https://ntrs.nasa.gov/search.jsp?

R=20150009530 2018-09-27T04:44:11+00:00Z

Status of Turbulence Modeling for High-

Speed Propulsion Flow Problems
N.J. Georgiadis
NASA Glenn Research Center
Cleveland, OH 44135 USA
Georgiadis@nasa.gov

W.A. Engblom, Embry-Riddle Aeronautical University

R.A. Baurle, NASA Langley
J.R. Edwards, N.C. State Univ.
A.Uzun, Florida State University
D.A. Yoder. M.A. Vyas, & J.R. DeBonis, NASA Glenn

The first author?s work was Sponsored by the NASA Fundamental Aeronautics Program and the DoD Test
Resource Management Center? s (TRMC) Test and Evaluation /Science and Technology (T&E/S& T)
Program through the High Speed Systems Test (HSST) area.
 Introduction
• High speed aircraft are highly desirable for military and
commercial applications.
• "Hypersonic" speed is frequently defined as Mach No. > 5
SR71 (Mach 3+):
X15 (Mach 6.7):

2
 Introduction

• Hypersonic air-breathing propulsion is one of the most difficult

R&D challenges facing the worldwide aeronautics community.

X51 Waverider:

• 1 out of 3 flights resulted in sustained combustion.

• Hydrocarbon fueled.

3
 Introducti on

• Current R&D uses ground tests and computationa l fluid

dynamics (CFD) to prepare for flight tests.

X43:

• Flight 2 achieved Mach 7; Flight 3 achieved - Mach 10.

• Hydrogen fueled.

4
 Introduction
• High Speed I Hypersonic R&D uses flight tests, ground tests,
and CFD ----all have difficulties:

1. Flight tests: Expensive, difficult to instrument, very harsh

environments for experimental aircraft that naturally lead to high
failure rates.
2. Ground tests: Difficult to achieve actual flight conditions, facility
interference effects, difficult to instrument.
3. CFD: Physical modeling is highly unproven - at best. Modeling
of turbulence, chemistry, heat transfer, fluid-thermal-structural
interactions is subject to high uncertainty.

5
 Introducti on

• Turbulence models remains one of the key pacing technologies

in Computationa l Fluid Dynamics (CFD).

• An overview of key turbulence modeling areas for propulsion

flows is presented.
• Emphasis is placed on "practical" state-of-the-a rt today:
- Standard practices using primarily RANS.
- Promising new technology (i.e. LES, hybrid RANS/LES) that may be available for
production use in near future.
- Key shortfalls for which R&D is necessary.
• Focus is placed on high-speed propulsion systems (i.e.
scramjets); turbine engines are also addressed in less detail.

6
 Key Turbulent Features of
Scramjet Flowpaths

FUEL INJECTION

Laminar-to-turbulent
transition, Shock wave I COMBUSTOR AND
turbulent boundary layer EXPANSION SYSTEM:
interactions 30 compressible mixing,
reacting flow, turbulent I
chemistry interactions
ISOLATOR: Shock wave/
turbulent boundary layer
interactions

7
 Key Turbulent Features of
Turbine Engine Flowpaths

INLET: NOZZLE/MIXER
Transition, '
COMPRESSOR: TURBINE: PLUME
Separation 30 Turbulent Mixing,
Swirling 30 flow, COMBUSTOR: Transition, 30, very
wakes, shock- Compressibil ity,
30 reacting flow high heat transfer
interactions ' ' Acoustics
turbulent I chemistry film cooling
interactions, multi-phase

8
 Presentation Outline

• Overview of Turbulence Modeling in Use for Propulsion Flows

- RANS
- LES and DNS
• Boundary Layer Transition - Inlets and Turbines
• 3D Boundary Layer Effects
• Shock-Wave /Turbulent Boundary Layer Interactions
• Combustor I Reacting Flows
- Scalar Transport
- Turbulent I Chemistry Interactions
• Exhaust System Modeling
- Jet and Mixing - RANS
- LES-based Methods
• Experimental Validation Data Needs
• Conclusions

9
 Navier-Stokes Equations

op o
--:;-
ut
+ u.A (pui) = 0
;l,..
Continuity
l

C(pui) + 0 (pu .u .) = - OP + Orif Momentum

Ot & . }
1
Ox . Ox.
}
1 }

oE,
___;_+ o ( u ( E, + P)) -_ o ( U/CiJ ) - 0q1 Energy
a o.x .
}
1
o.x .
}
o.x.
}

10
.
..

