International Journal of Dynamics of Fluids

ISSN 0973-1784 Vol.2, No.1 (2006), pp. 55-71

Research India Publications
http://www.ripublication.com/ijdf.htm



Numerical Simulation of Shock-Free Shear Layer
Interaction in Reacting Flows


M. Deepu
1
, S. S. Gokhale
2
and S. Jayaraj
3

1
Lecturer in Mechanical Engineering
N S S College of Engineering, Palakkad, Kerala-678 008, India
Email: m.deepu@rediffmail.com
2
Professor in Aerospace Engineering
Indian Institute of Technology Madras-600 036, India
Email: ssg1@ae.iitm.ac.in
3
Professor in Mechanical Engineering
National Institute of Technology Calicut, Kerala-673 601, India
Email: sjayaraj@nitc.ac.in


Abstract

The performance of any supersonic combustor system depends on efficient
injection and complete burning. Computational analysis of the flow field
associated with supersonic combustors is presented. Results are obtained by
numerically solving unsteady, compressible, turbulent Navier-Stokes
equations, using Unstructured Finite Volume Method (UFVM) incorporating
RNG based - two equation model and time integration using three stage
Runge-Kutta method. The developed numerical procedure is based on the
implicit treatment of chemical source terms by preconditioning. Reaction is
modeled using an eight step H
2
-air chemistry. The code is validated against
standard experimental data. The analysis could demonstrate the effect of
interaction of oblique shock wave with hydrogen stream in its mixing with
coaxially flowing air and subsequent reaction.

Key words: SCRAMJET, Supersonic combustor, Reduced chemistry, Point
implicit method, FVM.


Introduction
The recent interest in single stage to orbit trans-atmospheric vehicle has lead to the
development of a hypersonic flight [1-5], which incorporates a supersonic combustor.
Supersonic Combustion Ramjet engine (SCRAMJET) benefits from the better
performance of air breathing propulsion system. Scramjets need a combustor that
56 M. Deepu, S.S. Gokhale and S. Jayaraj
should have efficient mixing and combustion of fuel with air at supersonic speeds
without much pressure loss [5, 6]. Many experimental and numerical analyses have
been reported during the last few decades regarding the characteristics of the complex
flow field resulting due to fuel air mixing and combustion. Many fuel injection and
flame holding techniques which can have efficient fuel oxidant mixing with accurate
burning rate have been developed but there is always a trade-off between mixing
effectiveness and pressure drop occurring inside the combustor affecting the total
propulsive thrust available at the nozzle.
Performance of a supersonic combustor system depends on efficient injection and
complete burning. Future hypersonic vehicles are expected to require the performance
and operability benefits from air breathing propulsion systems as it is providing high
specific impulse [3]. The use of supersonic combustors in such vehicles requires
efficient supersonic combustion in combustor lengths short enough to be compatible
with practical engine sizes.
Combustion modeling simulates the chemical and physical evolution of a complex
reactive flow system by numerically solving the governing time dependent
conservation equations of mass, momentum and energy. The reaction zone propagates
as a consequence of strong coupling between combustion chemistry and the
appropriate fluid mechanical process. The combustion model should accurately
predict the strong interaction between the energy released from chemical reactions
and the dynamics of fluid motion. The chemical reaction leads to the generation of
gradients in temperature, pressure and density, which in turn influence the transport of
mass, momentum and energy. Macroscopically these gradients may lead to the
development of vorticity or affect diffusion of mass and energy. On a microscopic
scale, these gradients develop turbulence and thereby influence mixing and flame
velocities. The interaction of fuel with oxidizer is very difficult to predict in such
systems due to high non linear dependence of various flow parameters and chemical
source terms. Therefore predicting strong interplay between chemistry and fluid
dynamics in reacting flows is the real challenge in combustion modeling. The use of
supersonic combustors in atmospheric vehicles requires efficient combustion within
combustor lengths which is short enough for practical engine sizes. Supersonic
reacting flow field can be simulated by adding finite rate chemistry to standard
compressible Navier Stokes equations. Both turbulence and chemical kinetics are
important, since the residence time inside the combustor is much smaller. Explicit
treatment of all conservation terms with reaction chemistry results in stiff equations
and it will degrade the performance of numerical method as flow field and chemical
kinetics with differing time scales need to be solved simultaneously. Existence of
several non-equilibrium states creates challenges in solution procedure. Governing
equations of turbulent shear layer flows involving finite rate chemistry are often
difficult to solve due to stiffness (Ratio of largest time scale to smallest). Stiffness
will degrade the performance of numerical methods. While handling two differing
time scales, both can be advanced equally in pseudo time. In other words, it can be
treated as way of rescaling the equation in time such that both phenomena evolve at
comparable pseudo time scales resulting in higher time steps for integration.
Comparison of the numerical result has been done with the standard axisymmetric
Numerical Simulation of Shock-Free Shear Layer 57
reacting free shear flow experimental measurements of Cheng et al. [7] The predicted
heat release and species production rates are found to have reasonable agreement with
the experimental results. The analysis has been extended to study the mixing and
reacting behavior of hydrogen injection issuing into hot supersonic air stream with
angled inlet.


