Вы находитесь на странице: 1из 18

NASA Technical Memorandum 1O

w
/

Accuracy Requirements and


Benchmark Experiments for
CFD Validation
Joseph G.Marvin
(NASA-TLI- 100087)
BCCUBACY R E Q U I R E H E N T S A N U
BENCHMARK EXPERIHENTS F08 CPD Y A L I D A T I O N
(NASA)
18 p
C S C L 200

N88-21423

Unclas
G3/34

NationalAeronautics and

space Administration

0140259

NASA Technical Memorandum 100087

Accuracy Requirements and


Benchmark Experiments for
CFD Validation
~~

Joseph G.Marvin, Ames Research Center, Moffett Field, California

National Aeronautics and


Space Administration

Ames Research Center


Moffett Field, California 94035

~~~

2- 1

ACCURACY REQUIREMENTS AND BENCHMARK EXPERIMENTS FOR CFD VALIDATION


Joseph G. Marvin
Chief. Experimental Fluid Dynamics Branch
NASA Ames Research Center
Moffett Field. CA 94035
SUFMARY
The role of experiment in the development of Computational Fluid Dynamics (CFD) for aerodynamic flow
prediction is discussed. CFD verification is a concept that depends on closely coordinated planning
between computational and experimental disciplines. Because code applications are becoming more complex
and their potential for design more feasible, it no longer suffices to use experimental data from surface
or integral measurements alone to provide the required verification. Flow physics and modeling, flow
field. and boundary condition measurements are emerging as critical data. Four types of experiments are
introduced and examples given that meet the challenge of validation: (1) flow physics experiments;
(2) flow modeling experiments; (3) calibration experiments; and (4) verification experiments. Measurement
and accuracy requirements for each of these differ and are discussed. A comprehensive program of validation is described, some examples given, and it is concluded that the future prospects are encouraging.

1.

INTRODUCTION

Mathematical approximations. limited computer capacity, and lack of understanding of physical modeling lead to uncertainties in the application of Computational Fluid Dynamics (CFD). Consequently, the
pace of introduction and the extent of reliance on CFD in the design process depends on validation'; and
experiments that verify CFD have become an essential element of its evolutionary development.2
Experimental validation is required for a number of different aerodynamic flows that occur over the
full range of flight speeds. Any effective, timely program to provide the necessary data will require
good planning and cooperation between various aerospace disciplines. Because of this situation the topic
of validation has been intensely debated within NASA during the past year. An outgrowth of that debate
resulted in the concepts of CFD validation and calibration and categories of experiments recommended by a
NASA ad hoc Cornittee on Validation introduced by B r a d l e ~ . ~And the first NASA CFD Validation Workshop
made further reccinmendatlons: (1) provide closer cooperation between CFD developers and experimentalists;
(2) provide detailed measurements of the flow field and boundary conditions in addition to model surface
and integral quantities; (3) provide new or improved nonintrusive measurement capabilities, especially for
hypersonic or reacting flow conditions; (4) provide redundancy in both measurements and experiments whenever practical so as to clarify accuracy and credibility; (5) provide dedicated large facilities for validation research activities; and (6) provide standardized test cases with accessible data bases.
The intent of the present paper i s to provide a perspective on validation using these ideas and to
introduce a synergistfc approach for timely accomplishment of validation. Detai 1s on experimental and
accuracy requirements will be discussed using the concepts of validation, calibration, and categories of
experimentation as defined in Ref. 3.
CFD code validation: Detailed surface- and flow-field comparisons with experimental data to verify
the code's ability t o accurately model the critical physics of the flow. Validation can occur only when
the accuracy and limltations of the experimental data are known and thoroughly understood and when the
accuracy and limitations of the code's numerical algorithms, grid-density effects, and physical basis are
equally known and understood over a range of specified parameters.
CFD code calibration: The comparison of CFD code results with experimental data for realistic geometries that are similar t o the ones of design interest, made in order to provide a measure of the code's
ability to predict specific parameters that are of importance to the design objectives without necessarily
verifying that all the features of the flow are correctly modeled.
Categories of experimentation: (1) Experiments designed to understand flow physics; (2) experiments
designed t a develop physical models; (3) experiments designed to calibrate CFD; (4) experiments designed
to validate CFD.
The categories of experiments will be explained first with the aid of some examples that represent
the work of the author and his colleagues at the Ames Research Center. Their cooperation in the use and
preparation of this material is greatly appreciated. Following that, accuracy, instrumentation and facility requirements, and future prospects for validation experiments will be discussed.
2.

2.1

EXPERIMENTAL REQUIREMENTS
Role of Experiments

A framework for describing the connection between experiment and


Ref. 4. That framework can be depicted with the aid of Fig. 1, taken
new developments and the various categories of experiments defined in
ment are shown in ascending order of maturity and each is linked to a

computation was presented in


from Ref. 4. and extended to reflect
Ref. 3. The stages of code developtype(s) of experiment.

2-2
Research codes r e f e r t o those developed by i n t e g r a t i n g new enabling technology such as supercomput e r s , algorithms, g r i d methodology. and new understanding o f physical modeling t o solve s p e c i f i c problems. One o r two researchers are involved i n developing the code, and l i m i t e d documentation i s a v a i l a ble. Experiments u t i l i z e d a t t h i s stage are r e f e r r e d t o as b u i l d i n g blocks. These provide the data
r e q u i r e d t o understand flow physics, t o guide flow modeling processes. and t o v a l i d a t e the computations
f o r a p a r t i c u l a r problem. Two types of experiments make up t h e b u i l d i n g blocks leading t o the development
