Supersonic Combustion and Mixing Characteristics of ...

Old Dominion University Old Dominion University

ODU Digital Commons ODU Digital Commons

Mechanical & Aerospace Engineering Theses & Dissertations Mechanical & Aerospace Engineering

Winter 2002

Supersonic Combustion and Mixing Characteristics of Supersonic Combustion and Mixing Characteristics of

Hydrocarbon Fuels in Screamjet Engines Hydrocarbon Fuels in Screamjet Engines

Ahmed A. Taha Old Dominion University

Follow this and additional works at: https://digitalcommons.odu.edu/mae_etds

Part of the Heat Transfer, Combustion Commons, and the Structures and Materials Commons

Recommended Citation Recommended Citation Taha, Ahmed A.. "Supersonic Combustion and Mixing Characteristics of Hydrocarbon Fuels in Screamjet Engines" (2002). Doctor of Philosophy (PhD), Dissertation, Mechanical & Aerospace Engineering, Old Dominion University, DOI: 10.25777/eqzz-7986 https://digitalcommons.odu.edu/mae_etds/158

This Dissertation is brought to you for free and open access by the Mechanical & Aerospace Engineering at ODU Digital Commons. It has been accepted for inclusion in Mechanical & Aerospace Engineering Theses & Dissertations by an authorized administrator of ODU Digital Commons. For more information, please contact [email protected].

https://digitalcommons.odu.edu/

https://digitalcommons.odu.edu/mae_etds

https://digitalcommons.odu.edu/mae_etds

https://digitalcommons.odu.edu/mae

https://digitalcommons.odu.edu/mae_etds?utm_source=digitalcommons.odu.edu%2Fmae_etds%2F158&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/300?utm_source=digitalcommons.odu.edu%2Fmae_etds%2F158&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/224?utm_source=digitalcommons.odu.edu%2Fmae_etds%2F158&utm_medium=PDF&utm_campaign=PDFCoverPages

https://digitalcommons.odu.edu/mae_etds/158?utm_source=digitalcommons.odu.edu%2Fmae_etds%2F158&utm_medium=PDF&utm_campaign=PDFCoverPages

mailto:[email protected]

SUPERSONIC COMBUSTION AND MIXING CHARACTERISTICSI MI

OF HYDROCARBON FUELS IN SCRAMJET ENGINES

by

Ahmed A. Taha B.S. July 1988, Cairo University, Egypt M.S. June 1994, Cairo University, Egypt

A Dissertation Submitted to the Faculty of Old Dominion University in Partial Fulfillment of the

Requirement for the Degree of

DOCTOR OF PHILOSOPHY

MECHANICAL ENGINEERING

OLD DOMINION UNIVERSITY December 2002

Approved by:

Dr. Surendra N. Tiwari (Director)

Dr. Taj O. Mohieldin (Co-Director)

Dr. Sushil Chaturvedi (Member)

DrrArthur C. TaylorDJfMember)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

ABSTRACT

SUPERSONIC COMBUSTION AND MIXING CHARACTERISTICS OF HYDROCARBON FUELS IN SCRAM JET ENGINES

Ahmed A. Taha Old Dominion University

Director: Dr. Surendra N. Tiwari Co-Director: Dr. Taj O. Mohieldin

A numerical study has been conducted to investigate the combustion and mixing

characteristics o f hydrocarbon fuels in scramjet engines. The three-dimensional Reynolds

Average Navier-Stokes equations have been used to numerically investigate the

supersonic combustion of ethylene and propane using different combustor configurations

and fuel injection schemes.

Four physical models are used in the current study: the unswept generic rearward-

facing step, the side swept rearward-facing step, the wedge with no pilot injection, and

the wedge with pilot injection configurations.

The combustion characteristics of gaseous propane in supersonic airflow using the

rearward-facing step that is swept inward from both end sides is studied. The effect of

sweeping the step on the flow field features o f propane combustion is investigated.

The study of the supersonic combustion o f ethylene is carried out using different

combustor configurations, different main fuel equivalence ratios, and different pilot fuel

equivalence ratios. The effect o f the combustor configuration on the temperature flow

field and the flow structure is investigated by simulating two different combustor

configurations, the rearward-facing step and the cavity. Then, the effect of the

equivalence ratio of the main fuel injection on the general flow field features and energy

field characteristics is studied. Two normal fuel equivalence ratios are used, 0.4S and

0.60. Lastly, the effect o f the pilot fuel equivalence ratio on the flame holding mechanism

and the combustion and flow field qualities concludes the study. Six pilot injection

equivalence ratios are studiedO.0,0.02,0.03,0.045,0.06, and 0.08.

The swept step shows the ability to hold the propane flame in the supersonic air

stream without extinction. It was found that the side sweeping o f the combustor exhibits

the high temperature and combustion products concentration in the far field domain while


the area downstream of the normal injection location characterizes lower temperature and

products concentration. It is recommended to optimize the combustor length to ensure the

complete combustion and consequently the full liberation of the chemical energy stored

in the fuel before the fuel exits the combustor.

The main findings from the ethylene study can be summarized in the following

points. The step configuration with no pilot injection can afford the flame holding

mechanism in the supersonic air stream by creating the flow recirculations in the step

base area and featuring permanent high temperature regions surrounding the normal fuel

injection. The step configuration showed good mixing capabilities in the far field domain.

The wedge configuration proved superiority over the generic rearward-facing step

configuration in holding the ethylene flame in the supersonic airstreams, producing

overall higher temperature medium throughout the combustor, and exhibiting lower flow

losses and higher combustor efficiency. The increase in the equivalence ratio o f the

ethylene normal fuel injection enhances the general flow field features and energy field

characteristics in the combustor except in the step base area where the lower equivalence

ratio features better temperature distribution and higher combustion efficiency.

Although the wedge with no pilot injection configuration presents the highest

level o f temperature distribution in the cavity and downstream regions, the 0.02-pilot

equivalence ratio increases the temperature of the upstream face of the normal injection

and enhances the flame holding mechanism. The 0.02-pilot equivalence ratio presented

the optimum pilot injection case that can promote the flame holding mechanism and keep

good combustion and flow field qualities. While further increase o f the pilot injection

equivalence ratio quenches the high temperature gases in the cavity region, which leads

to the deficiency in the flame holding mechanism, the excessive pilot fuel injection shows

its positive effect by increasing the average flow field static temperature and absolute

pressure in the far field domain.


IV

ACKNOWLEDGMENTS

I would like to express my sincere thanks and deepest appreciation to my advisor,

Professor Surendra N. Tiwari, for his continuous and everlasting support, encouragement,

and patience and for the constructive and fully-informative graduate courses he taught me

during my program. The affectionate parental care he used to offer his students is not less

than, by any way, the great academic and professional guidance he is sincerely providing.

I gratefully acknowledge the financial support o f the Old Dominion University’s Institute

for Scientific and Educational Technology (ISET) through NASA Langley Research

Center, Cooperative Agreement NCC1-349 under the supervision of Prof. Tiwari.

My deep thanks and faithful respect are due to my co-advisor, Professor Taj. O.

Mohieldin, for the valuable scientific knowledge I gained from the graduate courses he

taught me and the priceless discussions and guidance he offered me during the course of

my research. I sincerely appreciate the great help and support my advisors offered me to

overcome the health problem I suffered during my program.

My best words of thanks are due to the faculty and staff of the Mechanical

Engineering Department, Old Dominion University, and especially to Professor Sushil

Chaturvedi, the Department Chair and my committee member, for the wonderful and

persistent effort that he does not spare to facilitate and promote the research environment

in the Department. My appreciation extends also to my committee members Profs. A. S.

Roberts Jr. and A. C. Taylor III for their time and help.

I would like to thanks all my colleagues and fellows at Old Dominion University,

my friends who offered me the help and support and among whom I felt the comforting

spirit o f the one family.

I extend my thanks and gratitude to my former advisors and Professors, M M .

Elkotb and O. M. F. Elbahar, Cairo University, Egypt, who helped me build up my

scientific background and research skills.

Last, but not least, special thanks, acknowledgment, and appreciation are due to

my dedicated parents, my persistent wife, and my children for their everlasting love,

encouragement, and patience. Without their support, this work would have not been

possible.


V

TABLE OF CONTENTS

Page

LIST OF TABLES ................................................................................................. via

LIST OF FIGURES . ................................................................................... ix

Chapter

I INTRODUCTION AND CHALLENGES......................................................... I

L I Introduction . . . . . ........................................... ........................................ L

1.2 Challenges in Supersonic Combustion......................... ..................... 3

1.3 Objectives of Present Study................................................................ 5

1.4 Work Layout ..................................................... 6

H LITERATURE SURVEY.................................................................................... 7

2.1 Introduction......................................................................... 7

2.2 Injection Modes and M ixing................................................................ 7

2.3 Auto and Forces Ignition................................. .................................... 19

2.4 Vitiated and Clean Incoming A ir ............................................................. 21

2.5 Hydrocarbon Fuels............................................ .................................. 23

2.6 Flame Holding Mechanisms for Liquid Hydrocarbon Combustion... 33

2.7 Importance of Modeling and Simulation ........................................... 38

UL THEORETICAL FORMULATIONS............................................................... 51

3.1 Introduction.......................................................................................... 51

3.2 Governing Equations............................................................................ 51

3.2.1 The Mass Conservation Equation............................................. 51

3.2.2 Momentum Conservation Equation.......................................... 52

3.2.3 The Energy Equation................................................................ 52

3.3 Basic Physical Models for Flow and Heat Transfer............................ 53

3.3.1 Turbulence M odels......................................................... 53

3.3.2 Compressible Flow s................................................................. 55

3.3.3 Chemical Species Transport and Reacting F low ..................... 57

3.4 Thermal M odel....................................................................................... 60

3.4.1 Density: Ideal Gas Law for Compressible Flows.................... 60


v i

3.4.2 Composition-Dependent Thermal Conductivity for

Multicomponent Mixtures.............................................. 60

3.4.3 Composition-Dependent Viscosity for Multicomponent

Mixtures................................................................ 61

3.4.4 Specific Heat Capacity as a Function of Composition 61

3.5 Grid Generation ......................................................................... 62

3.6 Method o f Solution.............................................................................. 63

3.6.1 Basics of the Overall Solution Algorithm................................ 63

3.6.2 Residual Reporting............................................ 64

3.6.3 Judging Convergence.............................................................. 66

3.7 Physical M odels................................ 67

3.7.1 Rearward-Facing Step Configuration...................................... 67

3.7.2 Swept Rearward-facing Step Configuration............................ 67

3.7.3 Cavity with no Pilot Injection Configuration.......................... 68

3.7.4 Cavity with Pilot Injection Configuration................................ 68

3.8 Boundary and Initial Conditions ...... 68

IV PROPANE SUPERSONIC COMBUSTION............................................. 75

4.1 Introduction........................ 75

4.2 Physical Models and Flow Calculations Description........................ 75

4.3 Grid Independence Study..................................................................... 76

4.4 Results................................................................................................... 77

4.4.1 Symmetry Plane Results............................................................ 77

4.4.2 Cross Row Results.................................................................... 79

4.4.3 Pilot Injection Plane Results................................................... 80

4.4.4 Upper-Wall Average Results.................................................... 81

V ETHYLENE SUPERSONIC COMBUSTION............................................. 99

5.1 Introduction........................................................................................... 99

5.2 Physical Models and Flow Calculations Description......................... 99

5.3 Definitions............................................................................................. 101

5.3.1 Stagnation Pressure Efficiency and Entropy Production 101

5.3.2 Kinetic Energy Efficiency......................................................... 101


v i i

5.3.3 Combustion Efficiency............................................................. 102

5.3.4 Fuel Penetration and Rate o f Fuel D ecay ................ 103

5.3.5 Upper-Wall Average Absolute Pressure and Static

Temperature Distribution........................................................ 103

5.3.6 Flow Field Average Absolute Pressure and Static

Temperature Distribution............. . ........................................... 104

5.4 Results ................................................ 104

5.4.1 Effect of Configuration............................................................. 104

5.4.2 Effect o f Main Fuel Equivalence Ratio ................................ 110

5.4.3 Effect of Pilot Injection Equivalence R atio ............... ............. 114

VI CONCLUDING REMARKS................................................................. 165

REFERENCES.......................................................................................................... 169

V ITA ......................................................................................................................... 187


viif

LIST OF TABLES

Table2.1 Energy properties o f and hydrocarbon fuels............. ...........................

5.1 Values of the longitudinal distribution o f the area-weighted average flow

field static temperature for different pilot equivalence ratios ..............


Page44

121

ix

LIST OF FIGURES

Figure Page

2.1 Schematic diagram o f the scramjet engine configuration................ 45

2.2 Essential features of two-dimensional pr planar geometry scramjet

engines................................................................................................ 45

2.3 Airframe-integrated scramjet along with a partial cross-section

through a typical modular engine.......................................... 46

2.4 Schematic of the region surrounding a single perpendicular

injector............................................................................................... 46

2.5 Generation of counter-rotating streamwise vorticity by ramp fuel

injector...................................................... 47

2.6 Scramjet fuel-injector compression and expansion ram ps............... 47

2.7 Energy content per unit mass for different avionic fuels................. 48

2.8 Temperature limit for different avionic fuels ......................... 49

2.9 Energy content per unit volume for different avionic fuels 50

3.1 An unstructured face meshing unit “Tri-Pave” scheme.................... 70

3.2 T-Grid meshing scheme.................................................................... 70

3.3 Overview of the solution process...................................................... 71

3.4(a) Schematic diagram for the rearward-facing step configuration 72

3.4(b) Schematic diagram for the step configuration at the symmetry

plane with full-dimensions ............................................................... 72

3.5 Schematic diagram for the swept rearward-facing step

configuration...................................................................................... 73

3.6 Schematic diagram for the cavity with no pilot injection

configuration...................................................................................... 73

3.7(a) Schematic diagram for the cavity with pilot injection configuration 74

3.7(b) Schematic diagram for the cavity with pilot injection configuration

at the symmetry plane with full-dimensions................ 74

4.1 Comparison between longitudinal average upper-wall absolute

pressure distribution for both grid sizes, grid independence study.. 83


X

4.2 Absolute pressure contours at the symmetry plane........................... 84

4.3 Static temperature contours- at the symmetry plane.......................... 85

4.4 COz combustion product mass fraction contours at the symmetry

plane.................................................................................................... 86

4.5 Mach number contours at the symmetry plane................................. 87

4.6 Static temperature contours at X=0.03 lm cross flow plane 88

4.7 Velocity vectors and streamlines contours at X=0.03 lm cross flow

plane............................................................ .................... ................... 89

4.8 Velocity vectors and streamlines contours at X=0.070m cross flow

plane .......................................................................................... 90

4.9 Mach number contours at X=0.070m cross flow plane..................... 9 1

4.10 Absolute pressure contours at X=0.070m cross flow plane.............. 92

4.11 Static temperature contours at X=0.070m cross flow plane ............. 93

4.12 CO2 combustion, products mass fraction contours at X=0.070m

cross flow plane.................................................................................. 94

4.13 Velocity vectors and streamlines contours at the pilot injection

plane, Y=0.0304m.............................................................................. 95

4.14 Static temperature contours at the pilot injection plane,

Y=0.0304m......................................................................................... 96

4.15 C 02 combustion product mass fraction contours at the pilot

injection plane, Y=0.0304m.............................................................. 97

4.16 Average upper-wall absolute pressure and static temperature

distribution along the combustor length.............. ............................ 98

5.1 Velocity vectors and streamlines contours at the symmetry plane.. 123

5.2 Temperature contours at the symmetry plane................................... 125

5.3 Velocity vectors and streamlines contours at X=0.040m plane 127

5.4 Temperature contours at X=0.040m plane......................................... 129

5.5 Velocity vectors and streamlines contours at the pilot injection

plane, Y=0.0304m.............................................................................. 131

5.6 Temperature contours at the pilot injection plane, Y=0.0304m....... 133

5.7 Velocity vectors and streamline contours at the normal injection 135


x i

plane, X=0.070m ..........................................................................

5.8 Temperature contours at the normal injection plane, X=0.070m... 137

5.9 Temperature contours at X=0.220m................................................... 139

5.10 Comparison between the combustion efficiency for the step and

wedge configurations.......................................................................... 140

5.11 Comparison between the fuel penetration of the step and wedge

configurations ............................................. 140

5.12 Comparison between the rate of fuel decay of the step and wedge

configurations ............ 141

5.13 Comparison between the upper-wall average static temperature

distribution o f the step and wedge configurations ................... 142

5.14 Comparison between the flow field average static temperature

distribution o f the step and wedge configurations............................ 142

5.15 Comparison between the upper-wall average absolute pressure

distribution of the step and wedge configurations............................ 143

5.16 Comparison between the flow field average absolute pressure

distribution o f the step and wedge configurations ............. 143

5.17 Comparison between the fuel penetration of the 0.45-ER and 0.60-

ER wedge configuration cases............................. ............................ 144

5.18 Comparison between the rate of fuel decay of the 0.45-ER and

0.60-ER wedge configuration cases................................................... 144

5.19 Comparison between the combustion efficiency of the 0.45-ER

and 0.60-ER wedge configuration cases............................................ 145

5.20 Comparison between the upper-wall average static temperature

distribution o f the 0.45-ER and 0.60-ER wedge configuration

cases................................................................................................... 146


distribution, o f the 0.45-ER and 0.60-ER wedge configuration

cases........................... ........... .................. .... .................................... 146

5.22 Comparison between the upper-wall average absolute pressure



xii

cases.................................................................................................... 147



cases.................................................................................................... L47

5.24 Reversed flow at the pilot injection ports for 0.01-ERP, pilot inj.

plane................................... . ....................................... . .................. L49

5.25 Reversed flow at the pilot injection ports for 0.0l-ERp, sym. Plane. 149

5.26 Velocity vectors and streamlines contours at the symmetry plane

for different pilot equivalence ratios ..................................... 151

5.27 Effect o f the pilot injection equivalence ratio on the fuel

penetration ........ 152

5.28 Effect of the pilot injection equivalence ratio on the rate of decay

of the fu e l.............................................. 152

5.29 Temperature contours at the symmetry plane for different pilot

equivalence ratios....................................................... 154

5.30 Temperature contours at the pilot injection plane, Y=0.0304m, for

different pilot equivalence ratios...................................................... 156

5.31 Comparison between the temperature contours of 0.0-ERp and

0.02-ERp-cases focusing on the cavity region............... 157

5.32 Temperature contours at X=0.040m cross flow plane for different

pilot equivalence ratios...................................................................... 159

5.33 Temperature contours the normal injection plane, X=0.070m, for

different pilot equivalence ratios........................................................ 161


distribution for different pilot equivalence ratios.............................. 162


distribution for different pilot equivalence ratios .................. 162

5.36 Effect o f pilot injection equivalence ratio on the stagnation

pressure efficiency.................................. 163

5.37 Effect o f pilot injection equivalence ratio on the entropy

production............................................................................................ 163


XU1

5.38 Effect of pilot injection, equivalence ratio on the kinetic energy

efficiency.......................................... ..............................— -............. 164

5.39 Effect of pilot injection equivalence ratio on the combustion

efficiency............................................................................................. 164


L

Chapter I

INTRODUCTION AND CHALLENGES

1.1 IntroductionSupersonic Combustion Ramjet (Scramjet) engines are attractive for both cruise

and boost missions in the hypersonic portion of the flight corridor [1]. The flow speed is

high in scramjet engines and the mixing and reaction dmes are limited in the engine

combustor. Therefore, combustion is a critical factor in the design o f scramjet engines

[2].Sustained airbreathing hypersonic flight shows potential advantages for military

applications. Furthermore, airbreathing propulsion could be of particular interest for

future reusable launchers in connection with rocket engines. High-speed ramjets

(scramjet and dual mode ramjet) are a key technology for these two kinds o f military and

commercial applications [3]. Missile applications exist that require the performance

benefits offered by the supersonic combustion ramjet (Scramjet) propulsion system [4].

The scramjet engine is expected to be the most effective propulsion system, but

there are severe conditions for ignition such as low pressure, low temperature, and high

airflow speed due to the supersonic incoming air [5]. Constraints on system size and

weight have led to the need to improve technology for analyzing and designing such

system. Considerable fundamental research has been conducted in response to the

increased interest in the development of scramjet propulsion systems. Many experimental

and numerical studies of various aspects of fuel injection in the combustor are discussed

in [6].1

The development o f hypersonic airbreathing plane capable o f horizontal takeoff

and landing and acceleration to low earth orbit has attracted significant interest in the

aerospace community for some time. It has a significant weight advantage over presently

used rocket propulsion systems by eliminating the need to carry onboard oxygen tanks.

One o f the major challenges in developing such a plane is the design o f a suitable

The journal model used for this dissertation Is the AIAA Journal.


2

propulsion system. Current airbreathing engine designs, which utilize subsonic internal

flow velocities, such as turbojet or even ramjet engines, would be inappropriate because

o f the extreme conditions seen in hypersonic flight. Temperature recovery from the

hypersonic free stream would be in the order of 2000K or more, leading to engine

materials difficulties and loss o f usable energy due to dissociation of air molecules. The

induced drag on the vehicle due to strong shocks at the engine inlets is a strong function

o f the flight Mach number so that it rapidly becomes impractical to use subsonic engines

at high speeds. A supersonic combustion RAMJET or SCRAMJET has long been

considered to be a feasible engine concept for hypersonic vehicles [7].

Although hydrogen is generally a more energetic fuel than hydrocarbon fuels (i.e.,

the hydrogen has a greater energy density), some hydrocarbons offer the advantage that

they can be liquefied without the use of cryogenic cooling and can also be contained

within a smaller volume. The potential for the combustion of storable liquid fuels in

supersonic airstreams for efficient propulsion engines at hypersonic speeds has been

recognized since the late 1950s [8,9]. The need to increase the fuel energy density for

hypersonic flight has long been recognized [10,11]. Liquid hydrocarbons are attractive

solutions for the low end of the hypersonic flight regime because of their higher

volumetric energy content and the relative simplicity of operational logistics [10,12].

For a Mach number range of 4-10, hydrocarbon fuels provide sufficient thrust and

are also being considered. The high density o f hydrocarbon fuels, as well as ease of

handling, make them very attractive for volume-constrained systems such as missiles and

hypersonic research vehicle (HRV) [13].

Most of the scramjet combustion work, currently being conducted, uses liquid

hydrocarbon fuels or hydrogen [14]. The Air Force’s Hypersonic Technology 9-years-

Program (HyTech) is developing a technology base for Mach 4-8 liquid-hydrocarbon

fueled scramjet propulsion systems. The goal of the HyTech Program is to develop and

demonstrate the operability, performance and durability of a Mach 4-8 hydrocarbon

scramjet propulsion system [15].

The investigation of the liquid hydrocarbon fuel-based scramjet combustor is one

o f the major purposes o f the hypersonic flight technology program. A primary problem to

be solved is the longer ignition delay for hydrocarbon combustion. As an example, under


3

the typical working conditions in a scramjet combustor, the ignition delay for kerosene is

5~10 ms [16,17], while the typical residence time available for complete mixing and

combustion is extremely limited to less than 2 ms in scramjet combustor [18]. Therefore,

injection and mixing enhancement play important role in designing scramjet combustor

[19].Because the missile applications impose demanding volume constraints, a strong

motivation exists for the development of a hydrocarbon-fueled air-frame-integrated

scramjet. Although studies o f supersonic combustion of hydrocarbon fuels have been

performed intermittently over the past forty years [18,20], they yielded only a limited

design database [4].

Although the energy/mass density of liquid hydrocarbon is lower than that of

hydrogen, some o f the mass increase would be recovered by a small and lighter structure

of the vehicle. Liquid hydrocarbon fuels require substantial residence time to achieve

vaporization and complete exothermic reactions, and such time is unlikely to be available

in reasonable-sized supersonic combustors. The chemical kinetics o f hydrocarbons are

slow in comparison with gaseous fuels, such as hydrogen. For many realistic operating

conditions, exothermic reactions cannot be achieved within the available residence time.

Use o f a pilot flame with fast kinetics (i.e., gaseous hydrogen) can provide locally, at the

liquid injection location, the conditions necessary to accelerate the hydrocarbon reactions

rates and reduce the Damkohler number, defined as the ratio between the characteristic

fluid residence time in the reaction zone and the chemical reaction time scale, i.e., high

temperature and low fluid velocities. For such a system it is necessary to verify: (a) the

level of piloted energy required to ignite and maintain stable combustion of the liquid

fuel, and (b) the interaction between the pilot flame and the heat sink represented by the

injected hydrocarbon [21].

Extensive literature survey that reviews the previous related research work is

reported in Chap. 2.

1.2 Challenges in Supersonic CombustionThe high-speed propulsion community has to orient more research activities to

address some of the technical issues, which must be resolved to realize the full potential

o f the scramjets. Among these issues are the fuel-air mixing enhancement, the


4

structuraUy-compatible injection systems, the fuel densification to increase the

energy/unit volume o f the fuel used, and the turbulence modeling that is capable of

predicting transition in the inlet, heat transfer, shear and mixing in the combustor, and

possible relaminization in the nozzle [22].

Future direction of supersonic combustion research is discussed recently by

Tishkoff et al. [6]. There is still serious question as to whether or not stable supersonic

combustion is possible over the range of expected operating conditions. The successful

development of the supersonic combustion ramjet (scramjet) for use on future hypersonic

vehicles is reliant on detailed understanding of the complex flow field present in different

regions o f the system over a range of operating conditions. To model chemical reaction

o f fuel and air in an engine, reduced kinetic models must be developed to reduce

computational time required for solving the species equations, particularly for

hydrocarbon fuels. To support hydrocarbon-based scramjet engine development, a

comprehensive data set for C7 — C n aliphatic fuel components under scramjet conditions

should be developed. The aliphatic fuels database should be utilized to derive suitable

starting and reduced mechanisms for candidate fuels. Changes in the state of the fuel can

affect the reactivity o f the fuel and the resulting combustion efficiency significantly.

There is a lack o f understanding of the physical processes that may contribute to these

effects. To understand these phenomena changes in the fuel state must be studied in a

realistic simulation of the scramjet preheating and combustion processes.

Fuel-air mixing, flame holding, pressure losses, and thermal loading are among

the major issues that need to be resolved for the successful design and implementation of

hydrocarbon-fueled supersonic combustion ramjet (scramjet) engines. A successful fuel

injection scheme must provide rapid mixing between the fuel and oxidizer streams,

minimum total pressure losses, and have no adverse effects on flame-holding capability

or thermal/structural integrity of the device. These requirements place somewhat

conflicting constraints on the design of a viable fuel-injection scheme, and solutions to

these problems are being actively sought internationally. A need exists for the

development o f a system that effectively integrates fuel injection and flame holding for

supersonic combustion. Such a device would contribute significantly to the present

research and industrial technology base [23].


5

la the recent work of Powell et al. [24], the many technological challenges that

are still under investigation in the Hypersonic Technology (HyTech) program and the

extent of advancement achieved in each area were presented. Among the discussed

challenges are: the starting condition and mass capture in the inlet, the pressure ratio in

the isolator, the piloting and flame holding, the fueL injection and mixing, the total

pressure losses in the combustor, the frozen flow losses, the divergence losses, the nozzle

thrust vector in the nozzle, the regeneradvely-cooled structures and thermal management,

the expected coking (deposition) o f the hydrocarbon fuels being exposed to high

temperatures during the endothermic reaction, the change in the composition of the

cracked hydrocarbon fuels associated with the change in the ignition delay time that

affects combustion stability, and the significant difference in engine operation during the

ramjet and scramjet modes.

Combustor scaling represents another source of challenge, which should be

seriously compromised by the non-linearity o f chemical reaction times and dissipative

processes. Therefore, it is critical that supersonic combustor testing be accomplished at or

near full scale [25].

For the supersonic mixing studies, it is suggested to investigate the effect of the

injection angle of the main fuel on the supersonic mixing and combustion characteristics.

It is recommended to study the effect of the transverse normal ethylene and propane

injections on the flame holding and combustion characteristics in order to investigate the

relationship between the injection scheme and the flame holding and flow field and

energy field features. It is suggested to simulate the supersonic combustion of the normal

ethylene and propane injections with the pilot gaseous hydrogen. For the future studies, it

is suggested to dedicate more effort trying to carry out the simulation o f the supersonic

combustion o f liquid kerosene with gaseous hydrogen as a pilot fuel. The radiative

interaction is also suggested to be included in the future studies due to the existence of

the hydrocarbon combustion products, which are radiative participating medium.

1.3 Objectives o f Present Study

The primary objective of the current study is to numerically examine the

possibility o f igniting hydrocarbon fuels in supersonic airstreams and maintaining the

hydrocarbon flames in such hostile environments. The secondary objective o f the study is


6

to investigate the effect o f the combustor configuration, the main fueL equivalence ratio,

and the pilot fuel equivalence ratio on the flame holding mechanism and the flow field

and energy field characteristics.

The combustion characteristics of gaseous propane in supersonic airflow using the

rearward-facing step that is swept inward form both end sides is studied. The effect of

sweeping the rearward-facing step from both end sides, in the cavity configuration, on the

flow field features o f propane combustion is investigated.

Next, an extensive study for the flame holding and mixing and combustion

characteristics of gaseous ethylene in supersonic airstreams is carried out. In this study,

different combustor configurations, different main fuel equivalence ratios, and different

pilot fuel injection equivalence ratios were employed. First, The effect of the combustor

configuration on the temperature flow field and the flow structure is investigated by

simulating two different combustor configurations, the rearward-facing step and the

cavity. Next, the effect of the equivalence ratio of the main fuel injection on the general

flow field features and energy field characteristics is studied. Lastly, the effect o f the pilot

fuel equivalence ratio on the flame holding mechanism, and the combustion and flow

field qualities concludes the study. The equivalence ratios examined are ranged between

0 and 0.08.

1.4 Work LayoutThe current study is presented in the following logical manner. An extensive

literature survey for the related previous work is compiled in Chap. 2. The theoretical

* formulations and code description are provided in Chap. 3. Chapter 4 presents and

discusses the results for the supersonic propane combustion while the obtained detailed

results for the supersonic ethylene combustion are analyzed and discussed in Chap. 5.

Finally, concluding remarks on the supersonic combustion o f hydrocarbon fuels in

scramjet engines are provided in Chap. 6.


7

Chapter II

LITERATURE SURVEY

2.1 Introduction

A supersonic combustion RAMJET or SCRAMJET has long been considered to

be a feasible engine concept for hypersonic vehicles [7]. Figure 2.1 shows a schematic of

a SCRAMJET engine concept where the aircraft underbody and engine designs are

integrated together [26].

The shock waves formed by the vehicle forebody and the engine inlets serves as

the engine compressor section while the engine exhaust and vehicle afterbody serve as

the expansion nozzle to generate thrust from the gases burned in the combustor. The

surface geometry would be chosen so that supersonic flow would be maintained

throughout the internal passages of the engine avoiding large recovery of temperature and

pressure from the free stream air [27]. The essential features o f two-dimensional or planar

geometry scramjet engines are diagrammed in Fig. 2.2 [26].

Figure 2.3 shows the Airframe-Integrated supersonic combustion ramjet

(scramjet) along with a partial cross-section through a typical modular engine. The

sidewalls of the inlet continue the compression process, which began on the vehicle

forebody. The instream struts complete the compression process and also provide

locations for global distribution of fuel. The particular concept depicted here shows a

combination of perpendicular and parallel fuel injection, which would typically take

place behind a rearward-facing step and at the base of the strut, respectively [28].

2.2 Injection Modes and MixingThe thermodynamic conditions at the test section entrance of the Scramjet engine

are representative of the conditions o f the lower end o f the hypersonic flight regime. For

these vehicles it is anticipated that the fuel be used to cool components o f the vehicle and

o f the engine and, thus, enter the test section at high temperatures (even a supercritical

conditions), resulting in fast vaporization and mixing upon injection. The low range of

hypersonic flight regime is, therefore, of particular interest, since the level o f heating of

the fuel will be relatively low, and thus longer residence time will be required for the fuel


8

to achieve complete combustion within the engine combustor. The mixing-combustion

coupling at these conditions is o f great interest [21].

Supersonic combustion ramjet (scramjet) propulsion systems rely on injection and

burning of fuel in very high speed flow fields. In such an environment the fuel remains

inside the combustion chamber for such short durations that rapid mixing o f the fuel into

the freestream air is essential for an acceptable design [29]. Rapid mixing ensures the

shortest possible combustor length and maximizes the heat release from the combustion, .

whereas a uniform distribution of fuel within the combustor optimizes combustion

efficiency and minimizes the amount of air processed by the inlet that does not participate

in the combustion. Clearly, an understanding of the fundamental fluid mechanics of

compressible mixing is essential to a successful supersonic combustor design [30].

Fuel-air mixing in an airbreathing engine becomes increasingly inefficient at

higher velocity, hence requiring a longer combustor length. Although this is caused by

the reduced flow residence time inside the combustor and the compressibility effect that

adversely affects the rate o f mixing [31-33], a short combustor length is desirable because

the thrust-to-drag ratio of an engine is roughly proportional to the ratio between the

combustor diameter and length [34].

Injection and mixing enhancement play important roles in successfully designing

a hypersonic air-breathing propulsion system. This is especially true when a liquid

hydrocarbon fuel is used. Several fundamental studies have focused on liquid fuel

injection from the walls o f combustors or from bluff body flame-holders into subsonic or

supersonic crossflow [35-37]. Several fuel injection schemes, such as angled injection

[38], non-circular nozzles [39], and ramp injectors [40], have been studied in attempts to

create deeper fuel penetration into the air stream for better mixing and to generate smaller

droplets of the liquid spray for faster evaporation [41]. Various mixing enhancement

schemes have been proposed including contoured injection orifices, flow field and shock

oscillations, baroclinic vortex generation, and wall-mounted vortex generators [23].

If the heat release by the combustion process becomes greater than a critical value

in a scramjet engine, the Mach number falls to unity and the flow becomes thermally

choked [42]. This choked flow may, in turn, cause a normal shock to form at the engine

inlet. This process, known as “unstart,” creates large amounts o f drag and radically


9

reduces the performance o f the engine at high flight Mach number [43]. The numerical

results of Moon et al. [44] suggest that there would be two possibilities to trigger thermal

choking, such as 1) strong turbulent mixing, and 2) instability of shear layer between

incoming air and fuel jet. McDaniel and Edwards [45] studied the dynamic simulation of

thermal choking in a nominally two-dimensional scramjet isolator/combustor

configuration.

