(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

A01-34203

AIAA 2001-3460
A Unified Analysis of Ramjet Operation
in an Integrated Rocket Ramjet Engine
Part I. Transition from Rocket Booster
to Ramjet Sustainer
Hong-Gye Sung
Agency for Defense Development
P.O. Box 35, YooSung, Taejon, Korea

Shih-Yang Hsieh and Vigor Yang

The Pennsylvania State University
University Park, PA 16802, USA

37th AIAA/ASME/SAE/ASEE Joint Propulsion

Conference and Exhibit
8-11 July 2001
Salt Lake City, Utah

For permission to copy or to republish, contact the copyright owner named on the first page.
For AIAA-held copyright, write to AIAA Permissions Department,
1801 Alexander Bell Drive, Suite 500, Reston, VA, 20191-4344.
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

AIAA 2001-3460

A UNIFIED ANALYSIS OF RAMJET OPERATION

IN AN INTEGRATED ROCKET RAMJET ENGINE

PART I. TRANSITION FROM ROCKET BOOSTER TO RAMJET SUSTAINER

Hong-Gye Sung*
Agency for Defense Development
P.O. Box 35, YooSung, Taejon, Korea

Shih-Yang Hsiehf, and Vigor Yang*

The Pennsylvania State University
University Park, PA 16802, USA

Abstract EI Activation energy of reaction /

A comprehensive numerical analysis has been conducted to Ev x-directional diffusion-flux vector
study ramjet operation in an integrated rocket ramjet (IRR) F y-directional convective-flux vector
propulsion system. This study consists of two parts. Part I, Fv y-directional diffusion-flux vector
this paper, details the transition from rocket booster to / frequency
ramjet sustainer while part II, another paper1, consists of f damping factor
combustion dynamics and system diagnostics during the
ramjet operation. The computational geometry consists of H source term vector
the entire IRR engine, including the inlet, the combustion h specific enthalpy of mixture
chamber, and the exhaust nozzle. The theoretical / imaginary part of complex
formulation is based on the complete conservation k turbulence kinetic energy
equations of mass, momentum, energy, and species kfi reaction rate constants of the forward
concentration, with consideration of finite-rate chemical reactions
kinetics and variable properties. Turbulence closure is kf,i reaction rate constants of the backward
reactions
achieved using a low-Reynolds-number K-E two-equation
L total number of reaction step
model. The governing equations are solved numerically by
A/W* molecular weight of species k
means of a finite-volume, preconditioned flux-differencing
N total number of species
scheme capable of treating chemical reacting flow over a
p pressure
wide range of Mach number. Various important physical
Q conserved variable vector
processes are investigated systemically, including unsteady
q{ heat flux vector
inlet aerodynamics, fuel/air mixing, ignition and so on.
Re Reynolds number
Ru universal gas constant
Nomenclature S Flmae speed
Sxy two-sided, cross-spectral density function f x and y
empirical constants
T Temperature
species specific heat at constant pressure
t Time
empirical constants
Uk diffusion velocity of species k
empirical constants u Velocity
C2 x spatial coordinate
C speed of sound Yk mass fraction of species k
E specific total energy or x-directional y+ variable distance from wall
convective-flux vector Z pseudo-time variable vector

Greek Symbols
* Senior Research Engineer, Tec-4-2 j3 scaling factor
t
Research Associate, Department of Mechanical fa molar concentration of species k
Engineering A forward difference or filter width
?
Professor, Department of Mechanical £ turbulence dissipation rate
Engineering, Associate Fellow AIAA 0 equivalence ratio
Copyright ©2001 by the American Institute of r preconditioning matrix
Aeronautics and Astronautics, Inc. All rights reserved.
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

>l heat conductivity from the rocket to the ramjet phase. A nozzleless booster,
ju Viscosity utilizing the propellant grain as a nozzle, circumvents the
Tij viscous stress tensor for a Newtonian gas inherent problem of an ejectable nozzle system.2 During
I/ stoichiometric coefficients on the reactants rocket operation, the combustor functions as a
side for species k in the /th reaction conventional rocket combustor, closed at the forward end,
y* stoichiometric coefficients on the products with a suitable nozzle formed from the propellant grain at
side for species k in the /th reaction the aft end. During transition, the combustion chamber
0 phase angle must reconfigure itself for the ramjet operation, open at the
p Density forward end to allow the ram air to enter the combustor,
<j empirical constant and with a large throat nozzle at the aft end, suitable for
ramjet operation, as illustrated in Fig. I.3 The port cover,
(j empirical constant which seals the upstream end of the combustion chamber,
is ruptured at the end of the boost phase to allow ram air to
r pseudo-time
enter the chamber for the sustainer phase. After the tail-off
L
T turbulent time scale of the booster thrust, drag force acting on an airframe
i
t Kolmogorov time scale causes the ramjet to rapidly lose the forward speed,
typically on the order of Mach numbers 0.1 per second.2
Co net production rate of chemically reaction Hence, the transition to ramjet take-over must be
accomplished in a timely fashion. When the pressure in
Subscripts the booster rocket chamber decays to a value where
av average value positive vehicle acceleration approaches zero, the inlet
/ Formation port-cover separation system operates. The port cover is
/ spatial coordinate or reaction step index forced into the combustor when the inlet ram air pressure
j spatial coordinate index exceeds the residual chamber pressure. The cover is
k spatial coordinate or species index expelled through the ramjet nozzle by the force of the
L laminar property incoming ram air. Reliable ignition and stable burning of
M model value of experiment the ramjet fuel during this transient phase are important
p pressure or index of polynomial function concerns, complicated by uncertainties such as the mass
R reference value flow rate and velocity of the ram air, fuel mixing,
rms root mean square ignitionability of fuel, flame holding, and so on.
t turbulent property inlet port cover solid propellant grain