o f t h e research code. They a r e flow physics and flow modeling experiments. An a d d i t i o n a l new development
a t t h i s 1 v e l i s t h e use o f f u l l and large-eddy numerical simulations5 (FS and LES) and computational
chemistry t o develop data bases f o r understanding phenomena such as a t r a n s i t i o n , turbulence, and react i o n r a t e chemistry.
P i l o t codes r e f e r t o a more mature stage of development. Documentation i s more complete, t h e code i s
operated by others besides those involved i n t h e research code development. and the envelope o f applicat i o n i s expanded i n r e c o g n i t i o n o f t h e p o t e n t i a l advances afforded by t h e research code. Benchmark experiments are t h e key t o t h i s stage of development. They provide t h e parametric i n f o r m a t i o n leading t o t h e
i d e n t i f i c a t i o n o f t h e range o f a p p l i c a b i l i t y o f t h e code. C a l i b r a t i o n and v e r i f i c a t i o n are t h e o b j e c t i v e s
o f these experiments.
Subsequently t h e code would advance t o i t s u l t i m a t e development stage when i t could be used alone or
i n combination w i t h codes from o t h e r d i s c i p l i n e s such as s t r u c t u r e s o r p r o p u l s i o n and applied c o n f i d e n t l y
i n the design process. Configurational, performance and system i n t e g r a t i o n experimental d a t a would be
needed f o r v e r i f i c a t i o n a t t h i s stage.
The d e l i n e a t i o n o f t h e various stages o f development o u t l i n e d above i s idealized, and not always
evident i n p r a c t i c e , because of t h e dynamic nature o f CFD and i t s wide-ranging p o s s i b i l i t i e s f o r s o l v i n g
such a v a r i e t y and complexity o f problems; b u t t h e framework d e p i c t s how experiment and computation, worki n g together. could accelerate t h e pace of development. Without question the success o f such a framework
depends on close c o o r d i n a t i o n between experimental and computational f l u i d dynamicists and instrumentation
developers. For t h i s paper, t h e emphasis w i l l be on t h e requirements f o r t h e f i r s t two stages o f
development.

2.2

Flow Physics Experiments

The l a c k of understanding o f fundamental physical phenomena i s l i m i t i n g t h e pace o f CFD development. Important expamples are t r a n s i t i o n from laminar t o t u r b u l e n t flow, turbulence, and h i g h temperature
gas physics r e l a t e d t o hypersonic flows. Flow physics experiments are defined as those experiments t h a t
p r o v i d e fundamental understanding o f such phenomena so t h a t they can be accurately modeled i n t h e codes.
As mentioned previously, computation i t s e l f i s beginning t o supplement d a t a from f l o w physics experiments
through numerical s i m u l a t i o n s t h a t do n o t r e q u i r e any modeling o f t h e physics. The f o l l o w i n g examples
show how f u l l simulations o f t h e Navier-Stokes equations are being ccnnbined w i t h experiment t o p r o v i d e a
more fundamental understanding o f t u r b u l e n t boundary layers. These examples w i l l s u f f i c e t o i l l u s t r a t e
what i s meant by f l o w physics experiments.
The p h y s i c a l i n s i g h t obtainable from d i r e c t Navier-Stokes simulation7 i s shown i n Fig. 2. The f i g u r e
h i g h l i g h t s several imp r t a n t aspects o f t h e s t r u c t u r e o f turbulence i n a simulated f l a t - p l a t e boundary
layer. The s i m u l a t i o n was developed by P. R. S p a l a r t o f Allles, and employs no turbulence models o f any
kind. The complete, time-dependent Navier-Stokes equations were solved using s p e c t r a l methods a t each o f
t h e 9.4 m i l l i o n g r i d p o i n t s i n t h e computational domain. The Reynolds number based on momentum thickness
i s approxiniately 670. I n t h e figure. elongated w h i t e surfaces i d e n t i f y the low-pressure cores o f vort i c e s . Shaded regions show where s i g n i f i c a n t c o n t r i b u t i o n s t o the Reynolds shear stress, -E, are occurr i n g . Regions o f low-speed f l u i d e j e c t e d outwards. and high-speed f l u i d swept wallward are labeled.
These, and o t h e r r e a l i z a t i o n s show t h a t l a r g e hook-shaped v o r t i c a l s t r u c t u r e s are c l e a r l y present i n t h e
numerical t u r b u l e n t boundary layer, and t h e l o a c t i o n s o f s i g n i f i c a n t Reynolds s t r e s s c o n t r i b u t i o n are seen
t o occur adjacent t o these v o r t i c a l structures. T h i s l e v e l o f understanding o f t h e fundamental processes
involved i n t h e Reynolds s t r e s s generation i s e v e n t u a l l y expected t o a i d t h e development o f improved stat i s t i c a l models t h a t accurately r e f l e c t t h e underlying p h y s i c a l behavior o f turbulence.

'

To g a i n s i m i l a r physi a1 i n s i g h t , experimental techniques must generally employ m u l t i p o i n t measurement schemes. An example


o f such an approach i n a f l o w t h a t i s beyond t h e c u r r e n t c a p a b i l i t y o f f u l l
s i m u l a t i o n i s shown i n Fig. 3(a). Here, t h e large-eddy s t r u c t u r e o f a h i g h Reynolds number, compressible
t u r b u l e n t boundary l a y e r was i n v e s t i g a t e d by mounting a f l x e d hot-wire a t t h e w a l l i n conjunction w i t h
another, t r a v e r s i n g hot-wire mounted d i r e c t l y above t h e f i r s t . The o b j e c t i v e was t o map t h e s p a t i a l chara c t e r and extent o f t h e coherent eddies i n a Mach 3 axisymmetric boundary layer, and t o compare t h e
r e s u l t s w i t h low Reynolds number, incompressible flow experiments and simulations.
The long-time-averaged, space-time-cross-correlation functions between t h e near-wall sensor and various p o s i t i o n s o f t h e o u t e r sensor a r e shown i n Fig. 3(b). These curves show t h a t measurable c o r r e l a t i o n
between t h e two wires occurs up t o a separation distance of a t l e a s t h a l f t h e boundary-layer thickness.
These data suggest t h e presence of coherent o u t e r - l a y e r s t r u c t u r e s t h a t extend w e l l i n t o the near-wall
region. which may p r o v i d e an energy t r a n s f e r p a t h between t h e free-stream f l o w and t h e near-wall,
turbulence-producing region. The slope of these s t r u c t u r e s can be deduced from t h e c o r r e l a t i o n curves by