Governing Equations of Compressible Reacting Flows
The conservation form of equations, which govern a two dimensional turbulent
compressible flow can be expressed in a generic form for axisymmetric flows as
1 ( ) U F rG
S
t x r r

+ +

(1)
Where
( )
2
, F=
xx
r xr
r
xx xr r eff
i
i
u
u
uv
u
T
v
E u v K
x
U E
u
x
u
Y
x
uY

1
1
+
1
1
1
+ 1
1
1

1
1
+ +
1
1

1
1

1
1
1
1
1
1

1
1
]

1
1
]

( )
( )
2
0
0
1
G= and
0
r
r rx
r xr
rr r rx eff
r
r
i
r i
v
uv
v
T
P
E v u K
r
r
S
v
H
r
H
v
r
v Y

1
1
1
+
1
1
+ 1
1
1
1

+
1
+ + 1
1
1

1
1
1
1

1
1
1
1

1
1

1
]
1
1
]

Here
2 2 1
2
3 3
r
xx eff
u u rv
P
x x r r

1 _
+ +

1

, ]

r
xr rx eff
u v
r x


1
+
1

]

58 M. Deepu, S.S. Gokhale and S. Jayaraj
2 2 1
2
3 3
r r
rr eff
v u rv
P
r x r r

1 _
+ +

1

, ]

2 1
2
3
r r
eff
v u rv
r x r r

1 _
+

1

, ]

In the above versions of formulations, the effective viscosity of the flow field
t l eff
+ (2)
From Sutherlands law laminar viscosity (Ns/m
2
) for different temperature (K) is
found as
l

=1.458 10
-6
4 . 110
5 . 1
+ T
T

and
t
is found from turbulence model
The effective thermal conductivity
eff l t
K K K + (3)
In which
Pr
Cp
k
l
l

t
t
t
Cp
k
Pr

For the present analysis modified - model called Renormalisation Group (RNG)
model was used. Yakhot et al. [8] proposed this model, which systematically removes
all the small scales of turbulence motion from the governing equation by expressing
their effects in terms of large scales and a modified viscosity. Following are the RNG
- model.
( )
.( ) [ ]
k eff
u div grad H
t

+ +

(4)

( )
( ) [ ]
k eff
div U div grad H
t

+ +

(5)
Here the turbulence source terms are defined as
2
*
1 2
2 and H 2 .
t ij t ij ij
H E C E E C





Turbulent viscosity is defined as
2
t
C

(6)
Closure coefficients are evaluated as
C=0.0845,

=

=1.39, C
1
=1.42 , C
2
=1.68
/
ij ij
(2E .E )