Navier-Sto kes for Perfect Gas

P=pRT Ideal Gas

2 Ou1.
r ..
lj
= 2nS
r·
.. - - 1
lj
1 5..
3r a. lJ Newtonian Fluid
1

Rate of Strain

or
q.=-k- Fourier' s law
1 ar.1

11
Reynolds Averaging

u. == -1
l
J I
I+T
u.dt
l
Traditional
Reynolds-averaging
r

A 1
u. ==-
f f+T
pu.dt
Density weighting
1
pr --
I
1
(Favre averaging}

A pf
f ==-=-
p
(Favre averaging}

12
 Modeled Turbulent Terms - RANS

Reynolds Stress

Turbulent heat flux

These terms replace ALL turbulent effects in the momentum and energy
equations.

13
 Spatial Filtering

Filtering operation

Filter function

- pf
f ==-=-
p
(Favre filtering)

Note that the Favre spatial filtering here and Favre Reynolds averaging
look similar but refer to two very different operations.

14
 Modeled Turbulent Terms - LES

sg.!>
-p
- ( f"""t,.t
U .U . -
I j
,..., ,...,
U .U . - T ..
I } lj
) -
Subgrid Stress

Subgrid heat flux

These terms replace only turbulent effects that are smaller than the
numerical scheme and grid (hence called subgrid} can resolve.

15
 ,,_\) . I I£( ~~' , tv4Lo4 !!;>..

Comparison of RANS and LES Equations ~~ -~

1 - Continuity C(-_,_, DT .C,,--- ~~~Na. ,.,...,.o<><

RANS

LES

16
 ."'\, i'·•..'.\) f:\' 1lt:.,/ ! '\~1 •IV6lV4t~

Com paris on of RANS and LES Equa tions ~; ~

2 - Momentum r-':1"UnC'~ ~.nr:o..o-<f;<

RANS

LES

17
 Comparison of RANS and LES Equations
3- Energy

"
oE, o ,. . " - o ,. . _ ,. . -r o _ r)
-Ot+ &. )
(u .(E, + P)) ==
&.
(u.r ..
1)
+ u.r ..
1)
1
)- (q.
&. )
1
+ q
)
. RANS
) ) )

,...,

- sgs)
oE, + 8 (.....,u. (E,.., + P)) -_ o (-U.T_ .. +U.T .. - o (-
q . +q sgs )
. LES
Ot &. ) 1 & . 11) 11) &. } }
) } }

18
 Major Differences Between Running
RANS and LES
All turbulent stresses are replaced by
the averaged effect {numerically
diffusive)
Constant CFL- goal is to get to
convergence as fast as possible.
Grids are packed to regions of high
mean shear {stretch ing OK)
Numerical scheme designed for
reasonable accuracy, shock wave
capturing, convergence
characteristics.
Dominant turbulent stresses are
calculated {numerically resolve
unsteady behavior)- can't have too
much diffusion or turbu lence goes
away.
Must run time accurately to capture
time-varying turbulence.
Best grids are uniform, isotropic-
need to be of size to capture sca les
of interest.
Numerical scheme driven by need
for high order of accuracy for
resolving unsteady behavior.
19
 RANS Turbulenc e Modeling

• Reynolds-Ave raged Navier-Stokes (RANS)- ·replaces all

unsteady turbulent motion with modeled turbulent stresses.
• Practical State of the art is two-equation models: k-E , k-co ,k-t;.
Menter Shear-Stress Transport (SST) is popular "hybrid model"
combining k-E and k-co.
• For subsonic/tran sonic external aerodynamics , one equation
models such as Spalart-AIIma ras are popular- not used as
much in propulsion flows.
• Full Reynolds-Stre ss Models -offer more complete
representation of 3-D turbulent stress field, but have not lived up
to promise in terms of improved predictions.
• Explicit algebraic stress models (EASMs) solve 2-eqn models,
but used additional relations to obtain "Reynolds-str ess-like"
behavior.

