1. business and industrial
2. energy
3. oil
4. diesel fuel

SPRAY CHARACTERISTICS AND COMBUSTION PERFORMANCE OF UNHEATED
AND PREHEATED LIQUID BIOFUELS
by
HEENA VINODKUMAR PANCHASARA
A DISSERTATION
Submitted in partial fulfillment of the requirements
for the degree of Doctor of Philosophy
in the Department of Mechanical Engineering
in the Graduate School of
The University of Alabama
TUSCALOOSA, ALABAMA
2010
Copyright Heena Vinodkumar Panchasara 2010
ALL RIGHTS RESERVED
ABSTRACT
Recent increases in fuel costs, concerns for global warming, and limited supplies of fossil
fuels have prompted wide spread research on renewable liquid biofuels produced domestically
from agricultural feedstock. In the present research diesel, Vegetable Oil (VO), two types of
biodiesel produced from VO and animal fat are investigated as potential fuels for gas turbines to
generate power. Experiments are performed using a laboratory scale burner simulating gas
turbine combustor operated at atmospheric pressure. A commercially available airblast (AB)
atomizer is used to create the fuel spray. A parametric study of combustion performance (CO and
NOx emissions) and spray characteristics (droplet diameter, drop size distribution, and mean and
RMS axial velocities) is carried out by varying air to liquid mass ratio (ALR), and fuel inlet
temperature in cold spray and spray flame with/without swirl air and without/with enclosure.
The problems of high viscosity and poor volatility of VO (soybean oil) were addressed by
using diesel-VO blends with up to 30% VO by volume. Gas chromatography/mass spectrometry,
thermogravimetric analysis, and density, kinematic viscosity, surface tension and water content
measurements are used to characterize the fuel properties. Characteristics of the resulting spray
are measured using a laser sheet visualization system and a Phase Doppler Particle Analyzer
system (PDPA). However, several operational and durability problems of using straight VO’s for
direct combustion occur because of their higher viscosity and low volatility compared to diesel
fuel. The high kinematic viscosity of vegetable oil (VO) makes it unsuitable for direct
combustion using conventional fuel preparation systems. Thus, we preheat the fuel to reduce its
ii
kinematic viscosity and to improve fuel atomization.
Measurements are obtained for fuel inlet temperature varying from 40 to 100°C and for
ALR varying from 2 to 4. Results show that an increase in the fuel inlet temperature decreases
NOx and CO emissions, which can be attributed to improved fuel atomization resulting from
decreased kinematic viscosity at higher fuel temperatures. Results also show a decrease in Sauter
Mean Diameter (SMD) with an increase in VO temperature, regardless of the ALR at any given
axial location in the spray. A significant difference in the distributions of mean and root mean
square (RMS) axial velocity occurs with an increase in VO inlet temperature for a fixed ALR,
presence of swirling air, and presence of flame. In general, the radial profiles show larger
droplets distributed towards the edge of the spray and smaller droplets in the interior spray
region. Higher VO inlet temperature and higher ALR produced a narrower spray with smaller
diameter droplets and higher peak axial velocities. Swirling air flow and high temperatures in
flames facilitate secondary breakup of larger droplets to significantly reduce the SMD. Finally
the effect of enclosure is also studied since it represents a more realistic combustor design for
any continuous flow system. The insulated enclosure eliminated the ambient air entrainment and
minimized heat loss to the ambient air to create a fine spray flame with characteristics similar to
those of an open flame.
iii
DEDICATION
This dissertation is dedicated to my beloved parents: “Mr. Vinodkumar G. Panchasara”
and “Mrs. Sadhana Vinod Panchasara”.
iv
ACKNOWLEDGMENTS
Though only my name appears on the cover of this dissertation, many of my colleagues,
friends and faculty members have encouraged and supported its production. I owe my gratitude
to all those people who have made this dissertation possible and because of whom my graduate
experience at UA has been one that I will cherish forever.
At the outset, my sincere thanks and deepest gratitude is extended to my advisor, Dr.
Ajay K. Agrawal. I have been amazingly fortunate to have an advisor who gave me the freedom
to explore on my own and at the same time the guidance to recover when my steps faltered. Dr.
Agrawal taught me how to question thoughts and express ideas. He has challenged me to think
more critically about my work and made me a better graduate student. His patience,
perseverance, practical and technical knowledge, organization skills and support helped me
overcome many crisis situations and finish this dissertation. Without his continuous and untiring
support, this work could not have seen its timely completion. I would take the opportunity to
thank Southern Co., and U.S. Department of Energy (DOE) for supporting me during the course
of my doctoral studies.
I gratefully thank my committee members, Dr. Ashford, Dr. Baker, Dr. Daly and Dr.
Lane, Dr. Taylor for their guidance and their invaluable support of my academic progress as well
as dissertation. I wish to thank Dr. Woodbury for lending me the infra-red thermal imaging
camera and spending time with me on data processing. I am also Department for helping me
directly or indirectly in bringing this work to successful completion.
v
I extend my sincere thanks to Jim Edmonds, James, Sam, Ken Dunn and Barry Johnson
who have helped me to a great extent with their timely support in the machine shop as well as lab
setting. I am extremely thankful to Ms. Lynn Hamric, Ms. Pamelia Bedingfield, Lisa Hinton and
Ms. Betsy Singleton for their cooperation with the administrative affairs during my stay at U.A.
My graduate studies would not have been the same without the social and academic
challenges and diversions provided by all my colleagues-friends during my stay here. I am
particularly thankful to Daniel Sequera, Ben Simmons, Troy Dent, Pankaj Kolhe, Tanisha
booker, Vijaykant Sadasivuni and Lulin Jiang. We not only studied, relaxed, and traveled
together, but they were always willing to provide me with any help and support in the lab as well
as stylistic suggestions and substantive challenges to help me improve my presentation skills and
clarify my arguments. Their efforts and friendship has really made a difference during the course
of my graduate studies here. I also extend my gratitude to my friends Saskia Clayton,Aparajita
Sengupta, Cehtna Maini, Indu Ankareddi,Vishal Warke, Krishna Chetri,Yogin Patel, Jayraj
Sethji, Vishwas Gadhvi for being friends and for the support they have lent me during my stay
in Tuscaloosa.
My very special thanks to Yajuvendrasinh Parmar ,Suhani Shah and Hardik Singh for
their constant encouragement and support. Their unconditional love and friendship is invaluable.
It’s beyond words to express my gratitude towards my parents whom I owe everything I
am today, Mrs. Sadhana V. Panchasara and Mr. Vinod G. Panchasara. Their unwavering faith
and confidence in my abilities and in me is what has shaped me to be the person I am today.
Thank you for everything. Without their love and support, encouragement and firm faith in me
this would not have been possible. I also thank my brother Sumit Panchasara and my sister in
law Ankita Panchasara for their love and support.
vi
It is indeed a great pleasure to look back and take this opportunity to express my gratitude
towards all the people for their help in my efforts which have really made a difference in my life.
Thank you every one.
vii
CONTENTS
ABSTRACT........................................................................................................................ ii
DEDICATION................................................................................................................... iv
ACKNOWLEDGMENTS ...................................................................................................v
LIST OF TABLES............................................................................................................ xii
LIST OF FIGURES ......................................................................................................... xiii
1. INTRODUCTION ...........................................................................................................1
1.1 Background ..............................................................................................................1
1.2 Combustion Fundamentals.......................................................................................1
1.3 Liquid Fuel Combustion ..........................................................................................2
1.4 Objectives ................................................................................................................8
1.5 Overview..................................................................................................................9
2. PROPERTIES OF BIODIESEL AND DIESEL-VEGETABLE OIL BLENDS...........18
2.1 Background ............................................................................................................18
2.2 Fuel Property Measurements .................................................................................21
2.2.1 GC-MS Analysis............................................................................................23
2.2.2 Thermogravimetric Analysis (TGA)..............................................................23
2.2.3 Density, Viscosity and Surface Tension ........................................................25
2.2.4 Water Content ...............................................................................................28
2.3 Spray Characteristics .............................................................................................29
viii
2.3.1 Spray Angle Measurements...........................................................................29
2.3.2 Droplet Size Characteristics .........................................................................30
2.4 Conclusions............................................................................................................31
3. EFFECT OF FUEL PREHEATING ON EMISSIONS FROM COMBUSTION OF
VISCOUS BIOFUELS .....................................................................................................41
3.1 Background ............................................................................................................41
3.2 Experimental Set-Up..............................................................................................42
3.3 Results and Discussion ..........................................................................................44
3.3.1 Visual Flame Images.....................................................................................45
3.3.2 Emissions Profiles.........................................................................................45
3.4 Conclusions............................................................................................................47
4. CHARACTERISTICS OF PREHEATED NON-EVAPORATING BIO-OIL
SPRAYS............................................................ ..........................................................64
4.1 Background ............................................................................................................64
4.2 Experimental Set-Up..............................................................................................67
4.3 Results and Discussion ..........................................................................................69
4.3.1 Spray Images.................................................................................................69
4.3.2 SMD Contour Plots.......................................................................................70
4.3.3 Axial Velocity Contour Plots ........................................................................71
4.3.4 Radial Profiles of SMD.................................................................................72
4.3.5 Droplet Diameter Distributions....................................................................73
4.4 Conclusions............................................................................................................74
5. CHARACTERISTICS OF PREHEATED BIO-OIL SPRAYS.....................................97
5.1 Background ...........................................................................................................97
ix
5.2 Experimental Set-Up and Procedure....................................................................102
5.2.1 Combustion System .....................................................................................102
5.2.2 Fuel Injection System..................................................................................104
5.2.3 Flare System................................................................................................105
5.2.4 Emissions Measurement System .................................................................106
5.2.5 Phase Doppler Particle Analyzer (PDPA) Set-Up .....................................107
5.2.6 Traversing System.......................................................................................108
5.2.7 Infrared (IR) Imaging .................................................................................109
5.2.8 Test Conditions ...........................................................................................110
5.3 Results and Discussion ........................................................................................111
5.3.1 Effect of Swirling Air Flow on Open Cold Spray ......................................111
(a) Axial Velocity Contours (Mean and RMS) ............................................111
(b) SMD Contours.......................................................................................114
(c) Transverse Profiles of Mean and RMS Axial Velocity ..........................116
(d) Transverse Profiles of SMD ..................................................................118
(e) Droplet Diameter Distribution Profiles ................................................120
5.3.2 Effect of Flame on Open Spray ..................................................................121
(a) Velocity Contours (Mean and RMS)......................................................121
(b) SMD Contours.......................................................................................122
(c) Transverse Profiles of Mean and RMS Axial Velocity ..........................123
(d) Transverse Profiles of SMD ..................................................................124
(e) Droplet Diameter Distribution Profiles ................................................126
5.3.3 Effect of Enclosure on Spray Flame and Emissions...................................128
x
(a) Transverse Profiles of Mean and RMS Axial Velocity ..........................129
(b) Transverse Profiles of SMD ..................................................................130