u unburned fuel
y viscous

Superscripts
transpose of vector or temperature
fluctuation

I. Introduction b) Transition

An integrated rocket ramjet (IRR) combines a

rocket booster and a ramjet sustainer in one efficient
propulsion system with a common combustion chamber for
both the boost and sustain phases. This concept yields a
much more compact design than previous rocket/ramjet c) Ramjet Operation
systems such as a ramjet with a tandem booster rocket that
is jettisoned after burnout. The large difference in chamber Fig. 1. Conceptual diagram of a coaxial integrated rocket
operating pressure between the rocket booster (about 100 ramjet engine during the transition from booster
atmospheres) and the ramjet sustainer (around 6 to ramjet sustainer.
atmospheres at cruising flight) phases is a major IRR
design problem, which has been handled by using an Since French engineer Rene Lorin first introduced
ejectable booster nozzle ejected after the boost phase. the ramjet engine concept in 1913, substantial research has
However, this system creates additional reliability been pursued to understand ramjet internal flows,
problems such as dynamic impact on the vehicle during especially in the areas of supersonic inlet dynamics and
ejection of the booster nozzle and transition time delay
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

combustion instability. Notable are the numerical and

experimental investigations of unsteady diffuser flowfield
discussed in Refs. 4-15 and longitudinal combustion opening of
oscillations addressed in Refs. 16-24. All of these efforts inlet port cover