using an eddy convection v e l o c i t y , and I s shown i n Fig. 3(c) t o vary from 5" near t h e w a l l t o 30" i n the
o u t e r layer.

2-3
The nature o f these large disturbances can be studied i n more d e t a i l by computing ensemble-averaged,
mass-f low h i s t o r i e s around strong, r a p i d accelerations and decelerations. The r e s u l t s f o r accelerations,
shown i n Fig. 3(d). closely resemble those of S i m i l a r investigations performed i n low-speed flows and i n
numerical simulations. This suggests Confirmation of Morkovin's hypothesis t h a t the basic s t r u c t u r e o f
turbulent boundary layers i s not fundamentally changed by compressibility, a t l e a s t f o r moderate Mach
numbers.

2.3

Physical Modeling Experiments

Pr c t i c a l CFD applications i n v o l v i n g complex t u r b u l e n t flows r e l y on s t a t i s t i c a l modeling o f turbulence.''


Physical modeling experiments are defined as experiments t h a t provide guidance f o r and v e r i f i c a t i o n o f the modeling process.
An example o f a physical modeling experiment" used t o improve turbulence modeling f o r transonic
flows w i t h strong shock-wave/boundary-layer i n t e r a c t i o n i s shown i n Fig. 4. The t e s t model consisted o f a
c y l i n d r i c a l body f i t t e d w i t h a c i r c u l a r arc section s i m i l a r t o t h a t o f an a i r f o i l . Shock-wave i n t e r a c t i o n s of varying strengths were studied by varying free-stream Mach number. The choice of an axisymmetric
geometry was made t o e l i m i n a t e three-dimensional effects. Mean-flow v e l o c i t y and turbulence p r o f i l e s ,
obtained w i t h a Laser Doppler Anemometer System (LOA), and surface q u a n t i t i e s such as pressure and o i l streak data were documented.
Computations o f the f l o w f i e l d from a Reynolds-averaged, Navier-Stokes code revealed d e f i c i e n c i e s i n
the turbulence modeling. By using a model developed p r i m a r i l y f o r attached boundary layers, the shock
wave l o c a t i o n was predicted i n c o r r e c t l y and consequently the Pressure recovery was s e r i o u s l y overpredicted. The mean- and turbulence-profile data were used t o e x p l a i n the differences and t o guide modeling
improvement. The primary cause of the pressure recovery overprediction was the f a i l u r e o f the eddy v i s c o s i t y model t o adequately r e f l e c t the lag o f turbulence adjustment through the shock wave. Using new
modeling concepts i n conjunction w i t h the turbulence data r e s u l t e d i n a s i g n i f i c a n t model improvement."
I n p a r t i c u l a r , the " h i s t o r y e f f e c t s " o f the turbulence changes through the shock wave were accounted f o r
by prescribing and solving an ordinary d i f f e r e n t i a l equation f o r the maximum shear stress development.
The improved model r e s u l t s are shown.
2.4

C a l i b r a t i o n Experiments

C a l i b r a t i o n experiments are intended t o reveal a code's a b i l i t y t o p r e d i c t s p e c i f i c parameters. The


data, i n most instances, are l i m i t e d w i t h respect t o t h e i r a b i l i t y t o determine the completeness o f the
flow modeling. Code c a l i b r a t i o n i s prevalent and important t o developing codes f o r r e a l gas hypersonic
applications because i n t h i s f l i g h t regime I t I s extremely d i f f i c u l t t o provide ground t e s t data f o r exact
f l i g h t conditions and t h e i r attendant chemical and length scales. For example. f a c i l i t i e s may duplicate
f l i g h t energy l e v e l s bUL not match the a i r chemlstry. o r they may duplicate f l i g h t Mach number but n o t
match the energy l e v e l .
An example o f a c a l i b r a t i o n experiment i n nded t o determine the a p p l i c a b i l i t y o f the a i r chemistry
model used i n a parabolized Navier-Stokes codef5 i s shown i n Fig.
Drag data from 10" sharp cones f i r e d
down a B a l l i s t i c Range are shown as a function o f angle of attack.P4 The angle o f attack range represents
the v a r i a t i o n (uncertainty) i n launch and f l i g h t p a t h angle of the cones from various f i r i n g s done nomin a l l y a t zero angle o f attack. For these t e s t conditions the flow i s laminar: viscous-inviscid i n t e r a c t i o n i s small; and t h e temperature I n the viscous l a y e r 4s s u f f i c i e n t l y high t o cause d i s s o c i a t i o n o f the
a i r . Drag owing t o f r i c t i o n and pressure i s about the same magnitude, so comparisons o f the data w i t h
integrated pressures and s k i n f r i c t i o n from the computations provide a s e n s i t i v e measure o f how w e l l the
code p r e d i c t s s k i n f r i c t i o n i n a high-speed boundary layer. The favorable comparison w i t h the computat i o n s performed by A. W. Strawa serves t o i l l u s t r a t e t h a t the code can p r e d i c t drag i n t h i s chemically
r e a c t i n g flow f i e l d . More discussion on t h i s experiment and i t s r e s u l t s are presented i n Ref. 15.

2.5

V e r i f i c a t i o n Experiments

V e r l f i c a t l o n experiments provide the f i n a l v a l i d a t i o n o f the codes. As such they r e q u i r e f l o w - f i e l d


and surface measurements over a range o f conditions and i n s u f f i c l e n t d e t a i l t o ensure t h a t the flow physi c s i s properly represented. The following example i l l u s t r a t e s t h l s category of experiment.
The improved turbulence model shown previously has r e c e n t l y been introduced i n t o a transonic NavierStokes code and compared w i t h data from an a i r f o i l section. The a i r f o i l was mounted i n a s p e c i a l l y
designed t e s t section w i t h s o l i d walls. Boundary-layer suction was applied upstream o f the a i r f o i l on the
sidewalls t o minimize interference. To f u r t h e r minimize w a l l Interference, the upper and lower w a l l s were
contoured t o streamline shapes t h a t were predetermin d by computation t o account f o r the presence o f t h
model, which f u r t h e r minimized interference. Tests'& were performed a t chord Reynolds number o f 6
10
and angle o f attack and Mach number were varied over a range s u f f i c i e n t t o produce transonic f l o w covering
weak and strong shock-wave/boundary-layer i n t e r a c t i o n and attendant displacement e f f e c t s . The boundary
layer was t r i p p e d on the upper and lower model surface t o ensure t u r b u l e n t f l o w beyond 7% chord. Model