(c) Droplet Diameter Distribution Profiles ................................................131
(d) Effect of Enclosed Spray on Combustion Emissions.............................132
(d) Enclosure Exterior Surface Temperature Distributions .......................133
5.4 Conclusions....................................................................................................134
6. CONCLUSIONS AND RECOMMENDATIONS ......................................................251
6.1 Conclusions....................................................................................................251
6.2 Recommendations..........................................................................................253
REFERENCES ................................................................................................................255
APPENDIX A..................................................................................................................261
APPENDIX B ..................................................................................................................264
APPENDIX C ..................................................................................................................267
APPENDIX D..................................................................................................................270
APPENDIX E ..................................................................................................................272
xi
LIST OF TABLES
2.1 List of Fuels used in the Study.....................................................................................21
2.2 Results of GC-MS Analysis.........................................................................................22
2.3 Physical Properties of Diesel, Biodiesel # 1, Biodiesel # 2 and Vegetable Oil...........26
2.4 Water Content in the Fuel ............................................................................................28
3.1 Physical Properties of VO at 25°C .............................................................................44
5.1 Properties and Characteristics of the Insulating Material ..........................................103
xii
LIST OF FIGURES
1.1 (a) Airblast Injector Concept (b) Injector Details. ...................................................................11
1.2 Working Principle of the Flow Blurring Injector ....................................................................12
1.3 Effect of Atomizing Airflow Rate on Visual Images of Diesel Flames: FB Injector
(Top); AB Injector (Bottom)....................................................................................................13
1.4 Axial Profiles of Emissions in Diesel Flames; (a) CO Concentration, (b) NOx
Concentration. [Open symbols represent FB Injector, Closed Symbols represent AB
Injector]....................................................................................................................................14
1.5 Radial Profiles of Emissions in Diesel Flames; (a) CO Concentration, (b) NOx
Concentration. [Open symbols represent FB Injector, Closed Symbols represent AB
Injector]....................................................................................................................................15
1.6 Axial Profiles of Emissions in Kerosene Flames; (a) CO Concentration, (b) NOx
Concentration. [Open symbols represent FB Injector, Closed Symbols represent AB
Injector]....................................................................................................................................16
1.7 Radial Profiles of Emissions in Diesel Flames; (a) CO Concentration, (b) NOx
Concentration. [Open symbols represent FB Injector, Closed Symbols represent AB
Injector]....................................................................................................................................17
2.1 Results of Thermogravimetric Analysis. .................................................................................33
2.2 Kinematic Viscosity of Diesel and Biodiesel Fuels.................................................................34
2.3 Kinematic Viscosity of Diesel-VO Blends. .............................................................................35
2.4 Airblast Atomizer Details ........................................................................................................36
2.5 Spray Visualization Photographs. (a) ALR 2.52 BD-1, (b) ALR 4.35 BD-1, (c) ALR
3.92 VO....................................................................................................................................37
2.6 Spray Cone Angle vs. Atomizing Airflow Rate... ...................................................................38
2.7 SMD versus ALR. (a) Diesel, BD-1 and BD-2. .....................................................................39
xiii
2.8 SMD versus ALR for Diesel, VO, and Diesel-VO Blends......................................................40
3.1 Schematic Diagram of the Experimental Setup; all dimensions are in cm..............................48
3.2 Schematic Diagram (top) and Photograph of the Swirler (bottom) at the Combustor Inlet
Plane; all dimensions are in cm ...............................................................................................49
3.3. Effect of Temperature on Kinematic Viscosity of Diesel and VO fuel ..................................50
3.4 Effect of ALR on Fuel Inlet Temperature................................................................................51
3.5. VO Flame images at ALR 2.4 for different fuel temperatures (a) Unheated VO, (b) 58o
C, (c) 78 oC, (d) 99 oC, (e) 121 oC ...........................................................................................52
3.6 VO Flame images at ALR 4.0 for different fuel temperatures (a) Unheated VO, (b) 58o
C, (c) 78oC, (d) 99 oC, (e) 121 oC. ...........................................................................................53
3.7. [(a) and (b)] Axial profiles for CO and NOx for different fuel temperatures at ALR 2.4......54
[(c) and (d)] Axial profiles for CO and NOx for different fuel temperatures at ALR 3.5.......55
[(e) and (f)] Axial profiles for CO and NOx for different fuel temperatures at ALR 4.0........56
3.8. [(a) and (b)] Radial profiles for CO and NOx for different fuel temperatures at ALR
2.4.............................................................................................................................................57
[(c) and (d)] Radial profiles for CO and NOx for different fuel temperatures at ALR
3.5.............................................................................................................................................58
(e) and (f)] Radial profiles for CO and NOx for different fuel temperatures at ALR 4.0 .......59
3.9. [(a) and (b)] Radial profiles for CO and NOx for ALR 2.4, ALR 3.5 and ALR 4.0 at
Tf = 58oC..................................................................................................................................60
[(c) and (d)] Radial profiles for CO and NOx for ALR 2.4, ALR 3.5 and ALR 4.0 at Tf
= 78oC ......................................................................................................................................61
[(e) and (f)] Radial profiles for CO and NOx for ALR 2.4, ALR 3.5 and ALR 4.0 at Tf
= 99oC. .....................................................................................................................................62
[(g) and (h)] Radial profiles for CO and NOx for ALR 2.4, ALR 3.5 and ALR 4.0 at Tf
= 121oC ....................................................................................................................................63
4.1 Kinematic Viscosity of Diesel and Vegetable Oil. ..................................................................76
4.2. Schematic representation of the liquid break-up indicating geometry and different
lengths. .....................................................................................................................................77
xiv
4.3 Schematic diagram of the experimental setup. .......................................................................78
4.4 Air-Blast Injector Details.........................................................................................................79
4.5 Schematic of PDPA system. ....................................................................................................80
4.6 Laser sheet spray images for VO at 40 C, 70 C and 100 C at ALR 2.0. (a) Spray Images
at Exposure time 100 ms. (b) Spray Images at Exposure time 40 ms .....................................81
4.7 Laser sheet spray images for VO at 40 C, 70 C and 100 C at ALR 4.0. (a) Spray Images
at Exposure time 100 ms. (b) Spray Images at Exposure time 40 ms .....................................82
4.8 Contours of Sauter Mean Diamter for (a) T = 40 C, ALR = 2.0, (b) T = 70 C, ALR
2.0.............................................................................................................................................83
(c) T = 100 C, ALR = 2.0 and (d) T = 70 C, ALR = 4.0. ........................................................84
4.9 Contours of Axial Velocity for (a) T = 40 C, ALR = 2.0, (b) T = 70 C, ALR 2.0. .................85
(c) T = 100 C, ALR = 2.0 and (d) T = 70 C, ALR = 4.0.. .......................................................86
4.10 Contours of Axial RMS Velocity for (a) T = 40 C, ALR = 2.0, (b) T = 70 C, ALR 2.0.......87
(c) T = 100 C, ALR = 2.0 and (d) T = 70 C, ALR = 4.0. ........................................................88
4.11 Profiles of Sauter Mean Diameter for (a) Y =20 mm ............................................................89
(b) Y = 50 mm .........................................................................................................................90
(c) Y = 80 mm..........................................................................................................................91
4.12 Profiles of Sauter Mean Diameter versus Fuel Viscosity for (a) Y =10 mm.........................92
(b) Y = 20 mm .........................................................................................................................93
(c) Y = 40 mm..........................................................................................................................94
4.13 Profiles of Diameter Distribution for (a) T = 40 C, ALR = 2.0, (b) T = 70 C, ALR 2.0.......95
(c) T = 100 C, ALR = 2.0 and (d) T = 70 C, ALR = 4.0 .........................................................96
5.1 Flame structure of an airblast atomizer..................................................................................138
5.2 Vaporization of a typical droplet in an idealized spray flame ...............................................139
5.3 Schematic of the combustor experimental set-up ..................................................................140
xv
5.4 Photographic view of the enclosure without and with insulation and schematic of the top
view of the pentagonal enclosure……………………………………...................................141
5.5 Airblast Injector Details.........................................................................................................142
5.6 Schematic of Liquid Fuel Supply System..............................................................................143
5.7 Schematic of Gaseous Fuel Supply System...........................................................................143
5.8 Photographic representation of the Flare system. ..................................................................144
5.9 (a) Emissions Analyzer; (b) Emissions Measurement Traversing system.............................145
5.10 Schematic of the PDPA system. ..........................................................................................146
5.11 Experimental set-up of a PDPA system mounted on a 3-way traversing system................147
5.12 Plan view of the PDPA traversing mechanism in radial and axial coordinates...................148
5.13 Photographic view of the PDPA system integrated with the combustor assembly .............149
5.14 Photographic view of the Infrared Camera..........................................................................150
5.15 Mean axial velocity contour for cold spray without swirl. ..................................................151
5.16 Mean axial velocity contour for cold spray with swirl. .......................................................152
5.17 Mean axial velocity contour for cold spray with swirl at Tf = 100 C. .................................153
5.18 RMS axial velocity contour for cold spray without swirl....................................................154
5.19 RMS axial velocity contour for cold spray with swirl.........................................................155
5.20 RMS axial velocity contour for cold spray with swirl at Tf = 100 C. ..................................156
5.21 SMD contour for cold spray without swirling air................................................................157
5.22 SMD contour for cold spray with swirling air... ..................................................................158
5.23 SMD contour for cold spray with swirling air at Tf = 100 C...............................................159
5.24 Transverse profiles of mean axial velocity and RMS axial velocity for cold spray with
and without swirl at
(a and b) Y = 5 mm ................................................................................................................160
xvi
(c and d) Y = 10 mm. .............................................................................................................161
(e and f) Y = 15 mm...............................................................................................................162
(g and h) Y = 20 mm..............................................................................................................163
(i and j) Y = 25 mm. ..............................................................................................................164
(k and l) Y = 30 mm...............................................................................................................165
(m and n) Y = 35 mm.............................................................................................................166
(o and p) Y = 40 mm..............................................................................................................167
(q and r) Y = 45 mm. .............................................................................................................168
(s and t) Y = 50 mm. ..............................................................................................................169
(u and v) Y = 60 mm..............................................................................................................170
(w and x) Y = 70 mm.............................................................................................................171
(y and z) Y = 75 mm. .............................................................................................................172