20
 Commonly Used RANS Models

1. Zero-equation (Algebraic): Cebeci-Smith, Baldwin-lomax

2. "Half-equation": Johnson-King
3. One-equation: Baldwin-Barth, Spalart-AIImaras
4. Two-equation:
a) k-E : Jones-launder (standard), Chien, many others
b) k-ro: Wilcox, Menter (BSl and SST)
5. Explicit algebraic stress models (EASM)
6. Reynolds-Stress closures

• The first 4 categories are "eddy viscosity models" where:

r!.7J = -pu~u'. = ,,r 2S.. - 2

Ouk o.
7 1 r 7J 3& 7J
k

21
Dv
-==C
--V+ -1 o
8 l l(
V+ U- ) ov
Oxk
+cb-2 ov ov -C J~'
Oxk Oxk wl 1
(uJd-
2

Df bl Oxk (J (J w

Vr == ufvl d == wall distance

x3
fvl - --
= =3 v"'
x· + cvl x==-
v ,. , v-
6 S =S + 2 ? fv2
+ -1-
J v2 -
X,A' g== r + cw2 (r - r) 1( d-
1 + /U v l
6 116
s = ~2QHQ H lj lj

f w= { l+cw3J
g6c6
3
w cbl =0.1355,cb2 =0.622,cvl = 7.1,a=2 / 3,
- cbl
Cwl_--2 + (l+cb2 ) ,Cw2 -0
-
3 -2 -0
. ,Cw3- ,K- .41
K CY

22
...
,._..:.\) l l:·tLr · ~•"""'-"4rli>v

Jones-Launder Two-Equation k-c RANS Model~ ~

D(pE) E Oui E2 {}
Dt = C&! k rif dt . -Cs2Pk+ dt.
} }
k2
llt = CJip-
E

CJI = 0.09,o-k = l,o-8 = 1.3,C8 1 = 1.55,C82 = 2.0

• Jones-launder form is referred to as the "standard" k-s model.
• Works well for attached boundary layers. Underestimates size of flow
separation (early reattachments).
• Works reasonably well for mixing layers, jet flows. Compressibility
corrections sometimes used for mixing problems at high convective
Mach numbers. 23
 Menter Two-Equation k-co "Shear Stress
Transport" (SST) RANS Model (1 of 2)

D(pk)
- - == T ..
au.
1
fJ*pOJk + o( + 0" ) ac
Dt lj a 1. -
a 1. r
II II
krt a 1.
D(pm) = !!__ Tij Oui _a 2
~pOJ +
o ll( Jl+ (jcuflt ) om + (1- ~)2p(} 1 Ok om
w2 - - - - -
Dt vr a.1 a 1. a. 1
m a . a.
1 1

11 == min pk · a1pkJ
fA't ' fl D
()) ~ l.r2

~* =tanh(arg:)
8
F4 = exp(-(R.v /120) )

~ = 1nax(F;* ,F4 )
500vJ 4pa- 2 k l
112

arg1
.I {
= n11nl rna
k
* ; ; UJ 2 J
2
L j3 coy cvy CDkUJY
24
...

Menter Two-Equation k-ro "Shear Stress

Transport" {SST) RANS Model {2 of 2)

• SST term comes from eddy viscosity expression. The hybrid k-ro,
k-E model without this term is referred to as the "Baseline" or
BSL model.
• Inner model reduces to original Wilcox k-ro formulation. Outer
model comes from transformed k-E model and is supposed to
reduce to the "standard" k-E model but differs in cross diffusion
term and diffusion coefficients.
• Model works quite well for attached -boundary layers, mild
separations, mixing flows {including jets). Numerically stable.
• All k-ro models may have an issue with artificial turbulence
production behind strong normal shock.