The fuel-injection modes are used to control heat release. At lower speeds (below

Mach 5, and to some extent Mach 5-7, too much heat release too early in the combustor

will result in thermal choking and inlet unstart. Parallel fuel injection, which is shown to

have a slower-mixing process, is therefore used extensively to stretch out the combustion

zone. Above Mach 7, thermal choking is much less likely to occur, and the faster-mixing

perpendicular injection process is utilized to get faster combustion and higher

performance. In order to quantify this mixing-controlled combustion philosophy, a

considerable amount o f research has been done on perpendicular and parallel mixing

[24].

Different methods o f liquid fuel injection have been investigated. Parallel flow

mixes through the growth of the shear layer generated at the interface of the different

components. It is observed that compressibility plays a major rule in reducing the growth

of the shear layer, thus requiring longer fluid residence time to insure good mixing in

practical devices. As the compressibility increases, the spreading rate of the shear layer

was observed to drop to about one quarter of the compressible shear layer growth for the

same velocity and density ratios of the two mixing constituents. The convective Mach

number (defined as a measure o f the speed of the turbulent structures in the shear layer

relative to the free stream) increases the stability of the large-scale turbulent structures,

which are responsible for the turbulent shear layer growth. This is o f particular

significance in high-speed flows, where parallel injection is the preferred mode due to the

lower total pressure relative to the other fuel injection and mixing options. An increase in

the stability of the large vortical structures in the shear layers due to compressibility is

observed even in situation where the turbulence is intentionally intensified. Growth o f the

shear layer was forced by a pulsating shock impinging on the shear layer in parallel

compressible flows in high speed (i.e., Mach 3 and 5, respectively) as a destabilizing


10

mechanism leading to the growth of the shear layer and accelerated mixing. Only a

negligible effect was obtained both in the near and far field. However, when a shock of

the same strength impinged on the boundary layer o f the two flows prior to their

coalescence, an increase in the shear layer growth (14 to 26 percent) was experienced.

Thus, an increase in the turbulence in the boundary layer upstream of the mixing station

increased, causing an increase o f as much as 26 percent in the shear layer growth. This

indicated the significant role played by the small-scale turbulence in the mixing process.

Introduction of axial vorticity was shown to improve mixing, which was also attributed to

the small-scale turbulence effect on the growth of the shear layer and on the enhancement

o f turbulence in the shear layer. Mixing at a molecular level is, in particular, important in

combustion applications when stoichiometric mixing is a prerequisite to initiate chemical

reactions [21].

Mohieldin and Tiwari [46] studied numerically the advantage of using the tandem

injection on the regular single step tangential injection regarding the mixing aspect in the

flow field.

Studying the flow field behavior of the transverse injection o f the fuel is one of

the recent demanding research trends in the combustion field o f the scramjet engine. This

technique is used to increase the fiiel-air mixing in order to achieve the required heat

release pattern with the short combustion residence time associated with the high Mach

number condition [6].

Figure 2.4 shows a schematic o f the region surrounding a single perpendicular

injector. Because the mainstream flow is supersonic, there is a bow shock off the

underexpanded fuel jet and internal wave structure associated with the je t expansion. The

boundary layer ahead of the je t separates, and there is a recirculation zone downstream of

the injector as well. The details of these local zones are quite important to the ignition,

flame holding, and combustion processes and therefore important to real combustor

design [24].

Schetz et al. [47] presented a comprehensive review o f the mixing of transverse

jets and wall jets in supersonic flow. While the streamwise injection has the advantage of

adding to the thrust component o f the engine, many o f the approaches utilized to improve

the wall injection have been used successfully to enhance the fuel-air mixing [6].


11

A numerical study was conducted by Lee et al. [48] for the mixing and burning

augmentation o f the transverse injection in a scramjet combustor. Based on the fact that

the main factor controlling the mixing characteristics in transverse injection is the

backpressure around the injection hole, it was tried to make a flow expansion beside the

injection hole with a cavity in order to reduce backpressure around the front hemisphere

o f injection hole. It was concluded that the near field has convection dominated regime

whose mixing and burning processes are accelerated by convective flow, while the far

field has diffusion dominated regime whose mixing and burning processes are dominated

by mass diffusion.

The effect of the staged hydrogen fuel normal injection in supersonic air streams

was studied numerically by Drummond and Weidner [49]. In comparison with the single

normal injection, the staged normal injection produced a much larger region of separation

around the injectors. The separated region becomes significantly larger with increased

spacing, and the fuel-air ratio there moves closed to stoichiometric increasing the amount

o f reaction. Staged injections provided two important requirements for improved flame

holding, a large separation that increases residence time of the fiiel-air mixture, and a

more favorable fuel-air ratio for reaction. Increasing the spacing between staged injectors

was anticipated to improve these qualities even more.

The axially staggered laterally inline step-injector configuration was investigated

by McClinton [50]. Both the top- and bottom-wall blocks incorporate a rearward-facing

step located ahead o f the injection orifices. The top wall step is located upstream o f the

bottom-wall step. Four fuel injectors were equally spaced in each wall. The results are

compared with results o f previous investigations made in this combustor model by using

laterally staggered rather than opposed injectors. The opposed injector configuration

eliminated the adverse lateral pressure gradient experienced at the downstream injectors

in the staggered injector configuration and strengthened the gradient over that produced

by single-wall injection.

The experimental study of King et al. [51] showed that the combination of a

moderate pressure, sonic normal injection combined with a tangential slot flow is

preferred for the mixing enhancement over the tangential slot injection alone. This


12

scheme produces total mixing layer spreading angles up to 70% greater, and hence more

total mixing layer growth, than the slot injector alone.

Transverse or inclined injection introduces different mixing mechanism. At a

macro-scale, axial vorticity is produced at the edges and above the jet as the main flow

surrounds the jet in an axial and transverse direction. These vortices break the plume and

entrain the injectant into the core flow. The level o f mixing obtained by this strong

momentum exchange is traditionally quantified by the degree of penetration and

concentration decay of the jet injected into the main flow, usually air [2L]. Studies

performed by Mays et al. [52] and Wood et al. [53] addressed the effect of small-scale

turbulence on transverse injection and mixing. It was suggested that the onset and

divergence of instability in the jet is ultimately responsible for the plume fracture. This

instability propagates upstream and produces additional turbulence in the core flow. Even

in supersonic flows, an upstream interaction exists, as the instability at the jet boundary

induces oscillations of the bow shock generated in front of the jet via the separated region

in front o f the jet. In turn, these shock oscillations contribute to increase the vorticity in

the flow [21].

An important feature of the normal injection flow field is the gross penetration of

the jet into the flow. Baranovsky and Schetz [54] introduced the injection angle as a

separate parameter in the correlation, which was set to evaluate the penetration of liquid

jets. It was found that the upstream injection (injection angle is greater than 90° with

respect to the incoming air flow direction) can be used as a method to increase

penetration and residence time of a liquid fuel into a scramjet combustion chamber. It

was also found that the maximum penetration takes place at injection angle (0) =135

degrees.

Flow field and combustion/ignition characteristics o f a supersonic combustor with

a fuel injection strut were examined experimentally and computationally by Masuya et al.

[55]. Both normal and parallel injections were employed through the strut. Normal wall

injections were added to the strut injection. Pilot injections upstream of the strut and wall

injections were used. Plasma torch port is located on both sides o f the strut to force

ignition. Mixing and combustion efficiencies for a more disturbing strut were higher than

those for a less disturbing strut due to more severe deceleration and disturbance of the


13

airstream. Mixing and combustion efficiencies o f fuel injected from the struts were

almost linearly increased with an increase of the flow rate ratio of the perpendicularly

injected fuel to the total fuel. Fuel from the strut ignited at a lower air temperature than

that from the wall in both the autoignition and the forced ignition conditions. The twin

plasma torches with pilot injection could ignite perpendicular fuel jets of both paths

without time lag, and they could also ignite parallel jets from the base of the strut.

Ignition limits with the plasma torch were extended by increasing the pilot fuel flow rate.

Flame stabilization and combustion, along with wall friction and heat-transfer

losses, are the major processes of concern in the supersonic combustor in addition to the

fuel injection and mixing processes. Flame stabilization and combustion are intimately

tied to the fuel injection, vaporization, mixing, and ignition processes as well as

combustor geometry. As such, just about all of the experimental combustion data

available are configuration dependent. Thus, flame stabilization is achieved by using a

rearward-facing step and efficient combustion by initially eliminating and finally

restricting the acceleration (or expansion) rate of the flow to prevent quenching of the

reactions over a wide range o f initial conditions [11].

Using the rearward-facing step in the supersonic combustors and injecting the fuel

perpendicular or at least inclined to the main air flow have the advantage of holding the

flame because of the existence of the separated and recirculating regions in the flow field

as depicted in Fig. 2.4. This can help in solving, in part, the complication o f holding the

flame in supersonic engines due to the fact that the flow in such engines is not premixed

which means that if bluff bodies are to be used as fiame-holders they have to be large.

This is highly undesirable in supersonic combustors due to very high drag. Therefore, this

problem of flame holding could be solved using the rearward-facing steps in the

combustors and/or the perpendicular fuel injection. This issue emphasizes the importance

of flow field details around the fuel injectors [24].

Segal et al. [56] conducted an experimental study of transverse hydrogen injection

and combustion behind a reward-facing step into a Mach 2 high temperature clean-air.

The experiment was conducted in an electrically heated (not vitiated), continuous-flow

facility to evaluate the effects o f initial conditions and analyze the interaction between

mixing and combustion in supersonic reacting flows. This work and the earlier non


14

reacting data for the same combustor reported in [57] provide new experimental data for

supersonic combustion o f hydrogen using clean air. Matsuo and Mizomoto [5] studied

the flow structure of supersonic flow o f Mach 2.5 past a backward-facing step with

perpendicular sonic injection from the bottom wall. They examined the effect of both the

step height and dynamic pressure ratio of the je t flow to the main flow using the perfect

gas condition in two-dimensional space. The effect o f Hi-transverse injection on the

characteristics of the flow Held was investigated numerically by Berman et al. [58] using

both adiabatic and constant temperature wall boundary conditions.

The experimental study of Jian-Guo et al. [59] concluded that the 90° rearward-

facing step enhances the mixing pattern of the kerosene-hydrogen dual fuel system in

comparison to that of the 45° step. This is attributed to the lower recovery temperature in

the 45° step case, which leads to the poorer mixing of the pilot hydrogen and the

supersonic air. It needed more hydrogen to produce enough radicals to ignite the kerosene

spray.

The flow over a ramp induces a pair of counter-rotating stream-wise vortices as

shown in Fig. 2.5 [60]. Supersonic flow over a ramp generates both shock and expansion

waves. This leads to the generation of the vorticity by baroclinic torque. Baroclinic

torque is exerted on a fluid flow whenever the density and pressure gradients are not

parallel [61]. Numerous ramp designs have been investigated. In the work o f Stouffer et

al. [62] an array o f compression and expansion style ramps were investigated whose

geometry is shown in Fig. 2.6. In these studies, hydrogen was used as the fuel and air as

the oxidizer. In these concepts, rapid turns in geometry induce formation o f shocks,

which for compression ramps occur at the base and for expansion ramps at the trough.

Dramatic differences were found in mixing and combustion efficiency between these two

styles. It is well known that, for the same ramp angle, vorticity shed from a compression

ramp induces much stronger vorticity than that from an expansion ramps due to

differences in projected stream area. Despite significantly improved fuel/air mixing from

the compression ramps, the expansion ramps achieved higher combustion efficiency. This

result may be expected because good combustion efficiency is achieved with mixing at

small scales. The extremely strong vorticity o f the compression style ramp, while


15

increasing the fuel/air contact area also provides a centrifuge that diminishes opportunity

for mixing.

Ramp fuel injectors have also been designed with sweep to control shock and

expansion waves. Tests conducted at NASA Langley Research Center show that

combustion efficiency of swept ramp injectors is much improved compared to those

without sweep. The ramp with sweep approaches the combustion efficiency that is

achieved with normal injection. The effect of ramp sweep was numerically investigated

by Drummond, et al. [63] and Donohue et al. [29], who also performed experimental

planar laser induced iodine fluorescence measurements. Their calculations confirm the

importance o f using sweep for the ramp injectors.

There are a number of studies specifically related to the 3-D mixing flow field of

ramp injectors. Hartfield et al. [64] presented measurements o f injectant mole fraction in

the flow field o f a swept ramp with air injection using the same technique used by

Donohue [33]. The results illustrated the strong effect that ramp generated vortices have

on the mixing process. Waitz et al. [65] presented both numerical simulations and

experimental measurements of total pressure and fuel concentration for helium injection

behind an unswept ramp, which added the effect of baroclinic vortex generation when a

low density fuel is used. Riggins et al. [66] presented numerical simulations of the mixing

flow fields o f both swept and unswept ramps with hydrogen injection into high

temperature air and went on to calculate the reacting flow field as well using the SPARK

code. The study showed substantially higher mixing as well as flow losses for the swept

ramp case. It also showed increased penetration and spreading of the jet plume in the

reaction case. Daso and Gross [67] computed the flow field behind swept and unswept

ramps with air injection using the USA code and illustrated how side sweep of the ramp

significandy increased the mixing downstream. A numerical study designed to

investigate the relative importance of ramp generated vortices vs. baroclinic vorticity on

supersonic mixing behind swept ramps was performed by Donohue et al. [68].

An experimental investigation was conducted by Fuller et al. [691 to compare the

performance of the aero-ramp injector with a physical ramp injector to enhance mixing in

supersonic flow. The scope o f the investigation focused on jet penetration, mixing

characteristics, and total pressure losses. The aero-ramp exhibited a significant increase


16

in je t penetration when the jet-to-free stream, flux ratio was increased from 1.0 to 2.0;

however, the physical ramp showed very little change. The aero-ramp produced superior

mixing in the near field and slightly less than comparable mixing in the far field. Thus,

the mixing performance decreased with increasing jet momentum for the physical ramp

while it increased for the aero-ramp. The pressure losses induced by the physical ramp

were more severe.

A numerical parametric study of the cantilevered ramp injector for various

geometric parameters was conducted by Schumacher and Sislian [70]. Unlike the

conventional ramp injectors, the sweeping of the sidewalls o f the cantilevered ramp

injector does not translate into appreciably superior mixing performance, due mainly to

the increased strength of the reflected shock, which drives the fuel jet to the lower wall

into regions of low vorticity.

Cox et al. [71] proposed a new injector concept for secondary gaseous injection

into a supersonic crossflow to improve the fuel/air mixing. The basic idea is to arrange an

array of flush-wall injectors in such a way as to induce large vortical structures in the

main stream to increase entrainment and mixing. Both the experimental and

computational results they got demonstrated the creation o f large-scale vortices in the

mixing region, as intended.

Shock waves have been used to enhance mixing. The fuel is injected in the intake

region where the temperature and pressure are sufficiently low so that combustion does

not occur. By injecting in the intake region, the fuel has time to mix before its pressure

and temperature are finally raised as it enters the combustion chamber. This increase in

pressure and temperature is generated by passing a chock through the mixture.

Buttsworth [72] demonstrated that shock cannot only be used to promote ignition, but

they can also be used to generate vorticity as they cross the fuel-air mixing region, and,

thus, can be used to enhance mixing.

It is understood that there exists a strong interaction between mixing and

combustion, even in the limiting cases when one or the other becomes rate controlling.

Due to this coupling, observations from mixing in non-reacting flows are hard to extend

directly to reacting flows. Furthermore, in many practical situations, the Damkohler

number, defined as the ratio between the characteristic fluid residence time in the reaction


17

zone and the chemical reaction time scale, is 0(1). In such cases, the flow field may

become either mixing or kinetic limited in localized regions; however, in general,

combustion takes place in thick, highly turbulent burning regions. The mean and time

varying parameters responsible for mixing are affected by heat release, which, in turn,

affect the chemical kinetics; thus, mixing and combustion become closely coupled [56].

When heat release effects are included, significant modifications o f the turbulent

structures occur. It has been indicated that the heat released from chemical reactions both

generates and suppresses turbulence. It is shown that the turbulent dilation and viscous

dissipation in flames reduce small-scale turbulence, while the shear gradient effects,

mainly of pressure and density, generate new large-scale structures [21].

The effect o f small-scale turbulence on the mixing process in high-speed flows

requires more detailed investigation. In particular, mixing at a molecular level, which

indicates the ability to initiate chemical reactions, is of great interest. Concentration

distribution is not sufficient to provide this information. Although it is recognized that a

reacting flow field is substantially modified once exothermic reactions take place, mixing

of non-reacting flows needs further investigation to provide further insight o f the

characteristics o f the micro-mixing processes [21].

The effect of the chemical reaction on the flow field characteristics of the normal

injection of sonic hydrogen downstream of a rearward-facing step in supersonic airstream

was investigated in [73]. The comparison between the non-reacting and reacting flow

fields showed that the combustion heat release precluded the flow expansion around the

step edge; instead, the released heat was sufficient to cause a shock compression of the

main flow immediately behind the step. The reacting jet has a completely different

structure. In the non-reacting case, the je t has the characteristic barrel shape of an

underexpanded jet, ending with a normal shock. Because o f the heat release, the fuel jet

penetrate sufficiently deeper into the main flow, approximately 80% more than in the

non-reacting situation.

The effect o f the heat released by combustion on the fuel/air mixing distribution

in the supersonic reacting flow field was studied experimentally by Abbitt et al. [74].

Hydrogen was injected transversely as staged, underexpanded jets behind a rearward-

facing step into a ducted Mach 2 air free stream. The pressure rise associated with the


18

heat release involved in the reaction forces a rapid lateral displacement of the fuel

towards the side walls. A large amount of burning, therefore, occurs on the side walls.

This would suggest that the mixing distributions observed in the non-reacting flow field

was greatly altered. The wall limited the amount of spreading that would otherwise occur,

and possibly induced some recirculation or deceleration of the flow, which could aid in

flame holding.

Effervescent (or aerated-liquid, or barbotaged) atomization, characterized by

introducing gas bubbles directly into a liquid flow immediately upstream o f the injector

orifice to generate a two-phase flow, has been extensively studied for quiescent

environments [75-78]. Depending on the amount o f gases added into the liquid jet, two

primary mechanisms generate small droplets in the effervescent atomization process. It

has been observed that droplet size could be reduced as the amount of gas added to the

liquid flow increased. Based on this observation, there is considerable interest in

examining the role of effervescent atomization in enhancing the mixing and vaporization

processes in a scramjet combustor [79,80].

In their recent work on the atomization performance of aerated-liquid jets, Lin et

al. [81] applied the effervescent atomization technique to the problem of fuel injection

into a supersonic crossflow. The spray penetration heights of aerated-liquid jets were

studied experimentally using wide ranges of liquid properties, nozzle orifice diameters,

jet-to-air momentum flux ratios, and barbotage gas-to-liquid mass ratios. Two different

modes of spray structures for aerated-liquid jetrs in supersonic crossflows were identified

- pure liquid mode spray and barbotage mode spray. It was observed that the overall

breakup process for pure liquid mode spray is slow and that greater axial distance is

required to generate fine droplets. On the other hand, liquid breakup processes for

barbotage mode spray occur immediately after or even before the liquid is discharged.

This is primarily due to the highly turbulent features o f this inhomogeneous two-phase

flow and the expansion caused by high-pressure barbotage gas. These two modes o f spray

appear randomly and alternatively for low levels o f barbotage [41].

Avrashkov et al. [82] developed a system for the kerosene injection into a

supersonic flow. This system is based on two main principles: firstly, the use of tubular

micropylons, and, secondly, the saturation o f fuel with gas bubbles. The use of


L9

micropylon allows to arrange uniformly the fuel jets in the flow and to increase

considerably their number. Immediately before the liquid kerosene injection, the

saturation of fuel with gas bubbles (i.e., the bargotage) is made. On the whole, the use of

tubular micropylons and liquid fuel bubbling permit to increase significantly the process

o f mixture formation under conditions o f supersonic flow.

Critical issues regarding fuel injection and mixing in a scramjet combustor are

discussed in details in the literature [6,83,84].

2.3 Auto and Forced Ignition

Ignition delays of conventional hydrocarbon fuels must be significantly reduced

for successful supersonic combustion [85,86].

McClinton [87] showed that combination of the step-base and the perpendicular

injection downstream of the step was effective in enhancing the ignition ability. The

effect o f a variety of parameters on the autoignition behavior was studied. The free-

stream total temperature, wall temperature, and wall boundary layer energy deficiency

have the greatest impact on autoignition. Injectors located within three step heights

downstream of a step are shown to have very poor autoignition characteristics. But when

the injectors are moved downstream, autoigition behavior is improved. Sweeping the step

and fuel temperature produced only slight changes in autoiginition behavior. Injection

angle was shown to have a significant impact on autoigntion: upstream injection

improved ignition, and downstream angled injection impaired autoignition. Huber et al.

[88] suggested possible autoignition source in the supersonic combustors and proposed a

simple model to predict ignition limits in typical combustors. These candidates are the

base region behind steps and struts, recirculation regions upstream of the fuel jets injected

perpendicular to the airflow, and the stagnant region behind the bow shocks upstream of

the fuel jets. The study reported that autoignition of fuel from struts occurred easier than

that from the wall because the strut had a thinner boundary layer that resulted in a higher

surface temperature than the wall.

Tomioka et al. [891 investigated the flow field at pre-ignition phase in a

supersonic combustor with perpendicular injections behind a backward-facing step to

understand the mechanism o f the ignition enhancement observed in the ignition tests with

the same combustor. It was concluded that the interaction o f the separation region


20

upstream of the fuel j e t and. the two-dimensional step base-recirculating region, namely

the merging o f these two regions, caused enlargement of the ignition region and the

enhancement of the ignition ability. Ignition parameters were compared for the cases with

and without the step. As a result, the interacted region in case with the step was found to

be more preferable ignition source due to its enlarged size.

Autoignition o f hydrogen fuel injected into a scramjet combustor does not occur

at low flight Mach number and an igniter is required [87,90]. Several kinds o f igniters

[91-95], have been tested. Among them, a plasma torch is one o f the most promising

igniters for a supersonic combustor. It is not pyrophoric, toxic, or corrosive, and it is safe

and reliable to use in a propulsion system of the aerospace plane [94]. Kimura et al. [95]

first applied the plasma torch to ignite a fuel jet in supersonic airstream. Northam,

O’Brien, and their associates found that an argon-hydrogen plasma torch is an effective

igniter [96] and developed an uncooled long duration torch [97]. They showed a strong

sensitivity of the plasma igniter performance to the combustor geometry and/or fuel

distribution, and designed a new injector suitable for the plasma igniter [98,99]. Sato et

al. [100] developed and tested a new plasma torch with oxygen or air as a feedstock,

which may be advantageous from the view point of the total aerospace plane system.

The effects of various operational parameters of plasma torches as well as

combustor geometry and fuel injection location of the ignition of hydrogen were

experimentally studied in supersonic combustor by Masuya et al. [94]. The ignition

temperature was found almost linearly decreased as the input electric power increased.

Adding small fraction o f hydrogen to the argon plasma torch igniter was found to

increase its effectiveness. The chemical effect of the plasma torch was considered to be

the main reason of this improvement.

Two simultaneous ignition sources, a spark plug and a plasma torch mounted in

the cavity floor, were used in the combustion o f liquid JP-7 in the scramjet USAF

combustor used by Mathur et al. [41]. It was noticed that turning off the ignition sources

in the cavity did not have any detrimental effect on the combustion.

The issue o f the ignition delay of hydrocarbon endothermic fuels was investigated

by Hawthorn and Nixon [1011. They studied the ignition delay o f propane-air and

methycyclohexane-air mixtures and compared them to hydrogen-air and methane-air


21

mixtures. It was found that the ignition delays of propane and methycyclohexane fuels

are greater than those for hydrogen at temperatures below 1800 F, but significantly lower

than those for methane. For temperatures above 1800 F, it appears that these

hydrocarbons may have ignition delays comparable to that of hydrogen.

The ignition delay o f RP-l (kerosene) was reduced by adding an accelerant

(Alex® nano aluminum powder) as indicated in the work of Tepper and Kaledin [102].

This accelerant was found to reduce the ignition delay, while increasing the volumetric

energy density of the hydrocarbon, which is an added benefit.

Ignition characteristics of a scramjet combustor with a strut may be different from

those without a strut. The ignition characteristics of fuel from the strut were compared

with those of fuel injected only from the walls [90,94,100]. Installation of struts divide

the combustor into several flow paths, ignition o f fuel jets in each flow path would occur

independently. At least one igniter should be attached on each stream tube [55].

2.4 Vitiated and Clean Incoming A ir

In order to provide the proper ground-based facilities for testing the supersonic

combustors, the test stagnation conditions must be on the level that matches the flight

conditions. Therefore, it is necessary to heat the incoming supersonic air. Typical heating

techniques used to generate this high enthalpy include electric arcs, combustion of

hydrogen or methane fuels in air with oxygen replenishment, or storage heaters. The use

o f an electric arc heater results in air dissociation and the generation of significant

amount o f nitrogen oxides, as well as a depletion in the net level o f oxygen below 21%

by volume. In combustion heaters, the flow constituents are a function of the fuel used;

for example, with hydrogen burner H2O is a primary contaminant, or with a methane

heater a combination of both H2O and CO2 are the primary contaminants. Due to the

presence o f these test flow contaminants, the combustion characteristics in a ramjet or

scramjet engine can potentially be different than results obtained in clean air or actual

flight [1031.

Differences between testing in air and vitiated air could be due to differences in

thermodynamic properties and chemical kinetic effects. Thermodynamic effects include

differences between the heat capacities o f H2O and CO2 compared to N2. The heat

capacities for both HiO and CO2 are substantially higher than N2. Therefore, for the same


22

amount o f water formed, the temperature rise in vitiated air will be less than that in air.

The temperature rise in vitiated air would also be less than that in air due to dissociation

of some o f the additional H2O in the vitiated air. Chemical kinetic effects wilL also be

different due to the presence of CO2 and additional H2O in the vitiated air as well as

additional chemical kinetic reaction [104].

Various studies, [105-110], have explored the effect of NO, HiO, and CO2

contaminants on combustion processes, primarily with hydrogen fuel with application to

hypersonic propulsion systems. These include both experimental and computational

studies. A general trend from these works is that H2O and CO2 have a minimal influence

on the kinetics of hydrogen-air combustion while NO serves to reduce both ignition and

reaction times. The range o f temperature and pressure investigated and the database with

fuels other than hydrogen are limited. Lai and Thomas [103] investigated numerically the

effect of NO, H2O and a combination o f H2O and CO2 on combustion of various fuels

including hydrogen, ethane and methane with air. Srinivasan and Erickson [110] studied

the influence o f the effects of vitiation for a relatively simple flow configuration but with

a full 3-D computational flow model including mixing, frnite-rate reactions, and

appropriate thermodynamic properties. Srinivasan and Erickson [104] used the same

simple 3-D combustor model to simulate the influence of vitiated air, as compared to air,

in the combustion of hydrogen injected at an angle of 30° into a confined supersonic flow

with equivalence ratio approaching unity. Computations have been made for air and

vitiated air with the same set o f initial (except for the molar composition) and boundary

conditions. Many numerical works have reported the effect o f vitiated air via H2O and

radicals. Bakos and Morgan [111] conducted an experimental and numerical

investigation to study the effects of test air contamination by atomic oxygen on hydrogen

fueled scramjet combustion in hypervelocity environments with Mach 5 test flow.

In the past, many researchers have used vitiated air, numerically and

experimentally, to calculate ignition delay time, ignition temperature, and combustion

efficiency. Experimentally, vitiation often occurs in chemically heated facilities, in the

form o f water vapor or hydrocarbons reaction products. Mitani et al. [112] investigated

experimentally the effect of vitiation on the combustion performance of scramjet engines

a t Mach 6. Both Vitiated Air Heater (VAH, V-mode) and Storage Air Heater (SAH, S-


24

radical pool, which could decrease the overall timescales of hydrocarbon consumption*

Furthermore, endothermic fuel reactions can be expected to improve the fuel’s

combustion properties, if for no other reason than the fuel enters m a hot vaporized state.

Moreover, the mixing length o f a supersonic combustor is theoretically proportional to

the square root o f the molecular weight of the fuel. Therefore, the large amounts o f

hydrogen generated from endothermic reactions such as reforming might also be

expected to also improve the fuel’s combustion properties [119].

Quite a large number o f liquid fuels and fuel pilots for scramjets have been tested

in connected-pipe combustor and free-jet engine tests [120-124] in an attempt to find a

fuel that is energy-density efficient, would bum to near completion in the residence times

available (typically < 1.5 ms), and also be logistically suitable. Unfortunately, those fuels,

fuel blends, and pilots that did perform well were not logistically suitable, i.e., they

contained toxic, pyrophoric, or carcinogenic components that were unacceptable.

Monopropellant pilots, which are logistically suitable, were also tested, but could not

sustain the desired degree of heat release [125]. Thus, there still exist a need to develop a

high-energy density, storable liquid fuel or fuel that is both highly reactive and safe [14].

Silane (SiHO is a pyrophoric gas that can be added to hydrogen to decrease

ignition delay times of the fuel. Morris [126] made an experimental investigation into the

ignition limits o f different mixtures of silane and hydrogen. It was concluded that silane

is a useful additive to produce ignition in hydrogen when the intake temperature o f the

combustion chamber are below that where spontaneous combustion o f hydrogen would

normally occur. Furthermore, the addition of silane will also decrease the ignition delay

time.

2.5.1 The Problem of Cooling In High-Speed Vehicles and The Role of FuelAs air breathing-engine-propelled vehicle speeds increase, thermal problems

multiply because o f the effect of stagnation temperature. While total cooling needs

increase, the most critical regions are the leading edges and the engines. Although

thermal effects can be somewhat accommodated by improved materials and passive

cooling, sustained hypersonic flight in the atmosphere requires a substantial heat sink.

Compared to a mechanical refrigeration system or a non-combustible coolant, the fuel is

the best source o f cooling [127].


25

la the high Mach number region of the flight envelope (M>4), aerodynamic

heating is too great for conventional structural materials to survive without active

cooling. The fuel not only must have a good heat-of-combustion but also must provide

the necessary heat sink for the cooling system [128].

Higher aircraft speeds also have a direct impact on the operating environment a

jet fuel will encounter. The higher speeds mean higher air stagnation temperatures, which

increase the aircraft cooling requirements and prevents the use of air as a coolant. Thus,

increasing engine thrust-to-weight ratio and aircraft speed result in large heat loads thatt

must be managed with the main coolant available on the aircraft—the fuel [129].

The fuel in modem military aircraft is the primary coolant for on-board heat

sources. It is used to cool aircraft components such as the engine lubricant oil, hydraulic

fluid, environmental control system (ECS), avionics and electrical systems, and at high

Mach numbers, the airframe. As aircraft and engine technology have advanced from the

F-4 to tomorrow’s advanced fighter, the heat sink requirements of the fuel have increased

significantly. The engine is the main heat source for the aircraft at Mach 3 and below, but

the ECS is becoming an increasingly important part of the heat management problem.

Fuels used in high-speed aircraft will have to absorb large amounts of excess heat for

aircraft thermal management purposes. This will result in the fuels being heated to

supercritical conditions [129].

Fuel is circulated through hot sections of the aircraft, usually back into the tanks,

and finally to the engine where it is burned. This "recuperative" approach has the benefit

that thermal energy dumped into the fuel is then eventually recovered as additional thrust

[130].

2.5.2 Endothermic Fuels, The Promising Fuel for Hypersonic Combustion SystemsDefinitions of and differences between cryogenic and endothermic systems: The

"Cryogenic,r fuels are the fuels that have a negative, less than zero, "Boiling Point."

Therefore, in order to use these fuels as liquids, they must be cooled beyond the boiling

point and storage environments must satisfy this condition. Examples for the cryogenic

fuels and their corresponding boiling point temperature are the liquid hydrogen H2 (-423

F) and liquid methane CH4 (-259 F).


' £ -■ f - ' r ' - \ g ' r • 7-

26

The "endothermic" fuels are the fuels that extract heat while forming from their

initial components. Therefore, the decomposition of these fuels into the initial

components requires energy for its accomplishment and by that a quantity of heat can be

absorbed from the surrounding environment causing the required thermal loading

management of the vehicle. An example for the endothermic fuels is the

methylcyclohexane MCH (C6H1CH3). The boiling point of MCH is 213 F, which means

that it is a liquid in the normal ambient conditions.

Crvoeenic fuels: advantages and disadvantages: The cryogenic fuels have high energy

contents and thermal tolerance. Figure 2.7 shows that hydrogen has almost three times

higher energy per unit mass than other fuels [130].

The temperature limit of a fuel is a key consideration for its selection as a high

speed transport fuel. Figure 2.8 shows that hydrogen has the highest temperature limit

over other fuels [130]. On the other hand, Fig. 2.9 indicates that cryogenic fuels have

relatively poor energy contents per unit volume, which is an undesirable property that is

magnified by their storage requirements [130].

Cryogenic fuels must be viewed with extreme caution because of the severe

constraints that they place on the design of the aircraft and because of the high cost of

transporting, storing, and delivering them. Hydrogen or methane handling facilities at

just a few U.S. airports would be very costly to install and. maintain, and of course these

fuels raise serious safety questions. From the military standpoint, cryogenic fuels appear

impractical because, again, military operations must not be tendered by a need for costly

and exotic infrastructures, and by all means not by an uncertain and vulnerable supply of

mobility fuels. Nevertheless, serious consideration is being given to such option and it is

possible that successful large-scale applications using hydrogen or methane might

eventually be developed [130].

Endothermic fuels: the phenomenon and advantages: An innovative approach based on

more conventional types o f fuels takes advantage of the fact that some hydrocarbons,

when they degrade at high temperatures, do so cleanly, i.e., without forming

carbonaceous surface deposits. Instead, they break down to smaller hydrocarbons that are

then circulated to the combustor. The chemical breakdown itself absorbs heat, thus

providing another useful thermal management process. A number of these "endothermic"


27

fuels are known, and some are under serious study. All of the known endothermic fuel

reactions require a catalyst to make them proceed and absorb extra heat at moderate

temperatures. Hence, variants such as the catalytic heat exchanger are being considered in

future aircraft concepts. While the endothermic fuel technique introduces a new element

of complexity into aircraft design (and particularly new maintenance requirements and

reliability concerns), it has obvious advantages over the use o f cryogenic fuels. There are

indications that, free of certain environmental and chemical influences, most

hydrocarbons tend to have good high-temperature tolerance [130].