5.25 Transverse profiles of mean axial velocity and RMS axial velocity for swirling air
cold spray; unheated and VO at Tf = 100°C
(a and b) Y = 5 mm ................................................................................................................173
(c and d) Y = 10 mm. .............................................................................................................174
(e and f) Y = 15 mm...............................................................................................................175
(g and h) Y = 20 mm..............................................................................................................176
(i and j) Y = 25 mm. ..............................................................................................................177
(k and l) Y = 30 mm...............................................................................................................178
(m and n) Y = 35 mm.............................................................................................................179
(o and p) Y = 40 mm..............................................................................................................180
(s and t) Y = 50 mm. ..............................................................................................................181
(u and v) Y = 60 mm..............................................................................................................182
xvii
(w and x) Y = 70 mm..........................................................................................................183
(y and z) Y = 75 mm. ..........................................................................................................184
5.26 Transverse profiles of SMD for cold spray with and without swirl at
(a and b) Y = 5 mm and 10 mm. .........................................................................................185
(c and d) Y = 15 mm and 20 mm.. ......................................................................................186
(e and f) Y = 25 mm and 30 mm.........................................................................................187
(g and h) Y = 35 mm and 40 mm........................................................................................188
(i and j) Y = 45 mm and 50 mm .........................................................................................189
(k and l) Y = 60 mm and 70 mm.........................................................................................190
(m) Y = 75mm ....................................................................................................................191
5.27 Transverse profiles of SMD for swirling air cold spray; unheated and VO at Tf =
100°C
(a and b) Y = 5 mm and 10 mm..........................................................................................192
(c and d) Y = 15 mm and 20 mm........................................................................................193
(e and f) Y = 25 mm and 30 mm.........................................................................................194
(g and h) Y = 35 mm and 40 mm........................................................................................195
(i and j) Y = 50 mm and 60 mm .........................................................................................196
(k and l) Y = 70 mm and 75 mm.........................................................................................197
5.28 Droplet Distribution Profile for cold spray with and without swirl at
(a and b) Y = 5mm, X = 0 mm and Y = 10 mm, X = 0 mm. ..............................................198
c and d) Y = 40mm, X = 0 mm and Y = 60 mm, X = 0 mm. .............................................199
5.29 Mean axial velocity contour for flame spray. ......................................................................200
5.30 Mean axial velocity contour for flame spray for VO at 100°C............................................201
5.31 RMS axial velocity contour for flame spray........................................................................202
xviii
5.32 RMS Axial Velocity contour for flame spray for VO at 100°C. .........................................203
5.33 SMD contour for flame spray. .............................................................................................204
5.34 SMD contour for flame spray for VO at 100°C...................................................................205
5.35 Transverse profiles of mean axial velocity and RMS axial velocity for flame spray and
cold spray at
(a and b) Y = 5 mm.............................................................................................................206
(c and d) Y = 10 mm...........................................................................................................207
(e and f) Y = 15 mm............................................................................................................208
(g and h) Y = 20 mm...........................................................................................................209
(i and j) Y = 25 mm. ...........................................................................................................210
(k and l) Y = 30 mm. ..........................................................................................................211
(m and n) Y = 35 mm. ........................................................................................................212
(o and p) Y = 40 mm...........................................................................................................213
5.36 Transverse profiles of mean axial velocity and RMS axial velocity for flame spray and
cold spray for Tf = 100°C at
(a and b) Y = 5 mm. ............................................................................................................214
(c and d) Y = 10 mm...........................................................................................................215
(e and f) Y = 15 mm............................................................................................................216
(g and h) Y = 20 mm...........................................................................................................217
(i and j) Y = 25 mm. ...........................................................................................................218
(k and l) Y = 30 mm. ..........................................................................................................219
(m and n) Y = 35 mm. ........................................................................................................220
(o and p) Y = 40 mm...........................................................................................................221
5.37 Transverse profiles of SMD for flame spray and cold spray at
(a and b) Y = 5 mm and 10 mm. .........................................................................................222
xix
(c and d) Y = 15 mm and 20 mm. .......................................................................................223
(e and f) Y = 25 mm and 30 mm.........................................................................................224
(g and h) Y = 35 mm and 40 mm........................................................................................225
5.38 Transverse profiles of SMD for flame spray and cold spray for Tf = 100°C at
(a and b) Y = 5 mm and 10 mm. .........................................................................................226
(c and d) Y = 15 mm and 20 mm. .......................................................................................227
(e and f) Y = 25 mm and 30 mm.........................................................................................228
(g and h) Y = 35 mm and 40 mm........................................................................................229
5.39 Droplet Distribution Profile for flame spray and cold spray at
(a and b) Y = 5mm, X = 0 mm and Y = 10 mm, X = 0 mm. ..............................................230
(c and d) Y = 40mm, X = 0 mm and Y = 60 mm, X = 0 mm. ............................................231
5.40 Droplet Distribution Profile for flame spray and cold spray for Tf = 100°C at
(a and b) Y = 5mm, X = 0 mm and Y = 10 mm, X = 0 mm. ..............................................232
(c and d) Y = 40mm, X = 0 mm and Y = 60 mm, X = 0 mm. ............................................233
5.41 Transverse profiles of mean axial velocity and RMS axial velocity for enclosed flame
for Tf = 100°C and Tf = 150°C at
(a and b) Y = 5 mm.............................................................................................................234
(c and d) Y = 10 mm...........................................................................................................235
(e and f) Y = 15 mm............................................................................................................236
(g and h) Y = 20 mm...........................................................................................................237
(i and j) Y = 25 mm. ...........................................................................................................238
(k and l) Y = 30 mm............................................................................................................239
(m and n) Y = 35 mm..........................................................................................................240
xx
5.42 Transverse profiles of SMD for enclosed flame for Tf = 100°C and Tf = 150°C at
(a and b) Y = 5 mm and 10 mm. .........................................................................................241
(c and d) Y = 15 mm and 20 mm. .......................................................................................242
(e and f) Y = 25 mm and 30 mm.........................................................................................243
(g and h) Y = 35 mm...........................................................................................................244
5.43 Droplet Distribution Profile for enclosed flame for Tf = 100°C and Tf = 150°C at
(a and b) Y = 5mm, X = 0 mm and Y = 10 mm, X = 0 mm. ..............................................245
(c and d) Y = 30mm, X = 0 mm and Y = 35 mm, X = 0 mm. ............................................246
5.44 Transverse profiles of CO for enclosed VO flame at 100°C and 150°C... ..........................247
5.45 Transverse profiles of NOx for enclosed VO flame at 100°C and 150°C...........................248
5.46 IR Image of the enclosure surface temperature... ................................................................249
5.47 Profile plot for exterior surface temperature at different axial location on the
enclosure... .............................................................................................................................250
xxi
CHAPTER 1
INTRODUCTION
Background
Global interest in clean power generation has driven continued improvement of power
systems. Improvement efforts focus on reducing emissions, increasing efficiency and lowering
costs without sacrificing reliability. Most continuous combustion systems operate on a single
fuel, mostly gaseous fuel. However, fuel flexible power systems that can operate on both gaseous
and liquid fuels are needed to minimize the operation costs. The ignition and combustion of
liquid fuels plays a major role in diesel engines, gas turbines and industrial burners. Liquid fuels
are usually burned as sprays of small liquid droplets, the droplets first evaporating to produce a
cloud of vapor which then burns in the gas phase.
1.2. Combustion Fundamentals:
One of the most important parameter used to characterize combustion is the fuel to air
ratio (f/a), expressed either on a volume or mass basis. With precisely enough air to consume all
of the fuel, theoretically, combustion is said to have a stoichiometric f/a ratio. Adding more air
produces combustion that is fuel-lean, and operating on less air produces fuel-rich combustion.
Because differing fuels have different stoichiometric fuel/air ratios, it is convenient to normalize
the fuel/air ratio by the stoichiometric value,
1
leading to the well known equivalence ratio, Φ
φ=
(f / a )
(f / a ) stoich
Combustion using different types of fuel is readily described as
(i)
Fuel-Lean if Φ < 1
(ii)
Fuel-Rich if Φ > 1
Another important combustion parameter is the adiabatic flame temperature. In principle,
the highest temperature would be attained at Φ = 1, because all the fuel and oxygen would be
consumed. In practice, the effects of species dissociation and heat capacity shifts the peak
temperature to slightly above stoichiometric Φ ~ 1.05.
1.3. Liquid Fuel Combustion
Combustion of liquid fuels provides a significant fraction of current total energy supply.
For efficient combustion to occur the fuel and air should be intimately mixed. Since combustion
of solid fuels result in higher emissions like soot and nitric oxides (NOx), liquid and gaseous
fuels are becoming more prevalent. Moreover the liquid fuels offer a greater flexibility since the
fuel can be transported easily as compared to gaseous fuels requiring large scale stationary
operation. Combustion of liquid fuels in diesel engines, spark ignition engines, gas turbines,
rocket engines and industrial furnaces is dependent on effective atomization to increase the
specific surface area of the fuel, thereby achieving high mixing rates and evaporation. The
process of atomization is one in which a liquid jet or sheet is disintegrated by the kinetic energy
2
itself, or by being subjected to a high velocity gas. A spray serves as the heart of almost every
type of liquid-fueled combustion system. Recent emphasis on fuel conservation and stringent
environmental regulations has prompted increased research examining the processes involved in
spray production.
Ideally, to promote combustion with maximum efficiency and minimum harmful
emissions, an injector should deliver a fuel spray that rapidly disperses and evaporates to yield a
homogeneous mixture of gaseous fuel and air. The diffusion mode of burning droplets, which
promotes the formation of soot and high nitric oxide (NOx) and carbon monoxide (CO)
emissions (Lefebvre, 1989], is consequentially avoided. From a practical standpoint, the fuel
injector should deliver good atomization over a range of fuel mass flow rates and incur low
pressure drop. Fuel jets or sheets break up into sprays in various regimes that depend upon the
influence and interaction of the liquid’s kinetic energy, surface tension, viscosity, and
surrounding air. The mechanism by which the liquid jet breaks up into a spray was first broached
by Rayleigh who analyzed the stability of a low speed, inviscid jet (Lasheras & Hopfinger,
2000).
In what Lefebvre (1989) has defined as “classical” atomization, a liquid jet or sheet,
subjected to destabilizing forces, breaks up after it leaves the injector. In the classical
atomization regime, a liquid jet or sheet initially disintegrates into ligaments which in turn break
up into droplets, whose size depends on the characteristic length scale of the liquid jet or sheet.