25
 Direct Calculation Methods

• Direct Numerical Simulation (DNS) - calculate all turbulent

scales down to the Kolmogorov scale - impractical for
engineering flows.
• Large-Eddy Simulation (LES) - directly calculate largest scales
and reserve modeling for smallest "subgrid-scale" stresses-
active research showing promise in combustor and jet plume
.
reg1ons.
• Hybrid RANS/LES - has become popular in recent years - most
effective use has been for flows where RANS can be used in
attached boundary layers and LES away from walls.
- Demarcated or zonal hybrid RANS/LES - clear distinction is made between RANS
and LES regions. Some physical mechanism is responsible for transition to
turbulence. This was intent behind design of Detached Eddy Simulation (DES).
- Continuous modeling - RANS and LES regions are not clearly separated -
solution is expected to adjust, based on resolution. Desirable in theory, but difficult
to achieve due to competing natures of RANS and LES.

26
 Smago rinsky SGS Model

r..usgs = 2(Cs Ll. pn

A)2- l/2(S- );: ..) --2 1LlA2-pnu· ~ ..
..lj --31 s-kku c
. lj 3 ~/

JT = slj..slJ.
~ = (LU~y~zY 13
• A few possibilitie s
~ = max(LU,~y,~z)
for the subgrid
l/2
turbulent length
Ll = (LU )2 + (Lly )2 + (Llz)2
scale:
3
• Note similarity of functional form to mixing-len gth RANS model
(i.e. Cebeci-Smith); gradient-d iffusion formulatio n; eddy
viscosity that adds to laminar viscosity just as is done in RANS.
• The effect on N-S equations, however, is very different- only
replacing subgrid turbulent stresses.

27
 Transiti on Modelin g

• Several RANS-base d models tried over the past several years - some solving
additional transport equations for intermittenc y, Re 9 •
• Some success for flows with high freestream turbulence intensity- i.e. turbine
cascades where bypass transition is dominant mechanism .
• Modal growth situations not easily represented by RANS-base d techniques.
• Work shown here is with a model based on the Menter SST k-ro turbulence model,
with transition modification s by Langtry, Sjolander, & Menter.
• Our work with the baseline published model indicated difficulties: (1) inability to
reproduce experimenta lly observed transition, (2) significant grid sensitivity, (3)
inability to become fully turbulent beyond transition. New formulation described in
Denissen, Yoder, Georgiadis, NASA TM 2008-215451 .

ak)
. opk apU·k
TKE equat1on a : a < = PT M · Pk - .8 p:»k
t X
j(

..!.. aaX . ( (Jl + CJkJlt) a X~

1 1 1

PTlvf = 1- 0.94(PTA11 + PTi\12) Fa tanh ( (y·' /17) 2 )

II ):! 1
F:5 = e·~ ( f (1- P(Rt)) + 2.P(Rt)
Modified model 2.5 - R t - .1 2
P(Rt) = -e 2
formulation : $
,. _ C { '(3.28E - 4)Rev - (3.94E - 7)Re~..!.. (1.43E - 10)&;]; Rev < 1000
• 'J. 1 -
PT' 1 - P'l' •\ fJ
· [0.12 + (1E - 5)Rev]; R ev > 1000

. 28
 Boundary Layer Transition Model
Incompressible Validation
Incompressible Validation:
• Transition locations and skin friction examined for T3A
benchmark data (ERCOFTAC)
• Several freestream intensities investigated.
• Grid sensitivity is high for incompressible cases.

·3
Ct for FSTI =2o/o Ct Variation with FSTI
X 10
7
\ - ssTModel
"' T3A Exp. - - · Mayle 91
6 I

' ·'·.... ...... - ·-· Correlations

1000
\ I
\~
- - AG-S 82
Error in Correlations
5 '~ 0 1xPTM1
800
L Ji\, \ 0
2xPTM1
4
I \
(.)

3
"' ' .
' .._
2
....... . 400 ' ... ....._ () .....

