Page 1
AEDC-TR-76-119
~ ~ ~ A COMPUTER STUDY OF HYPERSONIC LAMINAR
BOU ND ARY-LAYER/SHOC K-WAVE INTERACTION
USING THE TIME-DEPENDENT COMPRESSIBLE
NAVIER-STOKES EQUATIONS
I
VON KARM/~N GAS DYNAMICS FACILITY ARNOLD ENGINEERING DEVELOPMENT CENTER
AIR FORCE SYSTEMS COMMAND ARNOLD AIR FORCE STATION, TENNESSEE 37389
September 1976
Final Report for Period July 1974 -- April 1976
Approved for pubhc release; distribution unhmited.
Prepared for
DIRECTORATE OF TECHNOLOGY (DY) ARNOLD ENGINEERING DEVELOPMENT CENTER ARNOLD AIR FORCE STATION, TENNESSEE 37389
Page 2
NOTICES
When U. S. Government drawings specifications, or other data are used for any purpose other than a definitely related Government procurement operation, the Government thereby incurs no responsibility nor any obligation whatsoever, and the fact that the Government may have formulated, furnished, or in any way supplied the said drawings, specifications, or other data, is not to be regarded by implication or otherwise, or in any manner licensing the holder or any other person or corporation, or conveying any rights or permission to manufacture, use, or sell any patented invention that may in any way be related thereto.
Qualified users may obtain copies of this report from the Defense Documentation Center.
References to named commercial products in this report are not to be considered in any sense as an endorsement of the product by the United States Air Force or the Government.
This report has been reviewed by the Information Office (OI) and is releasable to the National Technical Information Service (NTIS). At NTIS, it will be available to the general public, including foreign nations.
APPROVAL STATEMENT
This technical report has been reviewed and is approved for publication.
FOR THE COMMANDER
ELTON R. THOMPSON Research & Development
Division Directorate of Technology
o
ROBERT O. DIETZ Director of Technology
Page 3
UNCLASSIFIED REPORTDOCUMENTATION PAGE
I R E P O R T N U M B E R 2 G O V T ACCESSION N O
AEDC-TR-76 -119 I
4 T I T L E ( ~ d SubtJtte)
A COMPUTER STUDY OF HYPERSONIC LAMINAR BOUNDARY-LAYER/SHOCK-WAVE INTERACTION USING THE TIME-DEPENDENT COMPRESSIBLE NAVIER-STOKES EQUATIONS 7 A U T H O R { s )
B . K . H o d g e - ARO, I n c .
9 P E R F O R M I N G O R G A N I Z A T I O N N A M E A N D ADDRESS
A r n o l d E n g i n e e r i n g D e v e l o p m e n t C e n t e r (DY) A i r F o r c e S y s t e m s Command A r n o l d A i r F o r c e S t a t i o n , T e n n e s s e e 3 7 3 8 9 I t C O N T R O L L I N G O F F I C E N A M E AND ADDRESS
Arnold Engineer ing Development Center(DYFS A i r Force S y s t e m s Command A r n o l d A i r F o r c e S t a t i o n , T e n n e s s e e 3 7 3 8 9 14 M O N I T O R I N G A G E N C Y N A M E & A D D R E S S ( i f d i f fe ren t I rom C o n t r o l l l n ¢ O l h c e )
IS D I S T R I B U T I O N S T A T E M E N T ( o f th i s Repo t / )
R E A D I N S T R U C T I O N S B E F O R E C O M P L E T I N G F O R M
3 R E C I P I E N Y ' S C A T A L O G N U M B E R
S T Y P E OF R E P O R T & P E R I O D C O V E R E D
F i n a l R e p o r t - J u l y 1974 A p r i l 1976
6 P E R F O R M I N G ORG R E P O R T N U M B E R
8 C O N T R A C T OR GRAN T NJMBER(~ ' )
10 P R O G R A M E L E M E N T P R O J E C T T A S K A R E A & WORK UNI ~ N U M B E R S
P r o g r a m E l e m e n t 6 5 8 0 7 F
12 R E P O R T D A T E
S e p t e m b e r 1976 13 N U M B E R OF P A G E S
45 15 S E C U R I T Y C L A S S ( o ! thJa repor t )
UNCLASSIFIED
15~ D E C L ASSI FI C A T I O N ' OOWNORADIN G S C H E D U L E
N/A
A p p r o v e d f o r p u b l i c r e l e a s e ; d i s t r i b u t i o n u n l i m i t e d .
17 D I S T R I B U T I O N S T A T E M E N T ( o l the abstract e n t e r e d In B l o c k 20, I I d l f l e ren t from Report)
18 S U P P L E M E N T A R Y N O T E S
Available in DDC
t9 K EY WORDS (Conl#nue on reve rse s ide J! neceaeeW ~ d I d e n t t f y by b lock number)
c o m p u t e r s s h o c k w a v e - b o u n d a r y l a y e r i n t e r a c t i o n l a m i n a r b o u n d a r y l a y e r h y p e r s o n i c f l o w
m a t h e m a t i c a l a n a l y s i s e x p e r i m e n t a l d a t a n u m e r i c a l m e t h o d s
20 ABST R A C T ~Cont lnue on r e v e r s e aJde f f n e c e e m a ~ and t d e n t l ~ by b l o c k numbe~
T h i s r e p o r t p r e s e n t s t h e r e s u l t s o f an i n v e s t i g a t i o n o~ h y p e r s o n i c l a m i n a r b o u n d a r y - l a y e r / s h o c k - w a v e i n t e r a c t i o n s u s i n g t h e m e t h o d o f MacCormack t o s o l v e t h e t i m e - d e p e n d e n t c o m p r e s s i b l e N a v i e r - S t r o k e s e q u a t i o n s . C o m p a r i s o n s o f t h e n u m e r i c a l s o l u t i o n s w i t h e x p e r i m e n t a l d a t a w e r e made t o a s c e r t a i n t h e v a l i d i t y o f t h e n u m e r i c a l m e t h o d and t o i d e n t i f y r e g i o n s o f a n o m a l o u s b e h a v i o r . The a l g o r i t h m g a v e g o o d r e s u l t s when a p p l i e d t o h y p e r s o n i c l a m i n a r
FOR~' 1 4 7 3 EO,T,ON OF I NOV es ,s OBSOLETE D D , JAN 73
UNCLASSIFIED
Page 4
UNCLASSIFIED
20 . ABSTRACT ( C o n t i n u e d )
i n t e r a c t i o n s t h a t c a u s e d e i t h e r s m a l l o r no s e p a r a t e d r e g i o n s and m a r g i n a l p e r f o r m a n c e when a p p l i e d t o h y p e r s o n i c l a m i n a r i n t e r - a c t i o n s h a v i n g l a r g e r e g i o n s o f s e p a r a t e d f l o w . The e x t e n t s o f t h e s e p a r a t e d r e g i o n s i n i n t e r a c t i o n s h a v i n g l a r g e r e g i o n s o f s e p a r a t e d f l o w w e r e u n d e r p r e d i c t e d when c o m p a r e d w i t h e x p e r i m e n t a l d a t a . The p r e d i c t e d w a l l h e a t - t r a n s f e r r a t e s e x h i b i t e d t h e c o r r e c t q u a l i - t a t i v e t r e n d b u t n o t t h e e x p e r i m e n t a l l y m e a s u r e d q u a n t i t a t i v e v a l u e s . C o n s i d e r a t i o n o f t h e e f f e c t s o f n e e d e d s t a b i l i z i n g t e r m s a s w e l l a s g r i d r e s o l u t i o n s u g g e s t s i n a d e q u a t e mesh s p a c i n g i n t h e l o n g i t u d i n a l d i r e c t i o n a s t h e c a u s e o f t h e a f o r e m e n t i o n e d anom- a l i e s . I f i n a d e q u a t e mesh s p a c i n g ( a n d t h e c o r r e s p o n d i n g l a c k o f s u p p o r t f o r e v e r y t e r m o f t h e N a v i e r - S t o k e s e q u a t i o n s ) i s a p r i m e c a u s e o f t h e c ~ t e d d i s c r e p a n c i e s b e t w e e n t h e n u m e r i c a l r e s u l t s a n d t h e e x p e r i m e n t a l d a t a , t h e n t h e n e e d e d r e d u c t i o n o f s e v e r a l o r d e r s o f m a g n i t u d e i n ~x w o u l d i n c r e a s e CPU t i m e a n d c o r e s t o r a g e r e - q u i r e m e n t s t o u n t e n a b l e l e v e l s .