If the Weber number, defined as the ratio of the jet’s kinetic energy to its surface tension, is
small, surface instabilities control the break-up. Some of the characteristics (Lefebvre ,1989),
has listed for an ideal injector include good atomization over a wide range of either steady or
3
transient fuel flow rates, unaffected by flow instabilities, low power requirements, scalability,
resistant to blockages, and delivery of a fine spray. Lin and Reitz (1988) have provided a good
review along with a stability analysis to predict the breakup of liquid jets. Reviews by McDonell
and Samuelsen (1991), Radzan (1998),Mansour (2005), and Nakamura et al (2006) deal with fuel
sprays optimized for lean combustion.
Several approaches to atomize the fuel the have been developed in the past. For example,
a high pressure gradient across a nozzle will ensure a large velocity differential at the jet’s liquidair interface which, in turn leads to rapid break-up of the fuel jet. This simple arrangement of
“pressure atomizer” can be problematic if the fuel flow rate is to extend over a wide range. The
orifice size and pressure gradient must accommodate the largest anticipated fuel flow rate, and
hence, the required pressure gradient cannot be maintained for low fuel flow rates. A second
strategy for fuel jet breakup transfers energy from the surrounding air to the fuel jet through the
use of air-assist or air blast. The air-assist injector employs a low volume flow of high velocity
air to break up the fuel jet. The air-blast atomizer delivers a relatively larger volume flow of low
velocity air to both break up the fuel jet and to disperse the resulting spray to the combustion
zone. For the air-assist and air-blast atomization procedures, the production of either small
volumes of high pressure air or large volumes of low pressure air can be expensive in terms of
auxiliary power requirements. Prefilming the fuel represents a third type of atomization process
in which the liquid fuel leaves the injector surface as a sheet. Subsequently, the thin liquid sheet
is subjected to air-blast on both sides to affect a fine droplet spray. The above three atomization
concepts have been combined into wide varieties of fuel injectors designed to suit their particular
usage. A fourth process known as effervescent atomization injects high pressure air directly into
the liquid fuel before injection. Effervescent atomization, as described by Sovani et al (2001),
4
typically produces a spray with fine, non-uniform, droplets; however, the resulting spray angle is
small and air pumping costs are high since the air must be pressurized to the fuel’s pumping
level.
Recently, Ganan-Calvo reported the discovery of a simple, reproducible atomizer
configuration, the so called Flow blurring injector, with gas-liquid interactions that yield in high
atomization efficiency. For a specified liquid flow rate and total energy input, Ganan-Calvo
claims that the FB atomizer creates about five to fifty times more droplet surface area than any
other pneumatic atomizer of the “plain-jet airblast” type. The basic operating principle of the FB
atomizer can be seen in Figure 1. Conceptually, in a FB atomizer the air is forced through a small
gap between the exit of fuel tube with inside diameter of ‘d’ and a coaxial orifice in a
perpendicular plate of the same diameter as ‘d’. The nozzle’s wall is tapered at the outlet, and the
sharpened edge of the orifice plate is the same diameter as the inner diameter of the nozzle (d).
When the axial distance between the nozzle exit and orifice place (H) is small (i.e., H/d <0.25),
some of the gas flowing into the lateral cylindrical passageway (LCP) between the nozzle exit
and orifice plane is forced upstream a short distance into the liquid nozzle. Flow recirculation
and vigorous turbulent mixing of the liquid and the gas occur in this short nozzle length result in
a fine spray forming downstream of the orifice plate. The FB atomizer operates on the same
atomization principle as the effervescent process but overcomes some of the effervescent
atomizer difficulties.
A study was conducted by Panchasara (2009), and Simmons (2008), to demonstrate the
behavior of a FB and an AB atomizer in a swirl stabilized combustion system with diesel and
kerosene as fuels. The FB injector concept has never been demonstrated in combustion systems.
5
Some of the results from the study are presented in Figures 1.3, 1.4, 1.5 and 1.6. The
performance of these atomizers is shown in terms of emission results and qualitative flame
images at a given fuel and air flow rates. Figure 1.4 presents the NOx and CO emission profiles
along the axis of the combustor in diesel flames for both injectors. The axial distance (z) is these
profiles is measured from the injector exit plane; thus z = 45 cm refers to the combustor exit
plane Figure 1.4 (a) shows that the CO concentrations tend to increase and then decrease in the
axial direction. Initially, the CO is produced in the flame during the fuel breakdown process and
it is subsequently oxidized in the reaction and post-reaction zones. An increase in the atomizing
air flow rate decreases the CO emissions since the premixed mode of combustion becomes more
prevalent. The NOx concentrations presented in figure 1.4 (b) are nearly constant in the axial
direction. Evidently, most of the NOx is formed in the reaction zone within z = 12 cm.
Emissions measurements were also taken at the combustor exit plane to characterize
variations in the transverse flow direction. Figure 1.5 shows that the NOx and CO
concentrations are nearly independent of the radial coordinate at the combustor exit plane,
although flow asymmetry is evident at low atomizing airflow rates because of the turbulent
fluctuations in the resulting diffusion flame. Results show that the flow mixing within the
combustor forms a homogeneous product gas mixture at the combustor exit, especially at higher
atomizing airflow rate. Figure 1.5 shows that the CO and NOx emissions decrease with
increasing atomizing air flow rate. Emissions profiles in the axial and radial directions for the
kerosene flames presented, respectively, in figures 1.6 and 1.7, show the same trends as those
observed previously in diesel flames: (i) the CO emissions initially increase and then decrease in
the axial direction, (ii) much of the NOx is formed in a short reaction zone near the injector exit,
(iii) the CO profiles at the combustor exit display asymmetry for low atomizing airflow rates but
6
they are uniform at higher atomizing airflow rates when combustion occurs in the premixed
mode, (iv) the NOx emission profiles are nearly uniform at the combustor exit plane, (v) both
CO and NOx emissions decrease with increase in the atomizing airflow rate.
Combustion of liquid fuels is largely dependent on effective ways of atomization. Fuel
atomization plays an important role in combustion of liquid biofuels as well. Economic and
political factors make supplies of petroleum uncertain and have given rise to tremendous price
escalation, for developing countries where oil imports are imperative. Moreover, the depleting
resources of petroleum-based fuels, the increasing threat to the environment from NOx, CO,
hydrocarbon emissions and global warming, have given momentum to the search for alternative
fuels which are renewable. Current research on biofuels mostly focuses on their use in internal
combustion engine applications, but they do show promise in power generation sector in gas
turbines and process burners. As fuel candidates for power generation are mushrooming
worldwide; there is presently a surge of interest around the liquid biofuels. Among these biofuels
available today, biodiesel fuels are the natural candidates due to their compatibility with gasoil
and their increasing use in the transportation sector as well as other areas of combustion such as
gas turbines for power generation.
Among the many different types of alternative fuels available, edible and non-edilble oils
and their derivatives can be promising options. They are renewable as the carbon released by
burning of vegetable oils is used when the oil crops undergo photosynthesis. In United States, the
focus of vegetable oils research is on the most abundant oil-producing crop, the soybean oil.
Vegetable oils have long hydrocarbon chains with trace amounts of triglycerides, that makes
them highly viscous with low volatility and hence they show difficulty in atomizing compared to
7
diesel fuels. Different methods are used to reduce the kinematic viscosity and improve volatility
of VO fuel. One of the methods to improve the physical properties of original VO is by the
process of transesterification. This chemical process however requires additional energy, time,
and cost penalty to prepare the fuel for acceptable atomization. Second method is to preheat the
fuel to reduce the kinematic viscosity and thereby making it possible to atomize in combustion
systems. Although fuel properties are important, it has been shown experimentally that the fuel
atomization process is critical for a turbine combustor to minimize the emissions for a given fuel.
Previous studies on various diesel, bio-oils and biodiesel by [Sequera,et.al., 2007.] have found
that by tailoring the atomization process, the emissions of CO and NOx can be decreased by a
factor of 2 to 3.
1.4. Objectives
The present research focuses on detailed spray characteristics and combustion
performance of bio-oils. At first, the fuel properties are measured to study the physical properties
of biodiesel fuels, soybean oil and diesel fuels. Then the combustion performance is studied
using viscous bio-oils followed with detailed spray characteristics at different fuel inlet
temperature in the cold flow. Later part of the research involves the study of flame spray
characteristics of these viscous bio-oils and their effect on combustion emissions. Such
measurements in preheated viscous bio-oils are non-existent at this time. Experiments are
conducted to measure spray characteristics as well as provide an insight into CO and NOx
emissions using a commercial airblast atomizer.
8
1.5. Overview
1. In Chapter 2 experiments are reported to study characteristics of alternative fuels for gas
turbine combustion. The fuels used in this study are biodiesels produced from vegetable
(soybean) oil and animal fat (chicken), and blends of diesel-vegetable oil. Measurements are
obtained to characterize the physical and chemical properties of fuel by Gas
Chromatography and Mass Spectroscopy (GC-MS) analysis, Thermogravimetric Analysis
(TGA) analysis, density, kinematic viscosity, surface tension and water content. Spray is
characterized by spray angle measurements and drop size predictions of SMD from an air
blast atomizer. Measurements are used to compare the combustion performance of bio-oils
to conventional diesel fuel emissions. This part of the study provided motivation to
demonstrate the combustion performance and spray characteristics of viscous bio-oils by
improving the fluid mechanics of combustion process as compared to chemistry of the fuel.
2. Chapter 3 provides the emissions measurements for viscous biofuels combustion to study
the effect of fuel inlet temperature on emissions. A method to preheat the vegetable oil prior
to injection in the combustor is developed and implemented. Visual flame images and
emissions of CO and NOx showing the effect of fuel inlet temperature are presented.
3. Chapter 4 examines the effect of preheated soybean oil using an airblast atomizer. The
non-reacting spray is created at ambient conditions of temperature and pressure. Both
qualitative and quantitative measurements of the spray are presented. Laser sheet
visualization system is used to present the qualitative spray images and a Phase Doppler
Particle Analyzer (PDPA) system is used to present the detailed quantitative droplet size
9
data in the spray. Such measurements of the spray using bio-oils are nonexistent at this
time.
4. Chapter 5 presents the study of flame spray characteristics of soybean oil using an
airblast atomizer. The measurements are obtained to present the (i) effect of swirl on spray,
(ii) effect of flame on spay, (iii) effect of enclosure on spray and emissions of CO and NOx.
The results of this study is expected to provide some insights into understanding the effect
of droplet diameter and droplet diameter distribution in a cold spray to demonstrate the
effect of co flow swirling air. For the flame spray the results of spray characteristics
obtained show the effect on pollutant emissions since they are affected by the fuel spray and
atomization characteristics and the fuel properties and hence improve the combustion
performance. The measurements are taken at various locations in the spray (near and far
field of the spray) axially and radially.
