenerate
interference-free results at blockage ratiosapproaching 30%, the
size of the test section (and windtunnel) can be much smaller than
if a traditional testsection configuration is employed. This
represents animportant reduction in both capital cost and
operatingcost for the facility, for superior simulation
quality.
There are two main advantages of the contoured wall testsection
versus the adaptive wall test section. The first isthat the
contoured wall test section does not include theautomated, closed
loop control system that shapes thewalls of the test section. This
plus the reduced numberof wall slats creates a much simpler system,
which alsorepresents an increment in reduced capital cost. Thebasic
principle of the contoured wall test section is toobtain most of
the simulation benefit of adaptive walltechnology in a simpler
configuration.
The absence of an automated control system is mostideally suited
to motorsport applications. This isbecause a given team tests
essentially the same modelfor an entire season, if not multiple
seasons. The samewould hold true for various motorsport teams
within agiven classification using the same wind tunnel
facility.Hence the requirement for wall re-shaping is
greatlydiminished relative to that which would be present in anOEM
contoured wall wind tunnel devoted to passengercar development. In
summary, the contoured wall testsection is a good solution for
testing a single type ofvehicle where infrequent wall reshaping is
required.When an occasional wall redefinition is required, it
isachieved via manual actuation.
Simulation Quality in the Contoured Wall Test
SectionComputational simulations have been carried out
toinvestigate the simulation quality attainable with thecontoured
wall test section. A single sedan shape with afrontal area of
23.0ft2 (2.14m2) was employed as the testarticle for all of the
simulations. All simulations werecarried out with the PMARC
potential flow code using aparameter-based wake model to represent
the vehiclewake. The simulations were carried out for the full
vehicleoriented at 0 yaw and 7 yaw, and for the half-vehicle
oriented at 0 yaw, for each of the following
threegeometries:
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 269ft2),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 of126ft2), representing a model blockage
of 18.3%
Note that the solid wall test section area is 210% largerthan
the contoured wall test section area. Figure 3presents an
illustration of the zero yaw simulation resultsfor the open road
condition.
Table 1 presents the cumulative results of thesimulations. No
attempt has been made to correct theforce and moment coefficients
from the solid wall testsection for blockage effects. Examining
Table 1, thecorrespondence of the contoured wall test section to
theopen road test section is very good, with deviationsranging from
0.0% to approximately 3.5%. Comparingthe solid wall test section to
the open road, the deviationsrange from approximately 15% to more
than 100%. Thecontoured wall test section clearly provides
superioraerodynamic simulation quality to the solid wall
testsectioneven though the latter is more than twice aslarge!
Just as important, it should be noted that the contouredwall
simulations for the vehicle yawed at 7 were run withthe test
section wall shape that was optimized for 0model yaw. Despite not
having optimized the wallshapes for the 7 yaw condition, Table 1
shows that thereis essentially zero degradation of the lift and
dragcoefficients, and only very minimal degradation for thepitching
moment coefficient. This illustrates that wallreshaping is not
required for the range of yaw anglesexpected for motorsport
testing. Further, it directlysupports the fact that, for a given
model geometry, wallreshaping is required only infrequently for the
contouredwall test section.
In the final analysis, it is useful to compare the quality
ofsimulation attainable from adaptive wall wind tunnels
andcontoured wall wind tunnels. Our internal computationaland
experimental work, including the results presented inTable 1, shows
that the adaptive wall test section offersthe best possible
aerodynamic simulation. The adaptivewall test section has been
proven capable of providingforce and moment coefficients with
errors of no more than0.5% to 1.5%, with no data correction
required. Theseresults can be achieved for model blockages of at
least30%. The contoured wall test section provides force andmoment
coefficients with errors of 1% to 5%, with nodata correction
required. This level of performance canbe achieved for model
blockages in excess of 20%with20% blockage suggested as a
comfortable design point.As illustrated by the computational
exercise presented
-
here, these simulation accuracies are far superior to thatwhich
can be attained with traditional test sectionconfigurations.