Cryogenic fuels can contribute only sensible and latent heat, whereas certain

hydrocarbon fuels can in addition provide cooling through endothermic reactions.

Hydrocarbons can undergo both thermal and catalytic reactions. Theoretically, the total

heat-sink capacities for hydrocarbon fuels range from about 50% to 112% of the cooling

capacity of hydrogen (on a "%" heat of combustion basis) with laboratory proven

capabilities up to about 85 % [127].

One example of an endothermic reaction is the dehydrogeneration of

methylcyclohexane to toluene and hydrogen as shown here:

C-jH h ->C7Hs +3H, (2.1)

This reaction, which has a chemical endotherm of 940 Btu/lb, was identified as a

potential source of cooling in earlier investigations of endothermic fuels. However, a

catalyst is required for this reaction to occur at a rate that will produce the necessary

cooling. Catalysts composed of platinum supported on high surface area alumina have

been shown, to work well in this application [131,132].

The cracking reactions, which convert high molecular weight paraffins into

mixtures o f lighter olefins and paraffins, can be conducted either with or without solid

catalysts. Catalytic cracking is carried out at relatively low temperatures, 400-550°C, in

the presence of acidic compounds such as zeolites [133]. Thermal cracking, on the other

hand, is a gas phase reaction and requires higher temperatures, above 500°C, for the

reaction to occur at useful rates. Although catalytic cracking and thermal cracking

reactions both produce a mixture of low molecular weigh hydrocarbons; they occur by

different mechanisms and result in different product distributions. The catalytic

mechanism occurs by way of carbenium ion intermediates on the catalyst surface and


produces high concentrations o f C3 and C* products and very few Ci and Cz products. On

the other hand, the thermal cracking reaction occurs by a free radical mechanism and

results in high concentrations o f C2 and C3 compounds [134].

h i the laboratory investigations o f Wickham et al. [134] the effect o f adding a

chemical initiator on the rate of thermal cracking o f both normal and isoparaffin fuels. It

was found that at all temperatures the addition of up to 2 wt% percent of the initiator

produced measurable increases in cracking rate. For example at 500°C, this initiator

increased the rate of n-heptane cracking by a factor of six, whereas at 550°C the initiator

produced over a factor o f two improvements in rate. It was found that the initiator only

accelerates the cracking rate and does not alter the overall thermal cracking mechanism.

The thermal decomposition o f three high-energy density hydrocarbon fuels; RIM,

JP-10, and quadricyclane was examined by Wohlwend et al. [135]. Stability

measurements performed in this study indicated that RP-a was the most stable at the

conditions investigated, followed closely by JP-10. Quadricyclane is the least stable,

degrading at relatively low temperatures compared to Jp-10 and RP-1. It was concluded

that quadricyclane would not be a capable pf providing significant fuel system cooling

(due to its thermal instability) if reaction times or temperature were excessive.

The endothermic fuels have the following technical, commercial, and safety

advantages:

Continued availability

Relatively inexpensive

Condensed phase under standard temperature and pressure (STP)

Higher latent heat o f vaporization

Safe handling and storage

High energy content per unit volume

No ignition limitations under subsonic combustion conditions

Non-cryogenic;

Insulation not required

Vapor recovery not required

Conventional Army/Commercial (A/C) fuel system

A/C turn-around time not governed by fuel

permission of the copyright owner. Further reproduction prohibited without permission.

29

Conventional fuel handling/logistics;

Off-site production

Launch side hardening easier

Available enabling technologies

Potentially smaller vehicles

Hydrogen has very low density, and it has boil-off problems, leading to packaging

challenges that derive up structural mass and required overall volume for a given mission.

Cryogenics such as hydrogen and methane will also have obvious handling problems. In

contrast, hydrocarbons are slower burning and have between two-fifth and oneOthird the

energy per unit mass compared to hydrogen; however, with up to 11 times the storage

density, they have over 3.5 times more energy content per unit volume [136]. The

properties of various hydrocarbons are summarized in Table 2.1, contrasted with

cryogenic hydrogen in both liquid and slush form, [137-140]. Hydrocarbons contain

approximately 15-20% hydrogen, and so they actually store hydrogen at up to twice the

density of pure cryogenic hydrogen [141].

Conventional hydrocarbon fuels can be divided into two basic categories:

methane, with energy density of 50 MJ/kg and specific gravity of about 0.4 (and also

cryogenic storage requirements), and the JP hydrocarbons, with nearly twice the density

but only 40 MJ/kg specific energy content. Interestingly, even endothermic fuels such as

methylcyclohexane (MCH) have about the same storage properties and combustion

energy content as the JP fuels, through they may have advantages in recuperating

combustion heat. Hydrogen as either liquid or slush has about the same energy content,

with a slight difference due to the heat o f fusion, but slush has about a 15% higher

density than the liquid form.

Clear technological, economical, and operational advantages exist when liquid

hydrocarbon fuels, such as kerosene, are used in comparison with hydrogen-based

systems for the development of small hypersonic vehicles. However, the flow features in

a supersonic ramjet combustor, primarily the short residence time of a fuel-air mixture,

which is <10'3 S, along with the multistage physical-chemical mechanism of liquid

hydrocarbon fuel burning, increase the difficulties of ignition and flame stabilization. If

the selected fuel is amenable to operate at supercritical conditions, a significant decrease


30

in the time required for the liquid fuel breakup and vaporization can be achieved. Further,

if chemical decomposition accompanies these transformations, the potential formation of

hydrogen or other active radicals will result in increased reactivity o f the mixture and a

reduction in the combustion length [142].

Moreover, endothermic fuel reactions can be expected to improve the fuel’s

combustion properties, if for no other reason than that the fuel enters in a hot, vaporized

state. Furthermore, the mixing length of the supersonic combustor is inversely

proportional to the gas diffusion coefficient [22]. The gas diffusion coefficient is in turn

inversely proportional to the square root of the molecular weight of the diffusing

components. Therefore, the large amounts of hydrogen generated from endothermic

reactions such as reforming might also be expected to improve the fuel combustion

characteristics by lowering the average fuel molecular weight [143].

The use of hydrocarbon fuels in volume limited systems will require an ignitor

and some form of combustion enhancement. One approach is to use part o f the air to bum

all o f the fuel in a subsonic pre-burner. Another approach is to use form of pyrophoric

material as an ignitor at low temperatures and a flame-holder or combustion enhancement

aid at high temperatures where residence times become shorter and kinetics dominates

the flame propagation [14].

Over the past 15 years a significant amount of experimental research work has

been done on the injector performance and combustion stability using kerosene (RP-1)

as the fuel of rockets. This work encompassed a large variety of injection schemes,

chamber dimensions, and operating conditions [144].

In the analysis conducted by Lewis [141] the fundamental question of whether the

packaging benefits o f high-density hydrocarbon fuels outweigh the high-energy content

of hydrogen fuels. For a cruiser operating in the Mach 8-10 corridor, in which it is

desired to maximize cruise range for a given total takeoff weight, the results seems to be

that the aerodynamic and volumetric advantages o f storable hydrocarbons are about

equivalent or superior to the specific impulse (Isp) advantages of hydrogen in determining

cruise range.

In the work of Edwards and Maurice [119], several hydrocarbon fuels/fuel

systems challenges for hypersonic cruise and space access vehicles were discussed under


31

the HyTech program. These challenges include extension of fuel heat sink capability,

improvement of hydrocarbon combustion properties, fuel system fouling mitigation, and

fuel system integration.

Heneghan et al. [145] developed a high-thermal-stability JP-8+100 hydrocarbon

fuel, which provides a 55°C (100F) increase in the bulk maximum temperature (from

325F to 425F) and improves the heat sink capability by 50-percent over conventional JP-

8 fuel.

A general purpose CFD model has been proposed by Zhou and Krishnan [146] for

the heat/mass transfer analysis o f endothermic fuel flows under high pressures and

heterogeneous catalysis mechanisms. This model was incorporated into a multi

dimensional CFD code, CFD-ACE. The model was demonstrated for the catalytic

endothermic reaction of MCH dehydrogeneration to Toluene.

Special attention is directed towards the simulation of the combustion flow field

of gaseous methane and gaseous ethylene which are considered to be the light-

hydrocarbon fuels resulting from the cracking of the heavy-hydrocarbon fuel, kerosene.

Ethylene C2H4 is a primary fuel itself and is also produced in large amounts

during the combustion of methane CH4, ethane C2H6 and other higher hydrocarbons

[147]. Ethylene is chosen in the hydrocarbon-fueled scramjet engines because it is used

as a surrogate test fuel for hydrocarbon fuels.

Sun et al. [148] investigated experimentally the effects of the equivalence ratio

and static temperature of premixed air-gasoline mainstream on the flame propagation

speed and ignition delay.

In the experimental work of Vinogradov et al. [149], liquid JP and gaseous

ethylene were injected in the isolator upstream of the combustion chamber in a Mach 1.6

airflow behind a thin pylon to determine the ability o f the pylon to improve penetration

and mixing. The liquid fuel injection behind the pylon increased the fuel penetration to be

entrained by the high velocity airflow core resulting in improved mixing, offering the

possibility of reduced combustion chamber length. Injecting behind the pylon increased

the normalized wall pressure distribution indicating higher rates o f heat release in the far

field as an indication of improved combustion efficiency. The behind-pylon ethylene

injection indicated a substantial increase in penetration over the basic wall injection case.


33

2.6 Flame Holding Mechanisms For Liquid Hydrocarbon CombustionFuel-air mixing, flame holding, pressure losses, and thermal loading are among

the major issues that need to be resolved for the successful design and implementation of

hydrocarbon-fueled supersonic combustion ramjet (scramjet) engines. A need exists for

the development of a system that effectively integrates fuel injection and flame holding

for supersonic combustion [152]. A study by Yu et al. [153] in an unheated Mach 2 flow,

with fuel injection upstream of a variable cavity length-to-depth ratio (L/D), suggested

that small-aspect-ratio cavities provide better flame holding capability than longer

cavities with inclined aft ramp angles.

In the experimental study of Mathur et al. [152] the cavity-based flame-holder

with low-angle flush wall fuel injection upstream of the cavity was successfully

demonstrated in a model scramjet combustor using gaseous ethylene. The study showed

that ignition and combustion produced a precombustion shock train, resulting in dual

mode combustor operation. The shock train became stronger and the starting location of

the shock train moved progressively upstream with increasing fuel-air equivalence ratio.

The cavity-based flame holding concept proved very effective over a wide range of

operating conditions and combustor fuel-air equivalence ratios.

The cavity can be used to set up a pilot flame reducing the induction time

associated with ignition of fuel-air mixture. Such a method can be useful for maintaining

stable flame holding at low equivalence ratio in scramjet operation or minimizing the

effect of shock trains. In the work of Situ et al. [154], the preliminary tests o f fuel-rich

hot gas as reacting jets were performed without liquid fuel jets. A dual cavity was

employed as the main flame holding mechanism. The pilot energy of fuel-rich hot gas

injected into the supersonic combustor was produced by a dependent pre-burner. The

kerosene in the pre-burner was ignited to produce fuel-rich hot gas, which was discharged

via six injectors divided equally into two groups: right upstream of the first cavity pointed

at 35 degrees with respect to the bottom wall in flow direction, and at the bottom wall of

the first cavity parallel to the main stream respectively.

Another fuel-riched hot gas combustion study was conducted by Situ et al. [155].

The major goal o f the study was to investigate the characteristics o f kerosene-fueled

supersonic combustion in dual-combustors. An experimental study was carried out to


34

observe the effect o f the fuel air equivalence ratio on supersonic combustion efficiency

and total pressure recovery coefficient, and to investigate where and under which

conditions the flame zone is supersonic* Liquid kerosene was injected into a semi-

dependent subsonic dump combustor* Then, the fiiel-riched hot gas was injected in

parallel to the freestream through a slot nozzle injector located at the bottom wall o f

supersonic combustor.

The results o f the early tests conducted by Kay et al. [156,157] clearly

demonstrated that the supersonic combustion of various hydrocarbon fuels could be

achieved; although for many test conditions, special externally mounted piloting devices

were required to initiate and stabilize the flame.

Bonghi et al. [158] injected liquid toluene at 5H down stream o f a step of height

H in a Mach 2.5 flow at 300 K and 1000 K using a parallel hydrogen-pilot flame.

Toluene (C6H5CH3) is on of the components of catalytically cracker methylcyclohexane

(MCH), a candidate fuel for supersonic combustion applications. Segal and Young [21]

using the same previous configuration injected the toluene from a reservoir at 300 K, and

thus, absorbed energy from the surroundings to vaporize and further heat to the local

temperature. The large recirculation region formed by hydrogen-pilot combustion extends

beyond the liquid injection station, 5H, and thus a substantial amount of the injected

liquid-reaches the region of stabilization of the pilot flame. When the amount of liquid

increases above a certain quantity, it quenches the pilot flame. The amount of liquid fluid

that can be injected without inducing quenching was found to be dependent strongly on

the equivalence ratio of the pilot flame and, in smaller proportion, on the air stagnation

temperature. It was found that as the pilot energy level increases the amount of toluene

that can be injected (while maintaining a stable flame) can also be increased. Beyond a

certain value, however, the ratio of energy levels decreases again, as more liquid is

injected and the amount that reaches the region o f stabilization o f the pilot flame

increases, inducing quenching. This effect is enhanced by the effect o f heat release on the

structure of the flow field. A larger deposition o f heat in the combustion region caused a

pressure increase followed by an enlargement o f the recirculation region and a reduction

in the local velocity, which favors upstream interaction, further facilitating the arrival of


35

liquid in the recirculation region and quenching of the pilot flame. The increase in the

stable flame margin with the air stagnation temperature, T0, was evident.

Vinogradov et al. [159] used various combinations o f strut and wall injection of

both pilot-hydrogen and liquid kerosene to obtain stable combination at kerosene

equivalence ratios as low as 0.6 and pilot equivalence ratios as low as 0.1 for Mach 6

flight conditions. Bonghi et al. [158] evaluated the piloted energy needed to maintain

stable combustion of liquid toluene using a rearward-facing step as the main flame

holding mechanism with a gaseous hydrogen-pilot flame in a Mach 2.5 airflow. Kay et al.

[4] used a gaseous ethylene pilot to stabilize the combustion o f a primary fuel, injected

both upstream and downstream of the pilot, for a Mach 3 airflow.

The piloted-supersonic combustion experiments strongly recommended that

hydrogen piloting is highly effective for a variety of fuels including methane and

kerosene. A monopropellent, OTTO fuel, was successfully used as an effective pilot in

the scramjets [14]. Other results showed that silane/hydrogen mixtures were effective

pilots for ethylene and kerosene combustion [20].

It was shown in the work of Chang and Lewis [160] that piloting the Jet-A-fueled

scramjet engines with silane is beneficial in decreasing the combustor length necessary to

complete and in reducing the drag. The addition of 10% silane to jet-A fuel reduces the

reaction length to 5% of the original length required at 1 atmosphere, 1000 K, and 2000

m/s.

In their experimental study, Kay et al. [4] used three wall-mounted pilots, which

serve to initiate the combustion process and stabilize the flame. Staged combustion,

whereby the engine fuel is injected at three axial stations corresponding to the pilot

location and two downstream stations, is employed to distribute the combustor heat

release. This leads to high combustion efficiency while controlling the combustor

pressure distribution and preventing combustor-inlet interaction that could unstart the

inlet. Also, by tailoring the fuel distribution, it is possible to operate as a dual-combustion

scramjet where the combustion process is either all supersonic or mixed

supersonic/subsonic, depending on overall equivalence ratio and flight Mach number.

They used gaseous ethylene to simulate the ignition characteristics o f vaporized Jet-A

(JP-5) fuel and limited tests were made using Jet-A. Supersonic combustion tests were


3 6

performed using gaseous ethylene fuel with primary fuel injection along and with staged

fuel injection over a wide range o f spacings between the primary and secondary injection

stages. In addition to the ethylene-fueled combustor performance tests, they performed

significant combustor testing using Jet-A, a liquid hydrocarbon, as the primary fuel while

the pilot injected as a liquid from the parametric piloting test hardware. In later tests, the

fuel-cooled piloting hardware was used and the fuel was preheated within the pilot

cooling passage, as it would be in an actual engine, to a thermodynamic state such that it

flash-vaporized upon injection into the combustor.

Jian-Guo et al. [59] investigated the effect of the hydrogen pilot injection on the

ignition of liquid kerosene in Mach 2.5-supersonic combustor. They concluded that

without additional ignitor, a proper combination of pilot hydrogen and recessed flame-

holder may ignite kerosene and maintain sustained combustion. The minimum

equivalence ratio for pilot hydrogen to sustain combustion was found to be 0.03. The

effect o f increasing the total temperature for the incoming supersonic air from 1470 K to

L700 K was not found of a significant effect implying that the characteristics evaporation

time for kerosene is smaller than the mixing time and characteristics reaction time. The

evaporation o f kerosene is not the controlling factor. They also concluded that the 90°

rearward-facing step enhances the mixing pattern of the kerosene-hydrogen dual fuel

system in comparison to that of the 45° step. This is attributed to the lower recovery

temperature in the 45° step case, which leads to the poorer mixing o f the pilot hydrogen

and the supersonic air. It needed more hydrogen to produce enough radicals to ignite the

kerosene spray.

Through the experiments of Yu et al. [161] the combustion of kerosene in two

supersonic test combustors using pilot hydrogen was studied. The integrated modules of

pilot and flame-holder using recessed cavity with different geometry were designed and

tested. Effect o f pilot injection scheme, cavity geometry, and combustor scaling on

minimally required pilot hydrogen equivalence ratio were systematically examined.

Results show that the stream wise pilot injection is not the most effective scheme for the

promotion o f kerosene ignition. Also, the cavity depth and length have significant effect

on ignition and flame sustaining, while the slanted angle of aft wall is less important.


37

In an effort to reduce the combustor length, Yu et al. [34] conducted experimental

study to evaluate the flame holding and mixing enhancement characteristics of supersonic

reacting flow over acoustically open cavities. Several wall cavities having various size

and aspect ratios were subjected to supersonic open-flow flame experiments. The results

showed that cavities with a short aspect ratio provided good flame holding, whereas those

with a relatively long aspect ratio shortened that flame length substantially via acoustic

excitation. In all cases that utilized cavities, the combustor pressure and the exit recovery

temperature were increased suggesting enhanced volumetric heat release. The subject of

the cavity flows and their relevance to flame holding in supersonic combustion engines

was surveyed in [162-164].

An unsteady flow analysis on the effect of cavity length to depth ratio and cavity

aft wall angle has been studied by Baurle et al. [16S] to determine the important

characteristics that influence cavity flame holding effectiveness.

Bumes et al. [166] investigated the use of combustor wall cavities not only to

provide stable flame holding but also to enhance fuel-air mixing in supersonic

combustors. Several different configurations of acoustically open cavities were placed

inside a supersonic combustor. For certain cavity configurations, supersonic flow over

the cavity generated large coherent structures via flow-induced cavity resonance. On one

hand, fuel injection directly into the cavity suppressed the flow oscillations even in the

case o f the most unstable cavity and inhibited coherent structure formation downstream.

This injection scheme would be ideal for cavities that will function as stable flame-

holders. On the other hand, fuel injected into the wake of the cavity was entrained into

the coherent structures, and showed the evidence of mixing enhancement. The presence

of large coherent structures periodically occurring in the wake o f the cavity showed that

turbulent compressible mixing in an enclosed supersonic duct is significantly affected by

flow-induced cavity resonance. This opens up the possibility o f organizing the coherent

“vortical” structures in a scramjet flow field and controlling the combustion inside the

vortices to shorten the combustor length and improve engine performance.

The results o f Gruber et al. [167] indicates that the aft ramp angle plays a strong

role in determining the character of the shear layer that spans the length of the cavity.

Reduction in the aft ramp angle to below 90° yields more stable, two-dimensional flow


38

fields. Reductions in the aft ramp angle result in higher drag coefficient and shorter

residence times within the cavity.

A computational study of an ethylene-fueled scramjet combustor employing an

aerodynamic ramp fuel injector and a cavity flame-holder was performed in the work of

Eklund et al. [168]. It was found that the position and strength of the three-dimensional

pre-combustion shock train and the combustor heat release distribution were strongly

coupled, which complicated the modeling of the flow field.

2.7 Importance of Modeling and SimulationThe supersonic combustion ramjet (scramjet) is expected to be the most effective

propulsion system for the Single Stage To Orbit (SSTO) transportation vehicles and

hypersonic transportation vehicles o f the next Generation. Many studies on scramjet

components such as inlet, combustor, and nozzle have been carried out to obtain a better

understanding of the performances o f the individual components, [169-172]. However, it

is expected that intensive interactions among these components will occur in real engines.

For example, a rise in pressure due to combustion causes separation which propagates

upstream into the inlet and changes the inlet back pressure, which may result in unstart

condition in the inlet. In the case o f interactions, the total performance of the engine

cannot be evaluated based on the linear combination of the obtained performance of the

individual components. Thus, tests of the models of the whole engine are necessary to

elucidate the interactions among these components and the overall performance o f the

whole engine [173]. Testing of the models of whole engine requires rather big and

expensive wind-tunnel facilities. Thus, only a limited number of results on whole engine

tests have been reported [20,117].

The design o f future hypersonic propulsion system will depend heavily on

computational fluid dynamics (CFD) because o f the difficulties associated with testing

combustors in ground-based facilities. If used properly, CFD can be a strong tool in

conjunction with experiments. Computational solutions have the great advantages that

aspects o f the flow field can be switched on and off to isolate the mechanism of an

observed phenomenon. For instance, chemistry may be turned on in a computed flow to

determine its influence, boundary layers may be switched from laminar to turbulent, and

perturbations can be introduced in upstream calculations to isolate cause and effect.


39

Computational solutions also permit a flow field to be assessed without intrusion, and

numerical measurement can be made to the resolution of the grid. This capability has

been used very effectively to identify details o f shock-shock interaction, shock-boundary-

layer interactions, supersonic shear layers, and combustion systems that could not

possibly be measured experimentally. A particularly promising example is the used of

CFD investigations of ionized or electrically charged gases to explain reported

aerodynamic and propulsion benefits [25].

Multiphase combustion is such a complex phenomenon that it warrants focused

research using modeling and numerical simulations. Though physical experiments are the

ultimate test to study the performance of combustion/propulsion systems, they are often

extremely expensive and complicated, and at times are not even possible. This is due to

the hostile environments, complicated interactions, and couplings. On the one hand, these

couplings make it impossible to study the effect of one parameter at a time in a physical

experiment o f this nature. On the other hand, numerical experiments open an avenue for

isolating and understanding the influence o f different parameters on complex combustion

situations. Furthermore, these can serve as means o f evaluating scaling laws, which

would be extremely difficult and expensive to determine from physical experiments

alone. Modeling also provides a potential design tool for future combustion and

propulsion systems. However, this is not a simple or straightforward extension o f

computational fluid dynamics (CFD) by including a few more terms in the governing

equations [174].

Numerical simulation is time consuming even with ultrafast modem computers.

The fuels of choice, e.g., hydrocarbons or metal slurries, will have a large number of

kinetic steps and species involved in their combustion. The computational time increases

as the square of the number o f species, and calls for reduced chemistry based on sound

logic. The presence of solid particles, such as soof or- unbumed fuel droplets, and their

transport in the turbulent flow field complicates the situation even further, and requires

extremely small grid sizes. Ultimately, the results of these complex numerical

simulations should be made available in the form o f user-friendly models for the

designer.


40

Strong interaction of the computational combustion community with

experimentalists is needed in order to assure meaningful model validations, and to

generate more reliable and advanced models to aid in the design of complex

combustion/propulsion systems of the future.

Computational fluid dynamics-(CFD) simulations have become invaluable in the

design process due to the high costs and long turnaround times encountered in

experimental programs and due to the inability to reproduce all desired flow field

conditions in present laboratory facilities. Fundamental numerical techniques for the

calculation of compressible flow fields have been available for some time, but the

computational resources required for making accurate 3-D simulations of complex flow

fields have become more generally available for design purposes only relatively recently.

This is, for the most part, due to the many dramatic improvements that have occurred in

the performance of computer hardware, but is also due to progress in the development of

efficient and accurate numerical techniques and algorithms. Advances in areas such as

grid generation, development of higher order schemes, and turbulence modeling make

reasonably accurate simulations of extremely complex flow fields now possible. It is

important to point out that if CFD codes are to be incorporated into a design process, the

need for experimental validation is a necessity, a point that is sometimes under-

emphasized [23].

One of the advantages of CFD simulations over standard experimental data sets is

that all flow field properties are available everywhere in the flow field so that any desired

quantity can be extracted from the CFD results. Both detailed information about

particular flow field features and global information, such as conservation consistency

checks and overall performance quantities, can be calculated from the numerical data set

[23].

The design of fuel injection configurations like the ones considered in the

scramjet engines rely on the understanding of extremely complex flow fields with

features such as: turbulent mixing, combustion reactions, strong three-dimensionality,

compressibility effects, boundary layers, and flow separations. The ability to study such

flow fields experimentally has been limited because of the difficulty in gaining access

into the harsh, yet easily perturbed environment o f a supersonic flow field. Understanding


41

the geometrical complexity o f such flows has been limited by the fact that most

traditional quantitative measurement techniques are pointwise techniques. Theoretical

studies in the past have been stalled by the difficulties in dealing with the highly non

linear set of equations that describe the physics of these flow fields. Numerical

approaches give reasonable, and often very accurate, solutions to problems such as these

that are not solvable analytically. Recently, due to the development in non-intrusive laser

diagnostics measurement techniques and to advances in the field o f computational fluid

dynamics, the ability to study such flow fields has been greatly improved [23].

Recent advances in numerical simulation techniques and theoretical studies offer

new descriptions of the physical phenomena in high-speed flows including mixing and

combustion interactions and the onset and development of instabilities and voticity, both

in non-reacting and chemically reacting flows. Although numerical simulations offer

much physical insight and a great level of detail, the experimental validation of both the

theoretical models and the numerical algorithms is largely lacking due to difficulties in

measuring essential parameters in these flow fields, such as velocity, temperature, and

pressure characteristics. Furthermore, there are numerous modeling issues, which are

largely unresolved and need insight from experimental observation. Existing

experimental facilities are, in general, dedicated to study o f only a. limited range of topics

because o f extreme conditions of high enthalpy flow, including an overview o f the

demands of availability of existing facilities. Certain facilities, capable of reproducing

high enthalpy conditions, can operate for only short time durations, insufficient to

achieve stable thermal conditions. Other existing facilities are, in general, limited in

terms o f their experimental flexibility [21].

To efficiently simulate the piloted-scramjet engines, a reasonable kinetic model

has to be implemented in the CFD code. The simple global kinetic models may be

inadequate to study piloting problems, since they neglect intermediate combustion

products, which are essential to the phenomenon [14]. At the same time, the

computational cost o f a given reaction mechanism depends primarily on the number of

species included, rather than the number of reactions [175]. Therefore it is recommended

to use a kinetic scheme with a minimum number o f species and reactions that is

effectively yet capable of modeling the chemistry with reasonable accuracy for the


42

expected combustion conditions. In the dual-fuel cases, in which pilot hydrogen is

injected to support the ignition of ethylene and kerosene hydrocarbon fuels, the

combustion o f pilot hydrogen has to be accounted for in the chemistry part o f the

simulation. This necessitates inclusion of the hydrogen combustion in the used kinetic

scheme. Fulfilling this task makes the selection of the chemistry scheme being more

tightened to achieve both the minimum possible number of species and the involvement

of hydrogen combustion as one of the elementary reactions of the scheme.

Singh and Jachimowski [13] developed a reduced 10-species, 10-step (with

backward reactions) kinetic model for ethylene combustion, which compared favorably

with the results calculated with the detailed mechanism of Jachimowski [176]. This

reduced model is also suggested for assembling reaction mechanisms of heavy

hydrocarbons such as propane, butane, n-heptane, etc.

Different CFD techniques used for Scramjet engine analysis were presented in the

work of White et al. [177]. Various levels of approximation to the Reynolds-averaged

Navier-Stokes equations, by which the flow field through a scramjet engine can be

described, were mentioned. It included using the Parabolized Navier-Stoked (PNS)

equations, reducing the geometry assumption to two-dimensional planar or axisymmetric

flow, and using the Euler/Boundary-Layer Superposition approach. The application of the

CFD techniques to scramjet engine analysis was highlighted by presenting examples of,

and a discussion of, the techniques applied to the analysis of the various engine

components; the inlet, the combustor, and the nozzle.

For the support of the design of a fixed geometry scram intake two-dimensional

Navier-Stokes calculations have been performed by Bissinger [178]. Unstructured-grid

Navier-Stokes (UNS) methodology is well suited to the analysis o f scramjet combustor

flow fields in view of its ability to deal with complex injection geometries and to

adaptively refine the grid in critical regions such as flame zones. UNS applications to

scramjet combustor flow have been limited in view of the need to deal with complexities

associated with combustion and turbulence modeling, which are better established in

structured flow solvers. Lee et al. [179] used UNS code to investigate the scramjet

combustor field for swept ramp fuel injectors. They summarized the progress achieved in

implementation o f the grid adaptive methodology.


43

In the work of Kodera et al. [180] the hybrid grid method to compute supersonic

reacting flow rate was applied to internal reacting flows of a scramjet engine at flight

Mach number of 8. The details of reacting internal flow field such as boundary layer

separation, mixing, ignition and combustion as well as engine performance such as thrust

and combustion efficiency were investigated. Ungewitter et al. [181] upgraded the k-e

turbulence model by improving the compressibility correction and transitional modeling.

A more stable Gauss-Seidel implicit solver was incorporated into the unstructured Navier

Stokes code and a new variable element grid methodology was employed which

improved the resolution in several key areas while maintaining a minimum number of

cells.

Chakravarthy et al. [182] presented the progress they recently achieved in the

turbulence modeling closure. Their contribution focused on the improvements to the

classical models and the usage of the hybrid Reynolds Average Navier Stokes Equations/

Large Eddy Simulation (RANS/LES) methods. They presented their simulation data for a

couple of supersonic flow case studies that are: swept ramp injection into a supersonic

flow, and two-hole transverse sonic injections downstream of a rearward-facing step in

supersonic flow field. The calculations produced using the new closure compared

favorably with existing relevant experimental data.

Quantifying the level o f confidence, or accuracy, of computational fluid dynamics

(CFD) codes and numerical solutions has recently received increased attention in the

research and engineering application literature. The issues o f credibility, validation, and

verification o f the CFD codes were discussed intensively in [183-188].

A fine enough mesh with a good enough quality must be used to ensure that there

are not significant variations in. the solution when either the number or the placement of

the mesh points is varied. Otherwise it is impossible to distinguish modeling errors from

numerical errors. The mesh refinement and optimization were studied in [189-190].

The accessibility o f the CFD tools on the Internet was discussed in [191]. A

concept was presented for evolving the current RocketWeb™ Internet Analysis System

into a comprehensive design process, capable o f seamlessly integrating the diverse design

and analysis tools necessary to create robust designs. The resulting system would result in

a rigorous, controlled design process capable o f reducing both design cycle time and cost.


44

Table 2.1 Energy properties of H2 and hydrocarbon fuels data complied from Refs.[137-140]

Fuel

Energy/Mass,

MJ/kg

Energy/Volume,

MJ/1

STP Density,

Kg/m3

Liquid H2 116.7 8.2 71

Slush H2 116.6 9.8 82

Methane 50.0 20.8 424

n-Heptane 44.6 30.3 717

JP-4 43.5 33.1 760

JP-5 43.0 35.1 815

JP-7 43.9 34.7 790

JP-8 43.2 35.0 809

Jet A 43.4 34.6 799

Kerosene 42.8 34.2 800

MCH 43.4 33.4 770


45

Fl ight

Path

F o r ebodyObl iqueS h o c kWa ve s

A f t e r b o d y

A i r b r e a th i n gEngine

Fig. 2.1 Schematic diagram of the scramjet engine configuration [26].

ExhaustNozzleSu p erson ic

Burner

Fuel InfectoraFreestream Exhaust

Flow

ForebodyOblique

ShockWave

Fig. 2.2 Ftq»nHal features of two-dimensional or planar geometry scramjet engines

[26].


Fig. 2.3 Airframe-integrated scramjet along with a partial cross-section through a

typical modular engine [28].

Interaction shock

Ma

Separation region

Boundary layer

Jet boundary

Macn disc

\ \ \ \ \ \ \ \ \ \

Fig. 2.4 Schematic of the region surrounding a single perpendicular injector [28].


47

Lasar light scattering of silica Hjist

Hjjet w/silicaSwept ramp Injectors

Dimensions in inches \ K "Flash-lamp later with 50 psae pulse

Fig. 2.5 Generation of counter-rotating stream wise vorticity by ramp fuel injector [60].

Compression Expansion

Flow^ 3

Flow

Fig. 2.6 Scramjet fuel-injector compression and expansion ramps [60].


48


Fig.

2.7

Ener

gy

cont

ent

per u

nit m

ass

for d

iffer

ent

avion

ic fu

els [

130]

.

49

L4F-::

r tr.r -■J-LW;.

v

i& ii& £ }

ftf

Oo oCM

o©o

mm

1a0 ) u

3CCSt*

0 1S *>IO IH


Fig.

2.8

Tem

pera

ture

lim

it for

diff

eren

t av

ionic

fuels

[13

0].

50

cepm©

• pm

V ~ \

V ®

A . / T v y . ^

■ *•?.'. A <C!.V ’.v /i i / ' J t f XTTC9 -AX5X.

« t

Ooo0%oin

oo e§ f* " «

o

6 3

ooooin


Fig.

2.9

Ener

gy

cont

ent

per u

nit v

olum

e for

diff

eren

t av

ionic

fuels

[13

0].

51

CHAPTER m

THEOETICAL FORMULATIONS

3.1 Introduction

The Computational Fluid Dynamics “CFD” Code used in the present study is the

FLUENT commercial code. FLUENT is a general-purpose computer program for

modeling fluid flow, heat transfer, and chemical reaction [192].

FLUENT models a wide range o f phenomena by solving the conservation

equations for mass, momentum, energy, and chemical species using a control volume

based finite difference method. The governing equations are discretized on a curvilinear

grid to enable computations in complex irregular geometries. A non-staggered system is

used for storage o f discrete velocities and pressures. Interpolation is accomplished via a

first-order, Power-Law scheme or optionally via higher order upwind schemes. The

equations are solved using SIMPLE-like algorithms with an iterative line-by-line matrix

solver and multigrid acceleration.

3.2 Governing Equations

For all flows, FLUENT solves conservation equations for mass and momentum.