-·-·-·-·-·-·-· ..... , '

. . -~ c: -::,_.,~ ~--·--· ~~ ·=----....~~:- =:_~=- -- c:

. . . .....
·- ·-·-·- ·-·- ·-·-·- 200 0

2 3 4 5 6 7
Rex 5 0.5 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6
x10 FSTI

29
 Boundary Layer Transition Model
Hypersonic Validation
Hypersonic Validation:
• Mach 7.93, 7 degree straight cone investigated in AEDC Tunnel B, Tw I T 0 = 0.42.
• Heat transfer measurements by Kimmel, JFE 1997.
Integrated heat transfer: Transition-SST (6. 7°/o error), Fully turbulent SST (18 .5 °/o

l
•
error).
0.01
... Re/ m = 3.3E6
• Re/ m = 3.9E6
• Re/m = 6.6E6
=
+

- - -
Re/m 8.2E6
SST- fully turbulent
SST- transition, C PTMI = 1.0
I
- -- ---- SST - transition, C PTMI =2.0
' - - - -- --- __j

-
en .' ' --~- - ~--• •·...,..,.
0.001 r '
:.
~/
•
it•
...
~~ ot..t : •• I
~·'"'•••
... '*'- ~... • • ...' 1-~il I
.. ~-!t ......~· ••
....... -~+ I
+ ...... +
.......
.......

1E+06 1E+07

30
 Transition Modeling Conclusions

• RANS-based models only applicable for bypass transition

situations.
• Free-flight transition is normally modal growth -a reliable RANS-
based method is not likely promising.
• LES is not promising either because accurately capturing the
small disturbances is crucial -which LES will model/smear.
• Long Term Prospects- DNS, eN methods.

31
 Shock-Wave Turbulent Boundary
Layer Interactions (SWTBLis)
• Pervasive to the entire hypersonic propulsion flowpath.
• Major challenge to RANS, LES and hybrid RANS-LES techniques.
• Nominally 20 problems are inherently 30.

--~> # $"

32
•

UFAST- Mach 2.25 Test Case

• 2010 AIAA Workshop: UFAST and U. of Michigan cases, targeted
at representing supersonic aircraft inlets.
• Several organizations submitted results - RANS, LES, hybrids
--- -·----- . --------- -. -- - - ---
z

L,

I I I I I - I I I I I I I I I I I I I I I I I
200 250 300 350 400 450 500
X (mm)

33
 U Velocity Contours
Experiment:

300 320 x (mm) 340 360

SST: BSL:
20 1- 20 r- I
~-. . .~~~~:.;;;-~50;;5=
0 ~1~
5 fl_25_0 _35_0_450 5_ 5_0
e
.§.10
e • --~--~--~
.§.10 ~------------~
> >

.-~~-

300 320 x (mm) 340 360 300 320 x (mm) 340

SA: k-mASM:
20 I

• _ U LmJ~: -50 50 150 250 350 450 550 I

e
.§.1 0
>

0 28~0~----~3~
00~---·~3~2~
0~x-(_
m_m_)~34~0~----~
360 300 320 x (mm) 340 360

34
 UFAST Velocity Contours
Exploring minor change to Menter SST model's stress limiter.
Experiment Menter SST k-OJ, a 1 = 0.34
20

E' E'
.§.10 §.10
>- >-

300 360

Menter SST k-m Menter SST k-OJ, a 1 = 0.355

20 20

E' E'
§.10 .§.10
>- >-

0 28~
0--~--~
30~0~~--~
32~0--x-(m
_m_)~34~0------~
360
360

Menter BSL k-OJ Menter SST k-OJ, a 1 =0.37

20 20

E'
I U (m/s): -50 50 150 250 350 450 550 I E'
I U (m/s): -50 50 150 250 350 450 550 I
§.10 §.10
>- >-
 Mach 5 SWTBLI

'\ •0 Schulein, et al. (1996)

Schulein (2004) [optical]
I \
SST

r\ Wilcox k-oJ (1988 model)

Chien k-E
ASM k-E
0.008 l \, ASM k-w

<.) I . 0 0
c
•
-
.2 0.006
0
'i:
u.
0.004
I 0 0
•
() c
•
0

0.002

0.000

-0.002 L...-....I....---L.-~_,j_--L-..l-..__J_____J._.J___l...___J._..J._..J____.J._..._..J____.J._...l.,___J__j
300 350 400 450 500
x (mm)

37
 SWTBLI Modeling Conclusions

• k-c: models are generally overly optimistic on boundary layer

health -smaller separations than expt.
• k-co models usually work better for mild adverse pressure
gradients, small separations, Menter SST predicts larger
separations than expt.
• One equation models (i.e. SA) provide similar accuracy to multi-
equation models.
• EASMs offer minimal improvement.
• Some success using LES at AIM Workshop, inflow conditions &
matching Re are significant challenges.
• Hybrid RANS-LES also being investigated - however, where is
the switch from RANS to LES done?

