AIAA JOURNAL

VOL. 22, NO. 7, JUL Y 1984

921

,..
y

The Baldwin-Lomax Turbulence Model

for Two-Dimensional Shock-Wave/Boundary-Layer Interactions
M. Visbal* and D. Knightt

RUlgers UniversilY, New Brunswick, New Jersey

The algebraic lurbulenl edd)' viscosily model of Baldwin and Lomax has been crilically examined for lhe case
of Iwo-dimensional (2-D) supersonic compression corner inleraclions. The flowfields are compuled usin)! lhe
Navier-Slokes equalions logelher with three differenl nrsions of lhe Baldwin-Lomax model, induding lhe
incorporation of a relaxation lechnique. The lurbulence models are nalualed b)' a delailed comparison wilh
av'ailable expe~imental data for comprcssion ramp flows m'er a range of corner angle and Re)'nolds numher. The
Baldwin-Lomax outer formulalion is found 10 be unsuitable for separated 2-D supersonic inleractions due l;u;e

)v'\A

R0'\1
~

unphysical streamwise ,'ariation of lhe compuled len th scale in lhe v'iclnll)' orSelJaration. Minor modifications
~p!QlJose
to p..!Irtiall)'re~e~)' Ihi~ diff!cull)'. The use of relaxation prov'ides significant impro\ement in the
flowfield prediction upstream of the corner. However, the relaxalion length required is one-lenth of that employed in a prnious computalional stud)'. Ali of lhe turbulence models lesled here Jail 10 simulale lhe rapid
recovery of lhe boundaf)'la)~

I.

do_wn!~a~

r~t1achmenl.

Introduction

HE
numericalaerodynamic
solution ofnows
the isNavier-Stokes
equations
for complex
now possible as
a result
of increases in computer capability,
the development
of efficient numericaJ algorithms,I.2
and the recent advances in
grid generation
techniques.3
Praclical
high-speed
nows,
however, are usually turbulent and thus a suitable empirical
turbulence model musl be selected. Algebraic eddy viscosity
models slill represent the most common
choice for compressible
Navier-Stokes
codes since their implemel)tation
results in the minimum requirements
of compuler
time and
storage, which is particularly important in three-dimensional
(3-D) compulalions.
Several Iwo-layer algebraic lurbulence
models (such as
Cebeci-Smilh~)
require,
for their
implemenlalion,
lhe
delerminalion
of the boundary-layer
Ihickness
and edge
velocily.
This constitules
a praclical
disadvantage
in lhe
numerical
solution
of
the
Navier-Slokes
equalions.
Specifically,
IwO effects complicale any altempl 10 devise a
suitable algorithm
for determinalion
of lhe boundary-Iayer
edge, namely, I) lhe presence of nonuniform
inviscid regions
in which lhe inviscid now varies in lhe direclion normal 10 lhe
boundary, and 2) lhe presence of small spurious oscillalions in
lhe numerical
solution. As discussed by Hung and MacCormack5
for compression comer flows and Baldwin and
Lomax6 for Iransonic flow over an airfoil, large varialions in
the compuled OUler eddy viscosity can occur as a result of the
uncertainlies
in lhe boundary-Iayer
thickness.
ln an altempt to overcome this difficulty,
Baldwin and
Lomax6 recently proposed a new algebraic eddy viscosity
model, paltemed after thal of Cebeci and Smith. This new
model does nOI require the delerminalion
of the boundarylayer edge, and Iherefore eliminales a source of potenlial error

Submitled May 16, 1983; presented as Paper 83-1697 atthe AIAA

16th Fluid and Plasma Dynamics Conference, Danvers. Mass, July
1:!-14, 1983; rev'ision received 01.'1. 28, 1983. Cop)Tight American
Institute of Aeronautics and Astronautics, lnc., 1983. Ali rights
reserved.
'Graduate Student, Depl. af Mechanical and Aerospace Engineering; presently, Visiting Scientist. Air Force Wright AcronaUlical Lab .
\\'right Pallerson AFB. Member AlAA.
t Associate Professor. Depl. af Mechanical and Aerospace
Engineering. Membcr AIAA.

in the compuled outer eddy viscosity. In addition,

since it
employs the vorticily which is invariant
under coordinale
Iransformalions,
the model may be applied to 3-D configuralions.
Oue to the above advantages and ease of implemenlalion,
lhe Baldwin-Lomax
model is a popular
algebraic
eddy
viscosity model in compulationaI
aerodynamics.
lndecd, this
model has becn applied (somelimes quite uncrilical!y) to a
variely of 2-0 and 3-D flowfield
calculalions
(sec, for
example, Refs. 7-10). Baldwin and Lomax6 evaluated their
model in delail for the case of Iransonic now over an isolaled
airfoil. Hungll employed lhe Baldwin-Lomax
model (in its
original form) for the simulation
of several compression
comer
nows. Oegani and Schiffl2
recently applied
the
Baldwin-Lomax
model in the computalion
of supersonic
nows around cones aI high incidence. They found the model
10 be unsuitable for regions of cross-flow separation.
due to
ambiguilies in lhe delerminalion
of the length scale. Oespile
its increasing use, no additional evaluations
of the BaldwinLomax model have becn conducled and are therefore necded.
The presenl investigation is aimed at partial!y fulfilling this
need by performing
a critical examination
of the BaldwinLomax model for the case of shock/boundary-Iayer
inleraction in a supersonic
compression
ramp (Fig. I). This
work represents a more detailed and eXlensive evaluation than
that of Ref. 11. The major focus of this research is to identify
the merits and deficiencies of the Baldwin-Lomax
mode! for
supersonic interactions, and to develop, whcn possible, simple
'modifications
(wilhin the limilations
of the algebraic edd)'
viscosity concept) that could improve the overal! nowfie!d
prediction.
The present test now case has becn selected for
two main reasons: I) shock/boundary-laycr
interaction is an
important phenomenon in many practical high-specd flows,l3
and 2) sufficiently detailed experimental
mcasurementsl~.16
are available for lhe 2-0 compression
ramp configuration.
The compression comer flows \Vere simulale:d using the fuI!
2-0 mass-a\'eraged
Navier-Stokes
cquations 17 expressed in
slrong
conser\'ation
form 18 and in general
curvilinear
coordinates.
Several versions of lhe Baldwin-Lomax
algebraic
turbulence
modcl were invesligaled,
induding
lhe incorporation
of lhe relaxation
tcchnique
of Shang and
Hankey.19 The gowrning
equations
were: sol\'e:d in nc:arly
orthogonal
body-fillcd grids~ e:mploying the implicil, approximate-faclOrizalion
algorithm of Bcam and \\'arming. I