A F S C A r ~ l , d A i r s ' l r e ~
UNCLASSIFIED
Page 5
AE DC-TR.76-119
PREFACE
The work reported herein was conducted by the Arnold Engineering
Development Center (AEDC), Air Force Systems Command (AFSC), under
Program Element 65807F. The results of the research presented were
obtained by ARO, Inc. (a subsidiary of Sverdrup & Parcel and Associates,
Inc.), contract operator of AEDC, AFSC, Arnold Air Force Station,
Tennessee. The research was conducted from July 1974 to April 1976
under ARO Project Nos. V33P-04A and V33A-08A. The author of this
report was B. K. Hodge, ARO, Inc. The manuscript (ARO Control No.
ARO-VKF-TR-76-64) was submitted for publication on June 21, 1976.
Acknowledgment and appreciation are extemded to Dr. J. C. Adams,
ARO, Inc., for overall guidance of the project. Special appreciation
is due to Dr. Barrett Baldwin and Dr. R. W. MacCormack of the
Computational Fluid Dynamics Branch, NASA Ames Research Center,
Moffett Field, California, for providing a copy of the basic computer
code as well as much information pertinent to its use.
Page 6
AE DC-TR-76-119
CONTENTS
1.0 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . .
2.0 ANALYTICAL ANALYSIS
2.1 Time-Dependent Method of Solution Overview .....
2.2 The Explicit Time-Dependent Method of
MacCormack . . . . . . . . . . . . . . . . . . . . .
2.3 Initial and Boundary Conditions . . . . . . . . . .
2.4 Computational Grid . . . . . . . . . . . . . . . . .
3.0 RESULTS
3.1 Boundary-Layer/Shock-Wave Interactions .......
3.2 Numerical Results . . . . . . . . . . . . . . . .
4.0 CONCLUDING REMARKS . . . . . . . . . . . . . . . . . . .
REFERENCES . . . . . . . . . . . . . . . . . . . . . . .
5
8
12
13
15
20
37
38
ILLUSTRATIONS
Fisure
I.
.
.
.
.
Schematic of Boundary-Layer/Shock-Wave
Interaction . . . . . . . . . . . . . . . . . . . .
Computational Mesh System for Boundary-Layer/
Shock-Wave Interaction . . . . . . . . . . . . . . .
Schlieren Photograph of Shock Generator and
Receiver Plate in AEDC Tunnel B (Taken from the
Study Reported in Ref. 19) . . . . . . . . . . . .
Inviscid Pressure Ratios for Incident-Reflected
Shock Wave . . . . . . . . . . . . . . . . . . . . .
Schlieren Photograph of a Boundary-Layer/Shock-
Wave Interaction in AEDC Tunnel B (Taken from
the Study Reported in Ref. 19) . . . . . . . . . . .
6
12
16
18
19
=
3
Page 7
AE DC-T R-76-119
Fisure
6.
,
8.
.
10.
11.
12.
13.
14.
15.
16.
Page
Flat Plate Pressure Distributions Computed by
the MacCormack Algorithm and the Brailovskaya
Algorithm . . . . . . . . . . . . . . . . . . . . . 20
Flat Plate Velocity Profile . . . . . . . . . . . . 23
Streamlines near the Leading Edge of a Flat Plate
at Hypersonic Velocities . . . . . . . . . . . . . . 24
Leading-Edge Shock Shape for a Flat Plate at
Hypersonic Velocities . . . . . . . . . . . . . . . 25
Laminar Hypersonic Boundary-Layer/Shock-Wave
Interaction Using VKF Tunnel B Conditions ..... 26
Laminar Hypersonic Boundary-Layer/Shock-Wave
Interaction Using LRC Mach 8 Tunnel Conditions . . 28
Heat Transfer for a Laminar Hypersonic Boundary-
Layer/Shock-Wave Interaction Using VKF Tunnel B
Conditions . . . . . . . . . . . . . . . . . . . . . 29
Laminar Hypersonic Boundary-Layer/Shock-Wave
Interaction with Separated Region ......... 32
Velocity Profiles at the Point of Separation and
within the Separated Region . . . . . . . . . . . . 34
Streamlines for Hypersonic Laminar Boundary-
Layer/Shock-Wave Interaction . . . . . . . . . . . . 35
Streamlines for Computational Region ........ 36
TABLE
I. Parameters for Numerical Cases .......... 22
NOMENCLATURE . . . . . . . . . . . . . . . . . . . . 42
4
Page 8
AE D C-T R-76-119
1.0 I N T R O D U C T I O N
The interaction of an oblique shock wave with a boundary layer
is a phenomenon of considerable interest and frequent occurrence in
supersonic and hypersonic flows. The effects of shock-wave impinge-
ment from externally carried aircraft stores on the aircraft and the
effects of shock impingement from the aircraft on external stores
have long been of interest to the military. Vehicles such as the
TITAN IIIC and the Space Shuttle have geometries that make considera-
tion of boundary-layer/shock-wave interaction a necessity. Numerous
experimental studies have been undertaken to study, correlate, and
explain the phenomenological aspects of the interaction problem.
Concurrent with the interest shown by experimentalists in the
boundary-layer/shock-wave interaction problem, much effort has been
expended to develop theoretical techniques capable of accurately
predicting the salient features of the problem.