5. Finally, in chapter 6, important conclusions from this study are presented and
recommendations for future studies are provided.
10
(a)
Air Fuel
(b)
Orifice Disc
Distributor
Nozzle Body
Screw Pin
O-ring Seal
Figure 1.1 (a) Air blast Injector Concept (b) Injector Details
11
Spray
d
Atomizing
air
H
d
Fuel
Figure 1.2 Working Principle of the Flow Blurring Injector
12
-4 -2 0 2 4 -4 -2 0 2 4 -4 -2 0 2 4 -4 -2 0 2 4
10%
15%
20%
25%
45
40
35
30
25
20
15
10
5
0
-4 -2 0 2 4 -4 -2 0 2 4 -4 -2 0 2 4 -4 -2 0 2 4
10%
15%
20%
25%
Figure 1.3. Effect of Atomizing Airflow Rate on Visual Images of Kerosene Flames: FB Injector
(Top), AB Injector (Bottom).
13
200
0
5
10
15
20
(a)
180
10%
15%
20%
15%
20%
160
CO (ppm)
25
30
35
40
45
AA-FB
AA -FB
AA - FB
AA - AB
AA -AB
50
200
180
160
140
140
120
120
100
100
80
80
60
60
40
40
20
20
0
5
10
15
20
25
30
35
40
45
50
40
45
50
200
Axial Distance (cm)
200
0
5
10
15
20
25
30
35
(b)
175
175
NOx (ppm)
150
150
10% AA - FB
15% AA - FB
20% AA - FB
15%AA - AB
20% AA - AB
125
100
125
100
75
75
50
50
25
25
0
5
10
15
20
25
30
35
40
45
50
Axial Distance (cm)
Figure 1.4. Axial Profiles of Emissions in Diesel Flames; (a) CO Concentration, (b) NOx
Concentration. [Open symbols represent FB Injector, Closed Symbols represent AB Injector].
14
-4
200
180
-3
-2
-1
0
1
(a)
10%
15%
20%
15%
20%
160
140
CO (ppm)
2
3
4
AA- FB
AA- FB
AA- FB
AA- AB
AA- AB
200
180
160
140
120
120
100
100
80
80
60
60
40
40
20
20
0
-4
-3
-2
-4
200
-3
-2
-1
0
1
Radial Distance (cm)
-1
0
1
0
2
3
4
2
3
4
175
150
175
150
(b)
125
NOx (ppm)
200
10%
15%
20%
15%
20%
100
125
AA- FB
AA- FB
AA- FB
AA- AB
AA- AB
100
75
75
50
50
25
25
0
-4
-3
-2
-1
0
1
2
3
4
0
Radial Distance (cm)
Figure 1.5. Radial Profiles of Emissions in Diesel Flames; (a) CO Concentration, (b) NOx Concentration.
[Open symbols represent FB Injector, Closed Symbols represent AB Injector].
15
104
0
5
10
15
20
25
30
35
40
10%
15%
20%
10%
15%
20%
(a)
AA- FB
AA- FB
AA- FB
AA- AB
AA- AB
AA- AB
50 4
10
103
CO (ppm)
103
45
10
2
10
2
10
1
10
50
1
0
5
10
15
20
25
30
35
40
45
40
45
Axial Distance (cm)
200
0
5
15
20
25
30
35
(b)
180
NOx (ppm)
10
50
200
180
160
160
140
140
120
120
10%
15%
20%
10%
15%
20%
100
80
60
AA- FB
AA- FB
AA- FB
AA- AB
AA- AB
AA- AB
100
80
60
40
40
20
20
0
0
5
10
15
20
25
30
35
40
45
0
50
Axial Distacne (cm)
Figure 1.6. Axial Profiles of Emissions in Kerosene Flames; (a) CO Concentration, (b) NOx
Concentration. [Open symbols represent FB Injector, Closed Symbols represent AB Injector].
16
-4
104
-3
-2
-1
10
10
CO (ppm)
3
1
10%
15%
20%
10%
15%
20%
(a)
10
0
2
3
4
AA - FB
AA - FB
AA- FB
AA- AB
AA- AB
AA- AB
104
10
3
2
10
2
1
10
1
-4
-3
-2
-1
0
1
2
3
4
2
3
4
Radial Distance (cm)
-4
200
-3
-2
-1
0
1
200
(b)
175
175
150
150
10%
15%
20%
10%
15%
20%
NOx (ppm)
125
100
AA- FB
AA- FB
AA- FB
AA- AB
AA- AB
AA- AB
125
100
75
75
50
50
25
25
0
-4
-3
-2
-1
0
1
2
3
4
0
Radial Distance (cm)
Figure 1.7. Radial Profiles of Emissions in Diesel Flames; (a) CO Concentration, (b) NOx
Concentration. [Open symbols represent FB Injector, Closed Symbols represent AB Injector].
17
CHAPTER 2
PROPERTIES OF BIODIESEL AND DIESEL-VEGETABLE OIL BLENDS
Background
In recent years, two important factors have influenced the world’s energy market. First,
the increased demand for petroleum fuels has steadily raised their price to impose economic
burden on importing nations. Second, the concerns for global warming have increased scrutiny
on emissions of greenhouse gases such as carbon dioxide (CO2) formed by the combustion of
fossil fuels. The depletion of conventional fossil fuels is also a cause for concern. The above
factors have prompted research on alternative energy sources. Biofuels derived from agricultural
products (i.e., biomass) offer the potential for a sustainable energy future. Unlike nonrenewable
fossil fuels, the CO2 emitted from the combustion of biofuels is recycled during cultivation of the
biomass. Biofuels can increase energy security since the biomass feedstock is domestic, diffuse,
and hence, secure. However, for commercial viability, the land usage and fossil energy input for
biofuel production must be reduced through process integration/optimization and/or specialty
energy crops such as microalgae (Chisti, 2007).
Ethanol and biodiesel are the two commonly available liquid biofuels in the US. These
fuels are currently mixed with gasoline or diesel fuel in small amounts (up to 10% for ethanol
and up to 20% for biodiesel) to reduce the demand for fossil fuels for vehicular transportation.
Accordingly, several studies have been performed to evaluate performance of these fuels for
automotive internal combustion engines (Agrawal, 2007; Monyem, et.al, 2001 and Rakopoulos,
2006).
18
Although market demand currently favors the use of biodiesel mainly for vehicles, the liquid
biofuels could also serve as an alternative fuel for the power generation industry, for example, by
replacing the fuel in land based diesel or gas turbine engines.
Biodiesel is obtained from oils such as vegetable oil (VO) or animal fats through a
process known as trans-esterification. In this process, the vegetable oil or animal fat
(triglyceride) reacts with an alcohol in the presence of a catalyst such as sodium or potassium
hydroxide to produce biodiesel and glycerol (Demirbas, 2005). The most common form of
biodiesel in the United States is made from soybean oil (vegetable oil) and methanol and it is
known as soy methyl ester (SME) or methyl soyate. The trans-esterification process improves the
physical properties of the original VO, and, in particular, decreases the viscosity to improve fuel
atomization in the combustion system. This process however incurs a cost penalty and hence, the
price of biodiesel exceeds that of the VO use d to produce it. Furthermore, glycerol, co-produced
with biodiesel, is often a waste product since it is yet to find a commercial market. Thus, the
direct use of VO in a gas turbine to generate power can offer economic benefits, especially if the
problem of fuel atomization associated with high viscosity can be alleviated by blending VO
with diesel or biodiesel fuels.
The continuous flow operation of gas turbines offers greater flexibility, unlike the
unsteady processes of internal combustion engines imposing constraints on fuel chemistry in
terms of octane and/or cetane numbers. Field tests in recent years have demonstrated the
technical feasibility of gas turbine engine operation on fuels such as bio-oil, ethanol, and
biodiesel (Lupandin, et. al, 2005, Chiang, et. al, 2007, Moliere, et. al, 2007, and Bolszo, et.al,
2007). However, these tests have produced limited data on emissions of nitric oxides (NOx) and
carbon monoxide (CO). Detailed NOx and CO emissions measurements in an atmospheric
19
pressure burner simulating typical features of a gas turbine combustor were reported by Sequera
et al. (2007) for flames of diesel, biodiesel, emulsified biooil, and diesel-biodiesel fuel blends.
(Sequera et al. 2007) found that the NOx and CO emissions are determined mainly by the fuel
atomization process. Although fuel properties are important, the atomization process itself could
be tailored to minimize the emissions for a given fuel.
The objective of this part of the study is to document physical properties of biodiesel and
diesel-VO blends. Biodiesels produced from vegetable oil (soybean) and animal fat (chicken)
are considered in this study. In case of VO, the fuel must be heated to reduce its kinematic
viscosity so that the pressure drop in the fuel supply line is reasonable, atomization occurs with
fine droplets, and combustion emissions are acceptable. To circumvent the need for heating the
VO, and thus, to avoid hardware changes to the fuel supply system, diesel-VO blends are
investigated. The emissions measurements to demonstrate the combustion performance of
different fuels are reported by Panchasara, et.al, (2009).
Table 2.1 lists the fuels used for the experiments. The physical and chemical properties of
fuel are characterized by gas chromatography and mass spectrometry (GC-MS),
thermogravimetric (TGA) analysis, and kinematic viscosity, surface tension, density, and water
content measurements. Spray characteristics are determined by spray angle measurements and
predictions of the Sauter mean diameter from an air blast atomizer. The following sections report
the fuel property measurements and the spray characteristics followed by the conclusions.
20
Table 2.1.
List of Fuels used in the Study
Fuel
Diesel
Description
Commercial grade # 2
Biodiesel # 1 (BD-1)
Methyl soyate made from soybean oil
Biodiesel # 2 (BD-2)
Methyl soyate made from chicken fat
Vegetable Oil (VO)
100% refined soybean oil
90-10 Blend
90% Diesel – 10% VO
80-20 Blend
80% Diesel – 20% VO
70-30 Blend
70% Diesel – 30% VO
2.2. Fuel property measurements
It is important to characterize chemical and physical properties of various fuels because
they were obtained from different sources. First, the fuel composition was determined using the
GC-MS technique. The GC-MS analysis of biodiesel fuels (BD-1 and BD-2) was used to
determine the fuel’s low heating value (LHV) and an equivalent chemical formula. The TGA
analysis was performed to characterize the fuel volatility. Fuel kinematic viscosity and surface
tension measurements were measured to determine an acceptable operational range and to
compute the Sauter mean diameter.
21
Table 2.2.
Results of GC-MS Analysis
Fatty Acid
Formula
Methyl Esters
%Mass
%Mass
Molecular Gas Heat of
LHV
BD-1
BD-2
Weight
kJ/kg
Formation
KJ/kmol
(Turns,S.R.)