M. VISBAL

922

11.
GOH'rninl:

Equations

Mf:"lhod of Solution

and T",rbulence

Models

The governing equations

are the fuIl mean compressible
Navier-Stokes
equation~
in two dimensions
using massaveraged variables,17 strU1Jg conservation
form,18 and general
curvilinear coordinates.
The Ouid is assumed thermaIly and
caloricaIly perfecl. The TT/olecular dynamic viscosity J1 is given
by Sutherland's
law. Thc molecular Prandtl number Pr= 0.73
(for air) and the turbulent Prandtl number Pr, =0.9.
Three different version~ of an algebraic two-Iayer turbulent
eddy viscosity rriodel w~re employed
in the present computations.
The first turbulence
model is that proposed
by
Baldwin and Lomax.6 In lhe inner region the eddy viscosity is
given by the Prandtl- Van Driest formulation
(I)
of the vorticity,
where p is the density, rlw I the magnitude
K = 0.40 is von Krmu's
constant,
and Y is the normal
distance from the wall. The Van Oriest damping factor D is
given by
(2)

where T . is the waIl sheM stress. In the outer region, in order

to eliminate the need Df finding the boundary-Iayer
edge,
Baldwin and Lomax replaced Clauser's
formulation
by the
foIlowing relation
(3)
where k = 0.0168 is Clauser's
constant
and
ditional constanl. The Olllcr function F ..a . is

C,p is an ad-

(4)

where Fmu =max(YlwIlJ),

and
which Fmu oecurs. The Klebanoff
given by

For equilibrium
turbulent
boundary
layers, as weIl as for
transonic Oow over an airfoil,6 the outer function F typicaIly
displays
a single well-defined
extremum
and the determination of F mIV. and Y mu is straight forward. For separated
supersonic
Oow over a compression
ramp, however, the
present research indicates that F displays two peaks in the
vicinity of separation.
Similar behavior
was observed by
Baldwin and Lomax6 for a 2-0 oblique shock/turbulent
boundary-Iayer
interaction.
Since the values
of Y m.u.
assoeiated with each one of thcse extrema may differ by one
arder of magnitude, the sclection of the peak closer to the waIl
(at the streamwise locations where it represents lhe absolute
maximum)
results in an abrupt, unphysical reduction in the
compuled outer eddy viscosity. An additional problem in the
Baldwin-Lomax
model (for both the inrier and outer formulation) may be caused by the vanishing of the Van Oriest
damping factor D at the locations where ;T , approaches zero
(i.e., near separation
and reattachment
in 2-0 Oows). The
smaIl values of D result in an unphysical reduction of the
computed
eddy viscosity. This effeet is found to be more
pronounced
for the inner eddy viscosity in the vicinity of
reattachment,
and in some cases (see Seco 111) can prevent the
Oowfield from achieving a fully steady state in this region.
In order to avoid the above difficulties, a second turbulence
model, referred to subsequenlly
as the modified Baldv.inLomax model, was employed. This new verslOn J.!1coroorates
two modifications,
namely,
I) at the locations where F
displays two peaks, the values 01 F ma> and Y mu are obtained
fram the extremum farthest from the waIl (ou ter peak), and
Van shear
Oriest stress
damping
factor in[Eq.
(2)], ofthethe
local
\alue
~ 2)jn thethetotal
(defined
terms
velocity
component paraIlel to the wall) is used in place of T .
The third turbulence
model incorporates
the relaxation
technique of Shang and Hankeyl9 in an attempt to account
for upstream turbulence history effects. The relaxation eddy
viscosity E (for both the inner and the outer formulation)
is
given by
(6)

is the value Df Yat

intermittency
correction is

Ymax

(5)
In the original Baldwin-Lomax
where Ckleb is a constanl.
model,6 an alternate formulation,
applicable to wake Oows,
was included in the expression
for F"'rn- Although
this
formulation
has becn employed in previous computations6.7.1I
of shock/boundary-Iaycr
interactions,
it is the authors'
opinion
that its use is not justified
for the present investigation, and, therefore, was not considered. The turbulent
eddy viscosity is switchcd fram the inner to the outer formulation at the location where Ei > EO' Baldwin and Lomax
suggested the values for Cep = 1.6 and Clleb = 0.3 based on a
comparison
with the' Cebeci
and - S"mlth model4
for
equilibrium
boundary
layers at transonic
speeds.
In the
present research, these constants were found to be dependenL
on the Mach number of the Oow .. 1I can oe Shown21 that for
- layer,
an equilibriumincomprcssible
= Ql..turbulent
Crp = 1.2
which obeys the waIl/wa'e ~M""
law,u
the values boundary
and Clleb = 0.65 are reqllired in the Baldwin-Lomax
model.1O
-c>rderro-matd
the Clauser-Klebanoff
formulation.
In addition, a series of Oat plate near-adiabatic
(Twl Tadiabatic
= 1.12) turbulent boundary-Ia)'er
computations
at Mach 3.0
dictated
the use of CC'p = 2.08 for the prescnt
ramp
ealculations.23
The elimination
of the need to determine
the boundarylayer edge in the outer edd)' viseosity formulation
eonstitutes
a major advantage
of lhe Baldwin-Lomax
model over the
Cebeei-Smith
model. This is onl)' true, howe\"er, when the
outer
funetion
F=
\wlD
pro\"ides
an
unambiguous
e\"aluation of the veloeit)' seale F ma\ and the lengt h seale }'mo>'

AIAA JOURNAl

ANO O. KNIGHT

where Ecq is obtained

from the modified Baldwin-lomax
model and Eu!,,, denotes the value of the eddy viseosity at the
upstream location Xo where the surface pressure rise begins.
A relaxation
length equal to the incoming boundary-layer
The
thickness 00 was employed in the present computations.
determination
of the value of the relaxation
length is
discussed in Seco 111.