were motivated to improve the knowledge base of low- fuel injection

frequency combustion instability and its inlet flows. ignition of fuel
However most previous studies focused on either the inlet
or combustor under steady operating conditions; the direct
computing between the two subsystems was ignored in
spite of its significance in determining the overall engine
dynamics. Furthermore, no effort was devoted to the initial
flow and ignition transients of the ramjet operation in an
entire ramjet engine. The purpose of the present work is to 0.005 . 0.01 0.015 0.02
time(sec)N
conduct a unified analysis, which allows for a complete a) pressure history during transition
treatment of the engine flow development, ranging from
the leading edge of the inlet center body through the
exhaust nozzle. Emphasis is placed on the transient
sequence from the rocket booster to the ramjet sustainer
phase, such that the detailed mechanisms of combustion
dynamics from its inception can be investigated
systematically (Fig.l).
From the aspect of system development, the
rocket-to-ramjet transition testing has been routinely used
as part of the final verification for IRR systems in terms of
the functional and performance evaluations. This full
evaluation, however, is extremely costly and not available
for all functional operations.25 Very limited information ^0.012 0.014 a016 .
time(sec)
0.018
regarding transient mechanisms is available in the literature. b) Chamber pressure history zoomed
Results obtained here may shed much light on the research for combustion phase
of the transient operation during transition from rocket Fig. 2 Chamber pressure history during the transition from
booster to ramjet sustainer. The overall transient sequence booster to ramjet sustainer.
from the rocket booster phase to the ramjet sustain phase is
shown in Fig. 2 in terms of the pressure history calculated • Shock trains just after removing the port cover.
by this study. The tail of the booster phase, opening of the • Interaction of cold ram air and hot booster rocket gas.
inlet port cover, fuel injection, ignition of ramjet fuel, and • Fuel spreading mechanism and flame propagation.
steady combustion sequences are considered. Chamber • The establishment of a subsonic combustion
pressure drops exponentially during the tail of the booster environment during flame propagation.
phase due to exhaustion of the booster propellant, with the
inlet port cover separating the inlet and combustion The next paper1, Part II, consists of the
chamber during this period. After opening the inlet port combustion dynamics during ignition and stable
cover, ram air washes out the hot gas of the booster before combustion and diagnostics of a ramjet system.
ignition of the ramjet fuel, which is injected into the
combustion chamber during this time. After the combustor II. Theoretical Formulation
is charged with the fuel, a heat source is supplied to ignite
the fuel near the backward step of the combustion chamber. A. Governing Equations
Before ignition, the decaying pressure in the The governing equations based on the
combustion chamber balances with the ram air at around 3 conservation of mass, momentum and energy, and species
atmospheres. Once stable combustion of the ramjet fuel is concentration for a compressible, chemically reacting gas
achieved, the chamber pressure reaches the quasi steady- can be written as
state condition at 5.5 atmospheres.
This paper, Part I, discussed the flow transient (1)
mechanism during the transition from rocket booster to dt
ramjet sustainer such as: j d(puiuj+ p3tj
• How the flow field is established on placing an inlet (2)
port cover between the diffuser and the combustion
chamber.
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.
dpE
(3)
dt dx i dx,
(4)
dt dXj dx :
B. Turbulence Closure
The standard k-e model was proposed for high
Reynolds number flows and is traditionally used with a
wall function and the variable y+ as a damping function.
Universal wall functions do not exist in complex flows,
however, and the damping factor cannot be applied to
flows with separation. Thus, a low Reynolds number k-e
model was developed for near-wall turbulence by Jones
and Launder.26 Shin and Lumley27 observed that within
certain distances from the wall, all energetic large eddies
will reduce to Kolmogorov eddies (the smallest eddies in