Lenoleic Methyl
C19H34O2
50.1
20.3
294.3
-513,000
37,634
C19H36O2
35.8
45.0
296.3
-643,250
37,756
C17H34O2
10.3
19.3
270.2
-727,500
37,284
C19H38O2
3.4
8.6
298.3
-787,400
38,314
C15H30O2
0.0
0.2
242.2
-684,300
36,526
C17H32O2
0.0
6.6
268.2
-587,000
37,184
ester
Oleic Methyl
Ester
Palmitic Methyl
Ester
Steartic Methyl
Ester
Tetradecanoic
Methyl Ester
9-Hexadecanoic
Methyl Ester
22
2.2.1. GC-MS Analysis
The GC-MS analysis technique combines gas chromatography and mass spectrometry to
determine % mass (or % volume) of species present in a chemical (by gas chromatography)
together with the elemental composition of each species (by mass spectrometry). The analysis
was carried out at UA’s GC-MS Facility using HP 6890 GC connected with a Micro mass
AutoSpec-UltimaTM NT Mass spectrometer. GC was performed with column DB-1, inlet
temperature of 225 °C; split ratio of 30, oven temperature of 100°C, heating rate of 10 °C/min
and hold time of 25 min. The GC column did not produce any peaks for the VO sample because
of the low column temperature. Table 2.2 presents the analysis results for the two biodiesel
fuels. In BD-1, linoleic methyl ester is the major component, while that in BD-2 is oleic methyl
ester.
The last two esters, i.e., tetradecanoic methyl ester and 9-hexadecenoic methyl ester were found
in small amounts in BD-2, but they were not present in BD-1. Table 2.2 also lists the enthalpy of
formation of each component taken from (http://webbook.nist.gov/ chemistry).This information
was used to compute fuel component’s low heating value (LHV) assuming complete
combustion. Evidently, the LHV of different biodiesel components is within 2% of each other.
Based on these results, the equivalent chemical formulas for BD-1 and BD-2 are, C18.71H34.71O2
and C18.47H35.10O2, respectively.
2.2.2. Thermogravimetric Analysis (TGA)
Thermogravimetric analysis is an experimental approach to determine a material’s
thermal stability and/or its fraction of volatile components by monitoring the weight change as a
specimen is heated. A volatile sample loses mass with an increase in temperature and after a
certain temperature it loses maximum of its mass and what remains is the residue mass. The
23
volatility has a significant effect on fuel vaporization which in turn affects fuel-air mixing and
combustion processes. For example, poor vaporization can results in fuel droplets burning in the
diffusion mode. The resulting high flame temperatures are known to produce high concentrations
of NOx, CO, and soot emissions (Turns, S.R., 2000). The fuel samples (8 to 17 mg) were
characterized using TA 2950 analyzer from TA instruments. The maximum temperature was set
at 600°C, and testing was conducted from the ambient temperature to 600°C with a heating rate
of 5°C/min, in an atmosphere of air supplied at the rate of 90 cm3/min. Before starting the
experiment the sample loading pan was flame cleaned and tarred to zero. The microbalance was
calibrated with standard weights varying from 1 mg to 100 mg. Figure 2.1 shows the TGA
profiles for diesel, BD-1, BD-2, and VO. Results show that diesel is the most volatile fuel among
all the fuels listed in Fig. 2.1. Diesel evaporated rapidly when heated to 80°C and it completely
vaporized at approximately 180°C.
Both biodiesel fuels show similar volatility characteristics; negligible mass loss when
heated below 100°C, rapid mass loss (or vaporization) at temperatures above 150°C, and nearly
complete vaporization approximately at 225°C. The vegetable oil is the least volatile fuel,
showing negligible mass loss at 225°C. In this case, much of the mass loss occurs at
temperatures between 250°C and 500°C. Clearly, VO will require significantly higher thermal
feedback from the flame to the fuel droplets to fully pre-vaporize the fuel, and thus, to create a
homogenous fuel-air mixture prior to combustion. Results in Fig. 2.1 explain the choice of
diesel-VO fuel blends; blending a high volatility fuel (diesel) with a low volatility fuel (VO)
would improve fuel vaporization, which can result in lower combustion emissions. Figure 2.1
includes TGA data for BD-2 from (Cayli, 2007). Results show an excellent agreement between
present results and data reported in the literature. Minor differences are attributed to the
24
presence of an oxidizing environment in this study (as opposed to an inert gas), which tends to
increase the oxidation of the fuel, especially at higher temperatures.
2.2.3. Density, Viscosity, and Surface Tension
Table 2.3 lists the measured physical properties of various fuels together with the relevant
measurement uncertainties. The density measurements were performed gravimetrically at 25°C
using volumetric glassware. Three test runs were taken to ensure the repeatability and accuracy
of the data. Diesel is a lighter fuel compared to both biodiesels, while the density of the VO is
the highest. Density measurements are important to determine the volumetric flow rate of the
fuel to obtain a specified heat release rate in the combustor. Table 2.3 also lists the LHV of fuels
on a volumetric basis. The LHV of the two biodiesel fuels was determined from component
analysis presented in Table 2.2, while data from existing literature were used for the remaining
fuels. For a given heat release rate, biodiesel fuels will require nearly 13% higher volume flow
rate compared to the diesel fuel.
Kinematic viscosity is an important physical property affecting pressure drop in the fuel
line as well as the fuel atomization in the combustor. High fuel viscosity can result in excessive
pressure drop and produce spray with large droplets, which deteriorates the combustion
performance. The kinematic viscosity was measured by Cannon-Fenske Type 200 viscometer
(Cannon Instrument Co., State College, PA, USA) with viscosity range of 20-100 mm2/sec
(centi-Stokes or cSt). The viscometer was immersed in a water bath connected to a temperature
controlled heat exchanger. The temperature was controlled using a Fisher Scientific Isotemp
1016S recirculating water bath (Pittsburgh, PA, USA). The kinematic viscosity was measured
for temperature range from 20°C to 70°C. Table 2.3 lists the viscosity values at 25°C, while the
detailed profiles are shown in Figs 2.2 and 2.3.
25
Table 2.3 and Figure 2.2 show that the kinematic viscosity of biodiesels is 45-55% higher
compared to that of diesel. The kinematic viscosity of both biodiesels is nearly the same over the
entire temperature range. The higher fuel viscosity will require higher pumping power for
biodiesel compared to diesel to overcome the pressure drop in the fuel supply line. Furthermore,
the high fuel viscosity of biodiesel can also negatively impact fuel atomization, an effect that will
be quantified later on by droplet size predictions.
Table 2.3.
Physical properties of Diesel, Biodiesel # 1, Biodiesel # 2 and Vegetable Oil
Property
Mol. Weight
kg/kmol
Density at 25 0C,
kg/m3
Viscosity at 25 0C,
mm2/s
Surface Tension at
25 0C mN/m
LHV, kJ/kg
3
LHV, MJ/m
Diesel
Fuel
BD-1
BD-2
VO
142.2
291.45
289.06
-
834.0 ± 8.3
880.0 ± 8.8
868.0 ± 8.6
3.88 ± 0.016
5.61 ± 0.016 6.14 ± 0.016
28.2 ±0.6
31.1 ±0.6
38002.3
44601.7
37,198 (Turns,
S.R.,2000)
33,442
925.0 ±
9.2
53.74 ±
0.22
30.7 ±0.6
30.1 ±0.6
37659.7
37000
(Rakopoul
os,2006)
32,689
34,225
(Rakopoul
os,2006)
Table 2.3 and Figure 2.3 show that the kinematic viscosity of vegetable oil is higher than
that of diesel by more than an order of magnitude. Fuels with such high viscosity do not produce
acceptable atomization using existing fuel injectors, which was also verified experimentally.
Hence, blends of VO with diesel were prepared to reduce the fuel viscosity. Figure 2.3 shows the
26
measured and computed kinematic viscosity of the 70-30 diesel-VO blend (by volume). The
viscosity of the blend was calculated from Eq. 1 for a mixture (Ejim, C.E, 2007):
n
lnν = ∑ xi lnν i
Eq. 1
i =1
Where, xi is the volume fraction of the species i, and νi is the kinematic viscosity of the
species i. The excellent agreement between measured and computed values for the 70-30 dieselVO blend in Figure 2.3 verifies application of Eq. 1. Thus, Eq. 1 was used to predict the
kinematic viscosity of 80-20 and 90-10 diesel-VO blends as shown in Figure 2.3. Results show
that the diesel viscosity at 25 0C matches the viscosity of 90-10 blend at 40°C, 80-20 blend at
50°C and 70-30 blend at 60°C. At 25°C, the biodiesel viscosity is nearly the same as that of the
80-20 diesel-VO blend. Further, vegetable oil will require temperature above 80°C to replicate
the room temperature viscosity of diesel. For a given temperature, the viscosity of 70-30 blend is
more than twice of that of the diesel. Hence, this blend was considered as the upper operating
limit for VO content in the diesel fuel. The blends remained miscible for all volume ratios.
Accordingly, the present combustor system was operated with a variety of fuels in the viscosity
range of 3.5 mm2/s – 10.0 mm2/s at room temperature.
A “Kruss” K-12 model Processor Tensiometer with a Platinum ring was used to measure
the surface tension of the fuel samples at 25°C. Three test runs were taken for each fuel to ensure
the repeatability of the results. The surface tension was also calculated theoretically from the
surface tension data of individual methyl esters listed in Table-2.3. The surface tension (σm) for
the fuel blends was evaluated from Eq. 2 (Ejim, et.al, 2007)
⎡n
⎤
σ m = ⎢∑ yi (σ i )1 / 4 ⎥
⎣ i =1
⎦
27
4
Eq. 2
Here, yi is the mass fraction of species i, and σi is the surface tension of species i. For biodiesel
fuels, measured values of surface tension were within 0.5% of those computed using Eq. 2 and
published surface tension data for esters listed in Table 2.3 (Ejim, et.al, 2007).
2.2.4. Water Content
The water content of the fuels was measured to observe the percent volume of water
present in the fuels. The water content of each fuel (as received) was determined experimentally
using a volumetric Aquastar Karl Fischer titrator (EM Science, Gibbstown, NJ, USA) with
Composite 5 solution as the titrant and anhydrous methanol as the solvent. Each sample was at
least 0.1 g and triplicate measurements were performed on each sample at 25°C. The equilibrium
water content for each sample was determined by first contacting equal volumes of fuel and
deionized water for 24 hours followed by titration of the fuel sample using the above method.
Deionized water was obtained from a commercial deionizer (Culligan, Northbrook, IL, USA).
Table 2.4 summarizes results from the water content analysis.