For flows involving heat transfer or compressibility, an additional equation for energy

conservation is solved. For flows involving species mixing or reactions, a species

conservation equation is solved or, if the PDF model is used, conservation equations for

the mixture fraction and its variance are solved. Additional transport equations are also

solved when the flow is turbulent.

3.2.1 The Mass Conservation Equation

The equation for conservation of mass, or continuity equation, can be written as follows:

( 3 1 )

Equation (3.1) is the general form of the mass conservation equation and is valid

for incompressible as well as compressible flows. The source Sm is the mass added to the

continuous phase from the dispersed second phase (e.g., due to vaporization of liquid

droplets) and any user-defined sources.


52

3.2.2 Momentum Conservation EquationsConservation of momentum in the i direction in an inertiai (non-accelerating)

reference frame is described by [193]

where p is the static pressure. Tjj is the stress tensor (described below), and pgi and Ft are

the gravitational body force and external body forces (e.g., that arise from interaction

with the dispersed phase) in the i direction, respectively. F, also contains other model-

dependent source terms such as porous-media and user-defined sources.

The stress tensor Xjj is given by

f 3 it,, 3u . Y| 2 3m,

r * = " a ^ + fcT 0 3 )L V 1 JJ

where p. is the molecular viscosity and the second term on the right hand side is the effect

of volume dilation.

3.2.3 The Energy EquationThe general form of the energy equation is:

where iCefr is the effective conductivity (k + kh where kt is the turbulent thermal

conductivity, defined according to the turbulence model being used), and Jr is the

diffusion flux of species The first three terms on the right-hand side of Eq. (3.4)

represent energy transfer due to conduction, species diffusion, and viscous dissipation,

respectively. S& includes heat of chemical reaction, and any other volumetric heat sources

defined.

(3.2)

InEq. 3.4,

(3.5)

where sensible enthalpy h is defined for ideal gases as

h = '2 f mr hri

(3.6)


53

In Equation (3.6), mj is the mass fraction of species/ and

(3.7)

where Tref is 298.15 K.

Equation (3.4) includes pressure work and kinetic energy terms, which are often

negligible in incompressible flows.

The viscous dissipation terms, which describe the thermal energy created by

viscous shear in the flow, is included in Eq. (3.4). Viscous heating will be important

when the Brinkman number, Br, approaches or exceeds unity, where

and AT represents the temperature difference in the system. The effect o f enthalpy

The source o f energy, Sh, in Eq. (3.4) includes the source of energy due to

chemical reaction:

where h°r is the enthalpy o f formation of species / and Rr is the volumetric rate of

creation o f species j \

When one of the radiation models is used, Sh in Eq. (3.4) also includes radiation

source terms. It should be noted that the energy sources, Sh, also include heat transfer

between the continuous and the discrete phase.

3.3.1 Turbulence ModelsIn turbulent flows, the velocity at a point is considered as a sum of the mean

(ensemble-averaged) and fluctuating components:

transport due to species diffusion is includes in Eq. (3.4) through the term«- /

h,reaction (3.9)

33 Basic Physical Models for Flow and Heat Transfer

(3.10)


54

Substituting expressions of this form into the instantaneous momentum equations

(and dropping the overbar on the mean velocity, u yields the ensemble-averaged

momentum equations:

Equation (3.11) has the same form as the fundamental momentum balance with

velocities now representing ensemble-averaged (or mean-flow) values. The effect of

turbulence is incorporated through the "Reynolds stresses", pu-u'. FLUENT relates the

Reynolds stresses to mean flow quantities via one of three turbulence models:

1. The k-E model

2. The RNG k-E model

3. The Reynolds Stress Model (RSM)

The process of selecting a turbulence model for a given turbulent flow problem is

greatly facilitated when having a good understanding of the salient features of the flow in

question. Based on this understanding, it should then be considered which model would

be more suitable. To do so, the capabilities and limitations of the individual models must

be known. This section provides an overview of the models and general guidelines that

helps choosing the correct turbulence model for the flow to be modeled.

The standard k-E model proposed by Jones and Launder [194] as been the

workhorse of engineering turbulence models for more than two decades. It falls in the

category of "two-equation" turbulence models based on an isotropic eddy-viscosity

concept. As such, it is more universal than other low-order turbulence models such as

algebraic ("zero-equation") and one-equadon models. Robustness, economy, and

reasonable accuracy for a. wide range of turbulent flows explain its popularity in

industrial flow and heat transfer Simulations. It is a semi-empirical model, and the

derivation o f the model equations, including the various model constants, relies on

phenomenological considerations and empiricism.

The Renormalization Group (RNG) Jfc-e model also belongs to the k-E family of

models. The model equations in their RNG form are similar to those for the standard k-E

model. There are major differences, however, between the RNG and standard k-E models.


The RNG model was derived using a more rigorous statistical technique, and its model

constants are derived "analytically."

The Reynolds Stress Model (RSM) is the most elaborate turbulence model that

FLUENT provides. Eschewing the isotropic eddy viscosity hypothesis, the RSM closes

the Reynolds stresses by solving their transport equations (six additional equations in 3D,

in comparison with k-E models). As such, the RSM accounts for the history and transport

o f the Reynolds stresses in a rigorous manner. The effects of streamline curvature, swirl,

and rotation are all directly accounted for by the transport equations for the Reynolds

stresses.

3.3.2 Compressible Flows

Compressibility effects are encountered in gas flows at high velocity and/or in

which there are large pressure variations. When the gas flow velocity approaches or

exceeds the speed of sound or when the pressure change in the system (Ap/p) is large, the

variation of the gas density with pressure has a significant impact on the flow velocity,

pressure, and temperature. Compressible flows create a unique set of flow physics, which

should be well known to provide the special input requirements and solution techniques

that are described in this section.

Compressible flows can be characterized by the value of the Mach number

M = u / c (3.12)

where c is the speed o f sound in the gas:

c= JyRT (3.13)

and y is the ratio of specific heats (Cp/Cv).

When the Mach number is less than one, the flow is termed subsonic. At Mach

numbers much less than one (e.g., M < 0.3 or so), compressibility effects are negligible

and the variation of the gas density with pressure can safely be ignored in the flow

modeling. When the Mach number approaches unity, however, compressibility effects

become important. The flow will choke at Mach 1.0. When the Mach number exceeds

1.0, the flow is termed supersonic. Supersonic flows may contain shocks and expansion

fans which can impact the flow pattern significantly and which require compressibility in


56

the FLUENT model. FLUENT provides a full range of compressible flow modeling

capabilities for subsonic, transonic, and supersonic flows.

Compressible flows are typically characterized by the total pressure P0 (isentropic

stagnation pressure) and total temperature T0 (isentropic stagnation temperature) of the

flow. These quantities can be related to the static pressure and temperature via the

following relationships:

These unique relationships describe the variation of the static pressure and

temperature in the flow as the velocity (Mach number) changes under isentropic (toss-

free, constant enthalpy) conditions.

For a given pressure ratio from inlet to exit, for example, Eq. (3.14) can be used to

For air, Eq. (3.14) predicts a choked flow (Mach number of 1.0) at an isentropic pressure

ratio, Ps/Po, of 0.5283. This choked flow condition will be established at the point of

minimum flow area (e.g., in the throat o f a nozzle). In the subsequent area expansion the

flow may either accelerate to a supersonic flow in which the pressure will continue to

drop, or the flow may return to subsonic flow conditions, decelerating with a pressure

rise. When a supersonic flow is exposed to an imposed pressure increase, a shock will

occur, with a sudden pressure rise and deceleration accomplished across the shock.

Compressible flows are described by the standard continuity and momentum

equations solved by FLUENT, and there is no need to activate any special physical

models (other than the compressible treatment of density as detailed below). The energy

equation solved by FLUENT, Eq. (3.4), correctly incorporates the coupling between the

flow velocity and the static temperature, and should be activated whenever a

compressible flow is solved. In addition, you should activate the optional viscous

dissipation terms in Eq. (3.4), which become important in high-Mach-number flows.

For high-Mach-number flows, compressibility affects turbulence through so-

called "dilatation dissipation," which is normally neglected in the modeling of

r

(3.14)

(3.15)

estimate the exit Mach number (which would exist in a one-dimensional isentropic flow).


57

incompressible flows [195], Neglecting the dilatation dissipation fails to predict the

observed decrease in spreading rate with increasing Mach number for compressible

mixing and other free shear layers. To account for these effects in the K-e models in

FLUENT, the dilatation dissipation term, YM, is included in the tc equation. This term is

modeled according to a proposal by Sarkar and Balakrishnan [196]:

Yu = pe2M ; (3.16)

where Mt is the turbulent Mach number, defined as

(3.17)

where a (=^JyRT ) is the speed of sound. This compressibility modification always takes

effect when the compressible form o f the ideal gas law is used.

3.3.3 Chemical Species Transport and Reacting FlowFLUENT models chemical species transport and chemical reactions using the

reacting flow models described in this section. The mixing and transport o f chemical

species are modeled by solving conservation equations describing convection, diffusion,

and reaction sources for each component species.

When you choose to solve conservation equations for chemical species, FLUENT

predicts the local mass fraction of each species, m,-, through the solution of a convection-

diffusion equation for the species. This conservation equation takes the following

general form:

+Rr +S r (3.18)

where is the mass rate o f creation or depletion by chemical reaction and Sr , is the

rate o f creation by addition from the dispersed phase. An equation of this form will be

solved for N -l species where N is the total number of fluid phase chemical species

present in the system.

The diffusion flux, / £.£, may optionally be augmented by a thermal diffusion

1 3 Tterm, —DT. —- — (also called Soret diffusion).1 T dx£

where Dj. is the diffusion coefficient for species r \


58

In turbulent flows, FLUENT computes the mass diffusion in the form:

(3.19)

where Sct is the effective Schmidt number, (with a default setting of 0.7).

For many multi-component mixing flows, the transport of enthalpy due to species

when the species involved have significantly differing heat capacities, this term cannot be

neglected.

The reaction rates that appear as source terms in Eq. (3.18) are computed from

Arrhenius rate expressions or by using the eddy dissipation concept due to Magnussen

and Hjertager [197]. Models o f this type are suitable for a wide range o f applications

including laminar or turbulent reaction systems, and combustion systems including

premixed or diffusion flames.

The source o f chemical species f due to reaction, R,-, is computed as the sum of

the reaction sources over the le reactions that the species may participate in:

where Rr./t is the rate o f creation/destruction o f species f in reaction k. Reaction may

occur in the continuous phase between continuous phase species only, or at surfaces

resulting in the surface deposition or evolution o f a chemical species. The reaction rate,

Rrk , is controlled either by an Arrhenius kinetic rate expression or by the mixing of the

turbulent eddies containing fluctuating species concentrations.

The Arrhenius reaction rate is computed as:

where molar stoichiometric coefficient for species i’ in reaction k (positivet

diffusion (V .|X , can have a significant effect on the enthalpy field and should

not be neglected. In particular, when the Lewis number Le = is far from unity, andk f c v

(3.20)it

(3.21)

values for reactants, negative values for products)


59

Af f. - molecular weight o f species i ’ (kg/kmol)

fik = temperature exponent (dimensionless)

Ak = pre-exponential factor (consistent units)

Cy = molar concentration of each reactant species j ’ (kmol/m3)

v r k = exponent on the concentration o f reactant j in reaction k

Ek — activation energy for the reaction (J/kmol)

The values for v ’ , fik, Ak , v r k , and Ek are defined in the problem setup.

The influence o f turbulence on the reaction rate is taken into account by

employing the Magnussen and Hjertager model. In this model, the rate of reaction Rrk is

given by the smaller (i.e., limiting value) o f the two expressions below:

= - v n M rA p ? ~ p & — (3.22)Rsk ft

e ^ m*rjt =-v'r.kMrA B p - ^ p. [ (3.23)

k L Pv P*M P

where mp = mass fraction of any product species, P

mR = mass fraction of a particular reactant, R

R = reactant species giving the smallest value of Rr k

A = an empirical constant equal to 4.0

B = an empirical constant equal to 0.5

The eddy breakup model relates the rate o f reaction to the rate of dissipation of

the reactant and product containing eddies, k/e represents the time scale of the turbulent

eddies following the eddy break up model of Spalding. The model is useful for the

prediction of premixed and diffusion problems as well as for partially premixed reacting

flows.

In turbulent reacting flows, FLUENT calculates the reaction rates from the

Arrhenius expression (Eq. (3.21)) and the eddy breakup model (Eqs. (3.22) and (3.23)).

The limiting (slowest) rate is used as the reaction rate and the contribution to the source

terms in the species conservation and enthalpy equations are calculated from this reaction


- : v ■ . ■ . . .:

60j-'-'-*V '

rate. Energy released by or required for the chemical reaction is accounted for in the

source term o f the enthalpy equation as shown in Eq. (3.9).

The introduction o f the kinetic term into the rate expression for turbulent flows is

useful as it can act as a cut-off to the mixing controlled rate when chemistry is very slow.

However, in many practical situations, the eddy breakup model describes the limiting

rate.

If the fuel is introduced to an oxidant, spontaneous ignition does not occur unless

the temperature of the mixture exceeds the activation energy threshold required to

maintain combustion. In the simulation an ignition source has to be supplied to initiate

combustion. This ignition source may be a heated surface that heats the gas mixture

above the required threshold level. Often, however, it is the equivalent of a spark: an

initial solution state that causes combustion to proceed. This "fuel spark" or "initial

solution state" is supplied by patching a hot temperature into a region of the model that

contains a sufficient fuel/air mixture for ignition to occur. Often, there may be a need to

patch both the temperature and the fuel/oxidant/product concentrations to produce

ignition in the model. The initial patch has no impact on the final steady-state solution-no

more than the location o f a match determines the final flow pattern o f the torch that it

lights.

3.4 Thermal Model3.4.1 Density: Ideal Gas Law for Compressible Flows

For compressible flows, the gas law has the following form:

(3.24)RT

where p is the local relative (or gauge) pressure predicted by FLUENT and Pop is defined

as the Operating Pressure in the Operating Conditions panel.

3.4.2 Composition-Dependent Thermal Conductivity for Multicomponent MixturesWhen the ideal gas law is used with multi-component flow field, the composition-

dependent thermal conductivity based on kinetic theory is calculated as

(3'2S)r Z ,j-X r

where


and X r is the mole fraction o f species

3.4J Composition-Dependent Viscosity for Multicomponent MixturesWhen the ideal gas law is used with multi-component flow field, the composition-

dependent thermal conductivity based on kinetic theory is calculated as

X r fir

2 r x r*r.r(3.27)

where

0£T =

1+

l Ir w Af,El

M r ,V 1 /(3.28)

8M r\

and X r is the mole fraction o f species f .

3.4.4 Specific Heat Capacity as a Function of CompositionThe species heat capacity is defined as a function of temperature in a polynomial

formas follows:

Cp (r)=A , +A2T + A 3T z -k.. (3.29)

where coefficients A,’s are characteristic data for each species.

The mixture's specific heat capacity is calculated as a mass fraction average of the

pure species heat capacities as follows:

Cp = X mr c„.r (3.30)

3.5 Grid Generation


62

The grid is a discrete representation of the continuous field phenomena that is

modeled and the accuracy and numerical stability o f the simulation depend on the choice

of the grid. In other words, the density and distribution o f the grid lines determines the

accuracy with which the model represents the actual physical phenomena.

In FLUENT, the control volume method, sometimes referred to as the finite

volume method, is used to discretize the transport equations. In the discrete form of the

equations, values o f the dependent variables appear at control volume boundary locations

(cell faces). These values have to be expressed in terms o f the values at the nodes of

neighboring cells in order to obtain algebraic equations. This task is accomplished via an

interpolation practice, also called a "differencing scheme." The choice of differencing

scheme not only affects the accuracy of the solution but also the stability o f the numerical

method.

The default differencing scheme used in FLUENT is the so-called power-law

scheme. This scheme is derived from the exact analytical solution to the one-dimensional

convection-diffusion equation. The power-law scheme is very stable and gives physically

meaningful (bounded) solutions but, in certain situations, is susceptible to numerical

diffusion effects. These effects, as mentioned earlier, are maximums when the flow is

aligned at 45 degrees to the grid lines and there are significant gradients in the direction

normal to the flow. Higher-order methods which are less susceptible to numerical

diffusion, but also less stable compared to the power-law scheme, are also available in

FLUENT.

It should be noted that recent research in CFD has focused on unstructured grids,

where the grid points are placed in the flow field in a very irregular fashion; this is in

contrast to a structured grid, which reflects some type of consistent geometrical

regularity. Among the advantages of the unstructured grid meshes over the structured

ones are the allowance for the maximum flexibility in matching mesh cells with the

boundary surfaces and for putting cells where needed (193].

Both the structured and unstructured meshes are generated using “Gambit,” which

is the grid generation package of the flow solver “Fluent.”

The meshes used in the present study are unstructured. The faces are meshed

using the “Tri-Pave” scheme, which means that the mesh includes only triangular mesh


63

elements that are interconnected in an irregular fashion to construct the unstructured

meshing o f the faces. The volume meshing emanates from the mesh elements o f the faces

and branches downward filling the whole volume and connecting the faces using the “T-

Grid” scheme through which GAMBIT creates a mesh that consists primarily o f the

tetrahedral mesh elements but may include hexahedral, pyramidal, and wedge elements

where appropriate.

Figure 3.1 shows an example for an unstructured face meshing using “Tri-Pave”

scheme. Figure 3.2 shows the general shape of the tetrahedral mesh elements that are

created in the volume if none o f the faces are meshed prior to the application of the "T-

Grid” scheme or if all pre-meshed faces are meshed by means o f a Tri-Pave scheme.

3.6 Method of Solution3.6.1 Basics of the Overall Solution Algorithm

The coupled solver in FLUENT solves the governing equations of continuity,

momentum, and (where appropriate) energy and species transport simultaneously as a set,

or vector, o f equations. Governing equations for additional scalars will be solved

sequentially (i.e., segregated from one another and from the coupled set).

The system of governing equations for a single-component fluid, written to

describe the mean flow properties, is cast in integral, Cartesian form for an arbitrary

control volume V with differential surface area dA as follows:

J WdV+§[F -G \d A = \H dV (3.31)at v y

where the vectors, W, F, and G are defined as

p pv 0pu pvu+ pi txipv , F = p w + p j , G = zyipw pvw+pk vtiPE pvE+pv Jijvj+q

and the vector H contains source terms such as body forces and energy sources.

Here p , v , E, and p are the density, velocity, total energy per unit mass, and

pressure o f the fluid, respectively, t is the viscous stress tensor, and q is the heat flux.


64

Total energy E is related to the total enthalpy H by

(3.33)P

where2

H=h+2

(3.34)

Each iteration consists o f the steps illustrated in Fig. 3.3 and outlined below:

1. Fluid properties are updated, based on the current solution. (If the calculation has

just begun, the fluid properties will be updated based on the initialized solution.)

2. The continuity, momentum, and (where appropriate) energy and species equations

are solved simultaneously.

3. Where appropriate, equations for scalars such as turbulence and radiation are

solved using the previously updated values of the other variables.

4. When interphase coupling is to be included, the source terms in the appropriate

continuous phase equations may be updated with a discrete phase trajectory

calculation.

5. A check for convergence of the equation set is made.

These steps are continued until the convergence criteria are met.

3.6.2 Residual ReportingThe process of obtaining a converged solution is of great importance in FLUENT

simulations. So that in order to monitor this process, FLUENT provides a running report

o f the residuals for each equation at each iteration. The residuals are a measure o f how

closely each finite difference equation is balanced, given the current state o f the solution.

In this section, a definition of the residuals is given.

At each iteration o f the solution algorithm, FLUENT reports a residual for each

equation that has been solved. These residuals provide a measure of the degree to which

each equation is satisfied throughout the flow field. FLUENT computes residuals for

each conservation equation b y summing the imbalance in the equation for all cells in the

domain. A detailed description o f the calculation o f the residuals is provided below.


At the end of each iteration, the residual sum for each o f the conserved variables

is computed and stored, thus recording the convergence history. This history is also saved

in the data file. The residual sum is defined below.

On a computer with infinite precision, these residuals will go to zero as the

solution converges. On an actual computer, the residuals decay to some small value

(“ round-off1) and then stop changing (“ level out”). For “ single precision” computations

(the default for workstations and most computers), residuals can drop as many as six

orders o f magnitude before hitting round-off. Double precision residuals can drop up to

twelve orders o f magnitude.

A residual for the coupled solvers is simply the time rate of change of the

conserved variable (W). The RMS residual is the square root of the average o f the

squares o f the residuals in each cell of the domain:

Equation (3.35) is the unsealed residual sum reported for all the coupled equations

solved by Fluent's coupled solver.

In general, it is difficult to judge convergence by examining the residuals defined

by Eq. (3.35) since no scaling is employed. Fluent scales the residual using a scaling

R(W)iterarion N R(W)iteration5

The denominator is the largest absolute value of the residual in the first five

iterations.

The scaled residuals described above are useful indicators of solution

convergence. I t is sometimes useful to determine how much a residual has decreased

during calculations as an additional measure of convergence. For this purpose, Fluent

allows to normalize the residual (either scaled or unsealed) by dividing by the maximum

residual value after M iterations.

Normalization o f the residual sum is accomplished by dividing by the maximum

residual value after Af iterations:

(3.35)

factor representative of the flow rate <j of through the domain. This scaled residual is

defined as


66

m y / ) = i t e r a t i o n N (3.37)R(W)iteration M

Normalization in this manner ensures that the initial residuals for all equations are

of 0(1) and is sometimes useful in judging overall convergence. By default, Af = 5,

making the normalized residual equivalent to the scaled residual.

3.6.3 Judging ConvergenceA FLUENT calculation is converged when all governing equations are balanced

at each point in the solution domain. This section provides guidance on how to judge the

convergence of the solution via residual values and how to monitor the progress towards

convergence via residual histories and histories of solution variables.

The residuals for each flow variable give a measure of the magnitude of the error

in the solution at each iteration. As discussed in the preceding section, these residuals are

normalized unless being set as un-normalized. Generally, a solution is well converged

when the normalized residuals are on the order of IxlO'3. An important exception is the

enthalpy residual, which should be about IxlO'6.

In addition, the residuals o f the species transport equations need to decrease to 1 x

10'5 to I x KT6 when solving problems involving mixing of two species o f very different

molecular weights (e.g., Hi and WFg). If the residuals have decreased to this level, are

monotonically decreasing, and the flow field looks unchanged from the solution 50

iterations earlier, then the solution can be called "converged." Sometimes it might not be

needed to generate a completely converged solution, if the basic features of the flow field

could be picked up right away. When quantitative results are sought, however, complete

convergence of the solution is essential.

fit the present study convergence is primarily judged by reaching unchanged

values for the most important flow field parameters such as temperature, pressure,

velocity magnitude, and species mass fractions. These parameters are averaged over all

grid cells at different cross flow planes along the combustor length using the mass

weighted-approach. A t the convergence condition these averages show unchanged values

with the solution iterations.


67

The conservation o f mass and energy are checked as another means for examining

the convergence* The imbalance in the mass flow rate and the rate o f heat flux relative to

the total inputs o f mass flow rate and heat flux rate respectively are accepted if less than

1% under the convergence condition.

3.7 Physical ModelsFour physical models are used in the present study that includes; rearward-facing

step, swept rearward-facing step, cavity with no-pilot injection, cavity with pilot injection

models. The configurations used in this case study are that used by Owens et al.

[198,199] with slight differences in the dimensions applied. Figures 3.8-3.11 present the

four models with the full dimensions. The case studied feature the symmetry condition at

the mid-span plane; therefore, just one half of the physical domain is simulated to

minimize the computational time.

3.7.1 Rearward-Facing Step ConfigurationIn the configuration used in this study the rearward-facing step is located at the

upper longitudinal wall. The 3-D schematic diagram for this configuration is

demonstrated in Fig. 3.8(a). The step height (H) is lOmm. Gaseous ethylene, as the main

fuel, is injected at four step heights downstream of the step normal to the incoming

supersonic air stream. The onset of thermal choking was delayed by diverging the test

section starting at 19H downstream of the step with a 3-deg half angle. The total length of

the domain is 350mm with an inlet test section o f 25.4x25.4 mm . The full dimensions of

the combustor at the symmetry plane are shown in Fig. 3.8(b). An unstructured grid with

a size of 389,251 cells is used in simulating this case study.

3.7.2 Swept Rearward-Facing Step ConfigurationThis configuration resembles that used in the rearward-facing case study except

for some differences. The inlet test section dimensions, the step height and its location,

the main injection diameter and its location, and the pilot injections diameter are exacdy

as those used in the ethylene case. The total length of the domain is 220mm. The onset of

thermal choking was delayed by diverging the test section starting at 6H downstream of

the step with a 3-deg half angle. The step implemented in the propane case study is swept

from both end sides starting at 15 mm downstream o f the combustor inlet creating as

extra means of vorticity that aims at enhancing the fuel-air mixing. Gaseous pilot propane


68

is injected parallel to the incoming air stream via two holes with 1 mm diameter each that

are centered at the base o f the step. The incoming air is a 1.751-Mach flow to simulate

the enthalpy level o f the typical flight conditions at the combustor inlet. The incoming air

is clean (no vitiation used). The schematic diagram for the propane combustion case is

presented in Fig. 3.9. An unstructured grid with a size of 207,125 cells is used in

simulating this case study.

3 .7 J Cavity with No Pilot Injection Configuration

In the configuration used in the present study the rearward-facing step is located

at the upper longitudinal wall. A 15°-wedge is located downstream of the step and

upstream of the main normal injection. The combination of the rearward-facing step and

the wedge forms a cavity-like configuration that helps in enhancing the fuel-air mixing in

addition to initiate and stabilize the main flame. This is demonstrated in Fig. 3.10. The

unstructured grid used to simulate this case study is o f a size o f245,702 cells.

3.7.4 Cavity with Pilot Injection Configuration

The only difference between this configuration and the cavity with no pilot

injection configuration is in the injection o f the gaseous pilot ethylene parallel to the

incoming air stream via three holes that are equally distributed at the base o f the step. The

3-D schematic diagram of this configuration and the combustor full dimensions at the

symmetry plane are presented in Figs. 3.11(a) and 3.11(b) respectively. The unstructured

grid used to simulate this case study is of a size o f245,702 cells.

3.8 Boundary and Initial Conditions

Once the governing flow equations are defined, then the real driver for any

particular solution is the boundary conditions. This has particular significance in CFD;

any numerical solution o f the governing flow equations must be made to see a strong and

compelling numerical representation of the proper boundary conditions [193].

The inflow boundaries are supersonic, so the total and static pressures, the total

temperature, and the species mass fractions are specified. For these supersonic inflow

boundaries, uniform, conditions were assumed for the primitive parameters. The outflow

boundaries are supersonic so all flowfield variables are allowed to float. They are

calculated using linear extrapolation based on the flowfield values at the internal points


69

[193]. No-slip boundary condition is dictated at all walls and step boundaries; meaning

that zero relative velocity between the surface and the gas immediately at the surface is

assumed and the temperature o f the fluid layer immediately in contact with the surface

equals the wall temperature. All surfaces are assumed adiabatic with zero temperature

gradient at the walls.

To start the calculations, initial conditions for the pressure, temperature, species

mass fractions, and turbulence parameters in the whole flowfield must be stipulated. The

choice o f the initial conditions should not afreet the final steady-state answer. In theory,

these initial conditions can be purely arbitrary. In practice, they should be chosen

intelligently to expedite the final steady-state answer, and hence shortening the computer

execution time [193]. For all case studies in the present investigation, the initial

conditions were set by applying free-stream air inlet conditions throughout the entire flow

field.


Fig. 3.1 An unstructured face meshing using “Tri-Pave” scheme.

Fig. 3.2 T-Grid meshing scheme.


Converged?

Update properties.

Solve turbulence and other scalar equations.

Solve continuity, momentum, energy, and species equations simultaneously.

Fig. 3.3 Overview of the solution process.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission

7 2

Fig. 3.4(a) Schematic diagram for the rearward-facing step configuration.

Bhydne

QmaticreaeinnTn

Dawu iienottDsaie

Vitiated

Fig. 3.4(b) Schematic diagram for the step configuration, at the symmetry plane with full-dimensions.


73

Fig. 3.5 Schematic diagram for the swept rearward-facing step configuration.

Fig. 3.6 Schematic diagram for the cavity with no pilot injection configuration.


74

Fig. 3.7(a) Schematic diagram for the cavity with pilot injection configuration.

Bhydm

o

h

OnHticreaeinnTn

Daw^aeiilfosafc

L

33>

Fig. 3.7(b) Schematic diagram for the cavity with pilot objection configuration at the symmetry plane with full-dimensions.


76

Because of the turbulent nature of the case study, the influence of turbulence on

the reaction rate is taken into account by comparing the Arrhenius reaction rate and the

two eddy-dissipation reaction rates. Out o f the three mentioned rates the limiting

(slowest) rate is used as the reaction rate and the contributions to the source terms in the

species conservation and energy equations are calculated from this reaction rate.

The following global one-step reaction of propane is used to calculate the

Arrhenius reaction rate in the present study:

C3ff8 +50 2----->3C02+4H20 (4.1)

The Arrhenius reaction rate constants for the chemistry scheme are listed below,

Pre-exponential factor= 4.836e+09

Activation energy = 1.256e+08 [J/kgmol]

Temperature exponent= 0

The renormalized group (RNG) form of the tc-e turbulence model is used. The

turbulence near-wall is treated using the two-layer zonal model.

Because the flow is compressible and a multicomponent mixture, it is modeled as

an ideal gas and the density is calculated as a composition-dependent property. A

composition-dependent specific heat capacity is defined for the multicomponent mixture

material while temperature-dependent heat capacities for the individual species are

specified. The solver computes the mixture’s specific heat capacity as a mass fraction

average of the pure species heat capacities. The viscosity and thermal conductivity are

defined as composition-dependent properties.

The coupled solver is used with the explicit formulation and the steady state

approach. The second order upwind scheme is used for discretization of the flow,

turbulence kinetic energy, and turbulence dissipation rate equations.

4.3 G rid Independence Study

Unstructured grid mesh is used to simulate this case study. The full details for the

mesh elements, and the faces and volume-meshing scheme were previously reported in

Sec. 3.5.

For the sake of eliminating the factor of the effect o f the grid refinement on the

accuracy o f the results, the grid independence study was performed. Two grid sizes were

implemented. Unstructured grid mesh with 207,125 tetrahedral cells was used reaching


77

the converged solution followed by its adapted version that is composed of 402,873

tetrahedral cells. The converged solution was also obtained using the larger grid.

The longitudinal distribution of the calculated averaged upper-wall absolute

pressure along the combustor length is plotted for both grid sizes to judge for the

independence of calculated results on the grid refinement. Twelve cross flow lines

marking the intersection o f the combustor upper wall and the corresponding cross flow

planes are defined along the upper wall length. The twelve cross flow planes are defined

at X=0.001, 0.015, 0.020, 0.025, 0.030, 0.031, 0.070, 0.090, 0.120, 0.150, 0.180, and

0.219m. The points on the curves of Fig. 4.1 represent the area weighted averages of

absolute pressure at all grid cells of the twelve intersection lines for both grid meshes

used. The solid line represents the calculated values obtained using the small grid size

while the dotted line represents those of the large grid size.

The area-weighted average of a quantity (d>) is computed by dividing the

summation of the product o f the selected field variable and facet area by the total area of

the surface as follows:

j > |A | (4.2)A A ,-=l

Figure 4.1 shows a comparison between the averaged upper-wall absolute

pressures for both grids used. An excellent coincidence between the two curves is

manifested. The maximum pressure difference between the two curves with respect to the

maximum average upper-wall pressure value in the whole field is 2.89% at the axial

location of X — 70 mm. Achieving the grid independence study, the remaining of the

calculations was conducted using the small grid and these results are presented here.

4.4 Results4.4.1 Symmetry Plane Results

The flow field features and the combustion characteristics are discussed through

the presentation of flow contours at the symmetry plane as depicted in Figs. 4.2-4.5.

The contours o f the absolute pressure throughout the combustor symmetry plane

are presented in Fig. 4.2. The combination of the expansion and shock waves along the

combustor length is shown. The Prandtl-Meyer expansion waves are formed right at the

step comer (X=0.030m). The contour values show the pressure decrease associated with


78

these expansion waves. The high-pressure fuel injection in the low-pressure combustor

environment causes the fuel to expand, penetrating down the combustor height. The fuel

expansion is followed by the barrel shock; confining the fuel plume and raising up the

pressure through the normal injection plane. The interaction o f the fuel injection and the

incoming supersonic air brings about the bow shock that extends diagonally downstream

the step base area hitting the lower combustor wall at X=0.090m. A series of reflection

waves off the upper and lower combustor walls is observed by following the change in

the values of the pressure contours around X=0.136m for the upper wall and X=0.092m

and 0.184m for the lower wall. The pressure drop in the regions upstream and

downstream of the normal injection location is responsible for creating vortices and

recirculation zones that entrain the incoming supersonic air to be mixed with the pilot

fuel injection forming a combustible mixture. The ignition o f this combustible mixture

yields high temperature medium that facilitates the ignition of the normal injection and

supports and maintains the main flame extending downstream towards the combustor

exit.

Figures 4.3 and 4.4 show the temperature and CO2 contours respectively at the

symmetry plane. The step base area exhibits a medium o f high temperature and high

combustion products concentration that surrounds the upstream side of the normal

injection. This region is brought by the ignition of the combustible mixture formed in the

step base area and the consequent release of heat of combustion. This high temperature

region represents the main flame holding mechanism in the investigated configuration. It

is worth nothing that in the main flame area extending downstream o f the normal

injection, the flame possesses its highest temperature and combustion products

concentration in the far field domain towards the combustor exit and attached to the

upper wall. The area closer to the downstream o f the injection location characterizes

lower temperature and products concentration. This observation is one o f the inherent

effects o f the side sweeping of the step that pushes the flow field downstream

substantially releasing the heat of combustion in the far field domain. For such flow field

it is recommended to optimize the combustor length to ensure the complete combustion

and consequently the full liberation of the chemical energy stored in the fuel before

exiting the combustor.