turbulence), and all the important wall parameters, such as
friction velocity, viscous length scale, and mean strain rate
at the wall, can be characterized by the Kolmogorov micro
scale.
Yang and Shin28 proposed a time-scale-based k-e
model for the near-wall turbulence related to the
Kolmogorov time scale as its lower bound, so that the
equation can be integrated to the wall. The advantages of
this model are (a) no singularity at the wall, and (b)
adaptability to separation flow, since the damping function
is based on the Reynolds number instead of y+. The low
Reynolds number models have been designed to maintain
the high Re formulation in the log-law region and further
tuned to fit measurements for the viscous and buffer layers.
The low Reynolds number model used in this work is
based on Yang and Shin28. In the bulk flow region, the
distributions of the turbulence kinetic energy k and its
dissipation rate 8 are calculated from the following
transport equations:
dpk
~~dt~
where
ak = 1.0
It is noted that the above empirical constants (Gk,(7£) are
identical to those of the standard K-e model.
The turbulent viscosity is determined from the
following relation
where T( is the turbulent time scale defined as
(18)
and Tk is the Kolmogorov time scale defined as
(19)
The damping factor / is used to account for the wall
effect
-a5R5y)]l/2 (20)
where
Ry=kl/2yn/v (21)
The empirical constants in the above equations are
recommended27 as follows:
Cp = 0.09 , C]= 1 .44, C2 = 1 .92,
07=1. SxlO"4, tf3=5.0xlO'7, fl5=1.0xlO'10
III. Numerical Formulation
The conservation equations for moderate and high
Mach number flows are well coupled, and standard
numerical techniques perform adequately. In regions of
low Mach number flows, however, the energy and
momentum equations are practically decoupled and the
system of conservation equations becomes stiff. In the
entire ramjet engine, the flow fields are governed by a wide
variety of time scales (from supersonic flow in the inlet to
almost stagnation-flow near the backward-step corner of
the combustor). Such a wide range of time scales causes an
unacceptable convergence problem in the case of unsteady
calculations. Particularly, in regions with strong grid
stretching or with low Mach number flows, the
convergence may be severely impaired.
To overcome the problem, a dual time-integration
procedure designed for all Mach number flows is applied,
which may be constructed in two steps. First, a rescaled
pressure term is used in the momentum equation to
circumvent the singular behavior of pressure at low Mach
numbers. Second, a dual time-stepping integration
procedure is established.
The pseudo-time derivative may be chosen to
optimize the convergence of the inner iterations through
using an appropriate preconditioning matrix that is tuned to
rescale the eigenvalues to render the same order of
magnitude to maximize convergence. To unify the
conserved flux variables, a pseudo-time derivative of the
form rdZ/drcan be added to the conservation equation.
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

Since the pseudo-time derivative term disappears as The conservation equations, including finite chemical
converged, a certain amount of liberty exists in choosing reactions and a low-Reynolds number K-8 turbulent model,
the variable Z. We take advantage of this by introducing a have been solved using an ADI scheme with
pressure p as the pseudo-time derivative term in the preconditioned Chakravarthy-Osher TVD,29
continuity equation:
IV. Flow Transient Mechanism from Booster to
,dZ {dQtd(E-Ev)^(F-Fv) Sustainer
=H (22) In the physical configuration used for simulations,
the cowl radius Rc is 3.4 cm and the length of the inlet
where diffuser is 40.12 cm. The combustion chamber measures
(23) 38.93 cm in length and 7,786 cm in radius, and the nozzle
measures 8.16 cm in length. The throat area of the inlet
' 1/0 0 0 0 0 0 0
diffuser Ati is fixed at 0.615 Ar, and that of the exhaust
ul /3 p 0 0 0 0 0 nozzle Atn at 1.322 Ar, where Ac(=7rRc2) is the cowl area
vl /3 0 p 0 0 0 0 (see Fig. 3). The inlet flow conditions are set at the inlet
design condition, which has a Mach number of 2.1,
htj3-i pu pv p 0 0 0
temperature of 271.91 K, pressure of 0.74 atm., and
kl /3 0 0 0 / ? 0 0 altitude 2.5 km.
• e/0 0 0 0 0 p 0 (24)
Yi'0 0 0 0 0 0 p M=2.1 fuel injector \I combustion chamber
IRd i 1 Rtn|