38
• Several interacting phenomena - kinetics, turbulence, heat
transfer, thermal-structural effects.
• Practical state-of-the-art: Arrhenius form for reaction rates, 2 eqn
turbulence model, constant Prt, Set. Specified wall temperatures
or heat fluxes.
• Most practical scramjet experiments: only centerline pressures

University of Virginia Supersonic

Combustion Facility (UVA SCF):
u
available; More data and/or unit problems are desirable.
Mach 2 nozzle / ramp injector

Side View

~isolator
• Mach 5 enthalpy, Mach 2 isolator
• overall pressure ratio- 4
( I ""~ combustionI
• H2 fueled, clean air and vitiated air. duct
• Documented heat transfer rates and
wall temperatures.
• NASA-sponsored experiments
- - - - - - - - --- -- --- -El- -1------------- Top View
focused on mode transition behavior.
• Continuing experiments through
National Center.
X IH: -55.60 -47.72 -5.85 9.97 57.77

39
 Turbulent transport in energy
and species equations
Turbulent heat flux:

Turbulent Prandtl number:

Turbulent species flux:

Turbulent Schmidt number: SeT == ~

D
The turbulent Prandtl and Schmidt number are frequently set
equal to 0.9. However, it is believed that realistic values can
be significantly different for many flows- particularly in
extreme environments such as scramjets.

40
 Set Sensitivity for UVA SCF

~= 0.26, Clean Air

x/H = -45 Beginning of
isolator
x/H = 0 Fuel exit/
ramp base
x/H = 57 Nozzle exit to Expt., Clean Air
- - - Wind-US, Pr1:.9, Sc1:.5
ambient -·--------· Wind-US, Pr1=.9, Sc1=.7
- - - Wind-US, Pr1=.9, Sc1=.9

..
'

.
..
..
- 2.5
~
a.
........ .
a. 2 ..

1.5

x/H

41
 Prt and Set Sensitivity for USAF

Fuel Injectors Cavity Flameholder ~

--~------------------~~:;~~~_c~~~~~~----~------------~~1
~ I
Facility Nozzle lsola1or Combustor Aft Nozzle

300.0
275.0
250.0
225.0

1250.0

...~ 1000.0

750.0
100.0
75.0 Pr1::0.89, Sc.=0.5
500.0 Pr1=0.89, Sc1=0-5
50.0 Pr1=1.8, Sct=O.S Prtm1.8, Sct•O.S
250.0 Prt=0.89, Sc1z0.25
25.0 Pr,~0.89, Sct=0.25

O.Q1~2-L.I...I...i::-L..L..J.....I.~l....L..L.~...L.U:-'-'-~.L...L...I.~l....L..L.~..I...L.l~~

An ~optimized" Prt and Set for one case do

not guarantee optimal performance for Figure: Courtesy of Robert A. Baurle, NASA LaRC
other ¢ 's, turb. models, kinetics, etc. 42
 Prt Sensitivity for USAF Scramjet

Pr,= 1.8
Figure: Courtesy of Robert A. Baurle, NASA LaRC
43
 Burrows-Kurkov "Unit" Test Case

• Mach 2.4 vitiated air I sonic hydrogen experiment (1973).