Numerical Simulation of Shock-Free Shear Layer 59
and C
1
*
=
3
0
1
) / 1 (
1

0=
4.377, =0.012
Value of constant is adjustable which is found from near wall turbulence data. All
other parameters are explicitly computed as part of RNG process. As noted in the
source term of the governing equations contains the chemical source terms
.
. These
terms represent the generation or destruction of species i due to chemical reaction. For
any chemical reaction with N
R
number of elementary reactions involving N
S
number
of species
1 1
.
' ''
1 1
' ''
( ) ( '' ' )
S S
S S
ji ji
j j
N N
ji i ji i
i i
N N
i ji ji f b j
i i
C C
C k C k C

1

1
]


(7)
Where i=1,2,3. N
S
represents species and j=1,2,3 N
R
represents reactions
Net change in concentration of any species can be found as
. .
1
R
N
i i
j
j
C C

(8)
And the net production of species is given by
i i i
W C
. .
(9)
i
W represents the molecular weight of individual chemical species.

Reaction rates in the above equation can be found from Arrhenius law
exp
j
j
j
N
j
f j j
f j
b
eq
E
k A T
RT
k
k
k

(10)
Here
j
eq
k is the equilibrium reaction constant.

An eight-step reaction mechanism proposed by Evans and Schexnayder [9] has been
used here for modeling hydrogen-air chemistry for which the reaction steps and
reaction rates are summarised in Table 1.






60 M. Deepu, S.S. Gokhale and S. Jayaraj
Table 1: Reaction steps and reaction rates of eight step mechanism.

Forward Rates Backward Rates
Reaction A N E A N E
H
2
+M=H+H+M 5.510
18
-1.0 51987.0 1.810
18
-1.0 0.0
O
2
+M=O+O+M 7.210
18
-1.0 59340.0 4.010
17
-1.0 0.0
H
2
O+M=H+OH+M 5.210
21
-1.5 59386.0 4.410
20
-1.5 0.0
OH+M=O+H+M 8.510
18
-1.0 50830.0 7.110
18
-1.0 0.0
H
2
O+O=OH+OH 5.810
13
0.0 9059.0 5.310
12
0.0 503.0
H
2
O+H=OH+H
2
8.410
13
0.0 10116.0 2.010
13
0.0 2600.0
O
2
+H=OH+O 2.210
14
0.0 8455.0 1.510
13
0.0 0.0
H
2
+O=OH+H 7.510
13
0.0 5586.0 3.010
13
0.0 4429.0
Here M is a third body. The reaction rates are expressed in Arrhenius law form.
Units of A are multiple of cm
3.
mole
-1.
s
-1.
E is in Joules

Evaluation of thermodynamic properties specific heat at constant pressure, enthalpy
and entropy are respectively found from standard thermodynamic data from Mc Bride
et al. [10] as.
2 3 4 i p
i i i i i
c
A BT CT DT ET
R
+ + + + (11)
2 3 4
2 3 4 5
i i i i i i
i
H B C D E F
A T T T T
RT T
+ + + + + (12)
2 3 4
ln
2 3 4
i i i i
i i i
S C D E
A T BT T T T G
R
+ + + + + (13)
For each species two sets of coefficients (A
i
-G
i
) are used for the temperature intervals,
one applicable from 300K up to 1000K and the other applicable from1000K up to
3000K. Total energy of flow field is given by
( )
2 2
1
2
1
v u
p
h Y E
i
N
i
i i
+ +

(14)
Temperature is worked out from the above equation using the Newton-Raphson
method from the modified energy equation. The pressure is calculated from the
resulting temperature as follows
1
i
N
i
i i
Y
p R T
W

(15)


Numerical Modeling
To analyse the SCRAMJET combustor flow field unstructured finite volume method
is employed here. The unstructured grid techniques are able to idealise and analyse
geometrically complex problems with more ease compared to conventional structured
Numerical Simulation of Shock-Free Shear Layer 61
grid method. The finite volume Runge-Kutta scheme suggested by Jameson et al. [11]
is used in the present analysis.
The conservation equation for a control volume can be written as
.
. 0
i
i i i
faces
dU
V F ds S V
dt
+

(16)
V
i
is the cell volume and ds is the area of elemental sides. This is a system of ordinary
differential equation. For obtaining the solution this has to be integrated with respect
to time.