a) Physical domain with an inlet port cover in place

\
V
1 1R0 0 0 0 0 0
A/-1 ' P fuel injector
M=2.1 combustion chamber
Also the scaling factor (3 can be taken as:
2 T 2 2
«£, if W <«£ b) Physical domain in case of opening an inlet port cover
2 2 2
w , if W o < w <c (25)
Fig. 3 Schematic of physical geometry both with and
without an inlet port cover.
The above variable vector are defined below:
To comprehensively analyze the initial transient
Q = y[p, pw, /TV, /?E, />K A ,, pk, pe]T (26) mechanism, two geometrical domains are considered. One
E- y\pu,pu2 + p',puv,(pe+ p)u,puYk,puk,pue\ (27)
consists of the inlet and the booster combustion chamber,
separated by the inlet port cover (Fig. 3a), and the other is
F = y\pv, puv, pv2 + p', (pe + p)v, pvYk, pvk, pve\ (28) the entire ramjet engine without the port cover (Fig. 3b).
iT"
Computations are carried out sequentially on the two
0, different regimes (first with the port cover, then without it),
(29) as described in the following sections.
—
cr jdx
->
t
port cover
-\T
0, Tyx, TyytUTjy + V Tyy ~ ^ , ^ ,

(30) a) Computational domain with an inlet port cover in place

fuel injector
2 d(fjv) p '_^_J^L
+ 2-/ / —-—
v -£L
'" ' ~~ '" "~ ""'
b) Computational domain in case of opening an inlet port cover
3 ' (fa dv

x J^ dM dv Fig. 4 Schematic of computational domain both with and

without an inlet port cover.
(31)
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

The computational domain is divided into four the incoming air spills over the inlet cowl. This flow is
zones (see Fig. 4). Zone 1 is the external flow region, zone drastically expanded over the inlet cowl. The flow spilled
2 is the supersonic diffuser region, and zones 3 and 4 make along the cowl surface cannot keep following this surface
up the combustion chamber region. The dimensions of the which is turning almost 180 degrees. Thus, a separation
grid points for each zone are 88 X 50, 214 X 60, 170 X 60, bubble appears on the front outside surface of the cowl-lip
and 170X50, respectively. which looks like the transient bubble on the transonic
airfoil surface at high angles of attack. Fig. 7 shows the
A. Flow Fields With the Inlet Port Cover in schematic of these complicated flow structures.
Place
During rocket operation, the inlet diffuser is
isolated from the combustor by a port cover at the interface
as depicted in Fig. 3a, and acts as a long coaxial cavity. cowl
The combustion chamber functions as a conventional
rocket motor. Calculation is first carried out to determine
the flow structure in the case of placing the inlet port cover.
The computational domain is composed of two blocks for
inlet flow calculation (Fig. 3a). The air approaching the a) Streamlines
ramjet engine at supersonic speed passes several oblique
shocks and is stagnated at the front of the port cover.
The wedge-shaped supersonic diffuser consists of
two ramps which turn the airflow and introduce oblique
shocks which decelerate the flow. Since the pressure
behind the terminal shock keeps increasing due to the
closed end of the inlet, the terminal shock is pushed out
toward the inlet cowl to adjust to the flow conditions of the
front zone and the aft zone of the terminal shock. The center body
terminal shock is expelled from the inlet cowl, moves
toward the vertex of the ram core, and finally forms a bow b) Mach number contours
shock detached from the cowl lip due to the formation of a Fig. 6 Close-up view near the cowl
large subsonic region, accompanied by the reversed
subsonic flow. During the rocket boost phase, the stagnation
pressure of the ram air at the front of the inlet port cover is
balanced on the other side by the chamber pressure
cowl produced by the burning rocket propellant. At the end of
the boost phase, the rocket propellant is almost exhausted
except for some slivers, and the pressure in the combustion
chamber decays. During this period, the inlet port cover is
forced into the combustion chamber as the inlet ram air
pressure exceeds the residual pressure of the chamber.