The complexity of the interaction between a shock wave and a
boundary layer gives rise to phenomena not characteristic of either
a shock wave or a boundary layer. Figure I schematically illustrates
the physics of a typical boundary-layer/shock-wave interaction. The
boundary-layer equations are parabolic and hence can be integrated
(at least to a point near separation) in a step-by-step downstream
fashion once an impressed pressure gradient is specified. The
boundary-layer/shock-wave interaction problem is not parabolic since
the impressed pressure gradient is determined in part by the response
of the boundary layer to the shock wave. Hence no a priori computation of
the pressure is possible. Moreover, since separation and reattachment is
a possibility, conventional boundary-layer methods cannot be used because
a square-root-type singularity exists at separation for the boundary-layer
equations. Thus, any solution technique must appeal to a system of
governing equations more fundamental than either the Euler equations for
inviscid flow or the boundary-layer equations. The more general Navier-
Page 9
A E D C - T R - 7 6 - 1 1 9
8 --,..,.,. ~ ure Plateau 7
Distance from Plate Leading Edge
Y
ock
Compress ion Wave~
Expansion
S Leading-Edge Shock / J
Reflected ~ S h o c /
X
U =
Separation Point - ~ - - J Reattachment Point - -J
Figure 1. Schematic of boundary-layer/shock-wave interaction.
Page 10
AEDC-TR-76-119
Stokes equations, from which the Euler equations and the boundary-layer
equations are derived, are fundamental to'the subject of viscous fluid
flow and are valid throughout the entire flow field. The Navier-Stokes
equations will yield valid results at separation and reattachment, within
the separated recirculation region, across incident and reflected shock
waves, throughout expansion fans, and for any combinative influences of the
aforementioned. The equations require numerical solution in either a
spatially elliptic or a temporally hyperbolic domain.
The present report compares numerical results from an explicit
time-dependent compressible Navier-Stokes analysis developed by
MacCormack (Ref. I) with experimental data for hypersonic laminar
boundary-layer/shock-wave interaction on a flat plate under AEDC yon
K~rm~n Gas Dynamics Facility (VKF), Hypersonic Wind Tunnel (B), condi-
tions as well as'NASA Langley Research Center (LRC) Mach 8 Variable
Density Tunnel conditions. It is shown that numerical solutions of
the time-dependent compressible Navier-Stokes equations yield reasonable
results when applied to hypersonic laminar boundary-layer/shock-wave
interactions.
2.0 ANALYTICAL ANALYSIS
2.1 T IME-DEPENDENT METHOD OF SOLUTION OVERVIEW
The time-dependent method starts with a complete specification of
the flow field and then uses the governing equations of motion to
advance the flow field temporally until a steady state is reached. Thus,
the flow field evolves numerically in a process analogous with physical
reality. "The initial specification can be just a uniform flow with
appropriate boundary conditions (see Roache (Ref. 2) and Richtmyer
and Morton (Ref. 3) for general expositions on the time-dependent method).
?
Page 11
AE DC-TR-76-119
The physical basis of the time-dependent approach as well as the
advantages accrued by application of the time-dependent method to
solve the compressible Navier-Stokes equations were delineated by
Crocco in 1965 (Ref. 4). Kurzrock and Mates (Ref. 5) in 1966 used
the time-dependent approach to study analytically the flow in a shock
tube and hypersonic flow over a sharp flat plate. Skoglund and Gay
(Ref. 6) applied the time-dependent method to the computation of
laminar boundary-layer/shock-wave interactions in 1969.
2.2 THE EXPLICIT T I M E - D E P E N D E N T M E T H O D OF M A C C O R M A C K
The so-called method of MacCormack, since its introduction in
1969, has become one of the most widely used explicit second-order
accurate methods for numerical solution of hyperbolic partial differential
equations. The algorithm was first introduced by MacCormack in Ref. 7
and subsequently modified and extended by Refs. 8 through 14 as well as
Ref. I. It has been applied to obtain solutions of the time-dependent
compressible Navier-Stokes equations by Baldwin and MacCormack (Refs.
9 and 10), MacCormack and Baldwin (Ref. 11), and Deiwert (Refs. 15" and
16) among others.
The two-dimensional time-dependent Navier-Stokes equations,
neglecting body forces and heat generation, can be written in conservation
form as:
au aF aC at +~if +~-y = 0 (I)
where
U = pu
V
(2)
Page 12
A,E OC-TR-76-119
with
F =
G =
" p u
pu 2 + a x
puv + rxy
(e + ax)U + ryxV - k0T/0~
p v
puv + ry x
pv 2 + a y
(e + ay)V . r x u - k 0 T / 0 v j
(ff__~_ 0 3 O. a x = p - )~ + - 2 # ~ x x
(3)
(4)
(5)
r y = ry x = - t z + ( 6 )
A(au a'~ av (7) Oy = P - \ a x + a y ] - 2g ~y
(see the Nomenclature for terminology).
The procedure used to advance the dependent variables (p, 0u,
pv, e) from a time t to a time t + At at the interior points will be
examined first. The procedures for the boundary points are different
and will be reviewed subsequently. The MacCormack method is of the
predictor-corrector type and can be utilized in such a manner that a
single predictor and a single corrector application will advance the
dependent variables in time by an amount At. However, if the concept
of splitting is employed, simplicity and computational efficiency
result (Ref. 8). Basically the concept of splitting involves a
predictor-corrector pass driven by gradients in the x-direction and a
separate predictor-corrector pass driven by gradients in the y-direction.
Thus four sweeps, two predictor and two corrector, are required. This
may be written as:
Page 13
A E D C - T R - 7 6 - 1 1 9
and
un+l/2 n (Ay) t,.n _ Gn ~ ],1 = U i , j - l,j i,j-
U n+l/2 ' (U un+l/2,~ _ l(~t'~ [Gn+X/2 G n+ l /2 ) = _ n + 2 k ~ y ) k i , j+ l - , ,j ',J 2 l,j i,j /
U n+l U. n+l/2 (z-~t~(F n+ l/2 n+ 1/2~ i .J = , , j - - ~ A x / ~ ' .J -- F i--l ,j /
(8)
(9)
( lO)
un+] ' (un+] /2 u n + i ) ] (~t~( ' F n+' _ F in,~l) i , j = 2 \ i,j + i,j - ~k~x)k~ ,+],j (11)
Denoting by Lx the operation performed by Eqs. (8) and (9) and
by Ly the operation performed by Eqs. (10) and (11), the sequence
becomes
U n+l = i.,xLyU n (12)
This operation (Eq. (12)) is not of second-order accuracy but
(13)
retains second-order accuracy. The stability criterion (the Courant-
Friedrichs-Lewy condition) for the y sweep is:
Ay ,~ty-I v [ + c (14)
and for the x-sweep
~X At x - (15)
1 4 1 + c
I0
Page 14
AEDC-TR-76-119
The minimum of Eqs. (14) and (15) for the entire grid represents the
maximum At for which computational stability is ensured. The sequence
of operations defined by Eq. (13) is used to advance temporally the
interior points of the computational grid. The boundary condition
treatment and the initial condition specifications are needed to
complete the method.
The numerical solution of flows containing strong shock waves is
often hampered by numerical oscillations which can eventually cause
program failure. A fourth-order damping term, effective only in
regions of large pressure gradients, has been used to reduce the
numerical oscillations (Refs. 9, 10, and 11). Essentially, this technique
adds an additional "viscosity" proportional to the second derivative of
pressure to each of the steps represented by Eqs. (8) through (11).