The ability to achieve quality aerodynamic simulation formodel
blockages in excess of 20% with no need for datacorrection has
opened new possibilities for motorsportdevelopment. For instance,
this type of facility makes itpossible to do both model-scale and
full-scale testing inthe same moderately-sized wind tunnel. The
ability to domeaningful full-scale testing in the wind tunnel, even
if itis possible only intermittently (e.g. during the racing
off-season), could represent a major enhancement to ateams ability
to achieve performance on the track.
OVERVIEW OF MODERN ROLLING ROADTECHNOLOGY
Rolling road systems are a critical system for windtunnels
dedicated to motorsport testing, and represent asecond area where
modern wind tunnels are achievingdramatic gains in simulation
capability. As discussed inthe introduction, traditional systems
were limited to a topspeed of 40-50m/s. This is fundamentally a
limitation ofelastomer belt systems, which typically
experienceseverely degraded belt life for even intermittent
operationat high speeds.
The advanced rolling road systems of today have brokenthrough
this limitation, offering a top speed capability of upto 100m/s if
required. This has been achieved by using astainless steel belt in
place of the traditional elastomerbelt, a technology pioneered by
MTS Systems throughtheir experience with tire testing. For the
stainless steelbelt, routine operation at high speeds does not
translateto degraded belt life. A typical belt lifetime of 2000
testinghours per unit is achieved with these high-speed
systems.
The other primary breakthrough of modern rolling roadsystems is
the ability to measure forces transmittedthrough the belt by the
rotating tires. This evolution hasbeen driven by motorsport
applications.
Two primary configurations for these high-speed rollingroad
systems have emerged: a single, wide belt systemand a five belt
system including narrow center belt.
Single, wide belt system Figure 4 presents aphotograph of an MTS
Flat-Trac rolling road system.The belt is fabricated of stainless
steel. Belt widths canrange up to 10.5ft (3.2m). The flat length of
the rollingroad can range up to 33.0ft (10.0m). The systems
arecapable of top speeds up to 100m/s.
Figure 5 presents a photograph of an MTS Flat-Trac
downforce measurement system. These modules arelocated under the
belt directly beneath a rotating tire onthe model. The presence of
four downforce measurementsystems 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 fullvertical load at its
respective tire, including both the weightof the model and the
aerodynamic downforce. As shownin the figure, the force measurement
modules and rollingroad system are typically integrated into a
turntable toallow the complete system to be yawed. Measurementof
side and axial forces is achieved through independentmeans. The
measurement systems for these forcecomponents are integrated into
the vehicle restraintsystem.
As shown in Figure 6b, modern rolling road systems arenow
commonly being designed to handle both scalemodel testing as well
as full-scale testing. Vertical forcemeasurement through the belt
is possible for both.
Five Belt System The other primary configuration formodern
rolling road systems is the five belt system.Figure 7 presents a
schematic of the MTS Flat-Trac fivebelt system. As shown in the
figure, this system is idealfor use with a traditional external
force balance, whichmakes it an attractive option for upgrading
existing windtunnels. The system consists of four mini-belts,
onebeneath each tire, plus a narrow central belt that runsfrom
upstream of the vehicle to some downstream of thevehicle. As
suggested by its integration with an externalbalance, these systems
have so far been intended mainlyfor full-scale vehicle testing.
As shown in Figure 8, the vehicle is actually supported byfour
rocker panel support struts with integrated actuation.These struts
are connected directly to the vehiclechassis, and provide automatic
control of vehicle rideheight during testing. Both the vehicle
support struts andthe mini-belts are supported by an intermediate
supportframe which is connected to the external balance. Thenet
result is that all forces are transmitted to the externalbalance.
Thus, unlike the single belt system, the five beltsystem does not
alter the method with which forces aremeasured.
The narrow center belts typically range in width from 1.0mto
1.4m. The mini-belts beneath each tire typically rangein width from
approximately 240mm to 410mm. Similar tothe single belt system, the
five belt system is capable oftop speeds of up to 100m/s with
excellent belt lifecharacteristics.
Single belt versus five belt Organizations developingnew
motorsport wind tunnels have to this point generallypreferred the
single belt system. For example, this isespecially true in the
Formula One community. However,the five belt system is ideal for
upgrading a wind tunnelthat has an existing external balance. For
example, asimilar upgrade solution was successfully implemented
atthe Pininfarina Wind Tunnel.16 Even for a new wind tunnel,the
traditional force measurement configuration mayrepresent an
advantage.