b) Mach number contour

Fig. 5 Flowfield in the inlet with port cover in
place

Fig. 5 shows streamline and Mach number

contours in the inlet diffuser. The stagnation pressure and
temperature of ram air are about 10.3 atm. and 745 K
respectively. Fig. 6 shows the detailed flow fields near the
cowl. Two oblique shocks occurred due to the
BS : Bow Shock LS: Lambda Shock OS: Oblique Shock
discontinuity in the surface of the ram core to hit the bow SB: Separation Bubble SL: Sonic Line VT: Vortex
shock and make it much steeper. The interaction of the
bow shock and a boundary layer forms a Lambda shock on Fig. 7 Schematic of the flow fields near the cowl lip with
the ram surface. Since the bow shock makes the flow an inlet port cover.
subsonic near the ram core and the shock wave is detached,
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.
B. Flow Fields on Opening the Inlet Port
Cover
Fig. 3b and 4b show the physical and
computational domains, respectively, for simulating the
transient flow field of an entire ramjet engine upon opening
of the inlet port cover. The computational domain consists
of four blocks.
Fig. 8 Flow evolution in the entire ramjet engine
after opening the inlet port cover
By using the results of the previous section as initial
conditions, the calculation continues to analyze the initial
transient phenomena without the inlet port cover. The
operating time is set to zero at the moment of burst the port
cover (Fig. 2). As the inlet port cover is opened, the ram air
compressed blocked by the inlet cover starts to penetrate
into the combustor and the bow shock is weakened and
disappears as the ram air flows into the combustion
chamber (Fig. 8). To better observe the transition
phenomena, Figs. 9 though Fig. 11 present three different
(time) sequences of the transitional ramjet flow.
Fig. 9 Time evolution of terminal shock (TS) in the Mach
number field before ignition.
Fig. 9 shows the behavior of the bow/terminal
shock and the evolution of the terminal shock train. Fig. 10
shows the generation/movement/reformation of the
vortices and reversed flow regions in the combustor. Fig.
11 's series show the temperature contours as the incoming
cold ram air interacts with the residual hot gases in the
chamber. The flow patterns are specifically discussed in
this section.
B.I The Terminal Shock Train
As shown before, a detached bow shock is formed
outside the inlet cowl, and this bow shock functions also a
terminal shock. After the inlet port cover is opened, this
bow shock is weakened and disappears as the ram air flows
into the combustion chamber. The bow shock is changed to
a normal shock and is swallowed into the inlet cowl (Fig.
9). The terminal shock keeps moving toward the
combustion chamber and finally disappears in the inlet. As
the ram air penetrates further into the chamber, the
terminal shock is not reformed anymore. Instead of a
terminal shock, a series of diamond shocks are formed like
an extremely overexpanded supersonic jet, which intensify
the core flow supersonic in the combustor. It is noted that
the ramjet fuel might be injected into combustor before the
supersonic ram air reaches the nozzle wall. Otherwise it
may be too difficult to ignite the ramjet fuel since the
sufficient fuel to ignite ramjet fuel can not penetrate into
the recirculation zone, the location of the ignition of ramjet
fuel. The streamline patterns (Fig. 10), also, support this
claim since fuel streak follows the streamline.
tsO.05 ms
t~13 ms
t=2.5 ms
Fig. 10 Time evolution of a streamline pattern before
ignition
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

B.2 The Generation/ Movement/ Reformation of Vortex

temperature \ 300,0 eoo.o 000.0- 1200.0 j
On opening the inlet port cover, the stagnated ram
air suddenly expands, and a small vortex is generated just
upstream of the dump plane of the combustion chamber
(Fig. 10). This small vortex moves to the corner of the
combustion chamber, and a recirculation zone is formed
near the backward step of the chamber. The recirculation
zone grows as the ram air penetrates into the combustor,
and the core of the recirculation zone also moves toward
the nozzle. As the ram air penetrates into the chamber, its
pressure decreases and the air is overexpanded. The ram
air separates and other vortices are reformed on the upper
and/or center region of the inlet. These vortices shrink and
expand like unsteady bubbles under high-pressure
conditions, which generates acoustic waves and pushes
them into chamber.
The oscillating inlet flow affects the formation of
vortices in the shear layer. These vortices move
downstream until they reach the convergent walls of the
nozzle, whereupon some of the vortical fluid is convected
back upstream by the recirculation zone near the chamber
wall. In the shear layer, they become highly stretched due
to the high strain rate of the mean flow there. It is noted Fig. 11 Time evolution of the distribution of temperature
that this strain rate decreases in the case of chemically before ignition.
reacting flow, since the mean supersonic flow near the
backward-facing step becomes subsonic due to the enhances spreading of fuel into the recirculation zone.
reformation of the terminal shock by the forcing of Second, even after introduction of the cold ram air into the
combustor pressure as pointed out in section V. chamber, some high-temperature gas still remains near the
backward step to supply energy to ignite the fuel. Finally,
B.3 The Interaction of the Cold Ram Air and the relatively uniform temperature in the chamber due to
the Hot Booster Gas roll-up of the hot gas at the backward step is conducive to
Once the port cover has been opened, the highly stable ignition.
compressed, stagnated, and cold ram air pushes the hot
booster gases toward the nozzle and expands into the V. Ignition of Ramjet Fuel
recirculation zone (Fig. 11). The expansion of the ram air
aids in the spreading of fuel into the recirculation zone (Fig. In going from the rocket boost phase to the ramjet
12). As the ram air spreads into the chamber, hot gas is sustain phase, the transient mechanism of the IRR is
washed out through the nozzle. Although the ram air relatively complicated after the inlet port cover is opened,
displaces much of the residual booster gas, some hot gas since the flow must adjust from supersonic at the ramjet
still remains in the recirculation zone (Fig. 11). Since the inlet to subsonic in the combustor. An important task of the
pressure near the backward step is relatively high as hot IRR is to establish subsonic combustion by achieving a
gas in the center region of the chamber is washed out, the normal shock as the terminal shock in the diffuser with
hot gas remaining near the backward-facing step starts to sufficient pressure recovery. But the terminal shock
roll up with the cold ram air toward the center region of the disappears after opening the inlet port cover. Moreover, the
chamber. This roll-up of the hot gas makes the temperature supersonic ram air occupying the combustor decreases the
gradient of the chamber more uniform, which promotes residence time of the flow, which may aggravate the
stable ignition. However, ignition should be carefully ignition of the ramjet fuel and may allow flame flashback.
timed since the roll-up of the hot gas initially creates a The major concerns regarding stable ignition in this section
local hot spot in the chamber, which may cause some are how the fuel spreads uniformly in the combustor, and
overshoot of combustor pressure and dynamic flame how the subsonic combustion environment is established.
shedding, making the ignition unstable, as detailed further
in section V. A. Fuel Spreading in the Combustion
From the above observations, three potentially Chamber
positive mechanisms to assist the stable ignition of ramjet The fuel injectors are located at 4 cm ahead of the
fuel are summarized. First, the expansion of the ram air dump combustion chamber, and the fuel is propane (Cj//5).
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