For a control cell ABCD as shown in Fig. 4.1 above FVM formulation can be written
as
( )
ij ABCD ij
Ud fdy gdx Sd
t

(17)
Here f and g are the Cartesian components of flux vector. The area is evaluated from
the vector product of the diagonals as
1
2
ABCD AC BD
x x (18)
[ ]
1
( )( ) ( )( )
2
ABCD C A D B D B C A
x x y y x x y y (19)
The cell averages of the derivative of different flow variables for the surface
described by a quadrilateral ABCD can be evaluated as follows.
( )( ) ( )( )
1
2
A C B D B D A C
ABCD
Udy U U y y U U y y 1
]
(20)
Relation for the x derivative is
1 U U
d
x x

_



,

(21)
( )( ) ( )( )
( )( ) ( )( )
A C B D B D A C
A C B D B D A C
ABCD
U U U y y U U y y
x x x y y x x y y
_

,
(22)
A
B
C
D
Fig. 4.1. An arbitrary shaped plane quadrilateral
62 M. Deepu, S.S. Gokhale and S. Jayaraj
Relation for the r derivative is
1 U U
d
r r

_



,

(23)
( )( ) ( )( )
( )( ) ( )( )
A C B D B D A C
A C B D B D A C
ABCD
U x x U U x x U U
r x x y y x x y y
_

,
(24)

In order to avoid oscillations near shocks and other uncoupled errors, the numerical
solution procedure needs to incorporate artificial dissipation. For the present analysis
the numerical dissipation terms are formed with Laplacian and Biharmonic operators.
The dissipation operation can be written as
(2) (4)
( ) ( ) ( )
i i i
D U d U d U + (25)
where d
(2)
(U
i
) represents the contribution of the undivided Laplacian operator and
d
(4)
(U
i
) is the contribution of the Biharmonic operator. Biharmonic operator does the
background dissipation to damp high frequency uncoupled error modes.
Laplacian operator for the i
th
cell is given by

6
1
2
) (
k
i k i
U U U (26)
where, the summation in k is taken over all control volumes which have the common
interface with i
th
cell.
The Biharmonic operator is given by
) (
2
1
) (
2 2
6
1
) 4 ( ) 4 (
i k
k k
k
i
i
ik i
U U
t
V
t
V
U d

,
_

(27)
The Biharmonic operator is not operated near shock as this will produce pre and
post shock oscillations.
Here
( ) ) ( , 0 max
) 2 ( ) 4 ( ) 4 (
ik ik
k
With the addition of artificial dissipation terms the space discretisation will become
( )
( ) ( ) 0
i
i i i
d U
V R U D U
dt
+ (28)
Explicit treatment of all conservation terms with reaction chemistry results in stiff
equations and it will degrade the performance of numerical method. Flow field and
chemical kinetics with differing time scales need to be solved simultaneously.
Bussing and Murman [12] introduced the method of preconditioning the
conservation equations in conjunction with chemical source terms alone being treated
implicitly. Such a method has the advantage of both explicit and implicit methods.
While handling two differing time scales, both can be advanced equally in pseudo
time. In other words, it can be treated as way of rescaling the equation in time such
that both phenomena evolve at comparable pseudo time scales resulting in higher time
steps for integration.
Numerical Simulation of Shock-Free Shear Layer 63
Thus the modified equation is
1 ( )

U F rG
SJ S
t x r r

+ +

(29)

The point implicit formulation of the time stepping can be written as
( )
( )
( )
(0) ( )
0 (1) (0) (0) (0)
1
1 (2) (0) (1) (0)
2
3 (3) (0) (3) (0)
3
1 (3)
( )
( )
( )
n
i i
i i i i
i i i i
i i i i
n n
i i i
U U
ti
SJ U U R D
Vi
ti
SJ U U R D
Vi
ti
SJ U U R D
Vi
U U U

+
(30)
The point implicit scheme in which all the six chemical species terms are treated
implicitly and all other terms explicitly. The preconditioning Matrix SJ used for this
purpose is given by