Compulalionnl

Domain

and 80undnry

Condilions

The shape of the compulationaI

domain is sho\\'n in Fig. I.
The inOow boundary
was located ahead of the comer in a
region of no upstream inOuence. The outOow boundary "'-as
placed sufficiently
far from the comer, in a region of small
streamwise
Oow gradients. The height of the computatio:Ja!
domain
(3-4 00) was chosen so as to obtain freestrcam
conditions
along lhe upper boundary
and to ensure the
emergence of the shock through the downstream boundar).
On the solid surface, the nonslip, isothermal condilio ..s
u= v=O and T= T . were applicd along with a bound2..'}
condition
for the pressure derived from the normal co:nponent of the momentum
eqllation.:1
For the freestTc:2.."'I1
boundary, a no-reOection condition, 2~suitable for superso:-ic
Oow, was prescribed.
Alon" the outOow boundar).
:.'1e
conventional
extrapolation
eondition ala~ =0 was emp!Oy~j.
The IIpstream
boundary
eondilions
were obtained
by
ca1culating the de\"clopment of a Oat plate lurbulent bou:-:d~'}'
layer up to the locations where the computed mom~:1!::""'I1
thickness O matched the experimental
\"alue. At the S2::1~
loeations, the computed and measured \"elocit)', skin fri~:x>n
and displacement
thiekness \\'ere also eompared and fOl:.:1':in
\"er)' good agrecmenl.
F0r inst:1I11:e, at the rnatching s.::::.:i-.'n

SHOCK-\V A VE/BOUNDARY-LA YER INTERACTIONS

JUL Y 1984

where ReB =8.2 x 104, the computed and measured skinfriction coefficient cJ are 1.02 x 10 - 3 and 1.00 x 10 - 3,
respeetively. The value of cJ predieted by the Van Driest 11
theory and the von Krmn-Schoenherr
equation25 is
1.04 X 10-3. At the same station, the eomputed displacement
thickness is essentially equal to the experimental value (0.66
em). Exeellent agreement was also found21 between the
computed and measured velocity profiles and the law of the
wall. :u
The eom2utational grids were generated by the numerical
pi"ocedure ofRf-20:- A-n-onuniform mesh spacing was used
in both eoordinate direetions in order to provide sufficient
resolution of the turbulent boundary layer and the interaetion
region. In the direction normal to the surfaee, the grid points
were distributed using a eombination of geometrieaIly
stretehed and uniform spacing. The normal spacing at the
wall was ehosen in order to resolve the viseous sublayer, and
satisfied the requirement trY~in~ 2.5 at al1 loeations. The
typical number of grid pomIS wlthm the boundary layer was
25 to 30. lhe streamwise mesh spaeing in the interaction
reglOn ranged from 0.027 o for the 8-deg ramp to 0.077 o
(for the 24-deg ramp). The maximum streamwise spaeing
(outside the interaction region) was always less than 0.6 o
Numerical

923

experimental cases. For the first category (i.e., variable ramp

angle), the different versions of the algebraic turbulcnce
mode! were employed. For the second category (i.e., variable
Re6 ), only the relaxation model was applied. A detailed
co~parison of the computed flowfie!ds and the experimental
data was performed; however, only the most significant
results are presented below. Reference 21 contains a more
extensive comparison along with the details of ali computations.
Results for Variable

Ramp Angle

Resultsfor 16-deg Ramp

The 16-deg ramp flowfield was computed using each of the

versions of the turbulence mode\. This compression comer
flow eonstitutes, according to the experiments, an incipiently
separated interaction. The computed and measured surfaee
pressure is shown in Fig. 2. The results for the original
Baldwin-Lomax and re!axation models are in elose agreement

Mo::>
b

76x 1O(

= 2.9
R

e8<o 7.7x 10

Algorithm

The governing equations were solved using the implicit,

approximate-faetorization algorithm of Beam and \Varming. I This seheme was formulated employing Euler implicit
time-differencing and second-order, eentered approximations
for the spatial derivatives. The boundary conditions were
implemented in the explicit or lagged approaeh described by
Steger. 26 Fourth-order explicit damping terms, required for
the smoothing of the embeddedshocks,
were preseribed
aecording to the procedure of Ref. 27. The developed NavierStokes eomputer code was extensively validated,21 with excel1ent results, for several test cases including inviseid shocked
flows, laminar and turbulent boundary layers, and laminar
shoek/boundary-Iayer interactions.
Sinee the compression ramp flowfields are obtained by time
integration of the unsteady Navier-Stokes equations until a
steady state is reached, considerable care was exereised in
ensuring convergence. Separated interactions were run for
physieal times of up to 10le' where le is the time required for a
fluid parcel in the inviscid region to traveI from the upstream
to the downstream end of the mesh. The flowfields were
assumed eonverged when the maximum reIative variation of
the flow variables over Ile were less than 1.0070.It should be
noted that, while changes of order 1.0070 occur at a few mesh
points (in regions of shock smearing), the reIative ehanges
were much smal1er at most locations. The eorresponding
average variations over Ile were typical1y less than 0.05070.

IH.

C~

r-----I

ATIONA1. DOMAIN_~

.- /\

I
I

Fig. 1

Flow configuratlon

D
-

and computatlonal

domaln ..