The fuel mass flow rate is 0.12 kg/s and the equivalence The fuel near the center region is evolving due to the
ratio is about 0.8. Computationally, fuel is injected into the secondary vortex caused by the flow separation behind the
ram air flow at each computational cell in the injection ram core spike.
plane, at whatever rate is necessary to maintain a specified
fixed equivalence ratio (e.g., 0.8) given the current ram air vorticity
flow rate at any given time. In the physical equivalence IISOOO 22000 26000 -300001

ratio to this computational injection scheme, the fuel

injection orifices are also assumed to be chocked, so that
the fuel injection rate is not directly affected by the local
pressure.

Fig. 13 Vorticity fields before ignition

Fig. 12 Ramjet fuel spreading before ignition
Fig. 12 shows the fuel spreading in the combustor. As
expected, the fuel stream basically follows the trajectory of
the cold ram air, so the fuel spreads into the combustion
chamber in the same way as the high-speed ram air which
B. Flame Propagation
To ignite the fuel, a heat source is located near the
backward step of the combustion chamber, and heat is
supplied until the gas temperature reaches the auto-ignition
temperature (e.g., 1500 K).
temperature.^ 800 100014001800

dominates the combustor flow field during this time. t=2.5 ms x i'SJBtion location
Initially, the flow expansion of the ram air plays a major
role in the penetration of fuel into the recirculation zone.
After that, vortex roll-up from the edge of the
backward facing step is the next important factor in fuel
spreading (Fig. 13). Vortex roll-up provides large-scale
vortical structures which supply energy for the cascade of
turbulence kinetic energy from large eddies to small eddies, t=2.9 ms
and the turbulence motions more evenly distribute fuel in
the shear layer. Initially, in the flowfield created by
expansion of the supersonic ram air into the combustor,
penetration of fuel into the recirculation zone is also
enhanced by the convection of fuel-laden vortical fluid
from the downstream region back along the combustor
wall. Fluctuations in the fuel field are observed near the
ends of the injection plane (i.e., near the inlet wall and the
centerline). The fuel distribution fluctuates near the inlet
wall and center region as the oscillating terminal shock Fig. 14 Time evolution of a flame structure during ignition
modulates the vortical boundary layer flow there. transition
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.
Fig. 14 shows the temperature fields just after fuel ignition.
The highest temperature region at t=2.5 msec is the
ignition location, and there are other regions of the second
highest temperature near the wall, due to residual gas from
the booster propellant, and near the center region, due to
the roll-up of hot residual gas. It is noted that the ignition
of the ramjet fuel should be set very carefully to avoid the
local hot spot caused by the roll up of residual gas of the
booster propellant. The flame starts to propagate into the
entire combustion chamber, albeit not uniformly. Flame
speed is a strong function of temperature, pressure,
equivalence ratio, and convective velocity. The flame
propagates rapidly along the surface of the local optimal
equivalence ratio, which is constantly changing due to the
evolution of vorticies in the shear layer (Fig. 14). The
detail mechanism of flame propagation is addressed in Part
II1.
C. The Establishment of Subsonic
Combustion Environment During Flame
Propagation
As indicated in section IV, the terminal shock is
smeared out and a peristalsis of the supersonic core appears
after opening the inlet port cover, and the flow becomes
supersonic in the combustion chamber. This flow pattern in
the diffuser is due to the shock-induced separation as
pointed by Bogar et. al.5, with boundary layers influenced
by the intense adverse pressure gradient associated with
unsteady oscillation of the shock which modulates the
separated boundary layer flow. Because the ramjet operates
with a subsonic combustor, an interesting question to ask is,
"How is subsonic combustor flow established from the
supersonic regime which follows the opening of the inlet
port cover?" ___
*«*&« <
Fig. 15 establishment of the terminal shock after ignition
10
To answer this question, Fig. 15 gives several snapshots,
each of which shows the Mach number in the ramjet
engine, with each frame synchronized with the
corresponding frame from Fig. 14. As the flame propagates,
the pressure in the combustion chamber increases. In the
supersonic flow region of the inlet diffuser, since the wave
(u-c), only propagates toward a combustor, the information
regarding this pressure increase due to combustion cannot
be delivered upstream through the supersonic core of the
inlet diffuser flow. But the wave (u-c), carrying the
information can propagate upstream through the subsonic
boundary layer. The higher pressure delivered by the wave
increases the adverse pressure gradient in the diffuser,
which induces flow separation, and thereby decreases the
effective flow area in the inlet. So the peristalsis of the
supersonic core in the inlet becomes stretched, segregated,
and weak.
Finally the upstream-running acoustic wave is
able to propagate through the core flow of the inlet, and the
terminal shock is reformed due to the increase in the
combustor pressure, with subsonic flow downstream of the
shock. Another major factor which promotes subsonic
combustor flow is the rapid increase in temperature due to
combustion of the fuel. This temperature rise yields higher
sonic speeds, thus promoting lower Mach number flow.
Once a subsonic combustor environment has been
achieved in which stable combustion can occur, dynamic
flame behavior is still possible due to coupling between
vibration of the terminal shock and pressure oscillations in
the combustor. This coupling is primarily characterized by
the longitudinal acoustic mode as described in Part II.
VI. Summary and Conclusions
The physical mechanism of the transition from
rocket booster to ramjet sustainer of IRR was investigated