pressures, wall-boundary shapes and pressures, t o t a l drag, l i f t , and f l o w - f i e l d and wake v e l o c i t i e s from
an LDA system were documented. A data base o f t h i s type w i t h minimal interference from a tunnel w i t h
s o l i d w a l l s provides an i d e a l basis f o r evaluating the development o f codes f o r the transonic speed range
because the codes can include wall-boundary conditions more p r e c i s e l y than I n t e r f e r e n c e corrections can be
made t o the data sets.

2-4

An example o f some o f t h e comparisons i s shown i n Fig. 6. A t present t h e code does not include t h e
s o l i d wall-boundary conditions, but a p r e l i m i n a r y assessment using these benchmark data i n d i c a t e s t h a t t h e
code provides very good simul i o n f o r t h e strong i n t e r a c t i o n cases when the improved turbulence model
developed by Johnson and KinJi
i s employed. Results o f t h e comparisons f o r one strong i n t e r a c t i o n case
(where separation occurred a t t h e t r a i l i n g edge) are shown. The a i r f o i l pressures, f l o w f i e l d v e l o c i t i e s
a t constant heights above t h e model, and a wake p r o f i l e a t t h e t r a i l i n g edge are compared w i t h computat i o n s using two d i f f e r e n t turbulence models, a two-equation model.17 and the Johnson-King model. The
comparison shows t h a t the computations using t h e improved turbulence model simulate the measuremenLs very
w e l l . I t i s important t o emphasize t h a t t h i s conclusion could not have been drawn without t h e complete
data s e t composed o f t o t a l drag, l i f t , boundary conditions and f l o w - f i e l d surveys. (See Ref. 16 f o r
f u r t h e r discussion.)
3.
3.1

MEASUREMENT REQUIREMENTS

Completeness

Each of t h e types o f experiment discussed p r e v i o u s l y requires s p e c i f i c information t h a t w i l l enable a


c r i t i c a l assessment o f t h e code's c a p a b i l i t i e s a t each stage o f i t s development. Some examples o f these
measurements and t h e t e s t c o n d i t i o n s where they are needed are l i s t e d i n Fig. 7 taken from Ref. 4. I n
these examples t h e measurements are r e p r e s e n t a t i v e and are germane t o t h e development o f Reynolds-averaged
Navier-Stokes codes f o r f u l l y developed t u r b u l e n t flow.
B u i l d i n g block experiments must provide t h e data r e q u i r e d f o r phenomenological understanding and/or
modeling guidance and enable a c r i t i c a l t e s t o f t h e research code's a b i l i t y t o simulate important aerodynamic flows (e.g.. shock-induced Separation). Surface v a r i a b l e s and f l o w - f i e l d variables, i n c l u d i n g t u r bulence data, are e s s e n t i a l measurements. For t h e turbulence modeling problem, f l o w physics experiments
and f u l l numerical s i m u l a t i o n o f t h e Navier-Stokes equations c a r r i e d out f o r simple flows a t incompressib l e and compressible c o n d i t i o n s can be very h e l p f u l i n p r o v i d i n g fundamental understanding and guidance o f
s t a t i s t i c a l modeling. But t h e f l o w modeling d a t a must be obtained a t r e p r e s e n t a t i v e f l i g h t Mach and
Reynolds numbers where t h e codes are t o be applied t o ensure t h a t t h e physics i s modeled adequately.
Benchmark experiments must provide t h e parametric measurements necessary t o c a l i b r a t e o r v e r i f y p i l o t
code development. Surface and f l o w - f i e l d data a t c r i t i c a l l o c a t i o n s are t h e e s s e n t i a l i n f o r m a t i o n since
t h e o b j e c t i v e of v e r i f i c a t i o n i s t o ensure t h a t t h e code represents t h e c o r r e c t physics o r f o r c a l i b r a t i o n
t o ensure t h a t t h e code adequately p r e d i c t s some p a r t i c u l a r f l o w q u a n t i t i e s . I n order t o c l e a r l y i d e n t i f y
t h e a p p l i c a b l e range o f t h e code, parametric t e s t i n g over as wide a range o f f l i g h t Mach and Reynolds
numbers i s necessary. Experiments a t extremes i n such c o n d i t i o n s are now o f t e n l i m i t e d by instrumentation
and f a c i l i t y development, as i n hypersonic o r high Reynolds number regimes.
Design experiments a t t h e f i n a l stage provide t h e optimal c o n f i g u r a t i o n d a t a necessary f o r p e r f o r mance e v a l u a t i o n and t h e experiments should be c a r r i e d o u t as close t o f l i g h t c o n d i t i o n s as p r a c t i c a l .
CFD i s expected t o expedite t h e executlon of these by e l i m i n a t i n g t h e need f o r f i n e increments i n parametr i c v a r i a t i o n s , by h e l p i n g t o r e s o l v e a n m a l w s data sets, and by e x t r a p o l a t i n g t h e design performance
data t o f l i g h t c o n d i t i o n s when f a c i l i t i e s a r e unable t o achieve them.
For each category of experiment Careful measurements o f boundary c o n d i t i o n s are r e q u i r e d because they
may i n f l u e n c e t h e flow f i e l d around t e s t models. Moreover, they may be needed t o i n i t i a t e computations.
Free-stream o r i n i t i a l conditions, Wall-boundary p h y s i c a l l o c a t i o n ard necessary measurement variables,
and p r e c i s e model l i n e s a r e examples of these measurement requirements.
3.2

Accuracy

Accuracy assessments f o r both computational procedures and experiments are essential. Otherwise
there i s no q u a n t i t a t i v e means f o r determining t h e l i m i t s and ranges o f a p p l i c a b i l i t y f o r t h e codes.
Uncertainty a n a l y s i s i s a well-established method f o r determining experimental data accuracy and should be
a p r e r e q u i s i t e f o r a l l l e v e l s o f experiment used t o develop CFD. It i s u s e f u l during t h e planning and
developmental phases of experiments, f o r e v a l u a t i n g data obtained w i t h d i f f e r e n t instruments, and f o r
comparing data from d i f f e r e n t experiments.
(See Ref. 18 f o r more discussion on accuracy.)
E r r o r estimates f o r t e s t geometry dimensions. t e s t operating and free-stream conditions, model and
f l o w - f i e l d measured variables, and instrumentation should a1 1 be s p e c i f i e d and t h e method used documented
s u f f i c i e n t l y t o a1 low independent assessment.
Reliance on s i n g l e experiments o r measurement procedures f o r code v a l i d a t i o n purposes should be