79

The distribution of the Mach number contours at the symmetry plane is presented

in Fig. 4.5. The dominance o f the supersonic flow in the combustor is exhibited except in

few pockets in the upper part o f the combustor. The recirculations upstream and

downstream o f the normal injection in the upper area of the combustor slow down the

flow in these regions creating two subsonic pockets. The 3°-divergence of the upper wall

expands the flow in the upper part of the combustor creating a slight supersonic flow just

upstream o f the reflection waves off the upper wall at X=0.L36m. The compression of

this weak supersonic flow brings about the third subsonic flow area in the far domain

close to the upper wall. Further downstream the flow expansion continues and the flow

exits the combustor slightly supersonic in the upper domain with the exit Mach number

increasing towards the combustor lower wall. The previous flow description well matches

the typical aerodynamics o f such flow field.

4.4.2 Cross Flow ResultsThe flow field features and the combustion characteristics at two cross flow

planes, at X=0.031m and 0.070m, are depicted in Figs. 4.6-4.12. These cross flow planes

are used to present the flow nature inside the flame holding area, which is the step base

area, and at the normal injection station, respectively.

Figure 4.6 shows the temperature contours inside the step base area at X=0.031m.

The two pilot injections are fully surrounded by high temperature medium that extends

downstream to approach the upstream side of the normal injection as discussed in Fig.

4.3. The effect of the side sweeping of the step is apparent in keeping the high-

temperature-combustion gases produced by the ignition o f the pilot injections

concentrated towards the middle area of the combustor with relatively lower temperature

gases at the side walls o f the combustor.

Figures 4.7 and 4.8 show the velocity vectors and stream line contours at two

cross flow locations, X=0.03Im and 0.070m, inside the step base area and at the normal

injection location, respectively. The two plots show the interaction between the incoming

supersonic airflow, coming normal the page, and the pilot and main injections,

respectively. Due the symmetry o f the configuration the streamlines were added to just

one half o f the domain. The vortices formed around the fuel injection help in promoting

the air-fuel muting and consequently lead to more efficient combustion. The high-


pressure normal injection adds to the momentum of the recirculations surrounding it

creating two strong vortices as shown in Fig. 4.8.

Figure 4.9 shows the Mach number contours at the normal injection cross flow

plane. Once the fuel is injected at high pressures it expands in the surrounding lower-

pressure medium. This is evident from the increase of the Mach number in this area that

reaches a value around Mach 3. Due to the effect of the barrel and bow shocks

surrounding the normal fuel injection, the flow is compressed decreasing its Mach

number. Passing the area o f shocks and due to the interaction between the supersonic

incoming airflow and the normal fuel injection, the cross flow again accelerates and

increases the Mach number.

The effects of the expansion and shock areas on the cross flow absolute pressure

and static temperature distributions along the normal injection plane are shown in Figs.

4.10 and 4.11, respectively. The advancement o f the tv/o pilot flames towards the normal

injection plane causes the two high temperature regions to permanently surround the

normal injection as shown in Fig. 4.11. This is a very supportive tool for the ignition of

the normal injection. The regions close to the sidewalls o f the combustor exhibit lower

temperature levels as previously explained in Fig. 4.6. Figure 4.12 presents the CO2 mass

fraction contours at the normal injection plane showing the high combustion products

concentration surrounding the normal injection.

4.4.3 Pilot Injection Plane ResultsThe velocity vectors and stream lines contours, the temperature and CO2 mass

fraction contours at the pilot injection plane, Y=0.0304m, are presented in Figs. 4.13-

4.1S. The interaction of the incoming supersonic air flow, that spills over the swept

sidewalls o f the combustor, and the pilot and normal fuel injections causes the flow

recirculations in the pilot and main injection zones as shown in Fig. 4.13. The pilot

flames initiated at the pilot injection locations and extended downstream surrounding the

main normal injection, at X=0.070m and Z=0, are manifested in Figs. 4.14 and 4.15. The

high-temperature and high combustion products concentrations featuring the pilot flame

area are clearly seen. As previously explained, sweeping of the combustor sidewalls

forces the hot and highly concentrated combustion gases towards the combustor

centerline while the combustor side areas characterize lower temperature and products


81

concentration as shown in Figs. 4.14 and 4.15. The high temperature and combustion

products concentration in the far field domain towards the combustor exit are clear,

which confirms the recommendation of using longer combustors with the swept step

configuration.

4.4.4 Upper-Wall Averaged ResultsThe average distributions of the static temperature and absolute pressure at the

combustor upper-wall are depicted in Fig. 4.16. The way the average temperature and

pressure are calculated is previously presented while discussing the grid independence

study at the beginning o f this chapter.

The slight increase in the upper wall temperature downstream of the combustor

inlet is due to the surface frictional heating. The flow expansion due to the side sweeping

o f the step decreases the temperature of the upper wall downstream reaching the step

edge where the pilot flame is initiated. This is responsible for the temperature increase

that reaches the first temperature peak around the pilot injection ports, X=0.031m. The

incoming supersonic air spills over the sidewalls of the swept step and expands mixing

with the high-temperature pilot flame combustion products in the step base area. This

leads to the reduction o f the average upper-wall temperature downstream the step area

approaching the low-temperature normal fuel injection location, X=0.070m. The ignition

o f the normal fuel injected leads to the second increase in the upper-wall average

temperature downstream of the injection station. The flow compression through the shock

wave reflected off the upper wall around X=0.136m contributes to increasing the rate of

the temperature increase at this location. The high temperature in the far field domain

towards the upper wall explains the increase of the temperature at the combustor exit.

The physics explained in the temperature field applies to the pressure field except

that the high-pressure peak coincides with the normal injection location, X=0.070m.

The propane study yielded the following conclusion. The swept step showed the

ability to hold the propane flame in the supersonic air stream without extinction. It was

found that the side weeping of the combustor exhibits the high, temperature and

combustion products concentration in the far field domain while the area downstream of

the normal injection location characterizes lower temperature and products concentration.

I t is recommended to optimize the combustor length to ensure the complete combustion


82

and consequently the full liberation o f the chemicaL energy stored in the fuel before the

fuel exits the combustor.


Aver

age

Upp

er-W

all

Abso

lute

Pres

sure

, kP

»

83

290

207125 Grid260

1— 402873 Grid

230

200

170

140

110

0 0.025 0.05 0.075 0.1 0.125 0.15 0.175 0.2 0.225Combustor Length, M

Fig. 4.1 Comparison between longitudinal average upper-wall absolute pressure distributions for both grid sizes, grid independence study.


84

Abs., Pres., Pa34 205081033 194336932 183592831 172848730 162104629 151360528 140616427 129872326 1191282 s25 1083841 . r>-24 976400 £23 868959 .BA22 761518 *321 654077 as20 546636 k19 439195 s

I18 33175417 22431316 116872 s15 9804014 8581413 84199 W12 8306911 7752210 729339 719198 697027 658566 624485 601364 582693 575882 546351 52927

0.04 -

0.035 -

Symmetry Plane

1

1I

I

0.015 -

0.005 -

0 0.02 0.04 0.05 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22

Combustor Length, M

Fig. 4.2 Absolute pressure contours at the symmetry plane.


8 5

Static Temp., K34 308233 299732 291231 282730 274229 265728 257227 248726 2402 s25 231624 2231 £23 214622 2061 *521 1976 X20 1891 faa19 1806 s

118 172117 163616 1551 s15 146514 1380 513 129512 121011 112510 10409 9558 8707 7856 6995 6144 5293 4442 3591 274

0.04 -

0.035 -

M 0.025

Symmetry Plane

0.015

0.005 -

0.02 0.04 0.00 0.08 0.1 0.12 0.14 0.10 0.18

Combustor Length, M2 0.22

Fig. 4.3 Static temperature contours at the symmetry plane.


86

C02 Mass Fract34 0.140633 0.136532 0.132431 0.128230 0.124129 0.120028 0.115827 0.111726 0.1076 §25 0.103424 0.0993 g23 0.095122 0.0910 *321 0.0869 as20 0.082719 0.0786 Q18 0.074517 0.0703 116 0.0662 JO15 0.0621 514 0.0579 &13 0.053812 0.049611 0.045510 0.04149 0.03728 0.03317 0.02906 0.02485 0.02074 0.01653 0.01242 0.00831 0.0041

Symmetry Plane

0.035

0.025

0.015

HiiJ m i

0.005

1 111 ’ 1 ' 1 '0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18

Combustor Length, M0.22

Fig. 4.4 CO2 combustion product mass fraction contours at the symmetry plane.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission

87

i

Mach No.36 3.13135 3.04234 2.95233 2.86232 2.77331 2.68330 2.59329 2.50428 2.41427 2.32426 2.23525 2.14524 2.05523 1.96622 1.87621 1.78620 1.75219 1.73318 1.69617 1.60716 1.51715 1.42714 1.33813 1.24812 1.15811 1.06910 0.9799 0.8898 0.8007 0.7106 0.6205 0.5314 0.4413 0.3512 0.2611 0.172

Symmetry Plane0.035

.O f 0.025

S 0.015

0.005

I t0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22

Combustor Length, M

Fig. 4.5 Mach number contours at the symmetry plane.


88

0.035Static Temp., K40 305539 298138 290837 283536 276235 268834 261533 254232 246831 239530 232229 224928 217527 210226 202925 195624 188223 180922 173621 166320 158919 151618 144317 137016 129615 123414 122913 122912 122711 122310 12139 12078 11507 10776 10035 9304 8573 7832 7101 637

0.025

’3 0.02

3 0.015

0.005 X=31mm cross plane

•0.005 0 0.005

Combustor Width, M0.01

Fig. 4.6 Static temperature contours at X=0.031m cross flow plane.

|Reproduced with permission of the copyright owner. Further reproduction prohibited without permission

Com

busto

r H

eight

, M

89

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0

Fig. 4.7 Velocity vectors and streamlines contours at X=0.031m cross flow plane.

, ill''I f ^

' I t V % v ^: ! | v ' *: \ \ \ \ ^ ^ ^ ^T, \ \ \ \ \ \ v

j , ‘ V \ V v \ \ \ > * ,l ' i m V \ \ \x-s j I /h ' 1 1 v I''\ \ v % t ,

- ^ ^ W ,r / 11— i . — /_ / i i

X=31mm cross plane] i>v W V > 't'l /

■ ' y " (H 1 1111' S / r t . I . I l((I '.I ■f . ! i ■ • 1 ■1 A1 A AACJL

- 0.01t

-0.005 0 0.005

Combustor Width, M


-0.01 -0.005 o 0.005 0.01

Combustor Width, M

Fig. 4.8 Velocity vectors and streamlines contours at X=0.070m cross flow plane.


91

Mach No. 0.03540393837363534333231302928272625242322212019181716151413121110987654321

2.9672.9072.8462.7862.7252.6652.6052.5442.4842.4232.3632.3022.2422.1822.1212.0612.0001.9401.8801.8191.7591.6981.6381.5771.5171.4571.3961.3361.2751.2151.1541.0941.0340.9730.9130.8520.7920.7310.6710.611

0.025

*S 0.02

3 0.015

0.005 ~ X=70mm cross plane

•0.01 -0.005 0 0.005

Combustor Width, Mo.oi

Fig. 4.9 Mach number contours at X=0.070m cross flow plane.


92

A b s.> P i c s . , .P a34 205081033 192415232 179749431 167083630 154417929 141752128 129086327 116420526 1037547 §25 910889 . r24 784232 4323 657574 jo t22 53091621 404258 9320 277600 u19 150942 a18 96731 a17 9341216 91030 a15 89992 ss

e14 88147 Q13 8729912 8465611 8103310 786379 751348 724367 667976 596675 548574 542823 537952 400981 30268

0.035

0.025 -

X=70mm c r o s s

0.005

0.015

0.005 -

•0.005

Combustor Width, M

Fig. 4.10 Absolute pressure contours at X=0.070m cross flow plane.


93

Static Temp., K 31 263430 255629 247728 239827 232026 224125 21632423222120191817161514131211

87654321

20842005192718481770169116131534145513771298122011411062

10 9849 905

827748670591512434355277

0.025

3 0.015

0.005 X=70 mm cross plane

- 0.01 -0.005 0 0.005 0.01

Combustor Width, M

Fig. 4.11 Static temperature contours at X=0.070m cross flow plane.


94

C 0 2 Mass F iac t 0.035 P35 0.1042034 0.1011333 0.0980732 0.0950031 0.0919430 0.0888729 0.0858128 0.0827427 0.0796826 0.0766225 0.0735524 0.0704923 0.0674222 0.0643621 0.0612920 0.0582319 0.0551618 0.0521017 0.0490316 0.0459715 0.0429014 0.0398413 0.0367812 0.0337111 0.0306510 0.027589 0.024528 0.021457 0.018396 0.015325 0.012264 0.009193 0.006132 0.003061 0.00003

0.025

5 0.02

3 0.015

0.005 X=70 mm cross plane

-0.005 0.005 0.01

Combustor Width, M

Fig. 4.12 CO2 combustion product mass fraction contours at X=0.070m cross flow plane.


95

0.0125

0.0075

0.005

0.0025

r> r» 5 >

» r.0-0.0025S o

w -0.005

0.0075

Pilot M ection Plane

£ c n x0.01250.04 0.05 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.2

Combustor Length, M

Fig. 4.13 Velocity vectors and streamlines contours at the pilot injection plane, Y=0.0304 m.


96

Static Temp., K 25 318224 306323 294322 282321 270420 258419 246418 234417 222516 218715 210514 198513121110987654321

18661746162615061387126711471027908788668549429

0.0125

0.0075

0.005

€ 0.0025

.0-0.0025S ©

W -0.005

0.0075

Pilot Iniiection Plane

0.01250.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22

Combustor Length, M

Fig. 4.14 Static temperature contours at the pilot injection plane, Y=0.0304 m.


97

C 0 2 M ass Enact.24 0.1433523 0.1373822 0.1314121 0.1254320 0.1194619 0.1134918 0.1075117 0.1015416 0.0955715 0.0895914 0.0836213 0.0776512 0.0716811 0.0657010 0.059739 0.053768 0.047787 0.041816 0.035845 0.029864 0.023893 0.017922 0.011951 0.00597

0.0125

0.0075

0.005

3 0.0025

I■Q-0.0025

W -0.005

0.0075

Pilot uuecfaon Plane

0.0125 i a r r r r0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22

Combustor Length, M

Fig. 4.15 CO2 combustion product mass fraction contours at the pilot injection plane, Y=0.0304 m.


Aver

age

Upp

er-W

all

Stati

c Te

mpe

ratu

re,

K

98

2902600

’Static Temperature 2602400

^ Absolute Pressure230

2200

200

2000170

1800■ 140

1600110

. ♦ 80

1200 50

10000.2250.075 0.125 0.1750.025 0.05 0.1 0.150

Combustor Length, M

Fig. 4.16 Average upper-wall absolute pressure and static temperature distribution along the combustor length.


Aver

age

Upp

er-W

all

Abso

lute

Pres

sure

, kP

a

99

CHAPTER V

ETHYLENE SUPERSONIC COMBUSTION

5.1 IntroductionA detailed description of the physical model, the boundary and initial conditions,

and the important flow parameters characterizing the case study are presented in the

physical model description. The definitions o f the terms used in the presentation o f the

results and their calculation procedures are then offered. The most significant results that

can to the best reflect the flow field characteristics o f the supersonic ethylene combustion

conclude the course o f this chapter. The results presented in this chapter are divided into

three sub-sections. First, the effect of the combustor configuration on the combustion

characteristics and flame holding of the supersonic ethylene flame is presented. Second, a

discussion on the effect of the equivalence ratio of the main fuel injection on the

combustion flow field is followed. The chapter is concluded by analyzing the effect of

the equivalence ratio of the pilot fuel injection on the flame holding mechanism, and the

general features of the flow and energy fields.

5.2 Physical Models and Flow Calculations DescriptionThe supersonic ethylene combustion flow field is investigated through three

physical models; rearward-facing step, cavity without pilot injection, and cavity with

pilot injection respectively. The configurations of the three physical models are described

in details in Sec. 3.7. In the present section, the flow field properties and the detailed

description o f the physical models used in the calculations are provided.

The three investigated physical models share the following flow features. The

incoming supersonic airflow is vitiated with H2O mass fraction of 0.17 and the typical Oi

concentration in air. The incoming air total temperature and pressure are 1800 K and

431.7 kPa respectively. The normal ethylene main injection is gaseous with static

temperature o f 300 K, and equivalence ratio of 0.6 calculated based on the total amount

o f the incoming supersonic airflow. For the pilot injection cases, the parallel ethylene

pilot injection is gaseous with, static temperature o f500 K, and equivalence ratios o f 0.02,


LOO

0.03, 0.045, 0.06, and 0.08 calculated based on the total amount o f the incoming

supersonic airflow.

The pressure inlet boundary conditions are specified for the incoming air inlet,

and both the pilot and main ethylene injections. The flow outlet is supersonic and

specified as pressure outlet boundary that uses first order extrapolation for all parameters

from the adjacent inflow cells. At the supersonic inflow boundaries, the total and static

pressures, the total temperature, and the species mass fractions are specified. For these

supersonic inflow boundaries, uniform conditions are assumed for the primitive

parameters. AIL walls and step boundaries are treated as no-slip adiabatic surfaces.

Applying ftee-stream inlet conditions throughout the entire flow field sets the initial

conditions.

Because of the turbulent nature of the case study, the influence of turbulence on

the reaction rate is taken into account by comparing the Arrhenius reaction rate and the

two eddy-dissipation reaction rates. Out of the three mentioned rates, the limiting

(slowest) rate is used as the reaction rate and the contributions to the source terms in the

species conservation and energy equations are calculated from this reaction rate.

The following global one-step reaction o f ethylene is used to calculate the

Arrhenius reaction rate in the present study:

C2ff4+302------>2C02Jt-2Hz0 (5.1)

The Arrhenius reaction rate constants for the chemistry scheme are listed below,

Pre-exponential factor = 2e+12

Activation energy = 1.256e+08 [J/kgmol]

Temperature exponent—0

The renormalized group (RNG) form of the K-e turbulence model is used. The

turbulence near-wall is treated using the two-layer zonal model.

Because the flow is compressible and a multicomponent mixture, it is modeled as

an ideal gas and the density is calculated as a composition-dependent property. A

composition-dependent specific heat capacity is defined for the multicomponent mixture

material while temperature-dependent heat capacities for the individual species are

specified. The solver computes the mixture’s specific heat capacity as a mass fraction


101

average of the pure species heat capacities. The viscosity and thermal conductivity are

defined as composition-dependent properties.

The coupled solver is used with the explicit formulation and the steady state

approach. The second order upwind scheme is used for discretization of the flow,

turbulence kinetic energy, and turbulence dissipation rate equations.

5.3.1 Stagnation Pressure Efficiency and Entropy ProductionAcross a scramjet combustor, the loss in the stagnation pressure is o f a

fundamental interest, and may introduce an efficiency based on such loss as follows

where r}^ , PT, and n are the stagnation pressure efficiency, the total pressure, and the

total pressure ratio respectively and the subscripts “inlet” and “outlet” refer to the

combustor inlet and outlet station respectively.

The relationship o f interdependence of the total pressure ratio n and the entropy

production AS is defined as [26]:

where 5 and R are the entropy and the gas constant respectively and the subscripts

“inlet” and “outlet” refer to the combustor inlet and outlet station respectively.

This means that the total pressure ratio n depends exponentially upon the

entropy increase. In other words, the total pressure ratio should be expected to decrease

very rapidly as entropy, the natural indicator of the effects o f dissipation or flow losses,

increases. From Eq. (5.3), the entropy production is defined as:

5.3.2 Kinetic Energy EfficiencyThe concept o f kinetic energy efficiency expounded in Refs. [200-202], can

be used to obtain the following definition:

5.3 Definitions

[143]:

(5.2)

(5.3)

(5.4)


103

where mCiW< ,nj is the injected fuel flow rate,

= \ p yc&< (V-n)dA,

p, YCiHt, V , and A are the local fuel density, fuel mass fraction, normal

velocity vector, and cross section area, respectively.

5.3.4 Fuel Penetration and Rate of Fuel DecayThe penetration of the fuel jet is measured at the point in the cross flow plane of

the normal fuel injection, X=0.070m, where the maximum fuel mole fraction has dropped

to 0.005 [52]. Since fuel mole fraction distribution at the normal fuel injection cross flow

plane marks the maximum fuel concentration at the line of symmetry, the fuel penetration

is defined by plotting the fuel mole fraction distribution versus the combustor height at

the cross flow line that marks the intersection o f the symmetry plane and the normal fuel

injection cross flow plane. For easily discussing the penetration results, the penetration

values are presented while measured from the upper-wall of the combustor at the normal

injection location.

The rate at which the maximum fuel concentration at each cross section decays

with axial distance is a good indication of the overall mixing rate o f the jet [52]. The fuel

mole fraction contours at sixteen cross flow planes, at X= 0.001, 0.015, 0.030, 0.031,

0.039, 0.048, 0.0566, 0.065, 0.070, 0.110, 0.150, 0.190, 0.220, 0.260, 0.300, and 0.349m

respectively, are plotted which showed the maximum fuel mole fraction to be located at

the combustor centerline. Accordingly, the combustor symmetry plane is used to define

the value of maximum fuel mole fraction at the sixteen lines making the intersection of

the symmetry plane with the corresponding sixteen cross flow plane that are mentioned

earlier. The longitudinal distribution of the maximum fuel mole fraction, a t each cross

flow plane, through the combustor length is plotted to evaluate the overall mixing rate o f

the fuel jet.

53.5 Upper-Wall Average Absolute Pressure and Static Temperature DistributionFor the heat transfer and design considerations, the longitudinal distribution of the

average absolute pressure and static temperature along the upper-wall of the combustor

are calculated. Sixteen cross flow planes are defined at X=0.001, 0.015, 0.030, 0.031,

0.039, 0.048, 0.0565, 0.065, 0.070, 0.110, 0.150, 0.190, 0.220, 0.260, 0.030, and


105

that extends to towards the upstream side of the normal infection. The temperature

contours depicted in Fig. 5.2(a) show the high-temperatiire gases in the step base area.

These high-temperature gases help in initiating the ignition of the normal fuel Injection

and maintaining its flame. This represents the flame holding mechanism of the step

configuration.

The interaction between the flow recirculations in the step base area and the

incoming supersonic airflow creates a two counter-flow vortices as shown in Fig. 5.3(a)

which presents the velocity vectors and streamline contours at a cross flow plane inside

the step base area, at X=0.040m. The vortices in the step base area direct the flow

outward in the direction o f the combustor sidewalls. These vortices are responsible for

promoting the fuel/air mixing in the lateral direction inside the step base area and

consequently maintaining an efficient combustion in this area. This is evident in the

temperature contours at the same cross flow plane that is depicted in Fig. 5.4(a), which

shows the high temperature area that extends throughout the width of the combustor

surrounding the step edge, Y=0.025m.

The flow field features and energy field characteristics at Y=0.0304m lateral flow

plane that lies in the middle of the step base area and extends longitudinally through the

combustor length are shown in Figs. 5.5(a) and 5.6(a). This lateral flow plane marks the

pilot injection plane for the cases with pilot infection. The flow field in this plane features

a pair of counter flow vortices created in the step base area and surrounding the normal

fuel infection as shown in Fig. 5.5(a). Downstream the normal fuel injection, another

vortical region exists that directs the flow at the wakes o f the injection location towards

the combustor sidewalls. As explained before, these sets of vortices are effective in

enhancing the fuel/air mixing in the regions both upstream and downstream of the normal

injection station. The permanent high temperature regions surrounding the normal

injection station are evident in Fig. 5.6(a), which are due to the vortices formed in their

respective areas as explained before.

The flow field features and energy field characteristics o f the step configuration

that are presented at the symmetry plane, at X=0.040m cross flow plane inside the step

base area, and at Y=0.0304m lateral flew plane exhibited the manifested flame holding

capability o f the step configuration in the supersonic air streams.


106

The injection of the under-expanded high-pressure normal fuel injection, at

X=0.070m, in the low-pressure supersonic air medium originates the strong two pair o f

vortices surrounding the sides o f the normal fuel injection as shown in Fig. 5.7(a). These

vortices push the flow towards the combustor sides. The interaction between the flow

recirculations surrounding the normal fuel injection and the incoming supersonic airflow

creates the two counter-flow vortices that are shown in Fig 5.7(a). Fig. 5.8(a) presents the

cross flow temperature contours at the normal injection plane. It shows the high-

temperature areas at the sides of the normal fuel injection especially towards the

combustor sidewalls. These hot gases contribute to the holding mechanism o f the flame.

When comparing the velocity vectors and streamlines contours for the wedge

configuration, Fig. 5.1(b), to that of the step configuration that is previously explained,

Fig. 5.1(a), it is shown that the existence of the wedge expands the recirculations to

predominate the step base area. This yields a better fuel/air mixing scheme and

consequently a more efficient combustion. This is manifested in the higher temperature

levels in the step base area o f the wedge configuration as depicted in Fig. 5.2(b)

compared to that o f the step configurations, Fig. 5.2(a). Another advantage for the wedge

configuration, which distinguishes it from the step configuration, is the pocket o f high

temperature gases that is trapped between the wedge straight surface and the normal fuel

injection. This area takes over the role o f a permanent energy source that supports the

ignition o f the normal fuel injection. The better combustion characterization for the

wedge configuration is not only limited to the step base area but is presented in the

downstream flow field as well by observing the higher temperature levels in the case of

the wedge configuration compared to those of the step configuration in the downstream

flow field.

The wider vortices in the step base area in the wedge configuration over that of

the step configuration, as shown in Fig. 5.1(a) and (b), stretches the cross flow in the step

base area towards the bottom wall. This is evident when comparing the velocity vectors

and streamline contours of both configurations at X=0.40m cross flow plane that are

presented in Fig. 5.3(a) and (b). The preference o f the wedge configuration over the step

configuration in supporting the ignition and the flame holding o f the normal fuel injection

is demonstrated when observing the more homogenous temperature distribution and


107

higher temperature levels pertinent to the wedge configuration in the cross flow plane,

X=0.40m, over that of the step configuration as shown in Figs. 5.4 (a) and (b).

When comparing Figs. 5.5(a) and (b), which present the velocity vectors and

streamline contours for both the step and wedge configurations at the Y=0.0304m lateral

flow plane, it is shown that the existence of the wedge downstream o f the step stretches

the two small counter flow vortices created in the step configuration to fully and laterally

occupy the entire cavity area. Therefore, the wedge configuration creates more

homogenous flow medium that enhances the combustion characteristics in this area. This

is obvious by observing the higher and more homogenous temperature distribution in the

step base area (the cavity area) in the case of the wedge configuration over that in the step

configuration as shown in Figs. 5.6(a) and (b), which present the temperature contours at

the Y=0.034m lateral flow plane. Again the influential effect of the wedge in both flow

regions upstream and downstream o f the normal fuel injection is presented by observing

the higher temperature levels in the entire flow field especially in the region surrounding

the normal fuel injection. This observation stresses on the superiority o f the wedge

configuration as a flame holding mechanism over that of the step configuration.

The comparison between the velocity vectors and streamline contours for both

configurations is done by observing Figs. 5.7(a) and (b), which present these vectors and

contours at the normal injection station, X=0.70m. The wedge existence enlarges the

recirculations area in the normal direction. This can be evaluated by locating the meeting

point of the two counter flow vortices that marks the interaction between the normal

injection and the incoming supersonic airflow. For the step configuration, this point is at

Y=0.01m compared to Y=0.007m for the wedge configuration. Comparing the

temperature contours for both configurations at the same cross flow plane, as shown in

Figs. 5.8(a) and (b) shows the positive effect of the wedge in increasing the temperature

of the gases surrounding the normal fuel injection plume compared to that of the step

configuration. This is translated in more potential for holding the flame in the supersonic

air streams.

One more comparison between the temperature flow fields o f the step and wedge

configurations in the downstream cross flow location at X=0.220m is presented in Figs.


L08

5.9(a) and (b). These figures show the overall higher temperature levels exhibited when

using the wedge configuration over that of the rearward-facing step configuration.

The advantage of the wedge configuration, over the step configuration, is clearly' * ' *shown when comparing the axial combustion efficiency for both configurations as

depicted in Fig. 5.10. The remarkable difference at the normal fuel injection location

emphasizes on the supportive effect of the wedge that is just upstream of the normal fuel

injection, on burning more quantity of fuel compared to that of the step configuration.

Figure 5.11 shows the fuel penetration for both configurations under

investigation. The wedge configuration exhibits more fuel penetration through the

combustor height than that offered by the step configuration. This can be attributed to the

effect of the wedge, which enlarges the recirculations in the normal direction, as shown in

Figs. 5.7(a) and (b), and consequently permits the deeper fuel penetration compared to

that o f the step configuration. This observation is another advantage that is added to the

wedge configuration.

The comparison between the rates o f decay of the fuel concentration for both

configurations is presented in Fig. 5.12. In the step base area, the rate of fuel entrained to

this area, due to the created vortices, is higher than the rate o f fuel depletion by

combustion. This is the reason for the increase in the fuel mole fraction in this area for

both configurations. Due to the fact that the general combustion efficiency for the wedge

configuration is higher than that of the step configuration, as explained in Fig. 5.10, the

rate o f fuel depletion by combustion in the wedge configuration is higher than that in the

step configuration and consequently its rate of increase of the fuel concentration is lower

than that o f the step configuration. This leads to the conclusion of the better mixing

capabilities o f the wedge configuration compared to that of the step configuration in the

main flame holding region, the step base area. In the downstream o f the normal injection

station the rate o f mixing in the middle o f the combustor is shows in favor for the wedge

configuration but in the far field domain the step configuration shows higher mixing

capabilities, which enables its fuel concentration to converge towards the respective value

of the wedge configuration at the exit o f the combustor. This shows the good mixing

capabilities o f the step in the far field domain.


109

The upper-wall average static temperature distributions across the combustor

length for both configurations are presented in Fig. 5.13. The wedge configuration

exhibits higher upper-wall temperature along the combustor length than that of the step

configuration. The same behavior is observed when comparing the flow field average

static temperature distributions for both configurations as depicted in Fig. 5.14. It is

worth of noting to observe the convergence o f the values o f the average temperatures for

both configurations, as compared in Figs. 5.13 and 5.14, at the combustor exit. This

observation stresses on the enhancement of the mixing capabilities of the step

configuration in the far field domain reaching that of the wedge configuration, as

discussed previously.

The longitudinal distributions o f the average absolute pressure for the two

configurations under investigation both at the upper-wall and in the flow field are

depicted in Figs. 5.15 and 5.16 respectively. The two figures show that, although the

pressure flow field is less sensitive to the configuration effect than the temperature flow

field, as presented in Figs. 5.13 and 5.14, but yet the wedge configuration shows higher-

pressure values in the region around the normal injection location.

Although the average absolute pressure values at the combustor exits for both

configurations do not have that much difference, but due to the higher average Mach

number at the combustor exit o f the wedge configuration over that of the step

configuration, the wedge-exit stagnation pressure is higher than that of the step. This

leads to the higher stagnation pressure efficiency for the wedge configuration, which

values 44.68%, while that of the step is 43.098%. This implies that the wedge

configuration exhibits lower stagnation pressure loss than that associated with the step

configuration. The same efficiency measure can be, noticed b y c o m p a r in g the entropy

production for both configurations, which amounts 254.97 and 261.65 J/kg. K for the

wedge and step configurations respectively. According to the definition o f the kinetic

energy efficiency presented in Sec. 5.3.2, the wedge configuration showed higher kinetic

energy efficiency than the step configuration. The kinetic energy efficiency for the wedge

configuration is 45.327% while its contemporary value for the step configuration is

43.679%.


n o

The comparisons presented in Figs. 5.1-5.16 and that are held between the

efficiencies’ values confirm the general advantage o f the wedge configuration over the

rearward-facing step configuration with regard to holding the fiame in the supersonic

airstreams, producing overall higher temperature medium throughout the combustor, and

exhibiting lower flow losses and higher combustor efficiency.

5.4.2 Effect of Main Fuel Equivalence RatioIn this section, the effect o f the equivalence ratio (ER) o f the normal fuel injection

on the general flow field and combustion characteristics of the wedge configuration is

presented. Two ER values o f 0.60 and 0.45 for the normal fuel injection, calculated based

on the mass flow rate of the total incoming supersonic airflow, are used with no pilot

injection. The ER value is set by changing the main fuel injection pressure. Increasing the

fuel injection pressure increases the ER through the increase in the fuel mass flow rate.

The comparison between the flow field features, the energy field characteristics, and the

efficiencies levels o f the two case studies is held. The only changing parameter in this

comparison is the ER value while all other boundary conditions and solution parameters

are set the same.

When comparing the velocity vectors and streamlines contours o f the ER of 0.60,

Fig. 5.1(b), with that of the ER of 0.45, Fig. 5.1(c), it is shown that, using the lower ER,

the fuel entrained in the step base area forms a single recirculation region that fills in the

entire step base area while increasing the fuel injection pressure, ER of 0.60, more fuel is

entrained in the step base area that intensifies the recirculations but confines it in a

smaller region that is close to the normal injection location. One more observation that

worth o f noting is the effect o f the higher injection pressure, higher ER, in stretching the

recirculation region in the vertical direction towards the combustor lower wall. This is

shown in the deflection o f the incoming supersonic air stream downward towards the

combustor lower wall. The reason for this behavior is the higher penetration for the

normal fuel injection in the case o f the higher ER value as shown in Fig. 5.17, which

depicts the comparison between the fiiel penetration patters o f the two ER values. The

temperature contours at the symmetry plane for the ER values o f 0.60 and 0.45 are

presented in Figs. 5.2(b) and (c) respectively. The comparison between the two

temperature fields shows the better temperature distribution in the step base area in the


I l l

lower ER case. It is also noticed that the higher injection pressure has a less advantageous

effect in reducing the temperature in the area close to the wedge angled-surface,

compared to that exhibited by the lower ER case, while it has a more advantageous effect

in the vertical stretching the high temperature medium backing the upstream side of the

normal fuel injection. This forms a supportive means of maintaining the main flame

while facing the high-speed incoming air. Another advantage for the higher ER case is

the pocket o f high temperature gases that is trapped between the wedge straight surface

and the normal fuel injection. This area plays the role of a permanent energy source that

supports the ignition of the normal fuel injection. In the downstream flow Held, the

higher ER case exhibited, in general, higher temperature levels compared to those of the

lower ER case.

The vertically widened vortices in the step base area in the higher ER case over

that o f the lower ER case, as seen in Figs. 5.1(b) and (c), stretches the cross flow in the

step base area downward towards the bottom wall. This would be clearly observed when

comparing the velocity vectors and streamlines contours of both ER cases at X=0.40m

cross flow plane as presented in Figs. 5.3(b) and (c). The additional fuel entrained in the

step base area, in the case o f the higher ER, with its relatively low temperature breaks the

temperature homogeneity featured in the lower ER case, Fig. 5.4(c), to produce a

temperature spectrum with low values in the middle of the combustor and high values

towards the combustor sidewalls. This is shown in Figs. 5.4(b) and (c). Although the high

ER case lacks the homogenous temperature distribution in the step base area, it posses

higher temperature levels in both the step base area and the lower half of the combustor

as shown in Figs. 5.4(b) and (c).