using the dual time stepping numerical algorithm with
preconditioning method. The conservation equations,
including finite chemical reactions and a low-Reynolds
number K-e turbulence model, were solved using an ADI
scheme with preconditioned Chakravarthy-Osher TVD.
The computational geometry consists of the entire IRR
engine, including the inlet, the combustion chamber, and
the exhaust nozzle.
The calculations for the rocket booster phase, in
which the inlet port cover is closed, show that the terminal
shock train reflected from the port cover forms a large
detached bow shock outside the inlet. The complex flow
structures are clearly captured. On opening the inlet port
cover, to begin the ramjet phase, the terminal shock is
smeared out and the flow throughout the entire engine
becomes supersonic. However, combustion of the ramjet
fuel builds up the pressure and temperature in the
combustor such that the combustor flow becomes subsonic
and a terminal shock is reestablished in the diffuser, which
are explored in theoretical base. Also ignition should be
(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.
carefully timed since the roll-up of the hot gas initially
creates a local hot spot in the chamber, which may cause
some overshoot of combustor pressure
REFERENCES
[I] Sung, H.-G , Hsieh, S, and Yang, V., "A Unified
Analysis of Ramjet Operation in an Integrated Rocket
Ramjet Engine: Part II. Combustion Dynamic of Ramjet
Engine" AIAA 2001-3192, 2001.
[2] Waltrup, P. J., White, M. E., and Gravlin, E. S.,
"History of U.S. Navy Ramjet, Scramjet, and Mixed-Cycle
Propulsion Development," AIAA Paper 96-3152, 1996.
[3] Myers, T. D., "Integral Boost, Heat Protection, Port
Covers and Transition," AGARD-LS-136, pp. 4-1-20, 1984.
[4] Chen, C. P., Sajben, M., and Kroutil, J. C, "Shock-
Wave Oscillations in a Transonic Diffuser Flow," AIAA
Journal, Vol. 17, No. 10, 1979, pp. 1076-1083.
[5] Bogar, T. J., Sajben, M., and Kroutil, J. C.,
"Characteristic Frequencies of Transonic Diffuser Flow
Oscillations," AIAA Journal, Vol. 21, 1983, pp. 1232-1240.
[6] Salmon, J. T., Bogar, T. J., and Sajben, M., "Laser
Doppler Velocimeter Measurements in Unsteady,
Separated, Transonic Diffuser Flows," AIAA Journal, Vol.
21, No. 12, 1983, pp. 1690-1697.
[7] Sajben, M., Bogar, T. J., and Kroutil, J. C., "Forced
Oscillation Experiments in Supercritical Diffuser Flows,"
AIAA Journal, Vol. 22, No. 4, 1984, pp. 465-474.
[8] Sajben, M., Bogar, T. J., and Kroutil, J. C.,
"Experimental Study of Flows in a Two-Dimensional Inlet
Model," Journal of Propulsion and Power, Vol. 1, No. 2,
1985, pp. 109-117.
[9] Bogar, T. J., Sajben, M., and Kroutil, J. C., "Response
of Supersonic Inlet to Downstream Perturbations," Journal
of Propulsion and Power, Vol. 1, No.2, 1985, pp. 118-125.
[10] Bogar, T. J., "Structure of Self-Excited Oscillations in
Transonic Diffuser Flows," AIAA Journal, Vol. 24, 1986,
pp. 54-61.
[II] Culick, F. E. C. and Rogers, T., "The Response of
Normal Shocks in Diffusers," AIAA Journal, Vol. 21, No.
10, 1983, pp. 1382-1390.
[12] Yang ,V. and Culick, F. E. C., "Analysis of Unsteady
Inviscid Diffuser Flow With a Shock Wave," Journal of
Propulsion and Power, Vol. 1, No. 3, 1985, pp. 222-228.
[13] Hsieh, T., Wardlaw, A. B., Collins, P., and Coakley, T.
J., "Numerical Investigation of Unsteady Inlet Flow
Fields," AIAA Journal, Vol. 25, No. 5, 1987, pp. 75-81.
[14] Hsieh, T., Bogar, T. J. and Coakley, T. J., "Numerical
Simulation and Comparison with Experiment for Self-
Excited Oscillations in a Diffuser Flow," AIAA Journal,
Vol. 25, No. 7, 1987, pp. 936-943.
[15] Oh, J. Y., "Numerical Study of Steady and Oscillatory
Flow Structures in an Axisymmetric Supersonic Inlet,"
Ph.D. Thesis, The Pennsylvania State University,
University Park, PA, August 1994.
11
[16] Schadow, K. C., Crump, J. E., and Blomshield, F. S.,
"Combustion Instability in a Research Dump Combustor:
Inlet Shock Oscillations," Proceedings of 18th JANNAF
Combustion Meeting, 1981.
[17] Abouseif, G. E., Keklak, J. A., and Toong, T.Y.,
"Ramjet Rumble: The Low Frequency Instability
Mechanism in Coaxial Dump Combustors," Combustion
Science and Technology, 1984, pp. 83-108.
[18] Yang, V. and Culick, F. E. C., "Analysis of Low
Frequency Combustion Instabilities in a Laboratory Ramjet
Combustor," Combustion Science and Technology, Vol. 45,
1986, pp. 1-45.
[19] Yang, V. and Culick, F. E. C., "Nonlinear Analysis of
Pressure Oscillations in Ramjet Engines," AIAA Paper 86-
0001, 1986.
[20] Kailasanath, K., Gardner, J. H., Oran, E. S., and Boris,
J. P., "Effects of Energy Release on High-Speed Flows in
an Axisymmetric Combustor," AIAA Paper 89-0385, 1989.
[21] Menon, S. and Jou, W.-H., "Large-Eddy Simulations
of Combustion Instability in an Axisymmetric Ramjet
Combustor," Combustion Science and Technology, Vol. 75,
1991, pp. 53-72.
[22] Menon, S., "Active Combustion Control in a Ramjet
Using Large-Eddy Simulations," Combustion Science and
Technology, Vol. 84, 1992, pp. 51-79.
[23] Bauwens, L. and Daily, J. W., "Flame Sheet
Algorithm for Use in Numerical Modeling of Ramjet
Combustion Instability," Journal of Propulsion and Power,
Vol. 8, No. 2, 1992, pp. 264-270.
[24] Hsieh, S. and Yang, V., "A Unified Analysis of
Unsteady Flow Structures in a Supersonic Ramjet
Engines," AIAA Paper 97-0396, 1997.
[25] Webster, F. F., "Ramjet Development Testing: Which
Way is Right?" Journal of Propulsion and Power, Vol. 5,
No. 5, 1989, pp. 565-576.
[26]Jones, W. P. and Launder, B. E., "The Calculation of
Low-Reynolds Number Phenomena with a Two-Equation
Model of Turbulence," International Journal of Heat and
Mass Transfer, Vol. 16, 1973, pp. 1119-1130.
[27] Shih, T. H. and Lumley, J. L., "Kolmogov Behavior of
Near-Wall Turbulence and Its Application in Turbulence
Modeling," NASA TM-105663, 1992.
[28] Yang, Z. and Shih, T. H., "New Time Scale Based k-e
Model for Near-Wall Turbulence," AIAA Journal, Vol. 31,
No. 7, 1993, pp. 1191-1197.
[29] Chakravarthy, S. R. and Osher, S., "A New Class of
High Accuracy TVD Schemes for Hyperbolic
Conservation Laws," AIAA Paper 85-0363, 1985.