Table 2.4.
Water Content in the Fuel
Fuel
% Mass of Water in
% Mass of Water in
Fuel
Equilibrated Fuel
Diesel
0.06
0.06
BD-1
0.11
0.17
BD-2
0.09
0.16
VO
0.05
0.20
28
2.3. Spray characteristics
2.3.1. Spray Angle Measurements
The spray angle measurements were done on the experimental set-up shown in figure 2.4.
The test set-up consists of combustor assembly and injector assembly. The details of the
combustor assembly will be explained in Chapter 3. Primarily the combustor assembly consists
of a plenum, mixing chamber and a swirler. The injector system runs through the plenum and the
mixing chamber. An O-ring within a sleeve is located at the bottom of the plenum to prevent air
leakage. A commercial air-blast atomizer (Delavan Siphon type SNA nozzle) with its details
shown schematically and photographically in Fig. 2.4 was used for the experiments. This
commercial version creates a swirling flow of atomizing air to breakdown the fuel jet as the
fluids exit the orifice plate. The liquid fuel was supplied by a peristaltic pump with reported
calibration error of ±0.25% of the flow rate reading ranging from 12 ml/min to 130 ml/min in
steps of 2 ml/min. Viton tubes were used to prevent degradation of the fuel lines. A 25µm filter
was placed in the fuel supply line to prevent dirt and foreign particles from entering into the
injector. The flow rate of injection air was varied for ALR of 2.52 and 4.35. The fuel flow rate
was kept constant at 12 mlpm. The spray images were taken for open cold spray conditions as
shown in figure 2.5 for BD-1, BD-2 and VO. Images were taken with a regular digital camera for
different ALR’s. The spray cone angle was calculated manually based on the slope of best curve
fit line on the edge of the spray from the pixel coordinate information obtained using image
editing software as shown in figure 2.6. The Figure 2.5 shows photographs of BD-1 spray for
different atomizing airflow rates. A photograph of the VO spray is also shown for comparison.
Visually, biodiesel produced a fine spray, while the VO spray contained large droplets around
the outer edge. These large droplets give the appearance of a larger cone angle with VO, a
29
problem that was addressed during data analysis using intensity profile to compute the cone
angle. Figure 2.6 shows similar spray cone angle for diesel and biodiesel fuels. The cone angle
for VO is smaller by about 5 degrees. For all cases, the cone angle increased with increasing
atomizing air flow rate, signifying increased penetration of fuel into the combustor.
2.3.2. Droplet Size Characteristics
Detailed characterization of the fuel spray requires measurements of drop size distributions,
for example, using phase Doppler particle analyzer and/or laser diffraction techniques. As the
first step, we performed droplet size calculations using the correlation of Rizk and Lefebvre,
(1984), for the Sauter Mean Diameter (SMD) from an air-blast injector:
0 .4
SMD
do
⎛ σ ⎞ ⎛
⎟ ⎜1 +
= 0.48⎜
⎜ ρ U 2d ⎟ ⎝
⎝ A R o⎠
⎞
⎟
ALR ⎠
1
0 .4
2
0 .5
⎛ μL ⎞ ⎛
⎟ 1+
+ 0.15⎜
⎜ σρ L d o ⎟ ⎜⎝
⎝
⎠
⎞
⎟
ALR ⎠
1
Eq. 3
with,
SMD: Sauter Mean diameter, μm
do :
Liquid discharge orifice diameter, m
σ:
Surface tension, N/m
μL:
Liquid fuel viscosity, m2/sec
ρA :
Density of air, kg/m3
ρL :
Liquid fuel density, kg/m3
UR :
Relative velocity of co flowing
ALR : Air to liquid mass flow ratio
Figure 2.7(a) presents the Sauter mean diameter versus the air to liquid mass flow ratio
(ALR) for diesel and biodiesel. Note that 15% AA pertains to ALR of 2.65 and 2.52,
30
respectively, for diesel and biodiesel fuels. The corresponding values for 25% AA are 4.35 and
4.12. Figure 2.7(a) shows that increasing the atomizing airflow rate (for a fixed fuel flow rate)
improves atomization by decreasing the SMD of the droplets produced. For a given ALR, the
biodiesel droplets are larger than diesel droplets by about 2 μm. The SMD of the two biodiesel
fuels is nearly the same for all ALR values. Figure 2.7(b) presents the SMD for diesel, VO, and
diesel-VO blends. For VO, the 15% AA and 25% AA correspond to ALR of 2.40 and 3.92,
respectively. For the entire range of ALR, the SMD of VO is two to three times higher than that
of the diesel. However, the SMD of diesel-VO blends is only moderately higher than that of the
diesel. Interestingly, the 70–30 diesel-VO blend produces nearly the same droplet diameter at
different ALR values as the two biodiesels (see Fig. 2.7(a)). The results in Fig. 2.7 have provided
a quantitative assessment of atomization behavior of different fuels, but an accurate knowledge
of the droplet size requires direct measurements in the spray as presented in later chapters.
2.4. Conclusions
Fuel properties and spray characteristics were measured for diesel, two types of biodiesel,
and diesel-VO blends. The composition of biodiesel fuels was determined using the GC-MS
technique, which also allowed us to compute the low heating value of the fuel. The main
constituent of VO biodiesel was linoleic methyl ester while that for animal fat biodiesel was
oleic methyl ester. In spite of the compositional differences, the volatility, kinematic viscosity,
and surface tension properties of the two biodiesel fuels were similar. The kinematic viscosity of
biodiesel fuels was about 50% higher than that of diesel. The kinematic viscosity of VO was
more than 10 times that of diesel, and hence, VO would require blending with a low viscosity
fuel. Diesel-VO blends with up to 30% VO (by volume) were used in this study to achieve an
acceptable range of fuel viscosity; about twice that of the diesel and equal to that of biodiesel.
31
The TGA analysis indicated diesel to be the most volatile, while VO was the least volatile fuel.
The Sauter mean diameter in biodiesel spray was estimated to be only slightly larger than that in
diesel spray. The estimated SMD of biodiesel and 70-30 diesel-VO blends was nearly the same.
32
100
200
120
% Mass
100
300
400
500
600
Diesel
BD-1
120
BD-2
Oleic Methyl Ester[Cayli,
[Ref. 13]
VO
100
80
80
60
60
40
40
20
20
0
100
200
300
400
500
Temperature (oC)
Figure 2.1. Results of Thermo gravimetric Analysis
33
0
600
Kinematic Viscosity (mm2 /s )
10
20
40
60
80
10
Diesel
BD-1
BD-2
8
8
6
6
4
4
2
2
0
20
40
60
Temperature (oC)
Figure 2.2. Kinematic Viscosity of Diesel and Biodiesel
Fuels
34
0
80
0
20
40
∗
2
∗
∗
80
Diesel
VO
70-30 blend ( Exp)
70-30 blend (Eq.1)
80-20 blend
90-10 blend
∗
102
Kinematic Viscosity, (mm /s)
60
∗
∗
102
∗
10
1
10
1
10
0
10
80
0
0
20
40
o
60
Temperature C
Figure 2.3. Kinematic Viscosity of Diesel –VO blends
35
Figure 2.4 Air Blast Atomizer Details
36
BD-1
ALR 2.52
BD-1
ALR 2.52
VO
ALR 4.35
Figure 2.5.Spray Visualization Photographs
37
Spray Angle (Degrees)
10
60
15
20
25
30
60
55
55
50
50
45
45
Diesel
BD-1
BD-2
VO
40
35
30
10
40
35
15
20
25
% AA
Figure 2.6 Spray Cone Angle vs Atomizing Airflow Rate
38
30
30
40
2
3
4
Diesel
BD-1
BD-2
35
Sauter Mean Diameter (μm)
5
40
35
30
30
25
25
20
20
15
15
10
2
3
4
5
Air to Liquid Mass Ratio (ALR)
Figure 2.7. SMD versus ALR for Diesel, BD-1 and BD-2.
39
10
Sauter Mean Diameter (μm)
80
2
3
4
5
Diesel
VO
70-30 Diesel-VO
80-20 Diesel-VO
90-10 Diesel-VO
60
80
60
40
40
20
20
2
3
4
5
Air to Liquid Mass Ratio (ALR)
Figure 2.8. SMD versus ALR for Diesel, VO, and Diesel-VO Blends
40
CHAPTER 3
EFFECT OF FUEL PREHEATING ON EMISSIONS FROM COMBUSTION OF VISCOUS
BIOFUELS
Background
Growing concern over effects of greenhouse gases on the environment has revived
interest in alternate fuels to replace the currently prevalent fossil fuels. Vegetable oils (VOs) and
their derivatives are potential biofuel alternative that have received significant attention in recent
years. The most common of all the vegetable oils in the U.S. is the soybean oil. Results from
Chapter 2 have shown that it is difficult to atomize VO because longer hydrocarbon chain
lengths with trace amounts of triglycerides results in high kinematic viscosity and low volatility
of VO compared to diesel fuel. At room temperature, the VO kinematic viscosity is about 20
times higher than that of diesel fuel [Chapter 2]. Poor volatility of VO also makes it difficult for
the fuel droplets to fully pre-vaporize before combustion.
One of the methods to improve the physical properties and fuel atomization of VO for
efficient combustion is the process of trans-esterification, which produces soy methyl ester
(SME) or biodiesel. This chemical process however incurs a cost penalty, and hence, the price
of biodiesel is higher than of VO. Furthermore, glycerol, a byproduct of the process remains to
be a waste product. The second method is to preheat the VO to reduce its kinematic viscosity,
and thus, make it possible to atomize the fuel. Much of the previous research with heated VO
has been conducted in diesel engines [Simmons, 2008; and Bari, 2007]. Heated VO could also be
used in stationary systems such as power generating gas turbines and furnaces.
41
Thus, this research seeks to investigate how heating the VO affects combustion emissions. A
commercially available air blast (AB) atomizer is used to atomize the fuel (soybean oil) for
experiments conducted in a swirl-stabilized burner operated at atmospheric pressure. Details of
the experiment set-up and some preliminary results of visual flame images are presented in the
following sections.
3.2. Experimental Set-Up
The test apparatus is shown schematically in Fig. 3.1. The test set-up consists of
combustor assembly and injector assembly. The injector assembly and the details of the injector
are explained in Chapter 2 and figure 2.4. The modification here is with the addition of fuel
heater to preheat the liquid fuel before atomization and modification of the fuel flow path to
overcome the heat losses. The combustion airflow path includes a plenum filled with marbles to
breakdown the large vortical structures, a swirler in the mixing section, and another swirler at the
combustor inlet (see Fig. 3.2) to enhance fuel-air mixing and improve flame stability. The bulk
inlet velocity of the combustion air is 2.7 m/s, which results in a Reynolds number varying from
3150 based on the equivalent diameter of the injector. An air-cooled quartz glass column of
inside diameter 8.0 cm and length 46 cm is used to enclose the flame.