EX1'ERt.e'lT

BAil>WN-LOMAX
M::OEl.
MOOtF1ED
B-L M::OEl.
- - . - - RB.AXA TION M::OEl.
--

.'4

~
~ 2.0

1. O

-2.0

-1.0

0.0

1.0

2.0

X/3o
Fig. 2

Surface pressure

distrlbutlon

for 16-deg ramp.

Results and Discussion

Figure I iIlustrates the eompression comer geometry

employed in the present evaluation of the Baldwin-Lomax
mode\. An extensive experimental study of this flow configuration, for a nominal Mach number AI"" = 2.9, has beco
conducted in recent years.14.16 The available experimental
date basel4 may be divided into two major categories:
1) surface and mean flowfie!d data for four comer angles
(a=B, 16, 20, and 24 deg) at a fixed Reynolds number
wal1
pressure
a fixed
ramp
of 20 deg at four different Reynolds numbers
(Rel>o=0.76x
106, 3.4x 106, 5.6x 106, and 7.7 x 106). The
first category ineludes measurements of surface pressure; skin
friction; and velocity, f\lach number, and static pressure
profiles at nine streamwise stations for each ramp angle.
Thesc flowfields encompass nominal1y attaehed (B-deg ramp),
as wcll as fully separated (20- and 24-deg ramps) interactions.
Computations were performcd for al1 cight of the above

I'

10.

Rel>o
= 1.6 x 106,and
andreattachment
2) surface data
(including
ano separation
locations)
for

angle

Flg.3

Ucfinltion

of Intcractlon

geometrlc

dlstances.

M. VISBAL AND D. KNIGHT

924

with each other and with the experimental data. However, the
computed surface pressure obtained with the modified model
~p
(see Fig, 3
displays an insufticient upstream propagation
for the definition
of geometric distances),
similar to the
results of Shang and Hankeyl9 and Horstman et al.,28 who
employed the Cebeci-Smith turbulence mode!.
The skin-friction
results are presented
in Fig. 4. Ali
calculations predict the existence of a separated region, which
is not displayed by the experiments.
The original BaldwinLomax
and relaxation
models are in reasonably
good
agreement
with the expe:riment in the region of sharply
decreasing cf' Howe:ver, both models seriously underpredict
the skin-friction
values in this recovery region (Le:., downstream of reattachment).
Also, the compute:d separation-toreattachment
lengths are toa large. The modified mode1
provides substantial improvement in this regard. Downstream
of attachment,
however, ali models approach the same skinfriction levei which is significantly below the measured value.
This behavior is similar to previous computations
using
algebraic eddy viscosity models. 28
The computed and expe:rimental velocity profiles at three
stations (upstream,
at the comer, and downstream
of the
interaction) are shown in Fig. 5. The original and relaxation
models give slightly bette:r results upstream and at the comer
(first two profiles). Downstream
of reattachment,
ali three
mode1s result in a ve]ocity profile that displays an insufficient
recovery or "filling out" near the wall. This observation
is
consistent with the underprediction
of skin friction diseussed
above.
Figure 6 shows the: evolution
of the outer funetion
F= Y IwlD across the interaction, for the ealculation with the
original Baldwin-Lomax
model. The corresponding
value of
Y max' and the: maximum value of fO at each streamwise: station
are given in Figs. 7 and 8. Upstream
of the interaction

2.0

D
- ---------'-

AIAA JOURNAL

(X /o = - 1. I), F displays a single:, well-define:d peak. lmmediately

before the: separation
point (X/o = - 0.27), F
exhibits two distinct extrema (re:ferred to subse:quently as the:
inner and outer peak). At this loeation the: outer peak. which
represents the absolute maximum. is still chosen by the model
to compule
F ma> and Y ma>' Downstream
of separation
(X/o= -0.19). the inner peak, whieh is very dose to the
wall, exceeds the outer one, and Y max abruptly decreases by
one order of magnitude: (Fig. 7). Despite the increase in F ma>.'
a net sudden drop in F.kt occurs. This reduetion in F.akt'
eombined
with the effeet of Ymu in the: Klebanoff
intermittency correction [Eq. (5)], produces a sharp decrease in
the computed
outer eddy viseosity (Fig. 8). As Fig. 6 indicates, lhe outer peak disappears further downstream,
and
the inner peak moves away from the: surfaee to a new
equilibrium position. These large streamwise variations in the
eomputed
length scale: Y ma> are unphysical and constitute a
major deficieney of the original Baldwin-Lomax
model for

.5E.o1

F/

0.3

EXPERr.e;T
BAl.DWlN-LOMAX MODEl.
MXJFED B-l MODEl.
RELAXATION MODEl.
CQNSTANT YMAX

--

,,
,
"

1.0 t a

0.2

,,.,.- .'
" ,.'.'
I'

--::-:.....

I
.1msu.

0.0

:1'
I
I

--

4.0

0.0

-2.0

1.0

MCOFED B-l '-COEI.

2.0

4.0

6.0

xjSo

Skinfriclion coeffieienl for 16-deg ramp.

x _

'-COEI.

0.0

2.0
X/So

Fig.4

- BAl.DWN-lOMAX

, I
O. t

0.0

-2.0

," - -

~o

.i

-1.0

Uc>

Fig. 6 E"olution of lhe Baldwin-Lomax OUler edd)' viscosil)' fuoelion F(Y) across 16-dcg ramp inleraclion.

li
li

0.0

Fig. 7

I-"

5.6

I"

III
~i

'o

"

'" :s..

~ 8

3.0
5.0 0.0

2.0

4.0

li

li

&0-

t.O
0.0
-2.0
I~

Variation of }'max across 16-deg ramp inleraeeion.

"
I

,'

I I
0.5

,,-

- -----

, -

- - BAl.Dwr./-LOMAX

--

MODFED B-l MODEl.

:(
I

0.0

2.0

4.0

6.0

X/~o
Fi~. 5 "doeil)" profiles aI sneral
\.6 X 106).

slalions

(a =

16 deg.