Figures 5.5(b) and (c) present the velocity vectors and streamline contours for

both ER cases at the Y=0.0304m lateral flow plane, it is shown that the additional

entrained fuel in the step base area, due to the higher injection pressure, develops a

couple o f vortices in the middle of the lateral plane. This couple o f vortices lumps

laterally towards the combustor sidewalls to form a recirculation that fully occupy the

entire step base area. Moreover, the higher injection pressure, higher ER case, lessens the

intensity o f the vortices formed in the down stream proximity o f the normal injection

station. These flow field differences are noticed when comparing Figs. 5.5(b) and (c).


L12

The temperature flow fields for the two ER cases, Figs. 5.6(b) and (c), present the same

observation reported in the X=0.040m cross flow plane, which states that although the

high ER case lacks the homogenous temperature distribution in the step base area, it

exhibits higher temperature levels in both of the sides of the step base area and the

combustor flow field downstream of the normal fuel injection location, as shown in Figs.

5.6(b) and (c).

The comparison between the velocity vectors and streamline contours for both

ER’s at the normal injection station, X=0.070m, is held by observing Figs. 5.7(b) and (c).

The low-pressure region formed in the neighborhood of the normal injection location

creates the strong flow recirculations that intensify with the increase of the injection

pressure, increasing ER, especially in the combustor upper-wall area surrounding the

injection port. This brings about a better vortical pattern in the normal injection plane in

the case of the higher ER as shown in Figs. 5.7(b) and (c). Comparing the temperature

contours for both ER cases at the same cross flow plane, as shown in Figs. 5.8(b) and (c),

it is shown that the area surrounding the normal fuel injection port features higher

temperature gases in the higher ER case compared to that of the lower ER value. This

helps in igniting the normal fuel injected and enhancing the potential for holding the

flame in the supersonic air streams. Moreover, the higher ER case presents higher

temperature levels in the bottom half o f the combustor that shares in elevating the

average temperature in the cross flow plane of the normal fuel injection.

The comparison between the temperature flow fields of the two ER cases in the

downstream cross flow location at X= 0.220m is presented in Figs. 5.9(b) and (c). These

figures show the overall higher temperature levels exhibited when using the higher ER.

This is attributed to the higher amount of heat released in the flow field as a result o f the

more fuel injected in the case o f the bigger ER.

In the comparisons held between the 0.60 and 0.45-ER cases in all planes; the

symmetry plane, the X=0.040m and X=0.070m cross flow planes, and the Y=0.0304m

lateral flow plane, the higher ER case showed less homogenous combustion

characteristics in the step base area that is presented in the less homogenous temperature

distribution in the entire step base region. At the same time, the comparisons held at the

symmetry and lateral planes show higher temperature levels in the downstream flow field


114

level of losses compared to those associated with the 0.45-ER case. The kinetic energy

efficiency o f the 0.60-ER case is 45.327% while the contemporary value o f the 0.45-ER

is 42.844%.

The comparisons presented in Figs. 5.1-5.9 (b) and (c) and Figs. 5.17-5.23 and

that held between the values o f the efficiencies shows the positive effect of increasing the

ER of the normal fuel injection on the general flow field features and energy field

characteristics in the combustor except in the step base area in which the lower ER

features better temperature distribution and higher combustion efficiency.

5.4.3 Effect of Pilot Injection Equivalence RatioThe effect o f the pilot injection equivalence ratio (ERp) on the flame holding

mechanism and combustion characteristics of ethylene in supersonic air streams is

studied. The only changing parameter in this study is the value of the ERp while keeping

all other boundary conditions and flow parameters the same. Changing the ERp is

achieved by changing the pilot injection pressure. The ERp is directly proportional to the

injection pressure, which means increasing the pilot injection pressures leads to the

increase in the ERp. There are seven values for the ERp investigated and these are 0 ,0 .01,

0.02,0.03,0.045,0.06, and 0.08. The ERp for each one of the three-fuel pilot injections is

calculated based on the total amount of the incoming supersonic airflow and the fuel

mass flow o f each individual pilot injection.

The flow field features and energy field characteristics o f the different ERp-case

are discussed and compared to determine the optimum ERp that achieves the aspired goals

o f holding the flame and enhancing the combustion characteristics. The injection pressure

that yields 0.01 ERp could not resist the backpressure in the step base area. The boundary

condition for the absolute injection pressure is 92705.8 Pa while the average

backpressure for the flow field surrounding in pilot injection ports in the cavity area is

95798.6 Par therefore, reversed flow through the pilot injection ports dominated the flow

field. This is demonstrated in Figs. 5.24(a), (b), and (c), which show the velocity vectors

and streamlines contours at the three fuel injection ports. The streamlines clearly show

the reversed flow getting out o f the injection ports. This observation nullified the

possibility o f using 0.01-ERp in the pilot injection study. Consequently, the 0.01-ERp-case


115

was excluded from the comparison study of the pilot equivalence ratio. Figure 5.25

presents the reversed flow through the central pilot injection port at the symmetry plane.

The comparisons between the velocity vectors and the temperature contours o f the

six pilot injection equivalence ratios of 0.0, 0.02, 0.30, 0.045, 0.06, and 0.08 are held at

different flow planes, the symmetry plane, the X=0.040m plane, the pilot injection plane,

and the X= 0.070m plane.

The effect o f ERp on the flow field features in the step base area is presented in

Figs. 5.26(a)-(f). The pilot injection adds to the momentum of the recirculations in the

cavity area. This stretches the vortical structure to dominate the entire cavity area and

intensifies the recirculations. There is a common pattern noticed in the figure, that is the

progressive vertical confinement of the vortices, which recirculate inside the cavity

region, with the increase of the value of the ERp except at 0.02-ERp. This can be

perceived by defining the vertical location of the tangent of the outer most recirculating

streamline that deflects and revolves inside the cavity region. For ERp o f 0.0, 0.02, 0.30,

0.045, 0.06, and 0.08, the vertical locations of the tangents are 0.01265, 0.01353,

0.01112, 0.0106, 0.0103, and 0.00926 m respectively, measured from the combustor

upper wall of the cavity region.

This observation demonstrates the passive effect for the pilot injection on the flow

field quality in the cavity region. This negatively affects the fuel/air-mixing scheme in the

region, which is expected to provide the holding mechanism for the flame of the normal

injection. The only exception for this general observation is the 0.02-ERp-case. This

stretches the recirculation region both axially, through the combustor length to occupy

bigger area in the cavity region, and vertically, through the combustor height to improve

the interaction with the incoming supersonic air flow. This is presented in Fig. 5.26(b).

The effect of the ERp on the fuel penetration at the normal injection plane is

presented in Fig. 5.27. The ERp decreases the penetration of the normal fuel in the

incoming supersonic airflow, which reduces the available chances for the fuel/air mixing

in the main-flame region downstream o f the normal injection station. This should

negatively affect the combustion characteristics o f the main flame extending downstream

towards the combustor exit. The only exception for this trend is the 0.02-ERp-case, which


116

features the highest fuel penetration value matching that of the 0.0-ERp-case. This is

another precursor for the superiority o f the 0.02-ERp-case in the range of ERp’s examined.

The effect of the ERp on the quality of the fuel/air-mixing scheme is shown in Fig.

5.28, which presents the rates of decay of the fuel along the combustor length for the six

ERp under investigation. This figure shows that the increase o f the ERp decreases the rate

o f decay of the fuel concentration (i.e., the slope of the lines) in the cavity region while it

does not affect it downstream the normal injection station. It is also shown that

downstream o f the normal injection location, the level of initial mixing is higher in the

cases of the smaller ERp (i.e., the lines are shifted downward in the vertical direction).

The progressive confinement of the recirculations in the cavity region with the increase

of the ERp, as shown in Fig. 5.26, leads to the decrease in the decay of the fuel

concentration and as a result affects the mixing rate in the cavity region negatively. The

0.02-ERp-case possesses the best mixing quality in the cavity region among all other

cases with pilot injection. In the region downstream of the normal injection and due to

the increased quantity o f pilot fuel injected, the initial mixing is seen to be higher in the

cases of lower ERp without affecting the rate of decay of the fuel concentration.

Comparing the rate of fuel decay for 0.0-ERp and 0.02-ERp-cases in the downstream flow

field shows that although the 0.02-ERp-case exhibits higher initial level of fuel mole

fraction, it offers a higher mixing rate (higher line slope) that converges its fuel mole

fraction value towards its lower counterpart value of the 0.0-ERp-case far downstream at

the combustor exit. This adds another feature for the 0.02-ERp-case.

The temperature contours at the symmetry plane for the investigated ERp-cases

are exhibited in Figs. 5.29(a)-(f). The no-pilot case presents the highest levels o f

temperature distribution in both flow field regions, the cavity region and the downstream

region. The excessive addition o f the relatively Iow-temperature-pilot injection in the

small cavity area quenches the flame entrained inside the cavity region, by the vortical

effect in this region. This effect reduces the overall temperature levels in the cavity

region that extends to surround the upstream part o f the normal injection. This leads to

deterioration in the combustion characteristics of the normal fuel injection. This is

obviously noticeable when comparing the levels of the temperature contours in the cavity

and downstream flow fields o f the 0.03-ERp, 0.045-ERp, 0.06-ERp, and 0.08-ERp-cases,


L17

with those of 0.0-ERp-case as presented in Figs. 5.29(a),(c)-(f). The 0.02-ERp-case is

excluded from the general observation mentioned in the previous statement. The

comparison between the temperature contours at the symmetry plane for the 0.02-ERp and

0.0-ERp-cases reveals an advantage for the 0.02-pilot injection equivalence ratio case

over the 0.0-ERp-case and all examined ERp-cases. This is shown in Figs. 5.29(a), (b).

Although the 0.02-pilot injection reduces the temperature in the area occupied by the

relatively lower-temperature pilot injection, the three pilot flames extend downstream

towards the normal injection and the energy liberated by the pilot flames elevates the

temperature levels in the upper part of the cavity region and at the vicinity o f the normal

injection. This behavior generally increases the temperature o f the upstream face of the

normal injection, which enhances the flame holding mechanism of the normal fuel

injection.

The effect of the pilot equivalence ratio on the general energy field at the

symmetry plane, which is discussed above, is now presented at the pilot injection plane.

This is presented by comparing the temperature contours depicted in Figs. 5.30(a)-(f) for

the six investigated pilot injection cases. The suppressive effect of the pilot injection on

the temperature field in both o f the cavity and downstream regions is clear when

comparing Figs. 5.30(a) and (c)-(f). The positive effect o f the 0.02-ERp on elevating the

temperature levels in the cavity area is shown in Figs. 5.31(a), and (b) which presents the

temperature contours in the pilot injection plane with a focus on the region surrounding

the normal injection. The 0.02-ERp-case exhibits an advantage over the 0.0-ERp-case by

offering a temperature increase o f200° in the combustor central area, which is supporting

the upstream side of the normal injection and around 120° temperature increase at the

sides o f the combustor. This observation represents the main advantage o f the 0.02-ERp-

case over the 0.0-ERp-case in promoting the flame holding mechanism in the cavity

region.

The temperature contours at X=0.040m inside the cavity for the six ERp-cases

under investigation are shown in Figs. 5.32(a)-(f). The superiority of the 0.02-ERp-case is

affirmed when com paring its temperature levels at this cross flow location to those o f the

0.0-ERp-case as presented in Figs. 5.32(a) and (b). The 0.02-ERp-case possesses higher

temperatures at the upper part and the sides o f the combustor. This high temperature flow


118

marches downstream to envelope the upstream side to the normal fuel injection. This

behavior stands in favor o f the 0.02-ERp-case while comparing i f to the 0.0-ERp-case in

regard to the quality o f the flame holding mechanism. The deterioration of the

combustion characteristics in the cavity area with the progressive increase of the ERp is

shown when comparing the decreasing temperature levels exhibited in Figs. 5.32(c)-(f),

which are pertinent to 0.03-ERp, 0.045-ERp, 0.06-ERp, and 0.08-ERp-cases respectively,

and those o f the base line case, 0.0-ERp, which are presented in Fig. 5.32(a).

The temperature contours at the normal injection plane, X=0.070m, for the six

ERp-cases under investigation are presented in Figs. 5.33(a)-(f). In this figure, the positive

effect of the pilot-injection equivalence ratio is revealed. Examining the values of the

temperature contours shows that increasing the ERp increases the temperature level in the

upper part of the combustor surrounding the normal injection. It is noticed that 0.03-ERp-

case exhibits the highest temperature level (3355 K) among the other investigated ERp-

cases. Although further increase in the ERp decreases the maximum temperature level, the

area inhabited by the high temperature values is getting larger especially in the 0.06-ERp-

case which reports the highest area-weighted average static temperature in the X=0.070m

cross flow plane.

The longitudinal distribution of the area-weighted average flow field static

temperature is plotted in Fig. 5.34 for the ERp under investigation. For easily analyzing

the effect o f the ERp on the temperature flow field along the combustor, the data of Fig.

5.34 is presented in Table 5.1. Examining the values of the flow field average

temperature reported in Fig. 5.34 and Table 5.1 shows that part of the relatively lower-

temperature pilot fuel injected in the small cavity region ignites releasing the energy of

combustion that raises the temperature in the cavity area. This is the reason for the

increase of the average temperature o f the 0.02-ERp over that o f the no-pilot injection

case, 0.0-ERp-case a t X=0.040m as shown in Fig. 5.34 and Table 5.1. Further increase in

the relatively low-temperature pilot injection does not sufficiently provide high quantity

o f energy o f combustion that can counteract the quenching effect o f the pilot fuel injected

in the cavity region. This leads to the continuous temperature decrease in the cavity

region with further increase in the pilot injection equivalence ratio, ERp, as presented in

Fig. 5.34 and Table 5.1 a t X=0.040, and 0.060m. The pilot fuel that does not ignite in the


119

cavity region marches downstream along the combustor and part of it starts igniting in the

high-temperature side areas o f the combustor at the normal injection station. The 0.06-

ERp-case shows the highest average temperature at the normal injection plane,

X=0.070m. Further increase in the ERp entrains more quantity of pilot fuel whose heat

absorption effect is higher than the heat liberated by its combustion and as a result the

average flow temperature decreases for the 0.08-ERp-case at X=0.070m plane. The partial

consumption through combustion of the remaining fuel pilot injection that is entrained in

the flow field narrows the temperature difference shown at X=0.110m plane. Further

downstream, the balancing effect o f heat liberation through the ignition of the pilot fuel

and the heat absorbed by the extra non-ignited fuel keeps the 0.02-ERp-case featuring the

highest flow field average static temperature reaching the combustor exit as shown in Fig.

5.34 and Table 5.1. A thorough examination of the average flow field temperature shows

that 0.02-ERp-case represents the optimum pilot fuel injection choice that can achieve the

highest temperature in the flame holding region, cavity area, and downstream of the

combustor.

The continuous liberation of the heat due to the partial combustion o f the pilot

fuel, that did not ignite in the cavity area and is entrained in the flow downstream the

combustor, interprets for the increase of the flow field area-weighted average absolute

pressure with the increase in the value of the ERp as depicted in Fig. 5.35. Although the

entrained pilot fuel keeps igniting through its way downstream the combustion, the

largest amount of pilot fuel ignites at the normal injection location. This brings about the

large difference in the average pressure values that take place at the normal injection

location, X=0.070m. This difference converges in the downstream o f the combustor. A

slight increase in the flow pressure at the combustor exit with the increase of the ERp,

leads to a slight increase in the stagnation pressure efficiency, as depicted in Fig. 5.36, a

slight decrease in the entropy production, as depicted in Fig. 5.37, and a slight increase in

the kinetic energy efficiency, as depicted in Fig. 5.38.

Increasing the pilot injection equivalence ratio, ERp, increases the amount of fuel

that does not ignite in the cavity region and is entrained in the flow field to be ignited in

part downstream the combustor. As the combustion efficiency is defined to be inversely

proportional to the quantity o f fuel unbumed, the combustion efficiency decreases with


120

the increase in the ERp, as shown in Fig. 5.39. This figure emphasizes on the advantage of

the 0.02-ERp-case over the other competing pilot injection cases by featuring the highest

combustion efficiency in the flame holding region, the cavity area. This combustion

efficiency is close to that o f the no-pilot injection case, 0.0-ERp.

Taking into consideration both of the flame holding capabilities and the

combustion and flow field qualities, the 0.02 case showed the optimum ERp value that

can achieve the trade off between the mentioned objectives. This conclusion is reached

after carefully analyzing the data presented in Figs. 5.24-5.39.

The main findings attained out o f the ethylene study can be summarized in the

following points. The step configuration with no pilot injection can afford the flame

holding mechanism in the supersonic air stream by creating the flow recirculations in the

step base area and featuring permanent high temperature region surrounding the normal

fuel injection. The step configuration showed good mixing capabilities in the far field

domain. The wedge configuration proved superiority over the generic rearward-facing

step configuration in holding the ethylene flame in the supersonic airstreams, producing


losses and higher combustor efficiency. The increase in the equivalence ratio of the






equivalence ratio increases the temperature of the upstream face of the normal injection

and enhances the flame holding mechanism. 0.02-pilot equivalence ratio presented the

optimum pilot injection case that can promote the flame holding mechanism and keep

good combustion and flow field qualities. While the further increase o f the pilot injection


to the deterioration of the flame holding mechanism, the excessive pilot fuel injected

expresses its effect by increasing the average flow field static temperature and absolute



121

£*5,

2IS■Ou&2I&S9

w•a*5Ci£9

C33)22S3

1■&

iia9

JB

9S3SJ9

•C

■O

S?9

9ss1

•U)9

2£

SCMO

s

©s

sCO

oCO

in

.1

29

I1sf f

s

0 !W

CM CO CO CO CM inO y— y— y— T— y—CO CO CO CO CO COCM CM CM CM CM CM

CO ^ h* CMCO 5 CO CO CO COCO W) CO CO CO COCM CM CM CM CM CM

in i - cm coT - i - O O ® C 0

<«3- ■<* T f 00 CO CM CM CM CM CM CM

co co ^ o i o inT - o CO (O (O M-CO CO CM CM CM CMCM CM CM CM CM CM

O)® O J N ^ N T - CO CM CM •’T CO CO CD 05 05 05 05

00 CO CO CMO 05 CO O in 05t— O O O 05 COCM CM CM CM t- r -

1 - CO CO r^. 05 CMo co co co r- coo) 05 co r* h* co

05 N ' t T - Wco in co cm

r^. r^. iv. co co in

Is*. CO CO CO 05 05 CO CO CO CO CO GO CM CM CM CM CM CM

•«- t- O O 1 - 0S S S S N SCM CM CM CM CM CM

CO 00 fv 1^. 00 h -in in in in in inCM CM CM CM CM CM

cm co cp coI S


122

0.035 R

0.03 -

s■O’

0.025 -

S 0.0201£ua

!a0.005

t fli 11

0.015 rC-—-T~—

0.01

0.025 0.05 0.075C o m b u s to r L e n g th , m

(a) 0.60-step, nopilot

0.035 R'v'W

0.025

0.01

0.005

0.025 0.05C o m b u s to r L e n g th , m

(b) 0.60-wedge, nopilot

0.075


Com

busto

r He

ight,

m

123

0.035

0.03

0.025

0 .02

0.015

0.0 L

0.005

Fig. 5.1 Velocity vectors and streamlines contours at the symmetry plane.

I I I: M 'J r / i i / 1 t f t w 111f W■41 ; /.y r / / f f / /J / Ur

C om bus to r Length, m (c) 0.45-wedge, nopUot


124

Static Tem p.

26 307725 297424 283323 279922 269321 255220 245819 241218 227117 213016 199015 184914 170813 156812 147911 142710 12879 12718 12597 11466 10055 8654 7243 5832 4431 346

Static Temp.26 307725 297424 283323 279922 269321 255220 245819 241218 227117 213016 199015 184914 170813 156812 147911 142710 12879 12718 12597 11466 10055 % 8654 7243 . 583'2 4451 346


0.04

0 .035

0.03s

|Ss 0.025

Xua 0.02

a 0.015

0.01

0.005

0.60>Step, No pilot

0.05

(a)

0.1 0.15 0.2 0.25C om bustorLeagth, m

0.60-step, nopilot

0.35

0.04 0.60-Wedge, Nopilot

0.035

0.03

u 0.025

0.015

0.01

0.05 0.1 0.15 0.2 iCombustor Length, m

0.25 0.3 0.35


125

Static Temp.26 307725 297424 283323 279922 269321 255220 245819 241218 227117 213016 199015 184914 170813 156812 147911 142710 12879 12718 12597 11466 10055 8654 7243 5832 4431 346

0.45-Wedge, Nopilot

0.035

u 0.025

0.015

0.005

0.05

(c)

0.1 0.15 0.2 0.25Combustor Length, m

0.45-wedge, nopilot

0.3 0.35

Fig. 5.2 Temperature contours at the symmetry plane.


Com

busto

r He

ight

, m

•0.01 >0.005 0 0.005 0.01C o m b u sto r W id th (m )


0.035

0.03

0.025

0.02

0.015

0.01

0.005

° -0.01 -0.005 0 0.005 0.01C o m b u sto r W id th , m



Com

busto

r He

ight

, m

127

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0

Fig. 5.3

i ,/ hi i iF H j I f '4 i L 1' I V ,; H o . f e - W c U e ‘l l , l x 4 ‘4 0 m k d r o ! siidPI^ O m m d r o s

is1 ;5I \ .. \ u V V.v \

u \ \ \ Wv\ , V 1» \ L v V 1,1

^ W ' V 1!.!

' \ vt\ ' l N

f t 1 ' 1 1. 11 i i i1 I M ill1

■ i ; , v ,

"1 f i l l • ‘ l . W V / / ' /I f 1 V ' h j A / ' J j z f r

,!fl

7 // >, * / r

- 0.01 -0.005 0 0.005C o m b u s to r W idth , m

0.01

(c) 0.45-wedge, nopilot

Velocity vectors and streamlines contours at X= 0.040m plane.


128

Static Temp* 0.035

0.03

0.025

31 327230 325129 322228 318327 314726 309625 302224 2897 s

. —23 283522 2773 -C21 2699 u20 2648 £

tm19 261118 2595 a

S'IT 252316 2398 M

S15 2273.14 2149 ■613 2024 V12 18991L 177410 16499 1524 - . *8 14677 14556 14375 14004 13273 12882 12781 1275

0.015 *,

0.01

0.005. 0.60-Step, Nopilot X= 40mm Cross Plane

-0.01 •0.005 0 0.005Combustor Width, m


0.01

Static Temp. 31 3272

3251 3222

0.035

30292827

24232221201918

31833147

26 309625 3022

13121110987654321

2897283527732699264826112595

IT 2523 16 239815 227314 2149

2024189917741649152414671455143714001327128812781275

0.025

*3 0*02

3J 0.015

0.60-Wedge, Nopilot X= 40mm Cross Plane0.005

•0.01 -0.005 0 0.005Combustor Width, m

Ob) 0.60-wedge, nopilot

0.01


129

Static Temp. 0.035 F"31 327230 325129 322228 318327 314726 309625 302224 289723 283522 277321 269920 264819 261118 259517 252316 239815 227314 214913 202412 189911 177410 16499 15248 14677 14556 14375 14004 13273 12882 12781 1275

T 7

0.025

S 0.02

0.015

0.45-Wedge, Nopilot X= 40mm Cross Plane

0.005

-0.01

(C)

-0.005 0 0.005Combustor Width, m

0.45-wedge, nopilot

o.oi

Fig. 5.4 Temperature contours at X= 0.040m plane.


130

0.005

-0.005

0.05 0.075 0.1 0.125 0.15 0.175 0.2 0.225 0.25 0.275 0.3 0.325 0.35Combustor Length, m


S*•o

la2

0.01

0.005

eu

-0.005

- 0.01






Fig. 5.5 Velocity vectors and streamlines contours at the pilot injection plane, Y=0.0304m.


132

Static Temp.26 336425 330124 318023 312522 307921 300420 298519 294018 286417 282216 278015 276214 275413 256912 238311 219810 20129 18278 16417 14566 12705 10854 8993 7142 528I 343

0.01

0.005 -

4

•0 .005 -

•0.01 -i

&

0.05 0.075 0.1 0 .125 0 .15 0 .175 0.2 0 .225 0 .25 0 .275 0.3 0 .325 0.35

C om b u stor L ength , m


Static Temp

0.0052940 g

2762 U

2383 g

&

2012 .0.005

- 0.01

K a0.05 0.075 0.1 0.125 0.15 0.175 0.2 0.225 0.250.275 0.3 0.325 0.35

Com bustor Length, m(b) 0.60-wedge, nopilot


133

Static Temp.26 336425 330124 318023 312522 307921 300420 2985 119 2940 s18 2864 417 282215 2780 i15 2762 i*14 2754 3(A13 2569 112 2383 s11 2198 <310 2012 •19 18278 16417 14566 12705- 10854 8993 7142 5281 343

0.005

0.01 »

-0.005



Fig. 5.6 Temperature contours at the pilot injection plane, Y=0.0304m.


134

0.035 F

0.025

J 0.015

0.005 ¥

- 0.01 -0.005 0 0.005C o m b u s to r W id th (m )

0.01


0.035

0.03

0.025S

-£? 0.02 su2

0.015 1 u

0 .0 1

0.005

- 0.01 •0.005 0 0.005C o m b u sto r W idth (m )

Oh) 0.60-wedge, nopilot

0.01


135

0.035

0.025

aJ 0.015

0.005

• 0.01 0 .01•0.005 0 0.005C om bustor W idth (m)


Fig. 5.7 Velocity vectors and streamlines contours at the normal injection plane, X=0.070m.


136

3MMT19 38 37 363334 33 32 31 30 29 28 27 26 25 24 23 22 2120 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 I

0.035

0.03

e08

£U3

0.025

0.02

0.015Sd

0.01

0.005

Q

0.60-Step, Nopilot X= 70mm Cross Plane

• 0.01 •0.005 0 0.005Combustor Width (m)

0.01

(a) 0.60-Step, nopilotgMMTiapnUw 39 336738 335537 332736 327235 322234 3147

0.035

333231302928272625242322212019181716151413121110987654321

309630222897283527732699264826112595252323982273214920241899177416491524146714551437140013271288127812751178619542446314211197

^ 0.025e

.BP£faaB3

£

0.60-Wedge, Nopilot X= 70mm Cross Plane

0.015

0.005

•0.005 0.005 0.01

(b)Combustor Width Cm)

0.60-wedge, nopilot


137

SMMTmpti39 338738 335937 332738 327238 322234 314733 309833 302231 289730 283929 277328 289927 * 264828 261129 259924 252323 239822 227321 214928 202419 189918 177417 184918 1524IS 146714 145513 143712 140011 132710 12889 12788 12757 11788 619S 5424 4463 3142 2111 197

0.035

0.025

•I? 0.02

aJ 0.015

0.45-Wedge, Nopilot X= 70mm Cross Plane0.005

•0.01

(C)

•0.005 0 0.005Combustor Width (m)

0.45-wedge, nopilot

0.01

Fig. 5.8 Temperature contours at the normal injection plane, X=0.070m.


138

Static Temp. 0.03517 3311

16 3203 0.0315 3090

14 29750.02513 2862 O

2748 £2634 4P 0.022521

2407 0.0152293

21790.01

2065

19521838 0.005 0.60* Step, Nopilot

X= 220mm Cross Plane1724

0.01 0.005 0.0114801 Combustor Width (m)

(a) 0.60-Step, nopilot

Static Temp. 0,03517 3311

14 3203 0.0315 3090

15-

29760.0252862

12 2748 JSw> 0.022634

2521

J 0.0152407

2293

21790.01

19520.0051838 X 6 0 - Wedge, Nopilot

X = 220mm Cross Plane1724

16100.005 0.005 0.01•0.01

14801 Combustor Width (m)(b) 0.60-wedge, nopilot


Static Temp. 0 .035" 3311

16 3203« 3090 00314 297613 2862 -v 0.02512 274811 2634

jg

*|f 0.02i10 2521 *&. I9 2407 3

J 0.01518 2293 8 17 2179 6 I6 2065 0.0 1 15 19524 1838 0.005 It32I

17241C101480

Nopilot220mm

fig . 5.9

•0.005 0 0Combustor Width (m)

0 .4 5 -W e d g e , nopilot TTcuge, ncemPerat»re Contours at X=0.220m.

Reproduced with permission

VP*■ 'V- . *

o f t he copyright« " e r . Further reproduction

prohibited without permission.

140

0.60-W edge, Nopilot '0.60'Step, Nopilot40.00*

0.00*ISO 200 250 300 35050 1000

Com b u ito r Length (mm)

Fig. 5.10 Comparison between the combustion efficiency of the step and wedge configurations.

0.036

0.03

0.0248

fs 0.018

0.60-Wedge, Nopilot '0.60-Step, Nopilot

0.012

0.006

0.T 0.8 10.4 0.60.10Fuel Mole Fraction

Fig. 5.11 Comparison between the fiiel penetration of the step and wedge configurations.


Fud

Moic

Frac

tion

141

► "0.60-W edge, Nopilot *0.60-Stcp, Nopilot0.8

0.7

0.6

0.4

0.2

0.1

0.05 0.1 0.15 0.2C om bustor Length, m

0.35

Fig. 5.12 Comparison between the rate of fuel decay of the step and wedge configurations.


142

3500

3000 •

2500 •

-0.60-W edge, Nopilot •0.60-Step, Nopllot

a 2 000 ■

1500 «

i 1000 •

500

200100 150 250 300 350500Com bustor Length, m m

Fig. 5.13 Comparison between the upper-wall average static temperature distribution of the step and wedge configurations.

2400 ■

2200 •

1 2000 •

'0.60-Wedge, Nopilot '0.60-Step, NopUot

< 1600

1400

1200250 300 350150 200100500

Combustor Length, mm

Fig. 5.14 Comparison between the flow field average static temperature distribution of the step and wedge configurations.


143

600

500 • 0.60-Wedge, Nopilot '0.60-Step, Nopilot

400 •

< 300 •

1 200 *

too

300 350150 200 250100500C om bustor Length, mm

Fig. 5.15 Comparison between the upper-wall average absolute pressure distribution of the step and wedge configurations.

200 •0.60-W edge, Nopilot '0.60-Slep, Nopilot

180 •

160 ■

1 « • '

< 120 •

100

80350200 250 30015050 1000

Com bustor Length, (mm)

Fig. 5.16 Comparison between the flow field average absolute pressure distribution of the step and wedge configurations.


!**"

144

0.03

0.024

0.010

0.60-W edge, Nopilot

■0.45-Wedge, Nopilol0.012

0.000

0.7 0.00.4 a s 0.6 10.10Fuel Mole Fraction

Fig. 5.17 Comparison between the fuel penetration of the 0.45-ER and 0.60-ERwedge configuration cases.

'0.60-Wedge, Nopilot ■0.45-Wedge. Nopllot0.8

*SI 0.4

0JL

0.05 0.L0Combustor Length, m

Fig. 5.18 Comparison between the rate of fuel decay of the 0.45-ER and 0.60-ER wedge configuration cases.


145

100.00%

80.00%

60.00%

J 0.60-Wedge, Nopilot '0.45-Wedge, Nopilot40.00%

20.00%

0.00%250 35050 150 200 3001000

Combustor Length (mm)

Fig. 5.19 Comparison between the combustion efficiency of the 0.45-ER and0.60-ER wedge configuration cases.


146

K1<Iia

100 200

Fig. 5.20 Comparison between the upper-wall average static temperature distribution of the 0.45-ER and 0.60-ER wedge configuration cases.

2400 •

2200 •

£ 1800 •

“0.60-Wedge, Nopilot •0.45-Wedge, Nopilot

< 1600 ■

1400

120035050 100 150 200 250 3000

Combuitor Length, mm

Fig. 5.21 Comparison between the flow field average static temperature distribution of the 0.45-ER and 0.60-ER wedge configuration cases.


147

600

500 ■ -0.60-Wedge, Nopilot •0.45-Wedge, Nopilot

400 <

< 300 *

1 200 ■

100 ■ •

150 200 250 300 35050 1000Combustor Length, mm

Fig. 5.22 Comparison between the upper-wall average absolute pressure distribution of the 0.45-ER and 0.60-ER wedge configuration cases.

200 •-0.60-W edge, Nopilot •0.45-W edge, Nopilot

180 •

160 ■

140 ■

< 120 •

100

80250 350150 200 30050 too0

C om bustor Length (mm)

Fig. 5.23 Comparison between the flow field average absolute pressure distribution of the 0.45-ER and 0.60-ER wedge configuration cases.


148

0.0044

0 .0042SI Llo t'T njectio n£ba 0.004

1E6

0.0038

0.0036

0.0 3 6 0.038 0.040.0340.03C o m b u s to r L e n g th , m

>«(a)

0 . 0 0 1 2

0 .0 0 0 8

0 .0 0 0 4

ttak P H o-t In^ejEtio -

© -0 .0 0 0 4

-0 .0 0 0 8

- 0 . 0 0 1 2

0.040 .0 3 80 .0 3 4 0 .0 3 0Combustor Length, m

0 .0 3 20.03

(b)


149

0 .0 0 3 8

0.0 0 4In ject io n

. 0 . 0 0 4 2

• 0 .0 0 4 4

•0.0 0 46

•0 .0 0 4 8

• 0.0 0 5 0.0 3 2 0.0 3 4 0.0 3 6C o m b u s t o r L e n g t h , m

0 .0 3 8 0.0 40.0 3

(C)

Fig. 5-24 Reversed flow at the pilot injection ports for 0.01-ERP, pilot inj. plane.

CombustorLength, m Fig. 5-25 Reversed flow at the pilot injection ports for 0.01-ERP, sym. plane.


151

0 . 1 3 9 F

• . 0 3

0 . 0 1 9

0 . 0 1

0 . 0 0 5

0 *0 5

0 . 0 I

0 .1

(d)<j»P- 0.0450 *0 3 5 F

0 . 0 3

0 .0 2 5

0 .0 2

0 . 0 0 5

0 . 0 1 9 — —T

0 .0 1

. . . "1----:---- -----r— :---- t -0 0 .0 1 9 0 .0 J

C o m b u s to r L e n g t h , m

(e) (j>p = 0.06

0 .0 7 5 0 _

• .t

(f) i |ip s0 .(Fig. 5.26 Velocity vectors and streamlines contours at the symmetry plane for different pilot equivalence ratios.