viewed w i t h c a u t i o n because o f t h e c u r r e n t l i m i t a t i o n s o f f a c i l i t i e s and instrumentation needed t o accomp l i s h v a l i d a t i o n . (These l i m i t a t i o n s are e s p e c i a l l y present i n hypersonic experiments.)
Therefore,
redundant measurement techniques and s l m i l a r experiments performed i n more than one f a c i l i t y may be
required. I n every case. c a r e f u l s u b s t a n t i a t i o n and s p e c i f i c a t i o n of experimental accuracy l i m i t s i s
crucial.
4.

WIND TUNNEL REQUIREMENTS

The requirements f o r t e s t f a c i l i t i e s used t o v a l i d a t e CFD were discussed i n Ref. 4. The most import a n t o f these requirements are: (1) v e r s a t i l i t y , along w i t h well-defined t e s t and boundary conditions;

2-5

( 2 ) appropriate scale and speed range; (3) a c c e s s i b i l i t y of nonintrusive instrumentation; (4) p r o v i s i o n


f o r high-speed data systems; and (5) dedication o f Use t o verification-experimentation.

5.

FUTURE PROSPECTS

During the past year NASA has embarked on a comprehensive CFD v a l i d a t i o n program. Coordinated experimental and computational studies have been i n i t i a t e d a t each o f the NASA OAST Research Centers by teams
comprised o f computational and experimental research s c i e n t i s t s .
A t the Ames Research Center, the major t h r u s t o f the a c t i v i t y 1s supporting the development of codes
employing the Reynolds-averaged Navier-Stokes equations. Data from in-house and university-funded experiments are expected t o be published i n the p u b l i c domain and made a v a i l a b l e t o other computational f l u i d
dynamicists c a r r y i n g out CFD v a l i d a t i o n . Some examples o f the benchmark experiments t h a t i l l u s t r a t e the
scope o f the program follow.

Turn-around-duct experiment: The experiment shown i n Fig. 8 i s under way t o help guide the development of a 3-D incompressible Navier-Stokes code (INS-30) ,19 including i t s turbulence model. The applicat i o n o f the code i s t o study the axisymnetric flow i n the Space S h u t t l e Main Engine turn-around-duct.
The
geometry consists o f a constant area aspect-ratio 10 duct which turns an a i r flow, a t high Reynolds number, through a 180" bend. The bend radius i s equal t o the duct height and some separation o f the flow
occurs on the inner corner w a l l near the end o f the turn. A planer r a t h e r than axisymmetric geometry was
chosen t o permit access f o r nonintrusive l a s e r instrumentation. Surface pressures, s k i n f r i c t i o n , veloci t y p r o f i l e s , and Reynolds-averaged normal- and shear-stress p r o f i l e s are being documented f o r a range o f
Reynolds numbers. Companion computations f o r t h i s geometry are planned t o v e r i f y the range o f a p p l i c a b i l i t y of the code and various turbulence modeling approximations.
Transonic Wing and Wing-Body Experiments: Transonic experiments have been erformed and others are
The approach t o the
now under w a y t o guide the development of a transonic Navier-Stokes code (TNS)."
experiments i s unique i n t h a t they are d e l i b e r a t e l y performed i n s o l i d - w a l l wind tunnel f a c i l i t i e s . This
t e s t technique was chosen because the code can use the tunnel w a l l s as boundary conditions and eliminate
uncertain corrections t o the data f o r w a l l interference. Once the code has been validated, i t can c o n f i d e n t l y be used f o r f r e e - a i r computations by appropriately changing the boundary conditions.
The f i r s t phase o f the experimental a c t i v i t y was conducted several years ago.21 A low-aspect r a t i o
wing w i t h a NACA 0012 p r o f i l e section i n the stream d i r e c t i o n was mounted on the sldewall o f a high
Reynolds number f a c i l i t y and tested over a range o f Mach numbers from 0.5 t o 0.84. Reynolds numbers from
2 a lo6 t o 8 x lo6, and angle o f attack from 0" t o 2'. Solid, s t r a i g h t , wind tunnel walls. sloped t o
c o r r e c t f o r "tunnel empty" boundary-layer growth and instrumented w i t h pressure taps were employed.
I n v i s c i d , n o - s l i p boundary conditions along a l l w a l l s were assumed f o r the computations, but t h a t may not
be e n t i r e l y adequate as discussed l a t e r . Model pressures, wall-boundary pressures, surface o i 1 flows, and
l i m i t e d v e l o c i t y p r o f i l e s obtained w i t h an LOA were documented. Thus f a r , the data have been used by
computational groups a t the NASA Ames and Langley research centers.
The Ames group used comparisons w i t h the data a t the lower Mach numbers and angles o f attack t o
e s t a b l i s h confidence i n the zonal techniques employed I n the TNS code. A t the higher Mach numbers and
angles
attack they used comparisons w i t h the data t o s o r t out g r i d refinement and turbulence modeling
Issues.
Results o f the comparisons w i t h the high Mach number data were s a t i s f a c t o r y only i n the sense
t h a t they reproduced many of the complex flow features, but it could not be determined whether the turbulence model was s o l e l y responsible f o r the differences w i t h the data. Recently the Langley group showed
the importance of i n c l u d i n g the viscous. n o - s l l p c o n d i t i o n along the mounting wall. Their r e s u l t s , taken
from Ref. 23, are shown i n Fig. 9. A perspective view of the surface streamlines shows the influence o f
the viscous sidewall. The streamline patterns, e s p e c i a l l y i n the side-wall region, are remarkably s i m i l a r
t o the experimental o i l flows.
The comparison o f computed and measured pressures on t h e tunnel w a l l s and
the wing shows good agreement except on the wing a t the span l o c a t i o n where a strong shock forms. These
differences r e f l e c t the inadequacy o f the turbulence model. E f f o r t s are under way t o improve the
model ing

$5

A follow-on experiment conducted i n a s o l i d w a l l transonic t e s t section i s under way. This experiment eliminates some of the shortcomings of the one previously described: the model and the t e s t f a c i l i t y
are larger; the Reynolds number range can be extended; a more r e a l i s t i c , low-aspect. h i g h - t a p e r - r a t i o wing
g e m e t r y i s being used; and the sidewall boundary layer w i l l be measured. Moreover, p r o v i s i o n i s made t o