Rt'ho

Fig.8

Variall"n

"f IOmn

Der"" t6-dcg ramp Inlcn~ell"n.

925

SHOCK-WA VE/BOUNDARY-LA YER INTERACTIONS

JULYl984

the predietion of supersonie shoek/boundary-layer

interaetions. Similar problems were eneountercd by Baldwin
and Lomax6 in the computation of a 2-D oblique shoek interaction, and by Knight23 in a 3-D interaetion ..
The behavior of the outer funetion F in the interaetion
region for the modified Baldwin-Lomax model is similar to
that deseribed above. The corresponding value of Y ma.. and
Om" are also shown in Figs. 7 and 8. Sinee in the modified
model the outer peak is always selected, an abrupt deerease in
Y ma' and Omax still oeeurs at the streamwise loeation where
the outer peak in F disappears.
For the 16-deg ramp eomputation using the original
Baldwin-Lomax model, a fully steady-state solution could not
be aehieved in a small region close to the wall in the immediate
vicinity of reattaehment. In this region the flow variables
exhibited bounded oscillations in time. This problem is apparently assoeiated with the very low values of the eomputed
inner eddy viseosity, eased by the Van Driest damping factor
D [Eq. (2)] approaching zero in the reattachment region .
This diffieulty was overcome in the modified version of the
model by employing the local total shear stress in the
evaluation of D.
In order to investigate the effects of the length seale Y max'
an additional computation was performed for the 16-deg
ramp utilizing the modified Baldwin-Lomax model with a
constant or "frozen" value of Ymax throughout the flowfield.
The eomputed surface pressure was essentially identical to the
results for the modified model. The eomputed skin-friction
coefficient, shown in Fig. 4, gave only a slight improvement in
the recovery region. The velocity profile at the downstream
station XI o = 5.4 (not shown) was very close to the previous
results (Fig. 5), and again failed to predict the rapid recovery
of the boundary layer downstream. This caIculation indicates
that the computed flowfield in the recovery region is not very
sensitive to changes in the ou ter eddy viscosity.

Downstream of reattachment, ali models seriously underpredict the reeovery of the boundary layer. The computed
separation-to-reattachment
length is toa large, although lhe
modified model does slightly better in this respeet.
Figure I I shows the computed and measured velocity
profiles at several streamwise loeations. The relaxation model
gives some improvement in the computcd vcloeity for the first
part of the interaction. This is consistent with lhe better
prediction of upstream propagation
discussed above.
Downstream of reattaehment, ali models fail to predict the
rapid recovery of the velocity near the wall. The results for the
stalie pressure are given in Fig. 12. ln the interaclion region,
the relaxation model displays the best comparison with lhe
experiment. The measured static pressure profiles downstream of reattaehment exhibit a normal gradient near lhe

o EXPER~
- - -SALOWIN-LOMAX
MODEl.
--MODIFIED S-L MODEl.
-. - - RELAXA TION MODEl.

2.0

"

1.0

0.0

-1.0
0.0

-2.0

2.0

4.0

Fig. 10

Sldofrielino eoeffieieol for :!O.degramp.

EXI'<'RMNf

--~-LCM.tJ<

--MOOFEO
- RElAXA

\.0
x _

~-

MCXJa.
91. MCXJa.
MCXJa.

T10N

0.5

0.0

Fig.
R~60

o.

11 "elodl~'

= t.6x

profiles

aI

senral

slallo05

(a

= :!O

de\:,

106).

OE~
- - -BAlDWIV-LOMA.X
"*DEI.
-MOOFEO B1. "*DEI.

RElAXA 110'< "*DEI.

0.0

1 0.16

: )\ 4-6

,eI.
....

2.0

. ot.
.....

,."

EXPffi~

BAl.DW1N-LOMAX
MODa.
~\X)F18) S-L MODa.
- - - - - - RELAXA TION MOOEl.

--

..

0.0

1.0
0.0

2.0

4.0

,'.'".,

,...
\

0.5
()

I.

6.0

X/t;o

Fig.

Surfare prt"SSul"l'dislribulioo for :!O-de):ramp.

R~6o

:1
I
.1
111

a'

3.0

Fi!:. 9

1.0

0.0

-0.44

'\.0

-2.0

6.0

X/t;o

Resu/rsfor 20-deg Ramp

The 20-deg compression ramp was simulated using the Ihree

different versions of the turbulence model. This case
represents, according to the experiments, a fully separated
interaetion. The computed and measured surface pressure is
shown in Fig. 9. The results for the modified Baldwin-Lomax
model significantly underprediet lhe extent of upstream
propagation and do not display the pressure "plateau" observed in the experiments. The use of relaxation, with a
relaxation length = o. substantially improves the wall
pressure predietion and gives the correct upstream influence.
However, the computed pressure plateau is more pronouneed
than in the experiments. The computed pressure for the
original model falls between the results for the modified and
relaxation models. The skin-friction results are shown in Fig.
10. The relaxation and original models provide a better
agreement with the experiment upstream of separation.

a
a

""

""

,
2.

Slalic presslIl'l' prtlriks

= 1.6

X 106).

. p/Pco
aI st'\er.ll sl:llinos Ia =:!O de\:.