152

0.0205

0.60-W cdge, with Pilots 0.60-Step, Nopilot 0.45-W edge, Nopilot

0.02

0.0195

SI

0.019

I0.0185

0.018

0.01750.070.04 0.06 0.08 0.090.01 0.02 0.030

Pilot-Injection Equivalence Ratio

Effect of the pilot injection equivalence ratio on the fuel penetration.Fig. 5.27

Fig. 5.28 the fuel.

0.35C om bustor L ength, m

Effect of the pilot iqjection equivalence ratio on the rate of decay of


153

S u i t e T• . 0 4 0 . 6 0 - W e d g e , N o p i l o t

«

A*

Comb U-* to r Lenfth.^ m(a) <j>p = 0.0

fi 0 * W e d g e , 0 . 0 2

B

#-• I

o.jC o n b m l o r L t o i t b , m

(b) <J)p = 0.020 .60 -W edge, 0.0 3

t . 0 3

• . II

• t

l . l f • .2-Conbnrtor Lcegth, nt

(c) <j>p = 0.03


\ r- ’>

154

S ta t i c Te r n p2 62 5 2 9 7 42 42 22 22 i2 9191 91 7161 51 4 1 7 9 9131 21 11 99976 1 9 9 55432 4 4 31 3 4 6

5 t a t l c T e a 9

2 6 3 9 7 72 5 2 9 7 42 4 2 9 3 32 3 2 7 9 92 2 2 4 9 32 I 2 5 5 22 0 2 4 5 9I 9 2 4 1 21 9 2 2 7 1I 7 2 1 3 91 6 1 9 9 91 5 1 9 4 9t 4 L 7 9 9I 3 1 5 6 9t 2 1 4 7 9I 1 1 4 2 71 9 1 2 9 79 1 2 7 19 1 2 5 97 1 1 4 69 1 9 9 55 9 6 54 7 2 43 5 9 32 4 4 3I 3 4 6

• »I4 I . C « * W * 4 | r r l . l 4 1

I •

M S I . SC o m b u s to r L e n g t h , m

t i e T * a 9 ►3 9 7 7 1 9 7 4 I D )3 T 9 V 3 9 9 3 3 5 9 3 3 4 5 9 3 4 1 33 3 T I 3 1 3 9 1 9 9 9 1 9 4 9 1 7 9 9 1 5 9 9 1 4 7 9 1 4 3 7 1 2 9 7 1 2 7 1 1 2 5 9 114 9 1 9 9 5 9 4 57 2 4 5 9 34 4 3 3 4 9

(d) <)>p = 0.0450 » ( 0 - W e d g e , 0 .0 69 »9 4

9 . 9 2

9 . 9 I

9 . 1 5 9 . 2

C o m b u s t o r L e n g t h , m

(e) <J>p at 0.060 . * « - W e d g e , t . l l9 . 9 4

9 . 9 2

9 . 9 I

IU..9 5 .15 9 . 2 5 9 . 3 5C

(f) <f>p =0.08Fig. 5.29 Temperature contours at the symmetry plane for different pilotequivalence ratios.

rL" V*Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

155

S t it le T e m p le25141 3 11 1114 t • I B t r i « l sI 4I 3 I 1 t I t I

0 .0 I

0 . 0 • 5

-0 .0 0 9

*0 *0 I

0 * 0 9 I J 7 5

S ta tic Temp,

0*119 0 *19 0 *1 7 9 0 *1 0.119 0 *1 9 0 *17 9C o m b u i l o r L e n g i h , m

(a) 6p =0.0

0 *3 0 *3 19 0 *J 9

2 4 3 3 4 42 5 3 3 9 12 4 3 I B B2 3 3 1 1 52 2 3 B T 92 1 3 B B 4I B 1 9 8 51 9 2 9 4 B1 • 2 8 4 4t T 2 8 2 21 4 2 T B Bt S 2 7 4 2I 4 2 7 5 41 3 2 5 4 9t 2 2 3 8 31 1 2 1 9 8t • I B 1 29 t B 2 T8 1 4 4 1T 1 4 5 44 I 2 7 BS 1 B B S4 8 9 93 ' 7 1 42 5 2 81 3 4 3

0 *0 1

0 * 0 0 9

- 0 * 0 0 9

-0 *0 I

0 *1 19 0 *19 0 *1 7 9 0 *1 0 *119 0 *1 9 0 *1 7 9C o m b u 1 to r L e n g t h , m

(b) <bp=0*02

0 *3 0 *3 19 0 *3 9

S t l a M c L T11 is1 41 3122 I 20 1 9 1 B I T I f1 51 4 t 3 t 21 t IB 9 Br4s432t

I

0 * 1 9 ' 0 - 2

C o m b n o l o r L e n g t h »

(c)<J)p = 0 .0 3

0 .3 0 *3 5


156

(d) ()>p = 0.045BT3v>

• • 5

• . 0 • 5

b u r to r L e n g t h , mC o B

S ta t i c T e aI t j j a 41 s n i t2 . 4 i n iX J j i i f1 2 J • T 9X I J • • 4X S 1 9 1 1I 9 2 9 4 1t • I S t 4t r l i t !1 1 X T • •tt t n tl A X T S 4t 3 X f 1 9t x X J S X1 1 X 1 9 •t • H U9 t i x rr t t 4 tr 1 4 5 4t t X T ts l l l f4 1 9 9S T 1 4I f t lt 1 4 3

I

• • 5

>1 .0 tr*£

C o o b v r t o r L e a | U » m

CO = 0.08Fig. 5JO Temperature contours at the pilot injection plane, Y=0.0304m, fordifferent pilot equivalence ratios.


157

^ T ? l

«n|0fMJai«tBO3<'*.c• ■ ^ rn m * m. ■* m m.

A • A

« *mPlM.J0Mnqai03 "a.IV *”


(b)

0.60-

Wed

ge, 0

.02.

Fig.

5. 31

Com

paris

on

betw

een

the t

empe

ratu

re c

onto

urs

of 0.

0-ER

p and

0.

02-E

Rp-c

ases

focus

ing

on the

cav

ity

regi

on.

158

5 U t l e T e n p3 I3 •2 9 3 2 2 22 V2 72 62 5 3 9 2 22 4 2 S 9 72 32 22 12 9I 9I 1171 6t 5 2 2 7 31 4L 3LI1 1 1 7 7 4I f9■7654321

S t i t l e T e m p3 3 2 7 23 9 3 2 5 129 3 2 2 22 9 3 19 32 7 3 1 4 72 6 3 9 9 62 5 3 9 2 224 2 9 9 723 2 9 3 52 2 2 7 7 32 1 2 6 9 92 9 2 6 4 9I 9 2 6 t It I 2 5 9 51 7 2 5 2 31 6 2 3 9 91 5 2 2 7 3t 4 2 ( 4 9I 3 2 9 2 4L2 t 9 9 9t I t 7 7 4I 9 1 6 4 99 1 5 2 49 L 4 6 77 L 4 5 56 1 4 3 75 1 4 9 94 ( 3 2 73 t 2 9 92 t 2 7 91 1 2 7 5

S t* t ic T e n3 I 3 2 7 23 9 3 2 5 12 9 3 2 2 22 9 3 L 9 32 7 3 1 4 72 6 3 9 9 62 5 3 9 2 22 4 2 1 9 72 3 2 9 3 52 2 2 7 7 321 2 6 9 929 2 6 4 9( 9 2 6 1119 2 5 9 51 7 2 5 2 31 6 2 3 9 115 2 2 7 314 2 1 4 913 2 9 2 412 1 9 9 9I t t T T 419 1 6 4 99 1 5 2 49 1 4 6 77 1 4 5 56 1 4 3 75 1 4 9 94 1 3 2 73 1 2 9 92 L 2 7 91 1 2 7 5

• I fI• .• 1

• I f

t i a e

C o n b i t f t o r W f d t h * m

(a) <|ip = 0.0• . • 3 f

• . • 2 5

• .• 2

• .•15

• . • • 5

C o a b m l o r W id th ,

(b) <{>p = 0.02• . • 3 $

• -0Z

g . o o sC o m b u s to r w Id to * n

• . • 1 5 -

• . • • 5 1 •

(c) <p s 0.03


159

S t i U e Tern p.3 1 3 2-7 23 4 3 2 5 12 9 3 2 2 2XI 5 1 4 32T 3 1 4 72 4 3 4 9 41 5 3 4 2 224 2 4 9 72 3 2 9 3 52 2 2 7 7 32 1 2 4 9 92 9 2 4 4 419 2 4 11I * 2 5 9 517 2 5 2 314 2 3 9 915 2 2 T 314 2 1 4 913 2 4 2 412 1 9 9 9t 1 I T 7 419 1 4 4 99 1 5 2 4

! 9 1 4 4 77 1 4 5 54 1 4 3 75 1 4 9 94 1 3 2 7

j 3 1 2 9 4! 2 1 2 7 4

t 1 2 7 5

S t a t i c T e r n3 t 3 2 7 239 3 2 5 t29 3 2 2 22 9 3 1 4 327 3 14 724 3 9 9 425 3 4 2 224 2 4 9 723 2 4 3 52 2 2 7 7 321 2 4 9 924 2 4 4 419 3 4 1 119 2 5 9 5IT 3 5 1 31 4 2 3 9 91 5 2 2 7 51 4 2 1 4 9t 3 24 1 4t 2 1 4 9 911 1 7 7 414 t 4 4 99 I 5 2 4

• 9 1 4 4 77 1 4 5 5

'4 1 4 3 75 1 4 4 44 13 2 73 L24 42 1 2 7 41 I 2 7 5

• .•35 F

4 .4 3

• .•15

• . • 2

# . • 1 5

• . • • 5 0 . 6 0 - W e d g e . 0 . 0 4 5 X > 4 0 m m C r o i r P l a n e

0 .0 1

(d) <{>p = 0.045

• . • 2 5

• . • 1 5

• . • • 5

(e) ()>p — 0.06S t a t i e

11)1

9 . * I S

(f) <J)p = 0.08Fig. 532 Temperature contours at X=0.040m cross flow plane for different pilot equivalence ratios.


160

■ I S M * I T X l a t *j a) «I T

i tJ 4J JJ *1a *x aX XX T I ,X Xx * j rX 4 %XX xX XX XX *i ax tXT1 4X IX 4XX1 XX XXXaar X4X43XX

• . I J f

t.t J

• J l f

5C o n b u r t o r w t d l b ( ■ )

(a) s 0.0

• 3

8

• .•15

• ►• I • ~W ed*e. 1.12 X * 7 X ■ m C r a i t P i t a *

• J LC o n b m l o r W i d l k , m

a i i N a t T e x a i r i i a r tX X 3 3 X 7I t 3 3 f tX T 3 3 X 7X X 3 X T X3 1 3 X X X3 4 3 1 4 T3 3 3 X X X3 X 3 X X X3 X X t X T3 X 1 X 3 *X X 1 T T 3xt xxxxI T 2 X 4 1X X X X X XX I x t x rX 4 2 * 2 3X 3 2 3 X tX X 2 2 . T 3X X X X X XX X X X X XXX xt xxL X X T T XI T X X 4 XX X 1 * 2 4i r 1 4 X TE 4 X X f SX 3 X X 3 TX T X X X XXX X 3 2 TXX x x x tX X X - T tt X X T tT X X T XX x t x3 t x x .4 X X X3 3 X 4X X X XX X X T

(b) <p = 0.02• .015

f

i

0 . 6 0 - W e d g e . 0 . 0 3 X * 7 0 m m C r o s s P l a n e

• -• LC o m b n s t o r W tdth. Cm )

(c) <|>p = 0.03


161

IJ9SI .• IC o m b u t Co r W Id tb ( m )

• If *■ r«111** 1 X 9 J i l t JUl J l I T *••*J i l t t l t f I I J I I T U t i l l 1141 <•11 t l t l t l t J X 1 * B X X 9 1X I 4 1 1 4 1 4 1 1 4 4 4 9 9 4 1 4 4 4 1 1 1 4 t 4 • T 1 4 4 1 1 4 J T I 4 • •4 1 X 9 t I B •I X 9 I t X 9 1 t 4 9 B 414 1 4 X * 4 •1 I «X I 4 1*9

(d) <|>p = 0.045

• J

7

• 2

• t S

• - • • 50 • 5n b u i t o r W id th ( m )C o

(e) <}>p = 0.06

X X T l

1

C o m b i i f t o r W [ d t h (m )

(f) 4>p = 0.08Fig. 5.33 Temperature contours the normal injection plane, X=0.070m, for different pilot equivalence ratios.


Combusto r Length, mm

Fig. 5.34 Comparison between the flow field average static temperature distributions for different pilot equivalence ratios.

200 ■ 0.60-0.00 .6 -0.020.60-0.030.60-0.046'0.60-0.06■0.60-0.08

180 '

£*t 160 '

1 140 •<

>•< 120 •

100

350250150 200 30050 1000Combustor Length, mm

Fig. 535 Comparison between the flow field average absolute pressure distributions for different pilot equivalence ratios.


163

“♦“ OfiO-Wedge, With pilots -#-0603tep , Nopilot -A -045-Wedge, Nopilot

006 007 on am003oaz 004aoioHlat-Ii cctfan Eqdtalm x Ratio

Fig. 5.36 Effect of pilot injection equivalence ratio on the stagnation pressure efficiency.

270

i

007 008 009006005003 OM0020010n * ^ K O n f i |M a n i W 9

Fig. 537 Effect of pilot injection equivalence ratio on the entropy production.


164

(160-Wedge, Nopilot 0.60-Step, Nopilot (145-Wedge, Nopilot

431500%

006 008001 002 003 OM 0090PUot-Injeclfcm Equivalence Ratio

Fig. 5.38 Effect of pilot injection equivalence ratio on the kinetic energy efficiency.

120.00%

100.00 %

80.00% ■

60.00% •

I “ ^ •0 .6 0 -0 .0■“ ^ 0 .6 0 - 0 .0 2“ ♦ “ 0.60-0.03“ ♦ ♦ 0 .6 0 -0 .0 4 5“ ♦ “0.60-0.06-e - 0 .6 0 -0 .0 8

40.00% - ■

20.00%

0.00%300 350200 25050 100 1500

C o m b u sto r L ength (mm)

Fig. 539 Effect of pilot injection equivalence ratio on the combustion efficiency.


165

CHAPTER V I

CONCLUDING REMARKS

A numerical study has been conducted to investigate the mixing and combustion

enhancements in scramjet engines, which are characterized by the supersonic combustion

flow fields. The primary objective o f the current study is to numerically examine the

possibility of igniting hydrocarbon fuels in supersonic airstreams and maintaining the

hydrocarbon flames in such hostile environments. The secondary objective o f the study is

to investigate the effect o f the combustor configuration, the main fuel equivalence ratio,

and the pilot fuel equivalence ratio on the flame holding mechanism and the flow field

and energy field characteristics.

The four physical models used in the current study are the unswept generic

rearward-facing step, the side swept rearward-facing step, the wedge with no pilot

injection, and the wedge with pilot injection configurations. The Computational Fluid

Dynamics “CFD” Code used in the present study is the FLUENT commercial code,

which is a control volume based finite difference method. FLUENT is a general-purpose

computer program for modeling fluid flow, heat transfer, and chemical reaction.

FLUENT solves the Reynolds average Navier-Stokes equations. It models a wide range

of phenomena by solving the conservation equations for mass and momentum. For flows

involving heat transfer or compressibility, an additional equation for energy conservation

is solved. For flows involving species mixing or reactions, a species conservation

equation is solved. Additional transport equations are also solved when the flow is

turbulent. The governing equations are discretized on a curvilinear grid to enable

computations in complex irregular geometries. A non-staggered system is used for

storage of discrete velocities and pressures. Interpolation is accomplished via a first-

order, Power-Law scheme o r optionally via higher order upwind schemes. The equations

are solved using SIMPLE-like algorithms with an iterative line-by-line matrix solver and

multigrid acceleration. The coupled solver is used with the explicit formulation and the

steady state approach. The second order upwind scheme is used for discretization o f the

flow, the turbulence kinetic energy, and the turbulence dissipation rate equations.


166

All physical models feature the following: the incoming airflow is supersonic, and

the main fuel is injected normal to the incoming airflow. For the cases with pilot

injections, the pilot fuel is injected parallel to the incoming airflow upstream o f the main

normal fuel injection. The pressure inlet boundary conditions are specified for the

incoming air inlet, and both the pilot and main fuel injections. The flow outlet is

supersonic and specified as pressure outlet boundary. All walls and step boundaries are

treated as no-slip adiabatic surfaces. Applying free-stream inlet conditions throughout the

entire flow field sets the initial conditions.

The influence o f turbulence on the reaction rate is taken into account by

considering the limiting (slowest) rate out o f the three rates of the Arrhenius reaction rate

and the two eddy-dissipation reaction rates. A global one-step reaction of the investigated

fuel is used to calculate the Arrhenius reaction rate. The renormalized group (RNG) form

o f the k-e turbulence model is used. The turbulence near-wall is treated using the two-

layer zonal model. Because the flow is compressible and a multicomponent mixture, it is

modeled as an ideal gas and the density, the specific heat, the viscosity, and the thermal

conductivity are calculated as composition-dependent properties.

A t the beginning o f the investigation, liquid kerosene was used, which is the

closest formula to the jet fuels that are most likely to be used in the hydrocarbon-fueled

scramjet engines. After using different reduced kinetic schemes for kerosene that have

the hydrogen combustion as one o f the elementary reactions, the difficulty o f igniting the

liquid kerosene while injecting pilot gaseous hydrogen ceased the study. The issue o f the

combustion simulation o f two different fuel injections with two different fuel types and

flow phases in supersonic airstreams was realized to be numerically hardly achieved.

Thereafter, attention has been directed towards the simulation o f ethylene and propane as

gaseous hydrocarbons in supersonic air flow fields.

The combustion characteristics o f gaseous propane in supersonic airflow using the

rearward-facing step that is swept inward form both end sides is studied. The effect o f

sweeping the step on the flow field features o f propane combustion is investigated.

Next, an extensive study for the flame holding and mixing and combustion

characteristics o f gaseous ethylene in supersonic airstreams is carried out. h i this study,

different combustor configurations, different main fuel equivalence ratios, and different


167

pilot fuel injection equivalence ratios were employed. First, The effect o f the combustor

configuration on the temperature flow field and the flow structure is investigated by

simulating two different combustor configurations, the rearward-facing step and the

cavity. Next, the effect of the equivalence ratio o f the main fuel injection on the general

flow field features and energy field characteristics is studied. Lastly, the effect o f the pilot

fuel equivalence ratio on the flame holding mechanism, and the combustion and flow

field qualities concludes the study. Six pilot injection equivalence ratios are studied, 0.0,

0.02,0.03,0.045,0.06, and 0.08.

The propane study yielded the following conclusions. The swept step showed the

ability to hold the propane flame in the supersonic air stream without extinction. It was

found that the side sweeping o f the combustor exhibits the high temperature and

combustion products concentration in the far field domain while the area downstream of

the normal injection location characterizes lower temperature and products concentration.

It is recommended to optimize the combustor length to ensure the complete combustion

and consequently the full liberation of the chemical energy stored in the fuel before the

fuel exits the combustor.

The main findings attained from the ethylene study can be summarized in the

following points. The step configuration with no pilot injection can afford the flame

holding mechanism in the supersonic air stream by creating the flow recirculations in the

step base area and featuring permanent high temperature regions surrounding the normal

fuel injection. The step configuration showed good mixing capabilities in the far field

domain. The wedge configuration proved superiority over the generic rearward-facing

step configuration in holding the ethylene flame in the supersonic airstreams, producing


losses and higher combustor efficiency. The increase in the equivalence ratio o f the






equivalence ratio increases the temperature o f the upstream face o f the normal injection


L68

and enhances the flame holding mechanism. The 0.02-pilot equivalence ratio presented

the optimum pilot injection case that can promote the flame holding mechanism and keep

good combustion and flow field qualities. While the further increase o f the pilot injection


to the deficiency o f the flame holding mechanism, the excessive pilot fuel injected shows

its positive effect by increasing the average flow field static temperature and absolute


For the supersonic mixing studies, it is suggested to investigate the effect o f the

injection angle o f the main fuel on the supersonic mixing and combustion characteristics.

It is recommended to study the effect o f the transverse normal ethylene and propane

injections on the flame holding and combustion characteristics in order to investigate the

relationship between the injection scheme and the flame holding and flow field and

energy field features. It is suggested to simulate the supersonic combustion o f the normal

ethylene and propane injections with the pilot gaseous hydrogen. For the future studies, it

is suggested to dedicate more effort trying to carry out the simulation o f the supersonic

combustion of liquid kerosene with gaseous hydrogen as a pilot fuel. The radiative

interaction is also suggested to be included in the future studies due to the existence of

the hydrocarbon combustion products, which are radiative participating mediums.


L69

REFERENCES

1. Swithenbank, J., Eames, I.W., Chin S.B., Ewan B.C.R., Yang Z.Y., Cao JM and

Zhao, X., ‘Turbulent Mixing in Supersonic Combustion Systems,” AIAA Paper

89-0260, January 1989.

2. Mitani, T., “Ignition problems in Scramjet Testing,” Combustion and Flame, Vol.

101, No. 3, May 1995, pp. 347-459.

3. Bouchez, M., Montazel, X., and Dufour, E., “Hydrocarbon Fueled Scramjets for

Hypersonic Vehicles,” AIAA Paper 98-1589, 1998.

4. Kay, I.W., Peschke, W.T., and Guile, R.N., “Hydrocarbon Fueled Scramjet

Combustor Investigation,” Journal o f Propulsion and Power, Vol. 8, No. 2, 1992,

pp. 507-512.

5. Matsuo, A., and Mizomoto, M., “Flow Structure o f Supersonic Flow Past

Backward-Facing Step with Perpendicular Injector,” AIAA Paper 98-0939,

January 1998.

6. Tishkoff, J.M., Drummond, J.P., Edwards, T., and Nejad, A.S., “Future Direction

o f Supersonic Combustion Research: Air Force/NASA Workshop on Supersonic

Combustion,” AIAA Paper 97-1017, January 1997.

7. Ferri, A., “Review o f Problems in Application o f Supersonic Combustion,”

Journal o f the Royal Aeronautical Society, Vol. 68, No. 45,1964, pp. 575-597.

8. Dugger, G.L., “Comparison o f Hypersonic Ramjet Engines with Subsonic and

Supersonic Combustion, ’Taper presented at Fourth AGARD Colloquium, Milan,

Italy; also Combustion and Propulsion-High Mach Number Airbreathing Engines,

edited by Jaumotte, Rothrock, and LeFebvre, Pergamon Press, New York, 1960.

9. Ferri, A., “Supersonic Combustion Progress,” Aeronautics and Astronautics, Vol.

2, Aug. 1964, pp.32-37.

10. Billing, F.S., “Research on Supersonic Combustion,” Journal o f Propulsion and

Power, Vol. 9, No. 4,1993, pp. 499-514.

11. Waltrup, P.J., “Liquid-Fueled Supersonic Combustion Ramjets: A Research

Perspective,” Journal o f Propulsion and Power, Vol. 3, No. 6,1987, pp. 515-524.


L70

12. Karagozian, A.R., “Fuel Injection and Flame Holding in High Speed Combustion

Systems,” Major Research Topics in Combustion, Springer-Verlag, 1992, pp.

237-252.

13. Singh, D.J., and Jachimowski, C.J., “Quasiglobal Reaction Model for Ethylene

Combustion,” AIAA Journal, Vol. 32, No. 1 ,1993, pp. 213-216.

14. Northam, G.B., “Workshop report: Combustion in Supersonic Flow,” 21st

JANNAF Combustion Meeting, October 1984.

15. Reddecliff, J.M., Weber, J., “Development and Demonstration o f a hydrocarbon

Scramjet Propulsion System,” AIAA Paper 98-1613, April 1998.

16. Schetz, J.A., Cannon, S.C., and Baranovsky, S., “Ignition o f Liquid Fuel Jets in a

Supersonic Airstream,” AIAA paper 79-1238, June 1979.

17. Segal, C., Owens, M.G., Tehranian, S., and Vinogradov, V., “Flame Holding

Configuration for Kerosene Combustion in a Mach 1.8 Airflow,” AIAA-97-2888.

18. Waltrup, P.J., “Liquid-Fueled Supersonic Combustion Ramjets: A Research

Perspective o f the Past, Present and Future,” AIAA 86-0158, January 1986.

19. Situ, M., Sun, Y.Y., Lu, H.P., and Wang, C., “Investigation on Fuel-Rich Hot Gas

as Piloted Energy for Kerosene Supersonic Combustion,” AIAA-2000-3587.

20. ^Northam, G.B., and Anderson, G.Y., “Supersonic Combustion Ramjet Researcht

at Langley*” AIAA Paper86-0159, January 1986.

21. Segal, C., and Young, C.D., “Development o f an Experimental Flexible Facility

for Nlixing-Combustion Interaction in Supersonic Flow,” Journal o f Energy

Resources, Vol. 118, June 1996, pp. 152-158. Heiser, W.H., and Pratt, D.T.,

‘Hypersonic Airbreathing Propulsion,” AIAA Education Series, Washington D.C.,

1994.

22. Billig, F.S., “Research on Supersonic Combustion,” Journal o f Propulsion and

Power, Vol. 9, No. 4, July-August 1993, pp. 499-514.

23. Waltrup, P.J., “Upper Bounds on the Flight Speed of Hydrocarbon-Fueled

Scramjet-Powered Vehicles,” Journal o f Propulsion and Power, Vol. 17, No. 6,

November-December2001, pp. 1199-1204.

24. Powell, O.A., Edward, J.T., Norris, R.B., Numbers, K.E., and Pearce, J.A.,

“Development o f Hydrocarbon-Fueled Scramjet Engines: The Hypersonic


L71

| 25.

i

i

26.

27.

28.

29.

30.

31.i

32.i!

33.|

34.

35.j

j


Technology (HyTech) Program,” Journal o f Propulsion and Power, Vol. 17, No.

6, November-December 2001, pp. 1170-1176.

Marren, D., Lewis, M., and Maurice, L.Q., “Experimentation, Test, and

Evaluation Requirements for Future Airbreathing Hypersonic Systems,” Journal

o f Propulsion and Power, Vol. 17, No. 6, November-December 2001, pp. 1361-

1365.

Heiser, W.H., and Pratt, D.T., “Hypersonic Airbreathing Propulsion,” AIAA

Education Series, Washington D.C., 1994.

Donohue, J.M., “Experimental and Numerical Study o f Ramp Injectors for

Supersonic Fuel/Air Mixing,” Ph J ) . Dissertation, Department o f Mechanical and

Aerospace Engineering, University ofVirginia, January 1995.

Hussaini, M.Y., Kumar, A., and Voigt, R.G. (eds.), “Major Research Topics in

Combustion," Springer-Verlag, New York, Inc., 1992.

Donohue, J.M., McDaniel J.C., and Haj-Hariri, H., “ Experimental and Numerical

Study of Swept Ramp Injection into a Supersonic How Field,” AIAA Journal,

Vol. 32, No. 9, September 1994, pp. 1860-1867.

Hollo, S.D., McDaniel, J.C., and Hartfield R.J., “Quantitative Investigation of

Compressible Mixing: Staged Transverse Injection into Mach 2 How,” AIAA

Journal, Vol. 32, No. 3, March 1994, pp. 528-534.

Papamoschou, D., and Roshko, A., “The Compressible Turbulent Shear Layer: an

Experimental Study,” Journal o f Fluid Mechanics, Vol. 197,1988, pp. 453-477.

Bogdanoff, D.W., “Compressibility Effects in Turbulent Shear Layers,” AIAA

Journal, Vol., 21, No. 6, 1983, pp. 926-927

Gutmark, E.J., Schadow, K.C., and Yu, K.H., “Mixing Enhancement in

Supresonic Free Shear Hows,” Annual Review o f Fluid Mechanics, Vol. 27, 1995,

pp. 31-43.

Yu, K B ., Wilson, K J., and Schadow, K.C., “Effect o f Hame Holding Cavities on

Supersonic Combustion Performance,” Journal o f Propulsions and Power, Vol.

17, No. 6, November-December 2001, pp. 1287-1295.

Yates, CX., ’’Liquid Injection into a Supersonic Stream,” Report AFAPL-TR-71-

79, Vol. 1, Aero propulsion Laboratory, Wright-Patterson AFB, OH, 1972.

of the copyright owner. Further reproduction prohibited without permission.

L72

36. Kush, E.A., and Schetz, J.A., “Liquid Jet Injection into a Supersonic Flow,” AIAA

Journal, Vol. I I , No. 9,1973, pp.1223-1224.

37. Wu, P.K., Kirkendall, K.A., Fuller, R.P., and Nejad, A.S., “Breakup Processes o f

Liquid Jets in Subsonic Crossflow,” Journal o f Propulsion and Power, Vol. 13,

No. 1,1997, pp. 64-73.

38. Fuller, R., Wu, P.K., Kirkendall, K., and Nejad, A., “Effect o f Injection Angle on

the Breakup Processes o f Liquid Jets in Subsonic Crossflows,” AIAA paper 97-

2966, July 1997.

39. Gruber, M il., Nejad, A.S., and Dutton, J.C., “Circular and Elliptical Transverse

injection into a Supersonic Crossflow—The Role of Large-Scale Structures,”

AIAA Paper 95-2150, June 1995.

40. Fuller, R.P., Wu, P.K., Nejad, A.S., and Schetz, J.A., “Comparison of Physical

and Aerodynamic Ramp as Fuel injectors in Supersonic Flow,” Journal o f

Propulsion and power, Vol. 14, No. 2, 1998, pp. 135-145.

41. Mathur, T., Lin, K.C., Kennedy, P., Gruber, M., Donbar, J., Jackson, T., and

Billig, F., “Liquid JP-7 Combustion in a Scramjet Combustor,” AIAA Paper

2000-3581, July 2000.

42. Curren, E.T., Heiser, W.H., and Pratt, D.T., “Fluid Phenomena in Scramjet

Combustion Systems,” Annual Review o f Fluid Mechanics, Vol. 28, 1996, pp.

323-360.

43. O’Byme, S., Doolan, M., Olsen, S.R., and Houwing, A.F.P., “Measurement and

Imaging o f Supersonic Combustion in a Model Scramjet Engine,” Proceedings o f

the 21st International Symposium on Shock Waves, Great Keepl Island, Australia,

Panther Publishing and Printing, 1988, pp.l 119-1124.

44. Moon, G-, Jeung, I-, and Choi, J-, “Numerical Study o f Thermal Choking Process

in a Model Scramjet Engine,” AIAA Paper 2000-3706, July 2000.

45. McDaniel, K. S., and Edwards, J.R., “Simulation o f Thermal Choking in a Model

Scramjet Combustor,” AIAA Paper 99-3411, June 1999.

46. Mohieldin, T.O., and Tiwari, S.N, “Effects o f Tandem Injection On Compressible

Turbulent Shear Layer,” AIAA Paper98-1638, April 1998.


L73

47. Schetz, J.A., Thomas, R.H., and Billig, F., “Mixing o f Transverse lets and Wall

Jets in Supersonic Flow,” in Separated Flows and Jets, V.V. Kozlov and A.V.

Dovgal (Edits.), Springer-Verlag, Berlin, 1991.

48. Lee, S-H., Kim., H-B., and Mitani, T., “Mixing and Combustion Augmentation of

Transverse Injection in Scramjet Combustor,” AIAA Paper 2001-0384, January

2001.49. Drummond, J.P., and Weidner, E.H., “A Numerical Study of Candidate

Transverse Fuel Injector Configurations in the Langley Scramjet Engine,” 17th

JANAAF Combustion Meeting, NASA Langley Research Center, Hampton, VA,

September 22-26,1980, pp. 561-586.

50. McClinton, C.R., “Interaction Between Step Fuel Injectors on Opposite Wall in a

Supersonic Combustor Model,” NASA Technical Report 1174,1978.

51. King P.S., Thomas, R.H., Schetz, J.A., and Billig, F.S., “Combined Tangential-

Normal Injection Into a Supersonic Flow,” AIAA Paper 89-0622, January 1989.

52. Mays, R.B., Thomas, R.H., and Schetz, J.A., “Low Angle Injection Into a

Supersonic Flow,” AIAA Paper 89-2461, July 1989.

53. Wood, C.W., Thomas, R.H., and Schetz, J.A., “Effect of Oscillating Shock

Impingement on the Mixing o f a Gaseous Jet in a Mach 3 Airstream,” AIAA

Paper90-1982, July 1990.

54. Baranovsky, S.I., and Schetz, J.A., “Effect o f Injection Angle on Liquid Injection

in Supersonic Flow,” AIAA Journal, Vol. 18, No. 6, 1980, pp.625-629.

55. Masuya, G., Komuro, T., Murakami, A., Shinozaki, N., Nakamura, A.,

Murayama, M., and Ohwaki,K., “Ignition and Combustion Performance of

Scramjet Combustors with Fuel Injection Strut,” Journal o f Propulsion and

Power, Vol. 11, No. 2, March-April 1995, pp. 301-306.

56. Segal, C., McDaniel, J.C., Whitehurst, R.B., and Kruass, R.H., “Mixing and

Chemical Kinetics Interactions in a Mach 2 Reacting Flow, ”Journal o f

Propulsion and Power, Vol. 11, No. 2, March-April 1995, pp. 308-314.

57. McDaniel, J., Fletcher, D., Hartfield, R., and Hollo, S., “ Staged Transverse

Injection Into Mach 2 Flow Behind a Reward-Facing Step: A 3-D Compressible


L74

^ Test Case for Hypersonic Combustor Code Validation,” AIAA Paper 91-5071,“ ' -/

December 1991....

58. Berman, H.A., Anderson, J.D., and Drummond, J.P., ’’Supersonic Flow over a

Rearward Facing Step with Transverse Non-Reacting Hydrogen Injection,” AIAA

Journal, Vol. 21, No. 12, December 1983, pp. 1707-1713.

59. Jian-Guo, L., Gong Y., Xin-Yu, Z., Qing-Sheng, H., “Combustion o f Kerosene in

a Supersonic Stream,” AIAA Paper 2000-0615, January 2000.

60. Rogers, R.C., Capriotti, D.P., and Guy, R.W., “Experimental Supersonic

Combustion Research at NASA Langley,” AIAA Paper 98-2506, June 1998.