The addition in regions of low-pressure gradients is negligible and
is of importance only where pressure gradients or pressure oscillations
are large. By using the arrow symbol to denote replacement, the damping
terms can be included in Eqs. (8) through (11) by
o ',j ,-- c. AG . 1,j + ,j ( 1 6 )
Fn+l ." 2 F n + l / 2 , n-r 1 /2 i,j *"" i,j + AFi , j ( 1 7 )
where
1 pi , i+ l - 2 Pi,] + Pi,j-1 Ui,] ) AGn, j = -~ (I v l + c)i , ] P i , j+ l + 2 Pi,j + P i , ] - I (O i ' ]+ l -
(18)
A F n + I / 2 = l ( [ u [ + c)i, ] P~+I,j - 2 Pi, j - P i - l , j (U~+I, j - Ui j) x,j P i+ l , j + 2 Pi, j + P i - l , j '
(19)
The quantities G~ ~n+i/2 1,j-I and ~i-l,j are treated in the same manner as indicated
in Eqs. (16) through (19) as are all the barred quantities; i.e., to each
F or G term in Eqs. (8) through (ii) is added the term analogus to Eq. (18)
or (19).
II
Page 15
AE DC-TR-76-119
2.3 INITIAL AND BOUNDARY CONDITIONS
Initially, the flow field must be specified completely in the
region under consideration, the computational plane. The computational
plane for the boundary-layer/shock-wave interaction is shown schemat-
ically in Fig. 2. All of the flow field except the upper boundary
(line AB in Fig. 2) need be specified only as a uniform flow. The
upper boundary is specified in such a manner that the incident shock
impacts the surface at the desired location. By specifying uniform flow
along line AE and Rankine-Hugoniot flow conditions along line EB, such
can be accomplished. The boundary layer will form normally on the
surface, the shock will spread downward, and the interaction will evolve
in the correct manner.
Coarse Mesh
Fi ne Mesh
A Incident S hock Wave
E
\ ! \ ,B
\ \ \
\ \
B
h
C D
Figure 2. Computational mesh system for boundary-layer/shock-wave interaction.
As the solution advances, only'the boundary conditions along AB,
AC, and CD must be specified. Along AEB, the same conditions initially
specified must be used. Segment BD cannot be defined as this would
overspeclfy the problem. A gradient condition such as ~/~x = 0 along
BD is used since a large portion of the flow field is supersonic and
will thus not propogate errors upstream. This specification of region
]2
Page 16
AE DC-T R-76-119
BD does not introduce spurious information into the remainder of the
flow field. Region AC can be taken as a uniform flow field if the
leading edge of the plate is to be considered or input as previously
determined profiles if the area of interest is too large for normal
processing. Both input profiles and input uniform flow have been used
with good results. The wall boundary conditions are defined along CD
in such a manner that flat plate results are obtained midway between
the first two grid rows; i.e.,
ui, 1 = - -u i , 2 ( 2 0 )
vi, 1 = - v i , 2 ( 2 1 )
ei , 1 = Cv(2T w - e i , 2 / C v ) ( 2 2 )
Pi,1 = Pi,2 - " ~ Y / i , 2 ( 2 3 )
Pi,1 P i , l - (y_ l ) e i , 1 (24)
2.4 C O M P U T A T I O N A L G R I D
Some consideration on the nature of the expected solution is
relevant at this point. The flow field may be viewed as being composed
of an outer region basically inviscid in nature and an inner viscous-
dominated region. The outer region will change only gradually as the
interaction evolves; and, therefore, will need to be computed less
frequently. Since the maximum allowable time step in an explicit
time-dependent method is proportional to the grid spacing (as indicated
by Eqs. (14) and (15)), the finer the grid the smaller the allowable
time step At. If a fine grid is used in the viscous-dominated region
and a coarse grid is used in the outer inviscid-like region, com-
putations will need to be performed in the coarse-grid region only once
for every M times in the fine-grid region. M is the smallest integer
plus one in the quotient of the smallest At in the coarse grid and the
smallest At in the fine grid, i.e.,
M = MOD(Atc , Atf ) + l ( 2 5 )
13
Page 17
AE DC-TR-76-119
Figure 2 illustrates the computational grid used, showing the fine and
coarse regions. Thus, most of the computational effort will be expended
in the viscous-dominated region. Caution must be exercised at the
boundary between the fine and coarse grids as accumulation or origina-
tion of conserved quantities can introduce computational anomalies to
the solution.
The basic digital computer code follows the University of Tennessee
Space Institute Short Course notes as given by MacCormack (Ref. 17).
All numerical results reported herein were generated using an IBM
370/165 digital computer. The computer program was executed using
single precision arithmetic and required 372,000 bytes of core for a
183 by 35 grid array using the IBM FORTRAN IV H-LEVEL 21.7 Compiler.
Typically about 100 sec of CPU time were required for one complete pass
through the program (advancement of the field from time t to time t +
At) for a 183 by 35 array. The code was modified to have restart
capability and to permit various initial-condition options to be exercised.
3.0 RESULTS
The von K~rm~n Facility at AEDC is concerned primarily with high
supersonic and hypersonic flows. Since only boundary-layer/shock-wave
interactions that are completely laminar are examined in this report,
relatively low-incident shock-wave angles are required at the hypersonic
Mach numbers of interest. Large-incident shock-wave angles at hyper-
sonic Mach numbers and the corresponding large pressure increases across
the incident/reflected shock-wave systems preclude the maintenance of
laminar flows throughout the interaction region. Laminar interactions
were considered since it was deemed desirable to validate the explicit
time-dependent method for numerically solving the Navier~Stokes equations
prior to any considerations of the complicating effects of turbulent
flow. Attempts at turbulent boundary-layer/shock-wave interaction
solutions by Baldwin and MacCormack (Refs. 9 and 10) and by Horstman,
]4
Page 18
AE DC-T R-76-119
et al. (Ref. 18), indicate that much development is needed in turbu-
lence modeling before transitional and turbulent processes can be
adequately treated.
3.1 BOUNDARY-LAYER/SHOCK-WAVE INTERACTIONS
Prior to discussing numerical solutions, it is appropriate to
examine typical wind tunnel models and inviscid solutions for shock
reflection. Figure 3 (taken from the study reported in Ref. 19) shows
typical VKF Tunnel B apparatus used to investigate boundary-layer/shock-
wave interactions. The upper wedge (generator) produces an oblique
shock wave that impinges on the lower wedge (receiver). Impingement
location as well as shock strength can be easily controlled using such
an arrangement. Pressure or heat-transfer data are taken from the
receiver plate. Because of viscous-induced effects near the leading
edge of the generator, the oblique shock angle produced is not the
"wedge" shock angle that the nominal generator angle of attack would
produce. The receiver plate leading edge also exhibits viscous-induced
effects, one obvious result being the leading-edge shock. The
aforementioned effects are such that, for nominal given angles, the
inviscid pressure ratios across the incident/reflected shock waves
are not obtained.
Numerical solutions corresponding to particular experimental
cases have been generated by using the experimental ratio p3/p] (see
Sketch I) and the free-stream Mach number (M) to define 'the inviscid
shock angle (e). This allowed the correct pressure ratio to be used
e
M ~ Pl P2
Sketch 1
]5
Page 19
AE DC-TR-76-119
Figure 3. Schlieren photograph of shock generator and receiver plate in AEDC Tunnel B (taken from the study reported in Ref. 19).