-
It is interesting to compare these advanced rolling roadsystems
to the common underbody simulation of aboundary layer treatment
system with blowing ordistributed suction, but no rolling roadthe
conditionsunder which a large portion of the wind tunnel testing
forstock car racing in North America occurs. Compared tothis, the
five belt system provides the overwhelmingmajority of the underbody
simulation benefit to be gainedfrom a single-belt system. The five
belt configuration isalso attractive for passenger car testing,
providing dualuse possibilities that can be important for
facilities thatexecute both OEM testing and motorsport testing.
Again,this type of situation is most logical for upgrading a
windtunnel with an existing external balance.
CONCLUSION
As the popularity of motorsport continues to grow, so todoes the
effort spent on wind tunnel testing to gaincompetitive advantage.
The purpose of this paper is toprovide an overview of two areas of
technology that arehaving a major impact on wind tunnel testing
dedicated tomotorsport.
Contoured wall test section technology has come to berecognized
as a cost-effective means of achieving qualitysimulation. The
technique is grounded in adaptive walltechnology, and provides the
bulk of the advantagesassociated with adaptive wall technology in a
simpler,less costly system. By enabling accurate
aerodynamicsimulation at large model blockage, the technology
allowstest objectives to be met in a much smaller facility
thanwould be required for traditional test sectionconfigurations.
These emerging test objectives includefull-scale testing in a
moderately-sized facility, which canbe achieved. The results
presented here demonstrate thesimulation advantage to be gained
from a contoured walltest section. Because the test section area
drives boththe size and power consumption of a wind
tunnel,contoured wall technology also provides a gain in
facilitycost effectiveness (capital and operational).
The other major impact area regarding wind tunnel testingis
rolling road systems. Single belt and five belt systemsare now
available that substantially enhance theunderbody simulation
capability of wind tunnel facilities,including routine operation at
speeds up to 100m/s. Forthe single belt system, vertical forces can
be measureddirectly through the rolling road, for both model scale
andfull scale testing. The five belt system can be
integrateddirectly into a traditional external balance, providing
anenhanced ability to simulate underbody effects whilemaintaining
the traditional force measurement system.
Both of these test section-focused technologies directlyenhance
wind tunnel simulation quality, and therefore dataquality. As a
result, the contoured wall test sectionconfiguration and high-speed
rolling road systems with
integrated force measurement are both gaining wideacceptance in
the motorsport community
ACKNOWLEDGMENTS
The authors would like to acknowledge the
substantialcontributions of David Meier of MTS
Systems([email protected]) regarding advanced rolling
roadsystems.
REFERENCES
1. Arnette, S.A., Martindale, W.R., Meredith, W.S., andHoffman,
J.H., Advances in Wind TunnelAerodynamics for Motorsport Testing,
SAE TOPTEC,October 1999.
2. Kelly, K. B., Provencher, L. G., and Schenkel, F. K.,"The
General Motors Aerodynamic Laboratory - AFull-Scale Automotive Wind
Tunnel," SAE Paper820371, 1982.
3. SAE Standards Committee on Closed Test SectionWind Tunnel
Boundary Corrections, "Closed-Test-Section Wind Tunnel Blockage
Corrections for RoadVehicles," SAE SP-1176, 1996.
4. Garry, K. P., Cooper, K. R., Fediw, A., Wallis, S. B.,and
Wilsden, D. J., "The Effect on Aerodynamic Dragof the Longitudinal
Position of a Road Vehicle Modelin a Wind Tunnel Test Section," SAE
Paper 940414,1994.
5. Duell, E., Everstine, D., Mehta, R., Bell, J., andPerry, M.,
"Pressure Sensitive Paint TechnologyApplied to Low Speed Automotive
Testing," to bepresented 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 ImageVelocimetry in the Korea
Aerospace ResearchInstitute Low Speed Wind Tunnel,
AIAA-2000-0411,2000.
7. Elliott, G.S., and Beutner, T.J., "Molecular FilterBased
Planar Doppler Velocimetry," Progress inAerospace Sciences,
35(1999), pp. 799-845.