61. Seiner, J.M., Dash, S.M., and Kenzakowski, D.C., “Historical Survey on

Enhanced Mixing in Scramjet Engines,” Journal o f Propulsion and Power, Vol.

17, No. 6, November-December 2001, pp. 1273-1286.

62. Stouffer, ST)., Baker, N il., Capriotti, D.P., and Northam, G.B., “Effects o f

Compression and Expansion Ramp Fuel Injector Configurations on Scramjet

Combustion and Heat Transfer,” AIAA Paper 93-0609, January 1993.

63. Drummond, J.P., Carpenter, M.H., Riggins, D.W., and Adams, M.S., “Mixing

Enhancement in a Supersonic Combustor,” AIAA Paper 89-2794, July 1989.

64. Hartfield, R.J., Jr., Hollo, ST)., and McDaniel, J.C., “Experimental Investigation

o f a Supersonic Swept Ramp Injector Using Laser-Induced Iodine Fluorescence,”

Journal o f Propulsion and Power, Vol. 10, No. 1, January-February 1994, pp.

129-135.

65. Waitz, I.A., Marble, F.E., and Zukoshi, E.E., “An Investigation o f a Contoured

Wall Injector for Hypervelocity Mixing Augmentation,” AIAA Journal, Vol. 31,

No. 6, June 1993, pp. 1014-1021.

66. Riggins, D.W., Mekkes, G.L., McClinton, C.R., and Drummond, J.R., “A

Numerical Study of Mixing Enhancement in a Supersonic Combustor,” AIAA

Paper90-0203,January 1990.

67. Daso, E.O., and Gross, B., “Analysis o f Supersonic Mixing in High Speed

Combustors,” Presented a t the Joint JANNAF Propulsion Meeting, Nov. 15,

1993.


176

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.


Thompson, M.W., “CIAM Reports for Phase 2 of the Joint US/Russian Scramjet

Program: Russian Hydrocarbon Fueled Scramjet Technology,” Report AFRL-PR-

WP-TR-1998-2017, Wright-Patterson AFB, Dayton, OH, 1997.

SabeL’nikov, V., Korontsvit, Y.P., Bain, W.L., Van Greithuysen, V.J.,

Ivanyushkin, A.K., and Ivanov, V.V., “Experimental Investigation o f Combustion

Stabilization in Supersonic Flow Using Free Recirculating Zone,” AIAA Paper

98-1515, April 1998.

Lin, K.-C., Kirkendall, K.A., Kennedy, P J ., and Jackson, T.A., “Spray Structure

o f Aerated Liquid Fuel Jets in Supersonic Crossflows,” AIAA Paper 99-2374,

June 1999.

Avrashnov, V., Baranovsky, S., and Levin, V., “Gas dynamic Features of

Supersonic Kerosene Combustion,” AIAA Paper 90-5268, October 1990.

Murthy, S.N.B., and Curran, E. T. (eds.), “High-speed R ight Propulsion System”,

Progress in Astronautics, AIAA, Vol. 137, Washington, DC, 1991.

Billig, F.S., “Propulsion Systems from Takeoff to High-Speed Flight,” Progress

in Astronautics and Aeronautics, AIAA, Vol. 137, Washington, DC, 1991, A.

Richard Seebass, Editor-in-chief, PP. 69.

Beach, H.L., “Supersonic Combsution Status and Issues,” in MajorReserach

Topics in Combustion, Hussaini, Kumar, and Voigt, eds, Springer Verlag, NY,

1992.

Northam, G.B., ed., “Workshop Report: Combustion in Supersonic Flow,” 22nd

JANNAF Combustion Meeting Proceedings, CPIA Publication 432, Vol. 1, 1985.

McClinton, C.R., “Autoignition of Hydrogen Injected Transverse to Supersonic

Airstream,” AIAA Paper, 79-1239, June 1979.

Huber, P.W., Schchenxnayder, C.J., and McClinton, R.C., “Criteria for Self-

Ignition of Supersonic Hydrogen-Air Mixtures,” NASA Technical Paper 1457,

1979.

Tomioka, S., Hiraiwa, T., Mitani, T., Zamma, Y., Shiba, K., and Masuya, G.,

“Auto Ignition in a Supersonic Combustor with Perpendicular Injection Behind

Backward-Facing Step,” AIAA Paper97-2889, July 1997.

of the copyright owner. Further reproduction prohibited without permission.

177

90. Sato, Y., Sayama, M., Masuya, G., Komuro, T., Kudou, K., Murakami, A., Tani,

K., and Chinzei, N., “Experimental Study on Autoignition ia a Scramjet

Combustor,” Proceedings o f $ h International Symposium on Airbreathing

Engines, Athens, September, 1989, pp. 569-576.

91. Spaddaccini, L.J., and Chinitz, W., “Supersonic Combustion o f Propane,” AIAA

Journal, Vol. 6,1967, pp. 1391-1393.

92. Beach, H.L., Jr. Mackley, E.A., Rogers. R.C., and Chinitz, W., “Use o f Silane in

Scramjet Research,” 17th JANNAF Combustion Meeting, Vol. I, CPIA

Publications 329, 1980, pp. 639-659.

93. Diskin, G.S., and Northam, G.B., “Evaluation of Storable Flourine Based Pilot for

Scramjets,” AIAA Paper, 86-0372,1986.

94. Masuya, G., Kudou, K., Murakami, A., Komuro, T., Tani, K., Kanda, T.,

Wakamatsu, Y., Chinzei, N., Sayama, M., Ohwaki, K., Kimura, I., “Some

Governing Parameters o f Plasma Torch Igniter/Flame-Holder in A Scramjet

Combustor,” AIAA Paper90-2098,1990.

95. Kimura, I., Aoki, H., and Kato, M., “The Use o f Plasma Jet for Flame

Stabilization and Promotion o f Combustion in Supersonic Air Flows,”

Combustion and Flame, Vol. 42,1981, pp. 297-305.

96. Northam, G.B., McClinton, C.R., Wagner, T.C., and O’Brien, W.F.,

“Development and Evaluation of a Plasma Jet Flame-Holder for Scarmjets,”

AIAA Paper 84-1408,1984.

97. Stouffer, S.D., O’Brien, W.F. and Roby, R.J., “Improved Plasma Torch for

Ignition and Flame Holding in Supersonic Combustion,” AIAA Paper 89-2945,

1989.

98. Wagner, T.C., O’Brien, W.F., Norrtham, G.B., and Eggers, JM ., “Design and

Evaluation o f a New Injector Configuration for Supersonic Combustion,” Proc.

S'* Int. Sympo. Airbreathing Engines, Cincinnati, June 1987, pp. 390-397.

99. Wagner, T.C., O’Brien, W F., Northam, G.B., and Eggers, JM ., “Plasma Torch

Igniter for Scramjets,” Journal o f Propulsion and Power, Vol. 5, 1989, pp. 548-

554.


178

100. Sato, Y., Sayama, M., Ohwaki, K., Masuya, G., Komuro, T., Kudou, K.,

Murakami, A., Tani, K., Wakamatsu, Y., Kanda, T., Chinzei, N., and Kimura, I.,

“Effectiveness o f Plasma Torches for Ignition and Flame Holding in Scramjet,”

Journal o f Propulsion and Power, Vol. 8, No. 4, July-August, 1992.

101. Hawthorn, R.D., and Nixon, A.C., “Shock Tube ignition Delay Studies of

Endothermic Fuels,” AIAA Journal, Vol. 4, No. 3, March 1966.

102. Tepper, F., and Kaledin, L.A., “Nano Aluminum as a Combustion Accelerator for

Kerosene in Air Breathing Systems,” AIAA Paper 2001-0521,2001.

103. Lai, H.T., and Thomas, S.R., “Numerical Study of Contaminant Effects on

Combustion o f Hydrogen, Ethane and Methane in Air,” AIAA Paper 95-6097,

April 1997.

104. Srinivasan, S., and Erickson, W.D., “Interpretation of Vitiation Effects on Testing

at Mach 7 Flight conditions,” AIAA-95-2719, July 1995.

105. Slack, M., and Grillo, A., “Investigation o f Hydrogen-Air Ignition Sensitized by

Nitric Oxide and by Nitrogen Dioxide,” NASA CR-2986,1977.

106. Laster, W.R., and Sojka, P.E., “Autoignition o f Hi-Air; The Effect o f No*

Addition,” Journal o f Propulsion and Power, Vol. 5, No. 4, 1989, pp. 385-390

107. Brabbs, T A ., Lezberg, E.A., and Bittker, D.A., “Hydrogen Oxidation Mechanism

with Applications to (1) the Chaperon Efficiency of Carbon Dioxide and (2)

Vitiated Air Testing,” NASA TM-100186, 1987.

108. Rogers, R.C., “Effect o f Test Facility Contaminants on Supersonic Hydrogen-Air

Diffusion Flames,” 23rd JANNAF Combustion Meeting, Hampton, VA, October

20-24,1986.

109. Erickson, W. D., and Klick, G.F., “Analytical Chemical Kinetic Study o f the

Effect o f Carbon Dioxide and Water Vapor on Hydrogen-Air Constant-Pressure

Combustion,” NASATND-5768, April 1970.

110. Srinivasan, S., and Erickson, W.D., “Influence o f Test-Gas Vitiation on Mixing

and Combustion at Mach 7 Flight conditions,” AIAA 94-2816,1994.

111. Bakos, R.J., and Morgan, R.G., “Scramjet Combustion and Thrust Measurements

in a Reflected Shock Tunnel with Varied Test Gas Atomic Oxygen

Contamination,” AIAA Paper 94-2520,1994.


L79

112. Mitani, T., Hiraiwa, T., Sato, S., Tomioka, S., Kanda, T., Tani, K., “Comparison

o f Scramjet Engine Performance in Mach 6 Vitiated and Storage-Heated Air,”

Journal o f Propulsion and Power, Vol. 13, No. 5, September-October 1997,

pp.635-642.

113. Hiraiwa, Ti, Sato, S., Tomioka, S., Kanda, T., Shirmura, T., and Mitan, T.,

“Testing o f a Scramjet Engine Model in Mach 6 Vitiated Air Flow,” AIAA-97-

0292, January 1997.

114. Owens, M., Sega, C., and Auslender, A.H., “Effects o f Mixing Schemes on

Kerosene Combustion in a Supersonic Air Stream,” Journal o f Propulsion and

Power, Vol. 13, No. 4, July-August 1997.

115. Edwards, T., ’’Combustion Challenges o f High Temperature Jet Fuels,” paper

204a presented at AIChE Annual Meeting, Nov. 15, 1996, Chicago, IL.

116. Billig, F.S., “Supersonic Combustion Ramjet Missile,” Journal o f Propulsion and

Power, Vol. 11, No. 6, 1995, pp. 1139-1146.

117. Jensen, J., Branendlein, B., ’’Review o f The Marquardt Dual Mode Mach 8

Scrmajet Development,” AIAA Paper 96-3037, June 1996.

118. Gurijanov, E.P., and Harsha, P.T., “Ajax: New Directions in Hypersonic

Technology,” AIAA Paper 96-4609,1996.

119. Edward, T., and Maurice, L, “HyTech Fuels/Fuel System Research,” AIAA Paper

98-1562, 1998.

120. Heins, A .EJr., Reed, G.J., and Woodgrift, K.E., “Hydrocarbon Scramjet

Feasibility Program, Part H: Experimental Investigation o f a Hydrocarbon Fueled

Scramjet Combustor,” AFAPL-TR-70-74, January 1971.

121. Kay, I.W., and McVey, J.B., “Hydrocarbon-Fueled Scramjet, Vol. XI. Combustor

Demonstration Program,” AFAPL-TR-68-146, May 1972.

122. Heins, A.E. Jr., Reed, G.J., and Woodgrift, K.E., “ Hydrocarbon Scramjet

Feasibility Program, Part HI: Free Jet Engine Design and Performance,” AFAPL

TR-70-74, Jan. 1971.

123. Billig, F.S., Pirkle, J.C., and Dugger, G.L., “Scramjet Fuels Evaluation,” AFAPL

TR-72-21, July 1972.

iI

i


180

124. Gerstein, M. and Choudhucy, P.R., “Use of Silane/Methane Mixtures for Scramjet

Ignition,” AIAA Paper 84-1407, June 1984.

125. Hunt, J.L., et al., “Hypersonic Airbreathing Missile Concepts Under Study at

Langley,” AIAA Paper 82-0316, January 1982.

126. Morris, N.A., “Silane as an Ignition Aid in Scramjet,” Ph.D. Thesis, University o f

Qeensland, Australia, 1989.

127. Lander, H., and Nixon, A.C., "Endothermic Fuels for Hypersonic Vehicles,1'

Journal o f Aircraft, Vol. 8, No. 4,1970, pp. 200-207.

128. Petley, D.H., and Jones, S.C., "Thermal Management For a Mach 5 Cruise

Aircraft Using Endothermic Fuel," Journal o f Aircraft, Vol. 29, No. 3, May-June

1992, pp. 384-389.

129. Edward, T., "USAF Supercritical Hydrocarbon Fuels Interest," AIAA Paper 93-

0807, January 1993.

130. JANNAF Ramjet Subcommittee Workshop, "Fuels For High-Speed Flight

Vehicles," CPIA Publications 518 (2 Vols.), June 1988, pp. 143-182.

131. Lander, H., and Nixin, A.C., “Endothermic Fuels for Hypersonic Aircraft,”

Journal o f Aircraf, Vol. 8, N o .4 ,1971, pp. 200-207.

132. Hawthorne, R.D., Ackerman, G.H., and Nixon, A.C., “A Mathematical Model of

a Packed-Bed-Heat-Exchanger Reactor for Dehydrogeneration of

Methylcyclohexane: Comparison o f Predictions with Experimental Results,”

AICHE Joura.nl, Vol. 14, No. 1,1968, pp. 69-76.

133. Gates, B.C., Katzer, J.R., and Schuit, G.C.A., Chemistry o f Catalytic Processes,

McGraw-Hill, New York, 1979, p.8.

134. Wickham, D.T., Engel, J.R., Hitch, B.D., and Karpuk, M.E., “Initiators for

Endothermic Fuels,” Journal o f Propulsion and Power, Vol. 17, No. 6,

November-December2001, pp. 1253-1257.

135. Wohlwend, K.L., Maurice, L.O., Edwards, T., Striebich, R.C., Vangsness, M.,

and Hill, A.S., “Thermal Stability o f Energetic Hydrocarbon Fuels for Use in

Combined Cycle Engines,” Journal o f Propulsion and Power, Vol. 17, No. 6,

November-December 2001, pp. 1258-1262.


18L

136. Kit, B., and Evered, D., “Rocket Propellant Handbook, ” ManMillan, New York,

1960.

137. “Survey o f Jet Fuels (1990-1996), ’’Defence Energy Support Center,” MIL-DTL-

5624T, June 1998.

138. Edwards, T., and Maurice, L.Q., “Surrogate Mixtures to Represent Complex

Aviation and Rocket Fuels,” AIAA Paper99-2217, June 1999.

139. Lefebvre, A., “Gas Turbine Fuels,” Gas Turbine Combustion, Hemisphere, New

York, 1983, Chapter 9.

140. Sobel, D.R., and Spadaccini, L.J., “Hydrocarbon Fuel Cooling Technologies for

Advanced Propulsion,” Journal o f Engineering Gas Turbines and Power, Vol.

119, No. 2, 1997, pp. 344-351.

141. Lewis, M.J., “Significance o f Fuel Selection for Hypersonic Vehicle Range,”

Journal o f Propulsion and Power, Vol. 17, No. 6, November-December 2001, pp.

1214-1221.

142. Owens, M, Tehranian, S, Segal, C., and Vinogradov, V., “Flame Holding

Configuration for Kerosene Combustion in a Mach 1.8 Airflow,” Journal o f

Propulsions and Power, Vol. 14, No. 4, July-August 1998.

143. Currant, E.T., and Murthy, S.N.B (eds.), “Scramjet Propulsion,” Progress in

Astronautics and Aeronautics, AIAA, Vol. 189, Reston Virginia, 2000, pp. 757-

822.

144. Pieper, J., Nguyen, T., and Hulka, J., “Performance and Stability Characterization

o f Liquid Oxygen/Kerosene Injection at Aerojet,” Recent Advances in Spray

Combustion: Spray Combustion Measurement and Model Simulation, Volume II,

Progress in Astronautics and Aeronautics, Volume 171,1996, pp. 349-368.

145. Heneghan, S.P., Zabamick, S.Z., Ballal, D.R., and Harrison HI, W.E., “JP-8+100:

The Development o f High-Thermal-Stability Jet Fuel,” Journal o f Energy

Resources Technology, Vol. 118, September 1996, pp. 170-179.

146. Zhou, N., and Krishnan, A., “A Numerical Model for Endothermic Fuel Flows

with Heterogeneous Catalysis,” AIAA Paper96-0650,1996.

147. Westbrook, C.K., and Dryer, F.L., “Chemical Kinetics Modeling of Hydrocarbon

Combustion,” Progress Energy and Combustion Science, 1984, Vol. 10, pp.1-57.


L82

148. Sun, Y.Y., Han, Z.Y., Luo, X.S., Dog, S.T., Xu, S i . , and Situ, M., “A

Fundamental Study foe Supersonic Combustion of Hydrocarbon Fuel,” AIAA-

2001-0522, January 2001

149. Vinogradov, V.A., Owens, M., Mullagiri, S., Segal, C., “Effects o f Fuel Pre-

Injection on Mixing in a Mach 1.6 Airflow,” AIAA-99-4918, November 1999.

150. Townend, L.H., “Domain o f the Scramjet,” Journal o f Propulsion and Power,

Vol. 17, No. 6, November-December 2001, pp. 1205-1213.

151. O’Brien, T.F.O., Starkey, R.P., and Lewis, M.J., “Quasi-One-Dimensional high

speed Engine Model with Finite-Rate Chemistry,” Journal o f Propulsions and

Power, Vol. 17, No. 6, November-December 2001, pp. 1366-1374.

152. Mathur, T., Gruber, M., Jackson, K., Donbar, J., Donaldson, W., Jackson, T., and

Billig, F., “Supersonic Combustion Experiments with a Cavity-Based Fuel

Injector,” Journal o f Propulsions and Power, Vol. 17, No. 6, November-

December 2001, pp. 1305-1312.

153. Yu, K., Wilson, K.J., Smith, R.A., and Schadow, K.C., “Experimental

Investigation on Dual-Purpose Cavity in Supersonic Reacting Flows,” AIAA

Paper 98-0723, January 1998.

154. Situ, M., Wang, C., Lu, H.P., Yu, G., and Zhang, X.Y., “Hot gas Piloted Energy

for Supersonic Combustion o f Kerosene with. Dual-Cavity,” AIAA Paper 2001-

0523,2001.

155. Situ, M., Sun, Y.Y., Zhang, S.D., and Wang, C., “Investigation o f Supersonic

Combustion o f Hydrocarbon Fuel-Riched Hot Gas in Scramjet Combstor,” AIAA

Paper 99-2245,1999.

156. Kay, I.W., Chiappetta, L., and McVey, J.B., “Hydrocarbon Fueled-Scramjet,

Combustor Investigation,” Technical Report AFAPL-TR-68-146, Volume IV,

May 1969.

157. Kay, I.W., McVey, J.B., Kepler, C.E., and Chiappetta, L., “Hydrocarbon Fueled

Scramjet, Piloting and Flame Propagation Investigation, Combustor

Investigation,” Technical Report AFAPL-TR-68-146, Volume IX, May 1971.


183

158. Bonghi, L., Dunlap, M., Owens, M., Young, C.D., and Segal, C., ‘Hydrogen

Piloted for Supersonic Combustion o f Liquid Fuels,” AIAA Paper 95-0730,

January 1995.

159. Vinogradov, V., Kobigsky, S., and Petrov, M., “Experimental Investigation of

Liquid Carbon-Hydrogen Fuel Combustion in Channel at Supersonic Velocities,”

AIAA Paper 92-3429, July 1992.

160. Change, J.S., and Lewis, M.J., “Development o f a Jet-A/Silane/hydrogen

Reaction Mechanism for Modeling a Scramjet Combustor,” AIAA Paper 99-

2252, June 1999.

161. Yu, G., Li, J.G., Chang, X.Y., Chen, L.H., Sung, C.J., “Investigation o f Kerosene

Combustion Characteristics with Pilot Hydrogen in Model Supersonic

Combustors,” Journal o f Propulsion and Power, Vol. 17, No. 6, November-

December 200 L.

162. Baurle, R.A., and Gruber, M.R., “Study o f Recessed Cavity Flow Fields for

Supersonic Combustion Applications,” AIAA Paper 98-0938, January 1998.

163. Ben-Yakar, A., and Hanson. R., “Cavity Flame Holding for Ignition and Flame

Stabilization in Scramjets: Review and Experimental Study,” AIAA Paper 98-

3122, July 1998.

164. Davis, D.L., “Numerical Analysis o f Two and Three Dimensional Recessed

Flame-Holders for Scramjet Applications,” Ph.D. Dissertation, Aeronautics and

Astronautics Dept., A ir Force Inst, o f Technology, W right Patterson AFB, OH,

Sept. 1996.

165. Baurle, R.A., Tam, C.-J., and Dasgupta, S., “Analysis o f Undteady Cavity Flows

for Scramjet Applications,” AIAA Paper 2000-3617, July 2000.

166. Bumes, R., Parr, T.P., Wilson, K J., and Yu, K., “Investigation o f Supersonic

Mixing Control Using Cavities: Effect o f Fuel Injection Location,” AIAA Paper

2000-3618, July 2000.

167. Gruber M il., Baurle, R A ., Mathur, T., and Hsu, K.-Y., “Fundamental Studies o f

Cavity-Based Flame-Holder Concepts for Supersonic Combustors,” AIAA Paper

99-2248, June 1999.


184

168. Eklund, D.R., Baurle, R.A., and Gruber, M.R., “Numerical Study o f a Scramjet

Combstor Fueled by an Aerodynamic Ramp Injector in Dual-Mode Combustion,”

AIAA, Paper2001-0379, January 2001.

169. Waltrup, P J ., Dugger, G.L., Billig, F.S., and Orth, R.C., “Direct-Connect Tests of

Hydrogen-Fueled Supersonic combustors,” 16th Symposium (Int’l) on

Combustion, Combustion Institute, 1976, pp. 1619-1630.

170. McCilinton, C.R., “Interaction Between Step Fuel Injector on Opposing Walls in

a Scramjet Combustor Model,” NASA TP-1174, 1978.

171. Trexler, C.A., and Sanders, S.W., “Design and Performance at a Local Mach

Number o f 6 o f an Inlet for an Integrated Scramjet Concept,” NASA TN-D-7944,

1977.

172. Nivkerson, G.R., “Optimized Scramjet Exhaust Nozzles for Hypersonic

Propulsion,” AIAA Paper 88-3161, 1988.

173. Tomioka, S., Hiraiwa, T., Sakuranaka, N, Murakami, A., Sato, K., Matsui, A.,

“Ignition Strategy in A Model SCRAMJET,” AIAA-96-3240, July 1996.

174. Roy, G.D., Propulsion Combustion: Fuels to Emissions, Taylor & Francis

Publishers, 1998.

175. Westbrook, C.K., and Dryer, F.L., “Simplified Reaction Mechanism for the

Oxidation o f Hydrocarbon Fuels in Flames,” Combustion Science and

Technology, Vol. 27, No. 1,1981, pp.31-43.

176. Jachimowski, C.J., “An Experimental and Analytical Study o f Acetylene and

Ethylene Oxidation Behind Shock Waves,” Combustion and Flame, Vol. 29, No.

1,1977, pp.55-66.

177. White, M E., Drummond, J.P., and Kumar, A., “Evolution and Application of

CFD Techniques for Scramjet Engine Analysis,” Journal o f Propulsion and

Power, Vol. 3, No. 5, September-October 1987, pp. 423-439.

178. Bissinger, N.C., “CFD Support for Scramjet Design and testing,” AIAA Paper 99-

4956, November 1999.

179. Lee, R.A., Hosangadi, A., Cavallo, PA ., and Dash, S.M., “Application of

Unstructured Grid Methodology to Scramjet Combustor Flow Fields,” AIAA



L85

180. Kodera, M., Sunami, T., and Nakahashi, K., “Numerical Analysis o f Scramjet

Combusting Flows By Unstructured Hybrid Grid Method,” AIAA Paper 2000-

0886, January 2000.

181. Ungewitter, R.J., Dash, S.M., Papp, J., and Hosangadi, A., “Design-Oriented

Analysis o f Scramjet Combustor Flow Field Using Combined UNS/PNS

Procedure,” AIAA Paper 2000-3616, June 2000.

182. Chakravarthy, S., Goldgerg, U., Batten, P., Palaniswamy, S., and Peroomian, O.,

“Some Recent Progress in Practical Computational Fluid Dynamics,” AIAA


183. Mehta, U.B., “Credible Computational Fluid Dynamics Simulations,” AIAA

Journal, Vol. 36, No. 5, May 1998, pp.665-667.

184. Rizzi, A., and Vos, J., ‘Towards Establishing Credibility in Computational Fluid

Dynamics Simulations,” AIAA Journal, Vol. 36, No. 5, May 1998, pp. 668-675.

185. Oberkampf, W.L., Blottner, F.G., “Issues in Computational Fluid Dynamics Code

Verification and Validation,” AIAA Journal, Vol. 36, No. 5, May 1998, pp. 687-

695.

186. Roache, P.J., “Verification o f Codes and Calculations,” AIAA Journal, Vol. 36,

No. 5, May 1998, pp. 696-702.

187. Oberkampf, W.L., and Trucano, T.G., “Validation Methodology in Computational

Fluid Dynamics (Invited),” AIAA Paper 2000-2549, June 2000.

188. Espina P J ., and piomelli, U., “Validation o f the NPARC Code in Supersonic Base

Flows,” AIAA Paper97-0032, January 1997.

189. Jameson, A., Martinelli, L., “Mesh Refinement an Modeling Errors in Flow

Simulation,” AIAA Journal, Vol. 36, No. 5, May 1998, pp. 676-686.

190. Habashi, W.G., Doinpierre, J., Bourgault, Y., Fortin, M., and Vallet, M.-G.,

“Certifiable Computational Fluid Dynamics Through Mesh Optimization,” AIAA

Journal, Vol. 36, No. 5, May 1998, pp. 703-711.

191. Johnson, C., and Muss, J., “Propulsion Analysis and Computational Fluid

Dynamics Tools on the Internet,” AIAA Paper 2000-3769, July 2000.

192. Fluent Version 5.5 User’s Guide, Fluent, Inc., Lebanon, New Hampshire, 2000.


186

193. Anderson, J.D., Computational Fluid Dynamics: The Basics with Applications,

McGraw Hill, Inc., New York, NY, 1995.

194. Launder, B.E., Spalding, D.B., Lectures in Mathematical Models o f Turbulence,

Academic Press, London, England, 1972.

195. Wilcox, D. C. ‘Turbulence Modeling for CFD,” DCW Industries, Inc., La

Canada, California, 1993.

196. Sarkar, S. and Balakrishnan, L. “Application of a Reynolds-Stress Turbulence

Model to the Compressible Shear Layer” ICASE Report 90-18, NASA CR

182002, 1990.

197. Magnussen, B.F., and Hjertager, B.H. "On Mathematical Models o f Turbulent

Combustion with Special Emphasis on Soot Formation and Combustion,” 16th

Sump. (Int’l.) on Combustion, The Combustion Institute, 1976.

198. Owens, M., Segal, C., and Auslender, A.H., “Effects o f Mixing Schemes on

Kerosene Combustion in a Supersonic Air stream,” Journal o f Propulsion and

Power, Vol. 13, No. 4, July-August 1997, pp 525-531.

199. Owens, M., Tehranian, S., Segal, C., and Vinogradov, V., “Flame-Holding

Configurations for Kerosene Combustion in a Mach 1.8 Airflow,” Journal o f

Propulsion and Power, Vol. 14, No. 4, July-August 1998, pp 456-461.

200. Curran, E.T., and Bergsten, M.B., “Inlet Efficiency Parameters for Supersonic

Combustion Ramjet Engines,” AR Aero propulsion Lab. Rep. APLTDR 64-61,

June 1964.

201. Curran, E.T., Leingang, J.L., Carriero, L.R., and Petters, D.P., “A Review of

Kinetic Energy Methods in High Speed Ramjet Cycle Analysis,” International

Symposium on Air Breathing Engines, Paper No. IS ABE 91-10.50-), AIAA,

1991.

202. Curran, E.T., Leingang, J., Carriero, LA., and Petters, DA., “Further Studies of

Kinetic Energy Methods in High Speed Ramjet Cycle Analysis,” AIAA Paper 92-

3805, July 1992.


187

CURRICULUM VITA

forAHMED A. TAHA

DEGREES:

• Doctor o f Philosophy (Mechanical Engineering), Old Dominion University,

Norfolk, VA, Dec. 2002

• Master o f Science (Mechanical Power Engineering), Cairo University, Cairo,

Egypt, June 1994

• Bachelor o f Science (Mechanical Power Engineering), Cairo University, Cairo,

Egypt, July 1988

PROFESSIONAL CHRONOLOGY:

• Mechanical Engineering Dept., Old Dominion University, Norfolk, VA. Research

and Teaching Assistant, Aug. 1996 - Dec. 2002

• Development, Research, and Technology Planning Center (DRTPC), Cairo,

Egypt. Consultant, Energy Dept., Feb. 1996 — Aug. 1996

• International Company for Engineering, Management, and Environment

Consultations (ICEMEC), Cairo, Egypt. Training Coordinator, Aug. 1994 - Oct.

1995

• Mechanical Power Engineering Dept., Cairo University, Cairo, Egypt. Research

and Teaching Assistant, Aug. 1988 - June 1994

HONORS AND AWARDS:

• Student Member, Virginia Academy o f Science (5/1999 - Present).

• Student Member, American Society for Mechanical Engineers (ASME) (1/1999 -

Present).

• Student Member, American Inst, for Aeronautics and Astronautics (AIAA)

(12/1996 — Present).

• Member, Mechanical Engineering Branch, Engineers Syndicate, Cairo, Egypt

(5/1988 — Present).


188

SCHOLARLY ACTIVITIES COMPLETED:

Refereed Journal Articles:

• Taha A.A., Tiwari S.N., Mohieldin T.O., “Mixing and Combustion o f Ethylene in

Scramjet Engines,” Journal o f Propulsion and Power. Vol. 18, No. 3, May-June

2002, pp. 716-718.

Selected Refereed Proceedings:

• Mohieldin T.O., Alfahaid A.F., Tiwari S.N., Taha, A.A., "On the Characteristics of

Turbulent Multiple Jets Diffusion Flame: Part 1 Cold Flow," International Thermal

conference, University o f Wales Swansea, Wales, UK, July 21-24, 1997.

• Tiwari S.N., Mohieldin T.O., Taha A.A., "Effects o f Supersonic Vitiated/Clean

Air on Ignition/Combustion Characteristics o f Hydrogen Fuel, Part I: Code

Validation," presented at the 37th AIAA Aerospace Sciences Meeting and Exhibit,

Reno, NV, January 11-14, 1999, AIAA Paper No. 99-0487, January 1999. Also

presented at the 77th Virginia Academy of Science Annual Meeting, Norfolk, VA,

May 25-28, 1999.

• Taha A.A., Tiwari SJST., Mohieldin T.O., "Study o f Supersonic Vitiated/Clean Air

Ignition/Combustion Characteristics o f Hydrogen Fuel," presented at the 9th

International Space Planes and Hypersonic Systems and Technologies Conference,

Norfolk, VA, November 1-5, 1999, AIAA Paper No. 99-4919, November 1999.

• Tiwari S.N., Taha A.A., Mohieldin T.O., "Numerical Investigation of Methane

Combustion in Supersonic Air Streams with Pilot-Hydrogen Injection," presented

at the 38th AIAA Aerospace Sciences Meeting and Exhibit, Reno, NV, January 10-

13,2000, AIAA Paper No. 2000-0320, January 2000.

• Taha A.A., Tiwari S.N., Mohieldin T.O., "Numerical Simulation of

Ignition/Combustion Characteristics o f Ethylene in Supersonic Air Streams,"

presented at the 36th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and

Exhibit, Huntsville, AL, July 16-19,2000, AIAA Paper No. 2000-3584, July 2000.

• Tiwari S.N., Taha A.A., Mohieldin T.O., "Mixing and Ignition/Combustion

Characteristics o f Hydrogen, Ethylene, and Kerosene Fuels in Supersonic Air

Streams," presented at the 34th National Heat Transfer Conference, Pittsburgh, PA,

August 20-22,2000; ASME Paper No. NHTC 2000-12149, August 2000.


189

• Taha A.A., Tiwari S.N., Mohieldin T.O., “Combustion o f Ethylene and Propane in

Scramjet Engines; Numerical Study,” for Presentation at the 39th Aerospace

Science Meeting and Exhibit, Reno, NV, January 8- i l , 2001.

• Tiwari S.N., Taha A.A., Mohieldin T.O., "Numerical Investigation o f Ethylene

Combustion in Supersonic Air Streams with Pilot-Hydrogen Injection," presented

at the 78th Virginia Academy of Science Annual Meeting, Radford, VA, May 23-

26,2000.

• Taha A.A., Tiwari S.N., Mohieldin T.O., “Combustion o f Hydrocarbon Fuels in

Scramjet Engines; Propane Combustion,” for Presentation at the 35th National

Heat Transfer Conference, Anaheim, CA, June 10-12, 2001.

• Taha A.A., Tiwari S.N., Mohieldin T.O., “Validation o f Fluent CFD Code in

Supersonic Flow Fields,” for Presentation at thel5lh AIAA Fluid Dynamics

Conference and Exhibit, Anaheim, CA, June 11-14,2001.

• Tiwari S.N., Taha A.A., Mohieldin T.O., “Effect of Pilot Injection and Combustor

Configuration on the Flame Holding and Stability o f Ethylene in Scramjet

Engines,” for Presentation at the 40th Aerospace Science Meeting and Exhibit,

Reno, NV, January 14-17,2002.

• Taha A.A., Tiwari S.N., Mohieldin T.O., “Pilot Injection and Flame

Characteristics o f Propane combustion in Scramjet Engines,” for Presentation at

the 38th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit,

Indianapolis, Indiana, July 7-10,2002.


Supersonic Combustion and Mixing Characteristics of ...

Documents

Transcript of Supersonic Combustion and Mixing Characteristics of ...