16
Page 20
AEDC-TR-76-119
without expending additional computational effort tb define viscous-
induced effects on the generator. Figure 4 was computed using inviscid-
flow wedge results and allows the rapid estimate of the appropriate
incident shock angle (@) for a given free-stream Mach number and pressure
ratio. Additionally, the minimum and maximum incident shock angle for
a normal reflection is given. It is interesting to note that, for a
diatomic gas (¥ = 1.4) and free-stream Mach number greater than about 3,
a regular reflection is not admissible for a shock angle greater than
40 deg. A more detailed examination was given by Zumwalt (Ref. 20).
Figure 5 (taken from the study reported in Ref. 19) is a schlieren
photograph of a typical laminar boundary-layer/shock-wave interaction.
The incident/reflected shock-wave system can be seen. The dashed
lines indicate the nominal region used in the computation. However,
for a region which does not contain the leading edge of the plate,
suitable profiles in density, velocity, energy, and pressure must be
available for use as initial and boundary conditions for the upstream
(left-hand side in Fig. 5) computational boundary. The required
profiles could have been generated by a typical parabolic boundary-
layer program, but the viscous-induced leading edge effects prominent
in hypersonic flow (see Fig. 3) would not have been present. Since
a compressible Navier-Stokes solution is capable of generating such
effects, the computer code without an incident shock wave possessed the
capability of calculating laminar flat-plate profiles including viscous-
induced leading-edge effects. Thus, whenever it was desired to start
an interaction computation without including the plate leading edge,
a suitable set of initial profiles was obtained by generating a
flat-plate Navier-Stokes solution including the leading edge and
retaining the required numerical information on a data storage device.
Baldwin and MacCormack (Ref. 9) used this technique in their turbulent
boundary-layer/shock-wave calculations as did Carter (Ref. 21) in
his laminar calculations on a supersonic ramp.
I?
Page 21
A E D C - T R - 7 6 - 1 1 9
Q -
o" Om
~eJ
300
200
I00
80
60
40
20
10
8
1 I '
0
0
M O O ' ~ " ~ \ \ %. \ %, "~ %. \
Moo 10
I 15 1
4
l ,I 3
~ / - - I ncident Shock-Wave Angle Limit for Regular Reflection
A 2
\ \ \
10 O,
Figure 4.
20 30 40 50 Incident Shock-Wave Angle
Inviscid pressure ratios for incident-reflected shock wave.
18
Page 22
AE DC-TR-76 -119
Figure 5. Schlieren photograph of a boundary-layer/shock-wave interaction in AEDC Tunnel B (taken from the study reported in Ref. 19).
19
Page 23
AE DC-TR-76-119
3.2 N U M E R I C A L RESULTS
As a check on the capability of the code, a low Reynolds number flat-
plate solution was generated using the conditions of Carter (Ref. 21).
The resulting pressure distribution is shown in Fig. 6 along with the
time-dependent compressible Navier-Stokes numerical solution of Carter.
Both solutions exhibit essentially the same behavior except very close
to the leading edge where the explicit MacCormack scheme is smoother
than the Brailovskaya scheme used by Carter.
4
3
2
f Moo - 3.0 Sym Re0~, L" 103 o Present Results
o To0,390o R o Carter (Ref. 21)
O )
0 I I I I I I I I I I 0 E1 E2 E3 E4 E5 E6 E7 E8 E9 L 0
Dis~ncefromL=dl~Edge, X¢
Figure 6. Flat plate pressure distributions computed by the MacCormack Algorithm and the Brailovskaya Algorithm.
Depending on the state of the boundary layer and the magnitude of
the pressure jump across the incident/reflected shock-wave system, flow
separation and reattachment with a recirculation region may occur. This
condition is schematically illustrated in Fig. I. The presence of a
separation region with recirculation and reattachment poses a more
severe test of the code's capability than an unseparated case. Thus,
to avoid possible complications resulting from separated regions, the
20
Page 24
A E D C - T R - 7 6 - 1 1 9
first results examined will be for non-separated hypersonic flows.
All the numerical results reported herein were obtained using a constant
wall temperature for the receiver plate. Data which would allow a more
exact specification of wall temperature were not generally available.
All of the VKF Tunnel B data examined in the present study had
the nominal shock interaction point one foot from the leading edge of
the plate. The parameters used in the numerical solution are given
in Table I, which contains tabular information for all the cases
examined. The first laminar boundary-layer/shock-wave interaction
presented (number 2 of Table I) was generated using a set of profiles
(density, velocity, internal energy) computed via the time-dependent com-
pressible Navier-Stokes code with no impinging shock wave. Figure 7 is a
partial reproduction of the input velocity ratio profiles (u/U) generated
and is typical of the profiles. Figure 8 shows the computed streamline
shapes for the leading edge of the flat plate, and Fig. 9 shows the shape
of the leading-edge shock. Shock-wave locations in the "shock-capturing"
MacCormack code were estimated by computing the first and second derivatives
of pressure with respect to longitudinal distance and locating regions where
dp/dx is a maximum and d2p/dx 2 is large. This is similar to the philosophy
of Grossman and Moretti (Ref. 22) in locating incipient shock waves in
inviscid time-dependent calculations.
By using the aforementioned input profiles and a shock-wave angle
of 8.6 deg, a boundary-layer/shock-wave interaction was generated.
This corresponded to conditions for which VKF Tunnel B data were
available (Ref. 19). Figure 10 presents the results of the computed
interactions as well as the VKF Tunnel B data for the corresponding
tunnel conditions. Agreement is satisfactory particularly near the
peak pressure location where the correct shape is generated and
in the mid-range where the correct pressure gradient is obtained. No
evidence of separation was obvious from the data, and no separated
region was predicted by the program. Because of the rather modest
2]
Page 25
AE DC-TR-76 -119
pressure rise across the incident/reflected shock-wave system and
because of the lack of separation, it is probable that this is a
completely laminar interaction. Overall agreement with the laminar
time-dependent Navier-Stokes solution tends to confirm this. Computer
time on an IBM 370/165 was 2 hr for the flat-plate solution from which
the initial profiles were secured and 3 hr for the interaction.
Table 1. Parameters for Numerical Cases
No.
1
2
Tunnel /V~
VKF B 7. 94
VKF B 7. 94
NASA LRC Mach 8 7.73
VKF B 7. 93
VKF B 7. 94
Rea~ft
o. 96x lO 6
0. xlO 6
0.48x lO 6
o.97x lo 6
0.96x 106
T(~, Peo, Tw, h f, h, 0, OR Iblft 2 OR ff ft deg
93.68 2.885 520.0 0.035 0.35 --- 46 40
93.68 2.885 520.0i0.035 0.35 8.6 66 40
106,6 2. 120 ~0.4 0.035 0.35 II. 1 60 40
94. 43 3. 140 525. 0 O. 030 0. 25 10. 7 171 25
93. 68 2. 885 520. 0 (3. 035 0. 35 16. 0 181 35
IMAX JMAX
22
Page 26
AE DC-TFI-76-119
0 . 4 0
0 . 3 2
0 . 2 4
y/h
0.16
0.08
0 0
I I 0 . 2 4 0 . 4 8
u /U c o
I 0 . 7 2
I 0 . 9 6
Figure 7. Flat plate velocity profile.