8. Arnette, S.A., Elliott, G.S., Mosedale, A.D., andCarter,
C.D., "A Two-Color Approach to PlanarDoppler Velicimetry,"
AIAA-98-0507, 1998.
9. SAE Standards Committee on Open Jet Wind TunnelAdjustments,
Aerodynamic Testing of Road Vehiclesin Open Jet Wind Tunnels,
SP-1465, 1999.
10. Mercker, E., and Wiedemann, J., "On InterferenceEffects in
Open Jet Wind Tunnels," SAE Paper960671, 1996.
11. Hoffman, J., Martindale, W., Arnette, S., Williams, J.,and
Wallis, S., "Effect of Test Section Configurationon Aerodynamic
Drag Measurements," to bepresented at the 2001 SAE Congress, March
2001.
12. Arnette, S.A., Buchanan, T.D., and Zabat, M.,
"OnLow-Frequency Pressure Pulsations and staticPressure
Distribution in Open Jet Automotive WindTunnels," SAE 1999-01-0813,
1999.
-
13. Whitfield, J.D., Jacocks, J.L., Dietz, W.E., and Pate,S.R.,
"Demonstration of the Adaptive-Wall ConceptApplied to an Automotive
Wind Tunnel," AIAA 82-0584, 1982.
14. Whitfield, J.D., Jacocks, J.L., Dietz, W.E., and Pate,S.R.,
"Demonstration of the Adaptive-Wall ConceptApplied to an Automotive
Wind Tunnel," SAE Paper820373, 1982.
15. Hucho, W.-H., ed. Aerodynamics of Road Vehicles,4th edition,
SAE, 1998, p. 641.
16. Cogotti, A., Ground Effect Simulation for Full-ScaleCars in
the Pininfarina Wind Tunnel, SAE Paper960996, 1996.
CONTACT
The primary contact for this paper is Stephen A. Arnette,Ph.D.,
Vice President, Sverdrup Technology,
Inc.([email protected]).
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%
Half Vehicle 0 Yaw
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%
(Solid WallOpen Road) / Open Road 133.28% -56.63% 121.30%
Complete Vehicle 7 Yaw
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%
-
Figure 1. Schematic of Sverdrups sub-scale adaptive wall test
section.
Figure 2. Vehicle surface pressure distribution obtained from
CFD fora sedan shape oriented at zero yaw for interference-free
conditions.
-
Figure 3. Drag coefficients for the MIRA fastback model from
various wind tunnels. The regressed curveserves as an extrapolation
to zero blockage for the four points obtained at various blockages.
The solidpoint at 11% blockage was obtained in the adaptive wall
test section with no wall shaping (i.e. straightwalls). The solid
point at 0% blockage was also obtained in the adaptive wall test
section at a modelblockage of 11%, but after optimizing the wall
contours.
Figure 4. Photograph of an MTS Flat-Trac rolling road
system.
0.20
0.21
0.22
0.23
0.24
0.25
0.26
0.27
0.28
0.29
0.30
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11
0.12
Blockage Ratio
Cd
College Of AeronauticsCranfield Institute of Technology
Straight Wall Test Sectionreported in SAE SP-1176
d*/H = 0.16Sverdrup Adaptive Wall Tunnel
Adapted WallGeometric Blockage = 0.11
d*/H = 0.15
Trendline
Sverdrup Adaptive Wall TunnelStraight Wall
Geometric Blockage = 0.11d*/H = 0.15
Passenger Car Shape
-
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 verticalforce
measurement systems located beneath the rolling road. (illustration
courtesy of MTS).
-
Figure 7. Schematic of a car installed on an MTS Flat-Trac five
belt rolling road system that hasbeen integrated with an external
force balance (illlustration courtesy of MTS).
Figure 8. Schematic of the MTS Flat-Trac five belt system
integrated with an external forcebalance (illustration courtesy of
MTS).
5-BELT CONCEPTMOVING BELT SYSTEM
ADD-ONAIR BEARINGPANELS
NARROW(CENTER) BELT
DISTRIBUTED SUCTIONFLOOR PANELS
ROTATING WHEELMINI-BELT
FORCEBALANCE
ROCKER PANELSUPPORT ACTUATORS
MINI-BELTS
NARROWBELT
TURNTABLE
FORCEBALANCE