2 2 2
2 2 2
2 2 2
2 2 2
2 2 2
2 2 2
. . .
. . .
. . .
1 . . . . .
1 . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . 1
H H H
i i i
H O H O
O O O
i i i
H O H O
H O H O H O
i i i
H O H
t t t
Y Y Y
t t t
SJ
Y Y Y
t t t
Y Y Y
1
1
1
1
1
1
1
1
1
1
1
1
1
]
O
(31)


Code Validation and Test Results
Coaxial Hydrogen-Air Supersonic Mixing and Reaction
Numerical simulations were performed on an axisymmetric configuration consisting
of coaxial jets (Hydrogen and vitiated air) exiting at supersonic velocities into
ambient air. Cheng et al. [7] made measurements of temperature and composition
using ultra-violet Raman scattering and laser induced predissociative fluorescence
techniques. These experimental results are used for validation. The flow conditions
have been summarized in Table 2.
The Details of computational domain is given in Fig. 1. The left face of the
computational domain is given with a supersonic inflow condition in the region of
supersonic air stream. The burner lips were assumed to non-catalytic. Hydrogen jet
with above mentioned conditions is introduced at respective position as separate
boundary condition. Remaining portion in left face is considered as no slip wall. Top
face is a no slip wall. A supersonic out flow condition is maintained at outlet. Since
64 M. Deepu, S.S. Gokhale and S. Jayaraj
the entire flow field is symmetrical about axis of the coaxial jet, a symmetry condition
is utilized there. The computational domain was discretised in to 104000 control
volumes. The initial condition for entire domain is ambient conditions. The grid was
added to allow specification of the entrainment boundary condition well upstream of
the coaxial jets. The grid is clustered to give more resolution near the lip regions in
the transverse direction, and near the burner exit in the streamwise direction. The
criteria chosen for steady state solution is the density error in successive steps falls
less than 1.010
-06
.

Table 2: Coaxial burner exit conditions.

H
2
Jet Air Jet

Ambient
Air
Diameter
(mm)
2.36 17.8 120
Mach
Number
1.0 2.0 0
Pressure
(MPa)
0.112 0.107 0.101
Temperat
ure (K)
545 1250 300
Mass
Fraction
2
1.0
H
Y

2
2
0.245,
0.175
O
H O
Y
Y

2
2
0.233,
0.01
O
H O
Y
Y

Figure 2 shows comparison between experimental data and numerical prediction of
temperature at exit which is 101 mm from exit. Reasonable match has been obtained
inside core region of the jet and a slight difference in the mixing layers.


Figure 1: Details of computational domain
Numerical Simulation of Shock-Free Shear Layer 65
The difference is due to initial turbulence level specification given at the inlet and also
due to the numerical dissipation effects. Figure 3 gives the quantitative comparison of
water mole fraction profiles at exit. The non-interference of fuel core to the outer
region can be noticed. The species production rates in the mixing region agree well
with experimental values. The water mole fraction values at core region could not
established. The prediction capability of the code developed is proven by this
comparison.


Figure 2: Temperature profile at exit.


Figure 3: Water mole fraction profile at exit.

The temperature plot for the entire domain is given in Fig. 4. The mixing becomes
more predominant in the region far from the jet outlets and heat release gradually
increases in the mixing layer between hydrogen and air. Non reacting fuel core
extends up to 30 mm from jet outlets. Outer layer between supersonic air stream and
66 M. Deepu, S.S. Gokhale and S. Jayaraj
ambient has almost same temperature through out the length, beyond this non reacting
core region.


Figure 4: Field view of temperature for the entire domain.

Though the typical eight step chemical kinetic mechanism employed here is
performing well when coupled with two dimensional turbulent compressible Navier-
Stokes equation; the introduction of seven species has created difficulty in
computational procedure. The total number of conservation variables increased to
thirteen including the chemical species which will pose a real challenge while
applying for scramjet combustors of larger geometry.