M. VISBALAND

926

wall, which is not duplicated in the computations. This ob

servation also applies to the 16- and 24-deg ramp flows.
Resu/tsfor 140deg Ramp

This fully separated compression comer interaction was

simulated using the modified BaldwinoLomax and relaxation
turbulence models. The comparison of computed and experimental results, contained in Figs. \3-15, exhibits the same
characteristics observed for the 20-deg ramp. As compared
with the modified model, the relaxation model (with = o)
provides a marked improvement in the prediction of surface
pressure distribution (Fig. \3), ineluding the pressure plateau
leveI and the extent of upstream influence. This agreement
dictated the use of =o in the present research. This value of
was also found by Horstman
et aI. 28 to predict X
reasonably well. On the other hand, Shang and Hankeyl~
required a value of = JOo in their compression comer
calculations. This will be~discussed in more detail below. The
use of relaxation also results in a eloser agreement with the
measured skin friction ahead of separation (Fig. 14). Both
models, however, overpredict the length of the separation
region, and do not provide the correct skin-friction leveI
downstream of reattachment. The higher values of c, obtained by Baldwin and Lomax6 for the 24-deg ra~ case are
~in_contrajCtion with the presnt results and those of Ref. 28.
Upstream of the corner, the relaxatian model praduces
again an improvement in the predicted velacity prafiles (Fig.
15). Downstream of reattachment,
bath models fail to
simulate the rapid recovery of the baundary layer near the
wall.
ln arder to examine the effects af the relaxatian length an
the computed flawfield, the 24-deg ramp flaw was also
calculated using a "frozen" eddy viscosity madeI. The term
"frazen" denotes that the eddy viscosity profile upstream af

D. KNIGHT

AIAA JOURNAL

the interaction was cmploycd at ali strcamwisc locations [i.e.,

= 00 in Eq. (6)]. Thc computalion was run for approximately three characteristic times, aI which point the
extent af upstream propagatian D.Xp had increased by almast
100070, as comparcd
with the previaus results for the
relaxation mode! with = o. Results by Harstman ct aI. 28 far
the same flow conditions indicated an increase of up to 500;0
in Xp when was changed from I o to 5 o. On lhe other
hand, camputations by Shang and Hankeyl9 for a 25-deg
ramp (with M -2.96 and Re6(J = 1.4 x 105) employing a frozen
and a relaxation ( = I Oo) model, did nOI exhibit this drastic
difference in Xp. The use of = 10o is then expected to
produce a grass overpredictian af the upstream influence far
the present compression comer flow. Since esscntially the
same baseline turbulence model (namely CebecioSmith4) was
used in Refs. 19 and 28, the discrepancy in the value of ,
required ta match the experimental pressure, is probably due
to the difference in the flaw Reynalds number Reho' In fact,
the measured separatian-to-reattachment
length for the
present 24-deg ramp flow is 1.7o while the corresponding
length for the 25-deg ramp cansidered in Ref. 19 is approximately 8 o. This could be interpreted as a dependence of
on the extent of the interaction, which far a given geometry
is a function of the flow parameters (Reho and M ). This
variatian of the relaxation length (which is intended to
represent the lagged response of the turbulent stress to sudden
mean flow gradients) is perhaps reasonable since for
separated shock/boundary-Iayer interactians large increases
in the Reynolds stress are observed29 before the reattachment
location.

Results for Variable Re)'nolds Number

Since the previous results far variable ramp angle indicated

that the use of relaxation pravides some improvement in the
predictian of the interaction upstream influence, it is of in-

5.0
C

EX.~

--

Sol ..x>El.

M(X)FID

- .... - RElAXA

MCX:El

I.D

4.0

-'.2

t=

~
Poo
3.0
D.S
2.0

D.

1.0

-4.0

-2.0

0.0

2.0

4.0

6.0

D.O

o.

U/Uco

X/tJo
Fig. 13 Surface pressure distribution for 24-deg ramp.

Fig.

IS

R~iO

= 1.6

Velocit)

[;)

E~

se\"eral

(a = 24 de'l.

slations

2.5

11

x,.

COM'.

[;)

li

.....
1..8 2.0
1.0

at

X 106).

2.0
-MOOFlED S-L MOOel.
- - - - - RELAXAT10N MOOEI.

promes

...

li.
<Jlc.<)
><1 O

..

O
..

"'

-t-

"'

1.5

"' "'

"'
...

---

-----

11

0.0
0.5

-1.0

-4.0

-2.0

0.0

2.0

4.0

6.0

X/tJo
Fig. 14 Sldnfriclion coeHicient for 24-del: ramp .
..

--

0.0 I

.SI:':.06

.SE+07
RecS o

Fil:. 16 ~rrl'<"1of Rt'ho on .:lXp'~Xs'

nnd [os (cr= 20 dc!:).


SHOCK-WA VE/BOUNDARY-LA YER INTERACTIONS

JULY 1984

927

c.

terest to evaluale lhe relaxation model for the case of variable

Reynolds number. For this purpose, several calculations were
performed for a 20-deg ramp over the Reynolds number range
0.76 x 106 Reho 7.75 X 106, using the relaxation turbulence
modelonly.
In order to illustrate the effects of Reynolds number on the
interaction, the values of t:.Xp, t:.X, (see Fig. 3 for definitions)
and the separation-to-reauachment
length L, are presented as
functions of Reho in Fig. 16. Although the computed results
for t:.Xp, t:.X,. and L, exhibit the correct Reynolds number
trend, only the upstream pressure innuence t:.Xp is predicted
with reasonable accuracy (see Refs. 16 and 28 for the scaUer
of the experimental data). The distance from the separation
location to the comer t:.X, as well as the overall separation
length L, are consistently overpredicted. The above observations are in agreement with the results of Ref. 28. The
relaxation model (with a constant relaxation length = 00) is
capable of predicting the upstream pressure innuence with
reasonable accuracy over the entire Reynolds number range
investigated.

IV.

Acknowlcdgmcnts
The authors gratefully acknowledgc many helpful and
stimulating discussions with S. Bogdonoff, D. Dolling, G.
Settles, and L. Smits. This research was sUPP0rled by lhe Air
Force Office of Scienlific Research under AFOSR Grant 820040, monilored by J. Wilson. Inilial development of lhe
numerical algorithm was supported by lhe Air Force Flight
Dynamics Laboralory-Air
Force Wright Aeronaulical
Laboratories under AFOSR Grant 80-0072, monilored by J.
Mace.