23
Page 27
AE DC-TR-76-119
Y Uoo
0. 05
0.04
0.03
y/h
0. 02
0.01
L I I I I 0 0.1 0.2 0.3 0.4 0.5
X, 'ft
Figure 8. Streamlines near the leading edge of a flat plate at hypersonic velocities.
24
Page 28
0 . 1 0 -
y, ft
0 ..05
0 0 0 . 2 0 . 4
AE DC-TR-76-119
x , f t
I ] 0 , 6
a. Shock wave shape
0 . 0 0 8
y~ : f t
O. 0 0 4
0 . I 0 0 . 0 5
I I 0 . 1 0 0 . 1 5
x , f t
Figure 9.
b. Leading edge detail
Leading edge shock shape for a flat plate at hypersonic velocities.
25
Page 29
2.0
o
ooo
8
1.0
~ . ~ . 6 deg
- - - ' ~ U
AE OC-TR-76-119
I
0 0
Figure 10.
I l f f =, Mo~ = 7. 94
Re~o/ft = 0. 96 x 105
Tw/T o = 0. 40
sym 0 AEDC Tunnel B, Reco/ft = O. 96 x 106, Moo = 7. 94
Present Results
I I
0.5 1.0 x, ff
Laminar hypersonic boundary-layer/shock-wave interaction using VKF Tunnel B conditions.
1.5
26
Page 30
AE DC-TR-76.119
An additional hypersonic laminar boundary-layer/shock-wave
interaction using as a basis the data of Kaufman and Johnson (Ref. 23)
taken in the Langley Research Center (LRC) Mach 8 Variable Density
Tunnel was attempted. The results of the computation as well as the
relevant experimental data are presented in Fig. ii. Agreement is
satisfactory particularly near the peak pressure location. The predicted
pressure gradient within the interaction region agrees well with the
experimental data. The major discrepancy occurs near the beginning
of the interaction region where the experimental pressure is about
50 percent higher than the computed pressure. It is tempting to
ascribe the anomaly to viscous-induced phenomena from the leading
edge of the plate, but the results of Fig. 6 give some confidence in
the ability of the code to properly compute viscous-induced effects.
Local variations in wall temperature as well as the overall model
temperature can affect the pressure in the vicinity of the leading
edge. Kaufman and Johnson suggest using a uniform model temperature
of 0.5 times the total temperature. This suggestion was followed
in making the computations and could have contributed to the discrepancy
as the exact model temperature distribution was not known. Six hours
of IBM 370/165 CPU time were required.
Heat transfer is another of the quantities commonly measured in
experiments and needed from calculations. Thus, the capability of
the code to predict the level of heat transfer in a laminar boundary-
layer/shock-wave interaction is also of interest. Figure 12 was
generated using as a basis heat-transfer data from the VKF Tunnel B
(Ref. 18). The agreement between the computed free-stream Stanton
number (St) and the measured free-stream Stanton number is disap-
pointing. The computed solution underpredicts the maximum heating
rate and overpredicts the minimum heating rate. The numerical
solution indicates no separation to be present, and an examination
of the data indicates little, if any, separation. Thus, the problem
appears to be one of grid resolution as qualitatively the correct
trends are observed in the computer-generated solution.
2?
Page 31
AEDC-TR-76-119
8
4
Y
"~" \ \ \ \ \1_ \ \ \ "~ 0. 889 ft -7 Moo = 7.73
Reoolft = 0. 481 x 106
TwiT o [] 0. 5
Z~
LRC Mach 8 Data (Ref. 23) Present Results
Figure 11.
0.5 1.0
Laminar hypersonic boundaw-layer/shock-wave interaction using LRC Mach 8 tunnel conditions.
28
Page 32
A E D C - T R - 7 6 - 1 1 9
x
8 u~
2.0
1.0
0
Figure 12.
" ~ - I0. 7 deg
/ f J
f
v I l ft
\ \ \ \ \ \
M m = 7.93 Rem/ft [] 0. 97 x 106 Tw/T 0 = 0. 411
sym l o AEDC Tunnel B
o - Present Results
~ O°°oo
J ~ ~ , ~ ~ O0 0
0
0 O0
0 0 o
I I I 0.5 1.0 1.5
X, ff
Heat transfer for a laminar hypersonic boundary-layer/shock-wave interaction using VKF Tunnel B conditions.
29
Page 33
AEDC-TR-76-119
The cell Reynolds number as an index of spatial resolution has
previously been examined by Roache (Ref. 2), Cheng (Ref. 24), and
MacCormack (Ref. 11). MacCWrmack, for example, suggests a cell
Reynolds number on the order of two for each coordinate mesh spacing
if every term of the Navier-Stokes equations is to receive adequate
support. For this particular case, a cell Reynolds number based on
Ay within the viscous-dominated region has a typical value of 6.0; i.e.,
ReAy = pvA..._, y = 6.0 (26)
and the cell Reynolds number based on Ax within the viscous-dominated
region has a typical value of 1,250; i.e.,
puAx ReAx - - 1,250 ( 2 7 )
Thus, the y-mesh should adequately support all terms, while the x-mesh
will not adequately support all terms. Obviously, Ax cannot be decreased
by the several orders of magnitude needed for adequate support of all
the terms in the Navier-Stokes equation as CPU time, and core storage
would become totally impractical. MacCormack (Ref. 11) suggests that
the inadequately supported terms are not necessary for boundary-layer
flow calculation, but anomalies between the experimental data and the
numerical solution show quantitative differences, which could be the
results of inadequate mesh resolution in the x-direction.
Additionally, at the hypersonic conditions used in this report, it
was necessary to add the damping terms previously discussed to stabilize
the solution. These terms, especially in regions of large gradients,
could cause "smearing" of the numerical solution as exemplified in
Fig. 12. It is not possible to say which of the two effects, grid
resolution or damping, caused the discrepancies between the experimental
and numerical results. Approximately 6 hr of IBM 370/165 CPU time
were required for the above solution.
30
Page 34
AEDC-TR-76-119
The last case to be examined is based on VKF Tunnel B conditions
and possesses a large separated region. Figure 13 summarizes the
tunnel conditions used and presents the results of a time-dependent
numerical solution. This solution was generated by computing a
hypersonic flat-plate solution and using this solution as initial
conditions for the upstream portion of the computational region.
Agreement between computed and experimental data was adequate in the
region about the peak pressure and good for the pressure gradient near
the end of the interaction. The middle portion of the interaction
exhibits the correct plateau of pressure. The largest area of dis-
agreement is the beginning of the interaction where the computed
plateau and separation regions are much shorter than those indicated
by experimental data.
Several attempts were made to rectify the situation, but none
were successful. A larger separation region was used in the initial
conditions for a subsequent series of runs, but as the solution evolved,
the separated region shrank to the initial expanse as seen in Fig. 13.
Another attempt was made using input profiles at 0.25 ft from the
leading edge instead of 0.50 ft. The final result after some hours of
computer time did not differ appreciably from that shown in Fig. 13.