Shock-free shear layer interaction in reacting flows
In real scramjet combustor, the flow filed is extremely complex. Various mixing
enhancement studies [13,14] used strut based fuel injectors. The parallel injection has
the advantage of momentum addition with minimum pressure loss. But the mixing
efficiency of such a system is very low when compared to transverse injections. The
mixing behavior of the planar jet can be enhanced by the artificial production of
recirculation zone due to shock wave-expansion wave pattern. For the present analysis
an oblique shock is created in the inlet of the flow passage by a 8
0
wedge. The oblique
shock emanated from the inlet will interact with coaxially injected hydrogen stream.
The interaction of shock with shear layer formed between air and hydrogen streams
will enhance mixing and thereby promotes reaction. The details of computational
domain are given in Fig. 5. The flow conditions have been summarized in Table 3.


Figure 5: Details of computational domain.
Numerical Simulation of Shock-Free Shear Layer 67
Table 3: Flow conditions.


H
2
Jet Air Jet

Diameter (mm)
6 60
Mach Number
1.0 3.0
Pressure (MPa)
0.2 0.107
Temperature (K)
400 900
Mass Fraction
2
1.0
H
Y
2
0.245
O
Y


The computational domain is discretised in to 54,000 control volumes. Figure 6
gives Mach contour plot. Low velocity regions can be identified along the path of
progress of the hydrogen jet. Alternate compression and expansion takes place for the
jet and is not enough to perturb the flow field much in the region near to the jet
outlets. But the shock wave reflections interfere with the upcoming jet and localised
low velocity regions are produced. Though these regions are responsible for pressure
loss of the jet, certainly enhances mixing and reaction. The computed pressure
contours are depicted in Fig. 7. Due to chemical reaction enormous amount of heat is
released at mixing layer. The variation of temperature along radial direction for
various axial locations is plotted in Fig. 8. The mixing and reaction is promoted by
shock induced vortex generation. Figure 9 shows the variation of the water mass
fraction along radial direction for various axial locations.


Figure 6: Mach contour plot.


Figure 7: Computed pressure contours.
68 M. Deepu, S.S. Gokhale and S. Jayaraj

Figure 8: Variation of temperature for various axial locations.


Figure 9:Variation of water mass fraction for various axial locations.

The effect of mixing and subsequent reaction is more down stream areas. The shock
reflections are responsible for blocking the development of jets in the shear layer and
thereby creating the low velocity regions. Since the residence time associated with the
flow field in such areas is more, the reaction is more predominant in such areas
leading to enhanced heat release. The numerical scheme could demonstrate such a
complex flow field.
Numerical Simulation of Shock-Free Shear Layer 69
Conclusion
A point implicit finite volume solver has been developed and tested for the supersonic
mixing and reaction of parallel streams of hydrogen and air. The implicit treatment of
the chemical species in the source term reduces the stiffness considerably and higher
CFL almost equal to that of non-reacting situations has been obtained. Reasonable
match has been obtained with experiment for the temperature inside core region of the
jet and a slight difference in the mixing layers. Though the species production in the
core of the jet is not predicted well, but it could almost accurately follow that in the
mixing layer.
Future works aims at the influence of initial level of turbulence in controlling
reaction. Though the typical eight step chemical kinetic mechanism employed here is
performing well when used along with two dimensional turbulent compressible
Navier-Stokes equation; the introduction of seven species has created difficulty in
computational procedure. The total number of conservation variables has risen to
thirteen including the chemical species. It will set a real challenge while applying for
scramjet combustors of larger three dimensional geometry with more number of grids.


Acknowledgments
The first author thanks T. Jayachandran, Head of Fluid Mechanics and Thermal
Analysis Division of Propulsion group, Vikram Sarabhai Space Centre,
Thiruvananthapuram, India, and his team members for their valuable suggestions for
this work.