• Used extensively for investigations/validation of H2 -air CFD methods
(kinetics, variable Prt, Set, hybrid RANS-LES ... ), perhaps overused.
• Measurements of species concentrations and temperatures.
Test section intermediate
meas uremer:1t station,
Test section initial 18. 3 em Test section exit
measurement station, measurement station,
0 em 35.6 em

~~~mn~~~k4~~
Static pressure ports
.----18.18.3 em-- · ·· ~
t--------35. 6em--------+

44
 Set Effects on Ignition Point for
Burrow-Kur kov Test Case
Prt =0. 7 (constant) for all cases
10

Set= 0.5
-
-
E
()
5
>

Oo~~:5~~~1:o~~1~5~~~~o:=::2:5::::3:o~~~35
x (em)
T(K) I
10 2200 I

-E
()
1950
• 11oo I
1450 i
12oo I
-> 5
Set= 0.7 ~~~ I
450

0o~~~5~~:1o~~~15~~~~o~==*-25~~3~o~.--j
200
- J

x (em)

T(K) I
10 2200

-E
()
1950
1700
1450
1200

Set= 0.9 -> 5 950

700
450

0 o~~~5~~:10~~~15~~;~0~~*-25~==3~0~~3~5
200

x (em)

45
 Hybrid RANS/LES Calcu lations of UVA
=
Dual-M ode Scram jet, <I> 0.17

____ __ - ...
- ~- --
Tempera ture

Eddy
viscosit y

Figure: Courtesy of Jack Edwards, NCSU

46
 Hybrid RANS/LES Calculations of UVA
Dual-Mode Scramjet, <I> =0.17
CARS comparisons (temperature): (X/H =6, 12, 18)

X/H=6

X/H=12

X/H=18

ZM

RANS LES/RANS CARS LES/RANS

(interpolated)
Figure: Courtesy of Jack Edwards, NCSU
47
 Compressible Mixing

• Most recent free shear layer mixing research has been in support of jet
aeroacoustics research (subsonics and supersonics).
• Practical state-of-the-art for RANS is also two-equation modeling.
• Some research in variable Prt for hot jet cases.

• Most research support is towards LES-based methods.

• Key LES issues:
1. Inflow boundary treatment
2. Grid resolution/sensitivity
3. Farfield noise propagation techniques.

48
 Jets and Mixing - RANS

RANS Findings:
• RANS underpredict mixing for incompressible jets - initial shear layer is difficulty.
• Uncorrected RANS models overpredict mixing rate for supersonic jets and mixing layers.
• Effects of temperature and 30 jet effects are not modeled correctly.
• Compressibility corrections (i.e. Sarkar) are highly empirical and do not reproduce correct
fluid dynamic effects.

Mach 0.5 Jet Mach 2.0 Jet

1.2
0 Seln er Data
1.1 ' - - SST
- - SST-V, CFL3D - - - - - - SST w ith Sarkar
~,W~~~.~ ---,
0.8 t-----+----'~
- -
c
- - SST-V, WINO
exp
0 '-'
- - Chten k·r.
- ----- Chien k· • with Sarkar
0
oo \:, ' - - - - -- -- - - . . . . . J
0.9 0 ,,
0 ..
0 \
0.6 0.8 €1 \
o;
Ci =:J- . 0 "
::::>- 0 '~,
.........
....... 0.7 0 ':,
:l 0 '',
0.4 0.6 0
0
',',
' '
0 ' .. :.....
0.5 o<S> ', ',
O.o ', ',

0.2 0.4 CJ 0
0
'->-'''
0 '
0
0.3 0

00 5 10 15 20 0.2 L-L-.L...-~...L....J......J.......L.~....L..J.-L...:J......J...._L......i...-.L..-J....::b....!....=
x/Dlot

49
 Jet Mixing - LES

• Acoustic Reference Nozzle (ARN) and Simple Metal Chevron (SMC)

configurations- tested at GRC, investigated by several LES researchers.
• Two Mach 0.9 jet simulations considered here: (1) DeBonis (GRC) DRP with 4
stage RK, 3.5 - 9.2 million points and (2) Uzun (FSU), 4th order compact scheme
with 4 stage RK, 50- 400 million points.
DeBonis {GRC) grid:

50
 ,.;.\) [1'11(·

ARN - Centerline Statistics (GRC} ~

Axial Turbulent Intensity Radial Turbulent Intensity

Mean Axial Velocity

1.2

0.2 - - - Coarse Grid 0.2 - - - Coarse Grid

- - - Medium Grid - - - Medium Grid
Fine Grid - - FineGrid
D Experiment o Experiment

0.15 0.15

0.8
-;:)-
:::-
:J 0.1
0.6

--- Coarse Grid

0.4
- -- Medium Grid
- -- - Fine Grid
c Experiment

0·2 o~--'---'----'---=-