t e s t a wing-body combination. A photograph o f the wing-body model mounted i n the tunnel i s shown i n
Fig. 10. The measurements t o be made are also l i s t e d . The half-model body i s mounted on the sidewall.
The TNS computations w i l l employ n o - s l i p boundary conditions along the mounting w a l l and s l i p conditions
on the other walls. Preliminary wing-alone and w a l l pressure data have been obtained recently.

3-0 Supersonic Shock I n t e r a c t i o n Experiments--Several experiments are under way t o study the interact i o n of shock waves w i t h t u r b u l e n t boundary layers. Reference 24 presented data f o r a series o f asymnetr i c separated flows on an ogive-cylinder-flare model. Shock unsteadiness was a major issue i n the experiments and the reader i s r e f e r r e d t o Ref. 24 f o r f u r t h e r discussion.
Another s e r i e s o f experiments on a swept-wedge p l a t e are being con c t e d by Settles. Figure 11 shows
the geometry, t e s t conditions and some r e c e n t l y published measurements.'!
The surface s k i n f r i c t i o n on
the p l a t e has been measured and compared w i t h a computation solving the Reynolds-averaged Navier-Stokes

2-6

equations. A two-equation turbulence m d e l w i t h w a l l functions was employed and t h e r e s u l t s compare w e l l


w i t h the data. I n Ref. 25. comparisons w i t h data f o r o t h e r wedge angles using both two-equation and algeb r a i c turbulence models show t h a t turbulence modeling i s n o t c r i t i c a l t o r e s o l v i n g the s t r u c t u r e physics
o f these flows, probably because they are dominated by I n v i s c i d e f f e c t s . However, the e f f e c t s of viscosi t y are e s s e n t i a l t o reproducing t h e structures, and Euler codes probably cannot represent these flows
adequate 1y

Hypersonic All-Body Experiment: The experimen depicted i n Fig. 12 i s being performed t o guide the
development o f a 3-0 Parabol ized Navier-Stokes c o d J 6 t h a t uses up-wind d i f f e r e n c i n g t o o b t a i n sharp
shocks. The geometry I s a 70" swept d e l t a w i t h an e l l i p t l c a l cross section. A t the t w o - t h i r d s body
l e n g t h s t a t i o n , an expansion surface forms t h e upper p a r t on the model. Some recent experimental r e s u l t s
taken from Ref. 27 are a l s o shown i n Fig. 11. Spanwise pressure d i s t r i b u t i o n s f o r 15" angle o f a t t a c k
and M = 10.3 over t h e forebody r e g i o n ahead o f t h e expansion are shown compared w i t h the computations
f o r a s i n g l e streamwise s t a t i o n , assuming e i t h e r laminar o r t u r b u l e n t f l o w from the leading edge. The
agreement i s good w i t h e i t h e r assumption because v i s c o u s - i n v i s c l d i n t e r a c t i o n has a small i n f l u e n c e on the
pressure d i s t r i b u t i o n a t t h i s Reynolds number. When t h e remaining measurements o f heating and v e l o c i t y
p r o f i l e s are completed. o t h e r v a l i d a t i o n issues such as aerodynamic heating w i l l be addressed.
6.

CONCLUDING REMARKS

Experiments p l a y a c r i t i c a l r o l e i n t h e development o f CFD. They provide phenomenological data t o


help understand t h e physics o f complex flows; they provide guidance i n t h e modeling process where the
physics I s unknown o r so complex t h a t computational procedures are n o t p r a c t i c a l ; and u l t i m a t e l y they provide t h e v e r i f i c a t i o n necessary t o e s t a b l i s h t h e l i m i t s o f a p p l i c a b i l i t y t o various aerodynamic flows.
Four types o f experiments supporting t h e development o f CFD were described: (1) f l o w physics experiments, (2) f l o w modeling experiments, (3) c a l i b r a t i o n experiments, and (4) v a l i d a t i o n experiments. The
f i r s t two types were broadly categorized as b u i l d i n g block experiments. They provide t h e phenomenological
and modeling data r e q u i r e d f o r research code development. An a d d i t i o n a l new technological advance cont r i b u t i n g t o t h e b u i l d i n g block data base i s f u l l - and large-eddy simulations and computational chemistry.
The b u i l d i n g block d a t a base i s more d e t a i l e d and o f t e n r e q u i r e s s o p h i s t i c a t e d instrumentation and
t e s t techniques. The second two types were broadly categorized as benchmark experiments. These experlments p r o v i d e t h e data needed t o i d e n t i f y t h e accuracy and l i m i t a t i o n s on t h e code's a b i l i t y t o compute
complex aerodynamic flows. The data requirements d i f f e r from t h e b u i l d l n g block experiments i n t h e sense
t h a t phenomenological and modeling issues are not i n v e s t l g a t e d i n d e t a i l .
The categories o f experiments and corresponding measurements lead t o s p e c i f i c requirements f o r f a c i l i t i e s used f o r v a l i d a t i o n . V e r s a t l l i t y , appropriate scale and speed range, a c c e s s i b i l i t y f o r n o n i n t r u s i v e
instrumentation, computerized data systems, and dedicated use f o r v e r i f i c a t i o n are t h e important
requirements.
A s y n e r g i s t i c , comprehensive approach t o v a l i d a t i o n was introduced. A program i s under way t o provide v a l i d a t i o n experiments t h a t can guide t h e development o f advanced computational procedures f o r a p p l i c a t i o n t o complex flows. Both c m p u t a t i o n a l and experimental f l u i d dynamicists are focusing on key aerodynamic problems whose s o l u t i o n s are paced by t h e l a c k o f adequate understanding o f t h e f l o w physics and
modeling and by t h e lack o f adequate v a l i d a t i o n data t o v e r i f y code development. The major challenge f o r
success o f t h e program depends on t i m e l y accomplishment o f t h e experiments, development and implementation
o f new instrumentation. and development o f appropriate h i g h Reynolds number and high Mach number, highenthalpy f a c i l i t i e s .

REFERENCES
1.

Bradley, R. G., Bhateley. I. C.. and Howell, 6. A.. "Computational F l u i d Dynamics - T r a n s i t i o n t o


Design Applications," Supercomputing i n Aerospace, NASA Conference. P u b l i c a t i o n 2454, pp. 69-767,
Mar. 1987.

2.

Current C a p a b i l i t i e s and Future D i r e c t i o n s I n Computational F l u i d Dynamics, National Research Counc i l , National Academy Press, Washington, D.C.,