Conciusions

A critical evaluation of the algebraic turbulence model of

Baldwin and Lomax was performed for the case of 2-D
shock/boundary-Iayer
interactions induced by compression
comers. Three different versions of this algebniic eddy
viscosity mode! were investigated, including the incorporation
A detailed compai-ison of the computed
of relaxation.19
nowfields with the available experimental data 14 was performed, and the capabilities and deficiencies of the turbulence
models were identified.
Regarding the characteristics of the Baldwin-Lomax model
for 2-D supersonic interactions, the following
specific
conclusions can be made:
I) The constants Ccp and Clltb, appearing in the BaldwinLomax outer formulation, were found to be dependent on the
now Mach number. These constants vary by a factor of two
over the Mach number range O sM '" 3.0, and therefore need
to be adjusted accordingly.
2) The Baldwin-Lomax outer function (F= Y Iw ID) is not
suitable for the determination of the length scale in the
separation region of the interactions investigated. This is due
to the appearance, near separation, of a double peak in F( Y)
which results in an abrupt (unphysical) decrease in the
computed length scale. This behavior constitutes a major
deficiency of the model for supersonic interactions, and could
perhaps be eliminated by the use of a different ou ter function.
In addition, for an incipiently separated interaction, the small
values of the eddy viscosity near the reattachment location
(caused by the vanishing of the Van Driest damping factor)
can prevent the solution from achieving a fully steady state.
The above difficulties can be panially overcome by using
the local total shear stress in the Van Driest damping term and
by the selection of the outermost peak of F( Y) in the computation of the length scale. These modifications provide a
better now prediction near reattachment.
3) Computations with the original and modified BaldwinLomax models exhibit an insufficient upstream propagation,
caused by the inability of the models to reproduce the lagged
response of the turbulence structure to the sudden adverse
pressure gradient. Significant improvement in lhe nowfield
prediction upstrcam of the comer can be obtained with the use
of relaxalion. A relaxalion length equal 10 the incoming
boundary-Iayer Ihickness was found 10 be suilable for the
range of Reynolds numbers and ramp angles considered. This
value, however, is one-tenth of thal suggesled by Shang and
Hankeyl9 and is expecled to depend on the extent of the interaction.
4) Ali of lhe turbulence modcls lested here fail to predicI
the_~a[1id recovery of=lhe boundary layer downstr~
reattachment. This is due 10 lhe inability of lhe models 10
simulatc the observed:~33 amplification of lhe turbulence

- 7'

nuctualions across a shock/boundary-Iaycr interaction. This

,deficiency would lead to a ralher poor prediction of nows
wilh multiple interactions, and means of improving the
present results have nOI yct been found. The fact Ihat
downslream of the interaction, the mean velocity profile
rapidly approaches its equilibrium shape, while the cnhanced
lurbulence nucluations relax very slowly loward equilibrium,3c}'J2points OUI lhe inadequacy of the algebraic eddy
viscosilY concept for these complex nowfields.

Refcrenccs
1 Beam, R. and Warming. R., "An Implicit Factored Scheme for
the Compressible Navier-Stokes Equations," AIAA Journal, Vol. 16,
April 1978, pp. 393-402.
2MacCormaek, R. \V., "A Numerical Method for Solving the
Equations of Compressible Viscous Flow," AIAA Paper 81-0110.
1981.
F., Warsi, Z. U. A., and Mastin, C. \V. "Bound3Thompson,
ary Fitled Coordinate Systems for Numerical Solution of Partial
DifferentiaI Equations-A
Review," Journal 01 CompUlarional

l.

Phrics. Vo\. 47,1982, pp. 1-108.

Cebeci, T. and Smith, A. M. O., Analysis 01 Turbulenr Boundary

La{,ers, Academic Press, New York; 1974.

Hung, C. M. and MacCormack, R.W., "Numerical Simulation of
Supersonic and Hypersonic Turbulent Compression Comer Flows."
AIAA Journa/, Vol. 15, March 1977, pp. 410-416.
6BaIdwin, B. and Lomax, H., "Thin Layer Approximation and
Algebraic Model for Separated Turbulenl Flows," AIAA Paper 78257,1978.
7 Hung, C.M.
and Chausee. D. S., "Comput3tion of Supersonic
Turbulent Flows Over an Inclined Ogi\'~Cylinder-Flare,"
AIAA

par,:,r 80-1410. 1980.

Deiwert, G. S. "Numerical Simulation of Three-DimensionaI
Boattail Afterbody Flow Fie1d," AIAA Paper 80-1347, 1980.
9Kutler, P., Chakravarthy, S. R., and Lombard. C. K., "Supersonic Flow Over Ablated Nosetips Using an Unsteady, Implicit
Numerical Procedure," AIAA Paper 78-213, 1978.
IOSteger, l., Pulliam, T., and Chima, R., "An Implicit Finite
Difference Code for Inviscid and Viscous Cascade Flows," AlAA
par.;r 80-1427,1980.
I Hung, C. M. 1980-198/ AFOSR-HTTMSTANFORD
Conlerence on Comp/ex Turbulenr F/ows. Vol. 111.edited by S. l. Kline et
ai., 1982, Thermoscienccs Div., Mechanical Engineering Depanment,
Stanford Univ., pp. 1283-84 and 1372-74.
12 Degani, D. and Schiff, L. B., "Computation
of Supersonic
Viscous Flows Around Pointed Bodies 8t Large Incidence," AIAA
PaRer 83-0034, 1983.
3Korkegi, R. H., "Survey of Viscous Interactions Associated with
High Mach Number Flight," AIAA Joumal, Vol. 9, May 1971,
pp.771-784.
14Settles, G., Gilbert, R., and Bogdonoff, A., "Data Compilation
for Shock \\'aveITurbulent Boundary Layer Intcraction Experiments
on Two-DimensionaI Compression Comers," Dep!. of Aerospacc and
Mechanical Engineering, Princcton University, Princeton, N. l.,
Re~t. MAE-1489, 1980.
~Settles, G., Fitzpatrick, T., and Bogdonoff, S., "Dctailed Study
of Attached and Separated Compression Comer Flo\\'fields in High
Reynolds Number Supersonic Flow," AIAA JOllrnal, Vol. 17, 1979,
pp. 579-585.
16Settles. G., Bogdonoff, S., and Vaso I.E., "Incipient Separalion
of a Supersonic Turbulcnt Boundary L1ycr aI High Reynolds
Numbers," AIAA JOllrnal, Vol. 14, Jan. 1976, pp. 50-56.
17 Rubesin,
M. and Rose. W., "The Turbulcnt /llean-Flow,
Reynolds-Stress and Heatl-lux 'Equations in Mass A\'eraged
Dependent Variables," NAS.-\ TMX-6:!:!48. M:udI1973.