5 __.___._.......___,__.,..,o=--'"---'----'---'---,1:'=-5_,__,__._~2o·
15 20
x/Di

Figure: Courtesy of Jim DeBonis, NASA GRC

51
 Turbulence Intensity

Axial Turbulent Intensity

0 0.04 0.08- 0.12
- --
0.16 0.2

1.5
Q) ·- 1
c: 0
';:::
t;: 0.5

5 0
1.5
,....;
a. 0-
>< ';::: 0.5
Q)

0
5
x/D;

Radial Turbulent Intensity 0 0.02 0.04 0.06 0.08 0.1 0.12

1.5
Q) ·- 1
c: ~
t;: 0.5

5 10 15
1.5 ~ --· - - --
...,a. 0- 1I
>< ';:::
Q) 0.5 ' l
01 --'
0 5 10 15
Figure: Courtesy of Jim DeBonis, NASA GRC x/Di
OL
 1.2
1.1
1
0.9
·--, 0.8
~
.......... 0.7
~

·--
~
:... 0.6
~
,..,
(;) 0.5
~~
0.4
SMCOOO Round Jet Experbnent
0.3 Coarse Mesh (50M Points) - Turbulent Inflow
0.2 Fine Mesh (400M Points) - Turbulent inflow
0.1 Fine Mesh (400M Points)- Laminar inflow
00 5 10 15 20
x!D.J

Figure: Courtesy of Ali Uzun, FSU

53
Figure: Courtesy of Ali Uzun, FSU
54
 Combustor/Exhaust System
Modeling Enhancement Needs
• RANS:
- Better prediction of 30, compressible mixing; highly separated/recirculating
flow in flameholder/cavity, SWTBLis, turbulent-chemistry interactions.
- More accurate boundary conditions for thermal state.
- Variable Prt and Set capability.
• LES:
- Capability to handle wall bounded and free shear layer regions. Hybrid
RANS/LES methods are under investigation - but location of RANS-to-LES
switch has significant effect.
- Significant uncertainty remains in how to best perform jet/mixing
simulations. Highly desirable to establish "best practices" if possible.
- Models for turbulent/chemistry interactions, i.e. Filtered Density Functions
(FDFs).

55
 Exper imenta l -Valid ation Data Needs

• Centerl ine pressu re distribu tions are not sufficie nt for validati on I
calibra tion of turbu lent flow CFD. There are too many interac ting
feature s in scramj et flowpa ths- unlike subson ic/trans onic
aerody namics .
• More comple te turbule nt statisti cs for momen tum, therma l, and
species transpo rt are needed .
• -A dvance d Diagno stics-: CARS, PLIF, PIV- for unit problem s, then
more comple x cases.
• Quanti fy uncert ainty- e.g. PIV is powerf ul techniq ue, but prone to
high uncerta inty in crucial regions such as initial mixing regions .
• Consid er revisitin g experim ents such as Burrow s-Kurk ov with the
advanc ed techniq ues.
• Design experim ents to avoid contam ination of focus region - i.e.
SWBLI cases - nearly all experim ents are in small tunnels where
sidewa ll separa tions domina te region of interes t.

56
 Con clus ions

• Many extre mely diffic ult chall enge s rema in in turbu lence mode ling
for air-b reath ing prop ulsio n flows .
• Statu s of RANS Mode ling for high spee d prop ulsio n flowp aths:
Not much adva ncem ent in pract ical state -of-th e-art in 2 deca des.
• Dom inant featu res of 3-D flow, large sepa ratio ns, SWT BLis,
chem ically react ing flow, comp ressi bility , turbu lent trans port of
heat and spec ies - overw helm the capa bilitie s of curre nt RANS
meth ods.
• Twea king one turbu lence mode ling para mete r while holdi ng all
other s fixed until cente rline press ure distr ibutio n matc hes
expe rimen tal data (typic al pract ice for scram jets) is of minim al
value .
• LES and relate d meth ods are demo nstra ting some prom ise, but
have their own mode ling issue s and (1) are not of suffic ient
matu rity for most prob lems , (2) comp uting powe r is not read ily
avail able to use in a prod uctio n engin eerin g envir onme nt, (3)
minim al cons isten cy betw een grou ps in how to achie ve most
accu rate resul ts.

57