1986.

3.

Bradley. R. G.. "CFD V a l i d a t i o n Philosophy." Paper No. 1. AGARO Symposium on V a l i d a t i o n o f Computat i o n a l F l u i d Dynamics, May 1988. Lisbon, Portugal.

4.

Marvin, J. G., "Wind Tunnel Requirements f o r Computational F l u i d Dynamics Code V e r i f i c a t i o n . "


Paper 34, AGARO Symposlum on Aerodynamlc Data Accuracy and Q u a l i t y , Requirements and C a p a b i l i t i e s i n
Wind Tunnel Testing, Naples. I t a l y , Sep. 28-Oct. 1, 1987.

5.

Moin, P. and Kim, J., "The S t r u c t u r e o f t h e V o r t l c i t y F i e l d i n Turbulent Channel Flow, Part I - Analy s i s o f Instantaneous F l e l d s and S t a t i s t i c a l Correlations." J. F l u l d Mechanics, Vol. 155, p. 441.
1985.

6.

Cooper. 0. M., Jaffe. R. L.. and Arnold, J. 0.. "Computational Chemistry and Aeroassisted O r b i t a l
Transfer Vehicles," Journal o f Spacecraft and Rockets, Vol. 22. No. 1, p. 60. Jan.-Feb. 1985.

2-7

7.

Kline, S. J.. and Robinson, 5 . K., 06Turbulence-ProducingCoherent Structures i n the Turbulent Boundary Layer: Progress of a Cooperative Evaluation." t o be presented a t the Zaric Memorial Internat i o n a l Seminar on Wall Turbulence, Dubrovnik. Yugoslavia. May 1988.

8.

Spalart, P. R.,
1986.

9.

Robinson, S. K. , "Space-Time C o r r e l a t i o n Measurements i n a Compressible Turbulent Boundary Layer,"


A I A A Paper 86-1130, Atlanta, GA, May 1986.

" D i r e c t Simulation o f a Turbulent Boundary Layer up t o Rtheta = 1410," NASA TM 89407,

10.

Marvin, J. G., "Turbulence Modeling f o r Computational Aerodynamics," AIAA Journal, Vol. 21. No. 7 ,
pp. 941-955. Jul. 1983.

11.

Bachalo. W. D.. and Johnson. 0. A.. "An I n v e s t i g a t i o n of Transonic Turbulent Boundary Layer Separat i o n Generated on an Axisymmetric Flow Model," AIM Paper 79-1479, Williamsburg, VA. 1979.

12.

Johnson, D. A., and King, L. S.. "Transonic Separated Flow Predictions Based on a Mathematically
Simple, Nonequilibrium Turbulence Closure Model." IUTAM Symp. on Turbulent Shear Layer/Shock Wave
I n t e r a c t i o n , Paris, France, Sep. 1985. NASA TM 86826, Oct. 1985.

13.

Prabhu. D. K.. Tannehill. J. C.. and Marvin, J. G.. "A New PNS Code f o r Three-Dimensional Chemically
Reacting Flows." (U). A I A A 87-1472. Honolulu, HI. Jun. 1987.

14.

I n t i e r e , P. F.. Kirk, 8. K.. Chapman. G. T., and Terry. J. E., " B a l l i s t i c Range Tests o f Ablating and
Nonablating Slender Cones," (U), AIM J w r n a l , Vol. 8, No. 3, pp. 558-564. Mar. 1970.

15.

Deiwert. G. S.. Strawa, A. W.. Sharma. S. P., and Park, C.. "Experimental Program f o r Real Gas Flow
Codes a t NASA Ames Research Center," Paper No. 21. AGAR0 Symposium on V a l i d a t i o n o f Computational
F l u i d Dynamics. Lisbon, Portugal, May 1988.

16.

Mateer, G. G., Seegmiller, H. L., Coakley, T. J.. and Hand, L. A., "An Experimental I n v e s t i g a t i o n of
a S u p e r c r i t i c a l A i r f o i l a t Transonic Speeds.'' A I A A Paper 87-1241, Honolulu. HI, Jun. 1987.

17.

Coakley. T. J., "Numerical Simulation o f Viscous Transonic A i r f o i l Flows." AIAA Paper 87-0416, Jan.
1987.

18.

M o f f a t t . R. J.. "Contributions t o the Theory o f Uncertainty Analysis f o r Single-Sample Experiments,"


1980-81 AFOSR/HTTM-Stanford Conference on Complex Turbulent Flows, Vol. 1 Objectives, Evaluation o f
Data, S p e c i f i c a t i o n o f Tests Cases, Discussion, and P o s i t i o n Papers, e d i t e d by 5. J . Kline, B. J.
Cantwell, and 6. M. L i l l e y .

19.

Chang, J. L. C.. and Kwak. D., "A Three-Dimensional Incompressible Flow Simulation Method and I t s
A p p l i c a t i o n t o the Space S h u t t l e Maine Engine, Part 11--Turbulent Flow," A I A A Paper 85-1670, Reno,
NV. Jan. 1984.

20.

Holst. T. L., Kaynak. U., Gundy, K. L.. Thomas, S. D., Flores, J.. and Chaderjian, N. M.. "Numerical
S o l u t i o n o f Transonic Wing Flows Using an Euler Navier-Stokes Zonal Approach," A I A A Paper 85-1640,
Cincinnati, J u l . 1985.

21.

Lockman. W. K., and Seegmiller. H. L.. "An Experimental I n v e s t i g a t i o n o f the S u b c r i t i c a l and Superc r i t i c a l Flow About a Swept Semispan Wing," NASA TM-84367, Jun. 1983.

22.

Kaynak, U.. Flores, J.. "Advances i n the Computation of Transonic Separated Flows over F i n i t e Wings."
A I A A Paper 87-1195, Honolulu, HI, Jun. 1987.

23.

Vasta. V. N., and Wedan. B. W., "Navier-Stokes Solutions f o r Transonic Flow Over a Wing Mounted I n a
Tunnel," AIAA Paper 88-0102, Reno NV. Jan. 1988.

24.

Brown. J. D., Brown, J. L., KUSSOY, M. I., Holt, M.. and Horstman, C. C.. "Two-Component LDV I n v e s t i g a t i o n o f 3-Dimensional Shock/Turbulent Boundary Layer Interactions," A I A A Paper 87-0553. Reno, NV.
Jan. 1987.

25.

K i m . K. S.. and Settles, G. S., "Skin F r i c t i o n Measurements by Laser Interferometry i n Swept Shock
Wave/Turbulent Boundary-Layer Interactions."

26.

Lawrence, 5 . L.,
Equations," (U),

27.

Lockman, W. K., Cleary, J. W., and Lawrence, S. L., "Flow V i s u a l i z a t i o n and Pressure D i s t r i b u t i o n s
f o r an All-Body Hypersonic A i r c r a f t , " (U), Paper No. 53. Fourth National Aero Space Plane Technology
Symposium, Honterey, CA, Feb. 1988.

Chaussee, D. S., and Tannehill, J. C. " A p p l i c a t i o n o f an Upwind Algorithm t o the PNS


AIM Paper 87-1112, Honolulu, H I , Jun. 1987.

2-8
PRODUCTION
CODES

EXPERIMENTS
CON FlGU RATION
PERFORMANCE
INTEGRATION
RESEARCH
CODES
VAL1DATION
EXPERIMENTS
TYPES:
FLOW PHYSICS AND SIMULATIONS
FLOW MODELING

Figure 1.

The role o f experiment in developing CFD.

Figure 2. Turbulent flow-physics obtained from a full simulation of the Navier-Stokes equations.
Re, = 670.

M, = 0,

2-9

7-

Me = 2.97
Ue = 594 m/sec
Re = 1.5 X lo7/,

.4

.0

1.2

2.0

1.6

xi6

IC)MEAN ANGLES OF DISTURBANCE FRONT

__
TRAVERSED

VITA DETECTION AT yt

29

FIXED
y+ = 29

la) HOT.WIRE PROBE INSTALLATION

/ I -

--

182
132

/o.o*

91

.5

71

R11

65
0
I

-2

TIME SHIFT, rUe/6

Ib) CROSS-CORRELATIONS BETWEEN TWO HOT-WIRES

1
4

-2

tUe/6
(d) ENSEMBLE.AVERAGED VOLTAGES FOR
LARGE POSITIVE EVENTS

Figure 3. Turbulent low-physics obtained from the experiment o f Ref. 9 employing m u l t i p l e hot-wires.
M, = 3; Re = 1.5 x 10 /m.

2- 10
23.65cm 28.73cm

61.00cm

MODEL SUPPORT

EXP., BACHELLO.JOHNSON
COMPUTATION, JOHNSON-KING MODEL
COMPUTATION, CEBECI-SMITH MODEL

.75
50

ya

.25
0
-.25
-50

'

.4

.6

.8

1.0

1.2

1.4

1.6

xIC

(b) PRESSURE COEFFICIENT

xlc

(c) VELOCITY, x/c = 1.0

"

.4

.8

1.0
1.2
xlc