As in the previous case, grid resolution effects and the damping
terms effects are cited as being possible causes of the discrepancies
between experimental data and the results of the numerical solution.
The cell Reynolds numbers ReAy and ReAx possess approximately the same
typical values for this case as in the case previously examined. Hence
the x-mesh will not adequately support every term in the Navier-Stokes
equations. The damping terms were used to stabilize the numerical
solution generated for this case. The gradients downstream of the
pressure plateau region are larger than the gradients upstream of the
plateau region. The values of the damping terms were examined for the
entire flow field, and in general, the values in the region downstream
31
Page 35
AE DC-TR-76 -119
Y
\ J
, I I l f t ---]-~ Moo = 7. 94
Rem/ft = O. 96 x 106
TwIT o = O. 4
21 -
m
1 7 -
m
1 3 -
PwlPm -
9 -
5 -
1 -
0
Figure 13.
Sym 0 Tunnel B
Present Results
O O
~ O O
I I I 0.5 1.0
X, ff
Laminar hypersonic boundan/-layer/shock-wave interaction with separated region.
32
Page 36
AEDC-TR-76-119
of the pressure plateau were an order of magnitude larger than the
values upstream of the pressure plateau. The magnitude of the damping
terms in the region upstream of the pressure plateau (in the region
about the point of separation) were larger than the remaining portion
of the flow field except for the previously delineated downstream region.
Thus, the damping terms are of less relative importance within the
separated recirculation region, while grid resolution effects are of
equal importance. Close to the point of separation, the boundary-
layer equations exhibit singular behavior, and within the separated
recirculation region, the Navier-Stokes equations hold. It then
follows that, within this region, terms not otherwise of importance
may be significant. Lack of grid resolution in the x-direction appears
to be a likely candidate for the anomalous behavior of the numerical
solution. Large separation regions in laminar flow are difficult to
predict numerically as evidenced by the results of Skoglund and Gay
(Ref. 6) and Hung and MacCormack (Ref. 12). It seems possible that
the difficulties encountered with accurate prediction of large separation
region are caused by inadequate grid resolution.
Although the region of separation was underpredicted in extent,
some interesting information was obtained from the numerical solution.
Figure 14 illustrates two velocity profiles: one at separation and
one in the separated region. The profile at separation exhibits the
expected behavior. The profile within the separated region shows
appreciable back flow, a "negative" wall shearing stress, and a
substantial height of the separation bubble. The enormous compres-
sion that a boundary layer goes through within the interaction is
well illustrated by Fig. 15. The streamlines leaving the interaction
are compressed to a small fraction of their height entering the
interaction. Figure 16 presents streamline patterns for the complete
computational region. The initial turning of the flow by the incident
shock wave as well as the location of the strong shock wave downstream
of the separated region can be seen. Approximately 16 hr of IBM 370/165
CPU time were required to obtain this solution.
33
Page 37
A E D C - T R - 7 6 - 1 1 9
O. 200 -
O. 160 -
Within 0. 120 - Separated
Region of Reverse 0.080 - I
0.040 - ~ 1
0 i -0.2 0 0.2 0.4 0.6 0.8 1.0
Velocity Ratio, ulUoo
Figure 14. Velocity profiles at the point of separation and within the separated region.
34
Page 38
AE DC-TR-76 -119
0.04
I
.Y ~ ~ - B = 16 deg
M~o = 7. 94
1 ft =, Reoo/ft = 0. 96 x 106
Tw/T o = 0. 4
0.~
y,~
O. 02
0.01
0 i 0.5 0.7 o.g 1.1 1.3
X,~
Figure 15. Streamlines for hypersonic laminar boundary-layer/shock-wave interaction.
35
Page 39
AEDC-T R-76-119
1.000
0.800
Oo 6 0 0
ylh
0.400
0. 200
|
0.5 , I i I =
0.7 0.9 x, ft
I 1.1
Figure 16. Streamlines for computational region.
36
Page 40
AEDC-TR-76-119
4.0 CONCLUDING REMARKS
The method of MacCormack has been used to solve numerically the
laminar compressible time-dependent Navier-Stokes equations for several
boundary-layer/shock-wave interactions in the hypersonic regime.
Solutions for each of the interactions were generated using conditions
corresponding to available experimental studies. Comparisons of the
numerical solutions with experimental data were made to ascertain the
validity of the numerical method and to identify regions of anomalous
behavior. Possible causes of the anomalies were examined.
The algorithm gave good results when applied to laminar hypersonic
interactions that possessed small or no separated regions. The pre-
dicted pressure distributions were in excellent agreement with experi-
mental data for cases with small or non-existent separated regions.
The predicted wall heat-transfer rates exhibited the correct qualitative
trends but not the experimentally measured quantitative values. Overall,
the code's performance should be considered as adequate for the predic-
tion of laminar unseparated boundary-layer/shock-wave interactions.
The method, when applied to hypersonic interactions having large
separated and recirculating regions, gave marginal performance. The
region of separation was underpredicted by a considerable amount
although the correct plateau pressure and the correct pressure gradient
near the end of the interaction were adequately predicted. Considera-
tion of the effects of damping as well as grid resolution suggested
inadequate mesh spacing in the x-direction as the cause.
If inadequate mesh spacing (and the corresponding lack of support for
every term of the Navier-Stokes equations) is a prime cause of the cited
discrepancies between the numerical results and the experimental data,
then the needed reduction of several orders of magnitude in Ax would
increase CPU time and core storage requirements to untenable levels.
37
Page 41
AEDC-TR-76-119
REFERENCES
I. MacCormack, R. W. "Numerical Solution of the Interaction of a
Shock Wave with a Laminar Boundary Layer." Lecture Notes
in Physics, Vol. 8, Springer-Verlag, New York, 1971, pp.
151-161.
2. Roache, P. J. Computational Fluid Dynamics. Hermosa Publishers,
Albuquerque, New Mexico, 1972.
. Richtmyer, R. D. and Morton, K. W. Difference Methods for
Initial-Value Problems. (Second Edition), Interscience Pub-
lishers, New York, 1967~
. Crocco, Luigi. "A Suggestion for the Numerical Solution of the
Steady Navier-Stokes Equations." AIAA Journal, Vol. 3,
No. 10, October 1965, pp. 1824-1832.
. Kurzrock, J. W. and Mates, R. E. "Exact Numerical Solutions of
the Time-Dependent Compressible Navier-Stokes Equations."
AIAA Paper No. 66-30 presented at the AIAA 3rd Aerospace
Sciences Meeting, New York, January 24-26, 1966.
. Skoglund, V. J. and Gay, B.D. "Improved Numerical Techniques
and Solution of a Separated Interaction of an Oblique Shock
Wave and a Laminar Boundary Layer." University of New Mexico
Bureau of Engineering Research Report ME-41(69)S-068, June
1969.
. MacCormack, R. W. "The Effect of Viscosity in Hypervelocity Impact
Cratering." AIAA Paper No. 69-354 presented at AIAA
Hypervelocity Impact Conference, Cincinnati, Ohio, April 30 -
May 2, 1969.
38
Page 42
AEDC-TR-76-119
. MacCormack, R. W. and Paullay, A. J. "Computational Efficiency
Achieved by Time Splitting of Finite Difference Operators."