Nomenclature
c Sound velocity
C
i
Concentration of a species
C
p
Specific heat at constant pressure
C
v
Specific heat at constant volume
D Artificial dissipation
E Total energy
E
ij
Deformation tensor
E
j
Threshold energy of a reaction F&G Flux vectors
G Gibbs free energy (kJ/kg)
H Enthalpy
K Thermal conductivity
k Reaction rate
p Pressure
Q Heat flux
r Radial direction
S

Source term
S
i
Entropy
t Time
T Temperature
U Vector of conservative variables
70 M. Deepu, S.S. Gokhale and S. Jayaraj
V Volume of the control volume
u Favre average velocity in axial direction
v
r
Favre average velocity in radial direction
Y
i
Mass fraction of species
Adjustable closure coefficient

Kinetic energy dissipation rate
Molecularity of a reaction
Kinetic energy of turbulence
Viscosity
Kinematic viscosity
Density
Shear stress
Surface area
Cell volume
Rate of production of species

Subscripts
f
Forward reaction
b
Backward reaction
i
Chemical species
j
Reaction step
l
Laminar
r
Radial direction
t
Turbulent

Superscripts

Product

Reactant


References

[1] Seiner, J, M., Dash, S. M., and Kenzakowski, D. C., (1999), Historical
Survey on Enhanced Mixing in Scramjet Engines, AIAA Paper 99- 4869.
[2] Tishkoff, J. M., Drummond, J. P., Edwards, T. and Nejad, A. S., (1997),
Future direction of supersonic combustion research: Air Force/NASA
Workshop on Supersonic Combustion, AIAA paper 97-1017.
[3] Voland, R.T., Rock, K.E., Huebner, L.D., Witte, D.W., Fischer, K.E., and
McClinton, C. R. (1998), Hyper-X Engine Design and Ground Test
Program, AIAA Paper 98-1532.
[4] Mercier, R. A. and Ronald, T. M. P., (1996), U S Air force hypersonic
technology Program, AIAA Paper, AIAA 7th International Space planes and
Hypersonics Systems and Technology Conference, Plenary Session.
Numerical Simulation of Shock-Free Shear Layer 71
[5] Gruber, M. R., and Nejad, A. S., (1995) New Supersonic Combustion
Research Facility, Journal of Propulsion and Power, 11(5), pp. 1080-1083.
[6] Tomioka, S., Murakami, A., Kudo, K., and Mitani, T., (1998), Combustion
Tests of a Staged Supersonic Combustor with a Strut, AIAA paper 98-3273.
[7] Cheng, T. S., Wehrmeyer, J. A., Pitz, R. W., Jarret, O., and Northam, J. B.,
(1991), Finite Rate Chemistry Effects in Mach 2 Reacting Flow, AIAA
paper 91-2320.
[8] Yakhot, V., Orzag, A., Gatski, T. B., and Speziade, C. B., (1992),
Development of Turbulence Model for Shear Flows by a Double Expansion
Technique, Physics of Fluids, 4, pp. 1510-1520.
[9] Evans, J. S., and Schexnayder, C. J., (1980), Influence of Chemical Kinetics
and Unmixedness on Burning Supersonic Hydrogen Flames, AIAA Journal,
18, pp. 188-193.
[10] Mc Bride, B. J., and Gordon, S., (1971), Computer Program for Calculation
of Complex Chemical Equilibrium Compositions, Rocket Performance,
Incident and Reflected Shocks, and Chapman-Jouget Detonations, NASA SP-
273.
[11] Jameson, A., Schmidt, W., and Turkel, A., (1981), Numerical simulation of
Euler equation by finite volume method using Runge-Kutta scheme, AIAA
Paper 81-1258.
[12] Bussing, T. R. A., and Murman, E. M., (1988), Finite Volume Method for the
Calculation of Compressible Chemically Reacting Flows, AIAA Journal, 26,
pp.1070-1078.
[13] Menon, S., (1989), Shock-Wave- Induced Mixing Enhancement in Scramjet
Combustors, AIAA Paper No. 89-0104.
[14] Fernando, E.M., and Menon, S., (1991), Mixing Enhancement in
Compressible Mixing Layers - An Experimental Study, AIAA, Paper No. 91-
1721.