(d) MAXIMUM TURBULENT SHEAR STRESS
.6

1.4

1.6

Figure 4. A flow-modeling experiment used to develop an improved turbulence model.

2-11

ORIGINAL PAGE

OF POOR QUALITY..1

------

c=20cm
zmax = 13% c

- c

tl

zlc
-.l

.2

.4

xlc

1.0

.E

.6

(a) TEST MODEL - MBB VA-2 AIRFOIL


TRAILING EDGE THICKNESS = 0.522% CHORD
1.5
1.o
.5

-cp

-.5
EXP., MATEER e

-1.0
-1.5

.2

.4

xlc

.6

1.o

.8

(b) PRESSURE COEFFICIENT

1.2

M
,

= 0.78

u/at
1.o

b-

4-

2.54 cm

.8

(a) GEOMETRY

.25
I-

P0

'20

.15 -

.2

.E

.6

.4

1.o

XIC

(c)

REALGAS
IDEALGAS
PRESSURE DRAG
(SIMS TABLES)

0
V
0

FLOW FIELD VELOCITIES


M,=

0.78

0.73
I 1

.4

.2

:x:.10 -

zlc
0

0
.5c

-.2
M = 15; Re = 400,000
-.4

ANGLE OF ATTACK, deg


(bl DRAG DATA FOR 10" CONE

Figure 5. A calibration experiment used to


evaluate a real gas chemical model in a parabolized Navier-Stokes code. loo cone; M, = 15; and
ReL = 400,000.

.2

.4

u/at

10

(d) AXIAL VELOCITY PROFILES

Figure 6. A benchmark airfoil experiment used to


verify development of an improved turbulence
model.

2-12

EXPERIMENT

MEASUREMENTS
(REPRESENTATIVEFOR
TURBULENCE MODELING)

BUILDING BLOCK
(PHENOMENOLOGICAL)

SURFACE QUANTITIES
INCLUDE TRANSITION PTS.

TEST CONDlTl,ONS

REPRESENTATIVE
FLIGHT M
,
Re,

FLOW FIELD QUANTITIES


TURBULENCE
INDIVIDUAL STRESSES
CORRELATION LENGTHS
STRUCTURE
BOUNDARY CONDITIONS
FREE STREAM
TUNNEL WALLS
MODEL SHAPE
SURFACE QUANTITIES
INCLUDE TRANSITION PTS.

BENCHMARK
(PARAMETRICAL)

VARY .M
,
Re, a OVER
FLIGHT RANGES

FLOW FIELD QUANTITIES


(SELECTED LOCATIONS)
BOUNDARY CONDITIONS
(SEE ABOVE)
~

DESIGN
(CONFIGURATIONAL)

DRAG, LIFT, MOMENTS,


HEAT LOADS, SHEAR LOADS

AS CLOSE TO FLIGHT
.M
,
Re, (Y AS PRACTICAL

BOUNDARY CONDITIONS
(SEE ABOVE)

Figure 7.

Experimental requirements.

(a) FLOW FIELD AND GEOMETRY

~ f OIL
,
FLOW
SURFACE: p,
FLOW FIELD: U, V, 2, 7 ,
REH = 0.1 - 3.0 X lo6, M = 0.1 - 0.3

(b) MEASUREMENTS AND TEST CONDITIONS

ORIGINAL PAGE IS
OF POOR QUALITY
(c) VIEW SHOWING LASER SKIN-FRICTION
INTERFEROMETER

Figure 8. 2-0 turn-around-duct experiment.

2-13
P

0 0

EXP. - LOCKMAN
COMPUTATION - VATSA

(SPAN) J = 0.25

.8
1.2 I

(a1 PERSPECTIVE VIEW OF COMPUTED STREAMLINES

ON WING AND WALL

COMPUTATION - VATSA
P~/P, = 0.975

'2

EXP. - LOCKMAN Z/C


0 0.750
A 1.500 0 0.375
0 1.125
h 2.000

UPPER WALL

CP

1.2 I

J = 0.50
I

-1.2,-

-.2
LOWER WALL

.I,,,,,
C = 0.78

-2

-1

x/c

(b) TUNNEL WALL PRESSURE COMPARISONS

1.2
0

.2

.4

x/c

.6

.E

1.o

(c) WING PRESSURE COMPARISONS

Figure 9. Comparison of Reynolds-averaged Navier-Stokes computations and data from a v e r i f i c a t i o n experiment. M, = 0.826, Re = 8 x lo6, a = 2". A/R = 3.

2-14
GEOMETRY
0

OGIVE-CYLINDER HALF-BODY

WING

SHARP FIN

NACA 64A008 STREAMWISE SECTION

INVISCID PLANAR
SHOCKWAVE

ASPECT RATIO = 3.2


TAPER RATIO = 0.25
L.E. SWEEP ANGLE = 36.9"
TEST CONDITIONS
= 0.5 TO 0.8

M
,

Re,+=

1 X IO6 TO 10 x lo6

a=O"T015"

'

MEASUREMENTS
0

FLOW VISUALIZATION

FLAT PLATE OR
WIND TUNNEL WALL

INITIAL 2D TURBULENT
BOUNDARY LAYER

SURFACE (OIL FLOW)


FLOW FIELD (VAPOR SCREEN)
0

PENN STATE EXPERIMENT:

SURFACE PRESSURES (WING, TUNNEL WALLS)

M
,

MEAN VELOCITY - FLOW FIELD


(LDV AND PROBES)

2.4

- 4.

a = 4 - 22", Rex

10 X106

MEASUREMENTS
SURFACE: OILFLOW, pw, cf

(a) GEOMETRY, TEST CONDITIONS AND MEASUREMENTS

FLOW FIELD: VAPOR SCREEN, p


'2'
(a) FLOW GEOMETRY

--,004

.003

,002

cf

.001 -

"15

YAW ANGLE

LISF WITH ERROR BARS


COMPUTATION, HO RSTMAN

a = 16"
20

25

30

35

40

45

P. deg
(b) PHOTO OF MODEL MOUNTED IN TUNNEL

(b) SKIN FRICTION ON THE PLATE, x = 3.5 in.

Figure 10. A low-aspect-ratio wing-body


experiment.

Figure 11. A 3-D shock-wave boundary layer


interaction experiment.

ORIGINAL PAGE 7s

OE POOR

QUALITY

50

55

2-15

(a) ALL-BODY HYPERSONIC AIRCRAFT MODEL IN NASA/


AMES 3.5-h HWT: LENGTH = 3 ft

EXP.-LOCKMAN; X/L
0.20
0 0.25

x 0.50

A 0.30
0.40

V 0.65

0 0.60

COMPUTATION-LAWRENCE; X/L

--- LAMINAR: 0.6


-TURBULENT; 0.6

-.05
n

LEEWARD

0
.05
.10

WINDWARD

h4

Ln

.30L

-1.0 -.9

-.8

-.7
-.6
-.5 -.4 -.3 -.2
SPANWISE STATION, Y/YLE

-.l

SPANWISE PRESSURE DISTRIBUTIONS FOR FOREBODY


= 5 X lo6

a = 15"; M, = 10.3; Re,,L

Figure 12.

A hypersonic all-body experiment.

Report Documentation Page


1. Report No.

2. Government Accession No.

3. Recipient's Catalog No.

NASA TM- 100087


4. Title and Subtitle

5. Report Date

Accuracy Requirements and Benchmark Experiments


for CFD Validation

6. Performing Organization Code

7. AuthorM

8. Performing Organization Report No.

May 1988

Joseph G. Marvin

A-88123
10. Work Unit No.

505-60-1 1

9. Performing Organization Name and Address

11. Contract or Grant No.

Ames Research Center


Moffett Field, CA 94035
13. Type of Report and Period Covered
2. Sponsoring Agency Name and Address

Technical Memorandum

National Aeronautics and Space Administration


Washington, DC

14. Sponsoring Agency Code

20546-0001

5. Supplementary Notes

Point of Contact: Joseph G. Marvin, Ames Research Center, MS 229-1


Moffett Field, CA 94035 (415) 694-5390 or FTS 464-5390
6. Abstract

The role of experiment in the development of Computational Fluid Dynamics (CFD


t
'
o
r aerodynamic flow prediction is discussed. CFD verification is a concept that
jepends on closely coordinated planning between computational and experimental
jisciplines. Because code applications are becoming more complex and their potential for design more feasible, it no longer suffices to use experimental data from
surface or integral measurements alone to provide the required verification. Flow
physics and modeling, flow field, and boundary condition measurements are emerging
3s critical data. Four types of experiments are introduced and examples given that
neet the challenge of validation: ( 1 ) flow physics experiments; (2) flow modeling
Ixperiments; ( 3 ) calibration experiments; and ( 4 ) verification experiments. Measurement and accuracy requirements for each of these differ and are discussed. A
Zomprehensive program of validation is described, some examples given, and it is
:oncluded that the future prospects are encouraging.

1 18. Distribution Statement

7. Key Words (Suggested by Authork))

Zomputer code validation


3xperiments
rurbulent flows

Unclassified-Unlimited
Subject Category

L Security Classif. (of this report)

20. Security Classif. (of this page)

Jnclass ified
NASA

'626

OCT

16
I

For sale by the National Technical Information Service, Springfield, Virginia 22161

34
22. Price

21. No. of pages

Unclassified
I

A02
I