AIAA Paper No. 72-154 presented at AIAA 10th Aerospace Sciences
Meeting, San Diego, California, January 17-19, 1972.
. Baldwin, B. S. and MacCormack, R. W. "Numerical Solution of the
Interaction of a Strong Shock Wave with a Hypersonic Turbulent
Boundary Layer." AIAA Paper No. 74-558 presented at the AIAA
7th Fluid and Plasma Dynamics Conference, Palo Alto,
California, June 17-19, 1974.
10. Baldwin, B. S. and MacCormack, R. W. "Interaction of Strong Shock
Wave with Turbulent Boundary Layer." Lecture Notes in Physics,
Vol. 35, Springer-Verlag, New York, 1975, pp. 51-56.
11. MacCormack, R. W. and Baldwin, B. S. "A Numerical Method for
Solving the Navier-Stokes Equations with Application to
Shock-Boundary Layer Interactions." AIAA Paper No. 75-I presented
at the AIAA 13th Aerospace Sciences Meeting, Pasadena,
California, January 20-22, 1975.
12. Hung, C. M. and MacCormack, R. W. "Numerical Solutions of Supersonic
and Hypersonic Laminar Flows over a Two-Dimensional Compression
Corner." AIAA Paper No. 75-2 presented at AIAA 13th Aerospace
Sciences Meeting, Pasadena, California, January 20-22, 1975.
13. Baldwin, B. S., MacCormack, R. W., and Deiwert, G. S. "Numerical
Techniques for the Solution of the Compressible Navier-Stokes
Equations and Implementatio~ of Turbulence Models." Paper 2
in "Computational Methods for Inviscid and Viscous Two-and-
Three-Dimensional Flow Fields." AGARD Lecture Series Pre-Print
No. 73, 1975.
39
Page 43
AEDC-TR-76-119
14. Baldwin, B. S. and Rose, W. C. "Calculation of Shock-Separated
Turbulent Boundary Layers." Paper 12 in "Aerodynamic
Analyses Requiring Advanced Computers." NASA SP-347, Conference
held at NASA Langley Research Center, Hampton, Virginia,
March 4-6, 1975, pp. 401-417.
15. Deiwert, G. S. "Numerical Simulation of High Reynolds Number
Transonic Flows." AIAA Paper No. 74-603, presented at AIAA
7th Fluid and Plasma Dynamics Conference, Palo Alto, California,
June 17-19, 1974.
16. Deiwert, G. S. "Computation of Separated Transonic Turbulent
Flows." AIAA Paper No. 75-829 presented at the AIAA 8th
Fluid and Plasma Dynamics Conference, Hartford, Connecticut,
June 16-18, 1975.
17. MacCormack, R. W. "Numerical Methods for Hyperbolic Systems."
Presented at the Short Course on Advances in Computational
Fluid Dynamics, The University of Tennessee Space Institute,
Tullahoma, Tennessee, December 10-14, 1973.
18.
19.
Horstman, C. C., Kussoy, M. I., Coakley, T. J., Rubesin, M. W.,
and Marvin, J. G. "Shock-Wave-Induced Turbulent Boundary-Layer
Separation at Hypersonic Speeds." AIAA Paper No. 75-4 presented
at the AIAA 13th Aerospace Sciences Meeting, Pasadena, Califor-
nia, January 20-23, 1975.
Haslett, R. A., Kaufman, L. G., II, Romanowski, R. F., and Urkowitz, M.
"Interference Heating Due to Shock Impingement." AFFDL-TR-72-66,
July 1972.
20. Zumwalt, G. W. "Weak Wave Reflections at near 90 deg of Incidence."
Tran. ASME, Journal Applied Mechanics, Vol. 41, Series E, No. 4,
December 1974, pp. 1142-1143.
40
Page 44
AEDC-TR-76-119
21. Carter, J. E. "Numerical Solutions of the Navier-Stokes Equations
for the Supersonic Laminar Flow over a Two-Dimensional
Compression Corner." NASA TR R-385, July 1972.
22. Grossman, B. and Moretti, G. "Time-Dependent Computation of
Transonic Flows." AIAA Paper No. 70-1322 presented at AIAA
7th Annual Meeting and Technical Display, Houston, Texas,
October 19-22, 1970.
23. Kaufman, L. G., II, and Johnson, C. B. "Weak Incident Shock
Interactions with Mach 8 Laminar Boundary Layers." NASA TN
D-7835, December 1974.
24. Cheng, S. I. "A Critical Review of the Numerical Solution of
Navier-Stokes Equations." Princeton University Department of
Aerospace and Mechanical Sciences Report No. 1158, February
1974.
4]
Page 45
AEDC-TR-76-119
NOMENCLATURE
C v
Constant volume specific heat, 4,290 ft2/sec2-°R
C Local speed of sound
Total internal energy per unit volume
Column vector containing convection and diffusion fluxes
in the x-direction defined by Eq. (3)
G Column vector containing convection and diffusion fluxes
in the y-direction defined by Eq. (4)
h Height of computational grid, ft
hf Height of fine mesh portion of computational grid, ft
IMAX Total number of mesh points in the x-dlrectlon
JMAX Total number of mesh points in the y-direction
Thermal conductivity
L Location of shock impingement point on-flat plate
Lx Operator denoting MacCormack algorithm x-direction
contribution defined by Eqs. (8) and (9)
Ly Operator denoting MacCormack algorithm y-direction
contribution defined by Eqs. (10) and (11)
M Fine grid passes for each course grid pass defined by
Eq. (25)
42
Page 46
AEDC-TR-76-119
M Free-streamMach number
P Static pressure
Re Reynolds number
ReAx Cell Reynolds number for x-mesh defined by Eq. (27)
ReAy
St=
Cell Reynolds number for y-mesh defined by Eq. (26)
Stanton number based on free-stream conditions and
adiabatic wall temperature
T Static temperature
t Time
U Column vector of conserved quantities per unit volume defined
by Eq. (2)
U Component of velocity in the x-direction
V Component of velocity in the y-direction
Coordinate along plate surface
Coordinate normal to plate surface
AF Damping term in Lx operator used to stabilize calculation
defined in Eq. (19)
AG Damping term in Ly operator used to stabilize calculation
defined in Eq. (18)
43
Page 47
AEDC-TR-76-119
At Time step used in temporal advancement
Ax Mesh spacing in x-direction
Ay Mesh spacing in y-direction
Ratio of specific heats, 1.4
Coefficient of bulk viscosity, taken as -2/3
Molecular viscosity coefficient
Mass density
Cx,Oy Normal stress fluxes in Navier-Stokes equations defined by
Eqs. (5) and (7)
~xy,~yx Shear stress fluxes in Navlef-Stokes equations defined by
Eq. (6)
Subscripts
Upstream of incident shock
Downstream of incident shock and upstream of reflected shock
3 Downstream of reflected shock
C Denotes coarse grid
Denotes fine grid
i General grid point in x-dlrectlon
44
Page 48
General grid point in y-direction
AEDC-TR-76-119
W Evaluated at wall
X x-direction
Y y-direction
0 Stagnation or total
Free stream
Superscript
II Time level
Evaluated during corrector pass using predictor values
45