928

M. VISBAL

18Pulliam, T. and Steger, J., "Implicit

Finite-Difference
Simulations of Three-Dimensional Compressible Flow," AJAA
Journal, Vol. 18, Feb. 1980, pp. 159-167.
19Shang, J. and Hankey, W. L., Jr., "Numerical Solution for
Supersonic Turbulent Flow Over a Compression Ramp," AJAA
JOllrnal, Vol. 13, Oc\. 1975, pp. 1368-1374.
2oVisbal, M. and Knight, D. D., "Generation of Orthogonal and
Nearly Orthogonal Coordinates with Grid Control Near Boundaries,"
AJAA Journal, Vol. 20, March 1982, pp. 305-306.
21Visbal, M., "Numerical Simulation
of Shock/Turbulent
Boundary Layer Interactions Over 2-D Compression Corners," Ph.D.
Thesis, Dep\. of Mechanical and Aerospace Engineering, Rutgers
Uni"ersity, New Jersey, Oc\. 1983.
22Sun, c.-c. and Childs, M. E., "Wall-Wake Velocity Profile for
Compressible Nonadiabatic Flows," AIAA JOllrnal, Vol. 14, June
1976, pp. 820-822.
23Knight, D. D., "A Hybrid Explicit-Implicit Numerical Algorithm
for the Three-Dimensional Compressible Navier-Stokes Equations,"
AIAA Paper 83-0223,1983
24Knight, D. D., "Calcul~lion of High Speed Inlet Flows Using the
Na~;er-Stokes Equations," AFFDL- TR-79-3130, Vol. I, 1980.
2..' Hopkins,
E. and Inouye, M., "An Evaluation of Theories for
Predicting Turbulent Skin Friction and Heat Transfer on Flat Plates
aI Supersonic and Hypersonic Mach Numbers," AIAA Journal, Vol.
9, June 1971, pp. 993-1003.

AIAA JOURNAL

AND D. KNIGHT

26Steger, J. L., "Implicil Finite-Diffcrcnce

About Arbitrary Geometries with Application

Simulation of Flow
to Airfoils," AIAA

PaRer77-665,1977.
7Thomas, P. O., "Boundary
Conditions and Conservative
Smoothing Operators for Implicit Numerical Solutions to the
Compressible Navier-Stokes Equations," LMSC-D630729, 1978.
28Horstman, c., Hung, c., Settles, G., Vas, 1., and Bogdonoff, S.,
"Reynolds Number Effects on Shock- Wavc Turbulent BoundaryLayer Intcraction-A
Comparison of Numerical and Experimental
Results," A1AA Paper77-42, 1977.
29Delery, J. M., "Experimental
In"estigation of Turbulcnce
Properties in Transonic Shock/Boundary-Layer Interactions," AJAA
Journal, Vol. 21, Feb. 1983, pp. 180-185.
30Settles, G. S., Baca, B. K., Williams, O. R., and Bogdonoff, S.
M., "A Study of Rcattachment of a Free Shc.1r Layer in Compressible
Turbulent Flow," AIAA Paper 80-1408,1980.
31Rose, W. c., "The Behavior of a Compressiblc Turbulcnl
Boundary Layer in a Shock Wave-Induced Adverse Pressure
Gradient," NASA TN D-7092, 1973.
32Hayakawa, K., Smits, A. J., and Bogdonoff, S. M., "Turbulence
Measurements in a Compressible Reattaching Shear Layer," AIAA
Pater 83-0?99, 1983..
"
,.
Hayakawa, K., Smlts, A.J., and Bogdonoff, S. M., Hot-\\Ire
Investigation of an Unseparated Shock- Wave/Turbulent Boundary
Layer Interaction," AIAA Paper 82-0985,1982.

Fro/n the AIAA Progress in Astronautics and Aeronautics Series

THERMOPHYSICS OF ATMOSPHERIC ENTRY-v.

82

Edited by T.E. Horton, The University of Mississippi

Thermophysics
denotes a blend of the classical sciences of heat transfer, Ouid mechanics, materiais, and clectromagnctic
theory with the microphysical
sciences of solid state, physical optics, and atomic and molecular dynamics. Ali of these
sciences are involved and interconnected
in the problem of entry into a planetary atmosphcre at spacenight speeds. At such
high speeds, the adjacent atmospheric
gas is not only compressed and heated to very high tcmpcratures,
but stongly
reactive, highly radiative, and electronically
conductive as well. At the same time, as a consequence of the intense surface
heating, the temperature of the material of the entry vehicle is raised to a degree such that material ablation and chcmical
reaction become prominent.
This volume deals with ali of these processes, as they are viewed by the research and
engineering community today, not only at the detailed physical and chemicallevel,
but also at the systcm cngineering and
design levei, for spacecraft intended for entry into the atmosphcre of the earth and thosc of othcr plancts. The twenty-(wo
papers in this volume represent some of the most important recent advances in this field, contributcd
by highly qualified
rescarch scientists and engineers with intimate knowlege of current problems ..

544 pp., 6 x 9, i/lIlS., $30.00 Alem., $45.00 List

TO ORDER WRITE: Publications Order Oep!., AIAA, 1633 Broadway, New \'ork, N. \'. 10019