Air from a compressor passed through a pressure regulator, a dehumidifier, and water
traps to remove any moisture present. Then, the air was split into combustion air supply and
atomizing air supply lines. The air flow rates were measured by the laminar flow elements
(LFE). The flow rates measured by the LFE’s were corrected for temperature and pressure as
specified by the manufacturer. The liquid fuel supplied by a peristaltic pump passes through a
pulse dampener, a fuel filter and an electric fuel heater (Infinity Fluids Corporations, CRES-ILB12/24 inline water/liquid heater). The heater uses a Proportional Derivative Controller (PID)
42
control unit to control the fuel temperature within the accuracy of ± 0.5 oC, measured at the
heater outlet by a K-type thermocouple.
Heated fuel enters fuel injector through 5 cm long tube of 5 mm ID to minimize the heat
loss. The fuel temperature at the injector inlet was measured by a K- Type thermocouple. The
experiment was started by supplying gaseous methane and then, igniting the methane-air reactant
mixture in the combustor. Next, the liquid fuel flow rate was gradually increased to attain the
desired value, while the methane flow rate was slowly decreased to zero. In this study, the
volume flow rate of total air (combustion + atomizing) was constant at 150 standard lpm.
Experiments were conducted for fixed volume flow rate of fuel at 12 mlpm which is measured
before preheating of the fuel. The atomizing airflow rate was varied to obtain ALR of 2.4, 3.5
and 4.0. The equivalence ratio during the experiments was kept constant at Φ = 0.78 for the fixed
volume flow rate of fuel and air.
The time required for the fuel to flow through the system and reach the atomizer was
about 30 minutes. Another 0-40 minutes were required before the fuel temperature reached the
set value. The product gas was sampled continuously by a quartz probe (OD = 7.0 mm) attached
to a three-way manual traversing system. The upstream tip of the probe was tapered to 1 mm ID
to quench reactions inside the probe. The probe was traversed in the axial direction at the center
of the combustor and in the radial direction at the combustor exit plane. The gas sample passed
through an ice bath and water traps to remove moisture upstream of the gas analyzers. The dry
sample was sent through the electrochemical analyzers to measure the concentrations of CO and
NOx in ppm. The analyzer also measured oxygen and CO2 concentrations, which were used to
cross-check the equivalence ratio computed from the measured fuel and air flow rates. The
uncorrected emissions data on dry basis are reported with uncertainty of +/-2 ppm.
43
Combustion performance is characterized by measuring CO and NOx emission profiles
within the combustor. Experiments are conducted using an AB atomizer for a fixed fuel flow rate
and the total air flow rate which will be discussed in the experimental set-up section. The total air
flow rate is split between the atomizing air and primary combustion air while air to fuel mass
ratio is varied through the injector to document the fluid dynamics effects on emissions. Results
and discussion section include the visual flame images for different fuel inlet temperatures as
increased on heater in the steps of 30°C at different atomizing air ratios, and axial and radial
emissions profiles.
3.3. Results and Discussion
Table 3.1 lists the key physical properties of VO taken from chapter 2. Figure 3.2 shows
the kinematic viscosity of VO, which is 20 times higher than that of diesel at room temperature.
Figure 3.3 shows that VO should be heated to about 80°C to replicate the room temperature
kinematic viscosity of diesel. In this study, the fuel temperature at the heater outlet (Th) was
varied from 70 to 160°C in steps of 30°C. It resulted in fuel temperature (Tf) at the injector inlet
from 57°C to 122°C, depending upon the ALR. Figure 3.4 illustrates this relationship between
the heater exit and injector inlet temperatures. For a fixed heater exit temperature, the fuel inlet
temperature increased by about 0.5 to 1.5oC as the ALR increased from 2.4 to 4.0. Fuel inlet
temperature for ALR = 3.5 is chosen to explain the results.
Table 3.1.
Physical properties of VO at 25°C
Kinematic Viscosity
Surface Tension
LHV
kg/m
2
mm /sec
mN/m
MJ/m3
925.0 ± 9.2
53.74 ± 0.22
30.1 ±0.6
34,225
Density
3
44
3.3.1. Visual Flame Images
Figures 3.5 and 3.6 show the photographic flame images at ALR of 2.4 and 4.0,
respectively, taken by a digital camera to obtain a qualitative understanding of the flame
characteristics. For ALR of 2.4, Figure 3.5 depicts a short blue premixed region near the injector
exit followed by a large yellow-orange diffusion flame zone. As the fuel inlet temperature
increases, the flame appears to be shorter and the blue premixed zone extends farther
downstream. At this low ALR the fuel droplets burn in diffusion mode attributing to the distinct
yellow flame since they don’t fully vaporize and premix with air before entering the reaction
zone. Further, poor atomization at low ALR produces large droplets that tended to accumulate
on the combustor wall and gradually formed a ring around the interior of the glass enclosure as
seen in the photograph. The ring can still be observed at elevated fuel temperatures since the
droplet accumulation started during the initial start-up when the fuel did not yet reach the set
temperature. Accumulated droplets subsequently pyrolized to form a distinct black ring around
the combustor wall. At a higher ALR of 4.0, as shown in Fig. 3.6, the flames show a distinct
blue color typical of premixed mode of combustion. These images suggest that most of the fuel
droplets prevaporize and premix with air before the combustion. The soot ring around the
interior of the glass enclosure resulting from the burn-out of the fuel accumulated during the
startup was still present. Large droplets were observed to reach the combustor wall, although
their number decreased significantly with increasing fuel temperature. The flames exhibited
stable operation for all test conditions.
3.3.2 Emissions Profiles
Figure 3.7 shows the effect of ALR and fuel temperature on axial profiles of CO and
NOx emissions. The axial distance (z) is measured from the atomizer exit plane, thus z = 45 cm
45
refers to the combustor exit plane. Figure 3.7 (a) shows the decreasing trend for CO emissions in
the flow direction for all conditions. This result signifies that CO is produced in the flame
initially during the fuel decomposition and is oxidized in the downstream zone. The CO
emissions are highest for the fuel inlet temperature Tf = 580C and lowest for Tf = 1210C for all
conditions shown in Fig. 8. The CO concentrations decrease with increase in ALR (see Figs 3.7c
and 3.7e) signifying that an increase in the atomizing air flow rate makes the premixed mode of
combustion more prevalent.
The corresponding NOx profiles are shown in Figs 3.7(b), 3.7(d), and 3.7(f). Figure
3.7(b) for ALR = 2.4 shows that the NOx concentration decreases with an increase in the fuel
inlet temperature. Most of the NOx is formed within the reaction zone, i.e., z < 20 cm. Figure
3.7(d) and 3.7(e) show that increasing the ALR decreases the NOx emissions for all fuel
temperatures. For a given ALR, an increase in temperature shows a decrease in NOx emissions
in the range of 20-25 ppm while for a fixed fuel temperature an increase in ALR from 2.4 to 4.0
decreases the NOx emissions by a factor of up to 15.
Figure 3.8 shows the radial emissions profiles taken at the combustor exit plane. Figure
3.8 (a) shows a decrease of about 1 ppm in CO concentration for each 20°C increase in the fuel
temperature. Data for the lowest fuel temperature show slight asymmetry in the flame, attributed
to larger turbulent fluctuations in the flame. The flow asymmetry reduces at higher ALR of 3.5
and ALR 4.0 (Fig. 3.8 (c) and (e)). At the highest ALR, the CO concentrations are small and tend
to overlap each other for different fuel inlet temperatures. Overall, the CO emissions were in the
range of 2 to 10 ppm for all fuel temperatures at all ALR conditions except for the worst
condition of low fuel temperature and low atomizing air flow rate, i.e., Tf = 58 C and ALR = 2.4.
Figure 3.8 (b), (d) and (f) show the radial NOx profiles for ALR of 2.4, 3.5 and 4.0 respectively.
46
The NOx emissions decrease significantly with an increase in the fuel temperature. The NOx
emissions also decrease with an increase in the atomizing air flow rate.
Figure 3.9 shows the same results plotted for a fixed fuel temperature, but with varying
ALR. Figure 3.9 (a) and (b) show the radial profiles CO and NOx concentrations for Tf = 58 C.
Results show a decreasing trend with an increase in ALR. Both CO and NOx emissions are
nearly uniform at the combustor exit, except for ALR = 2.4, which results in greater asymmetry.
NOx concentrations show a large effect of ALR for a given fuel temperature. At Tf = 580C, the
NOx concentration decreased by a factor of 6 as ALR increased from 2.4 to 4.0.
3.4. Conclusions
In this study, the effect of fuel temperature on NOx and CO emissions was studied using
high viscosity vegetable oil (soybean oil) as the fuel of interest. Experiments were conducted in
an atmospheric pressure, swirl stabilized burner. Results show that both CO and NOx
concentrations decrease with an increase in the fuel temperature. For a given ALR, the CO
emissions decreased by 1 to 2 ppm for each 20 C rise in the fuel temperature. The decrease in
NOx emissions was much greater, especially at higher fuel temperatures whereby a 20 C rise in
fuel temperature decreased NOx emissions by up to 20 ppm. For a given fuel inlet temperature,
increasing ALR from 2.0 to 4.0 showed a decrease in CO emissions by a factor of 2 to 3 and a
decrease in NOx emissions by factor of 6 to 15. The visual images were consistent with the
emission measurements, but they also reveal large wetting of the combustor wall by unburnt
droplets at all test conditions, although its extent decreased significantly with an increase in the
fuel inlet temperature and/or ALR.
47
Figure 3.1. Schematic Diagram of the Experimental Setup.
48
Figure3. 2. Schematic Diagram of Swirler
49
70
25
50
75
125
70
60
VO
Diesel
50
50
40
40
30
30
20
20
10
10
2
Viscosity ( mm /s)
60
100
0
25
50
75
100
125
0
o
Temperature ( C)
Figure 3.3 Effect of Temperature on Kinematic Viscosity of Diesel and VO fuel.
50
1.5
130
2
2.5
3
3.5
4
Measured Fuel Inlet Temperature ( o C)
120
4.5
130
120
o
70 C
o
100 C
130 oC
160 oC
110
110
100
100
90
90
80
80
70
70
60
60
50
1.5
2
2.5
3
3.5
4
ALR
Figure 3.4. Effect of ALR on Fuel Inlet Temperature
51
50
4.5
(a) (b) (c)
(d)
(e) Figure 3.5 VO Flame Images at ALR 2.4 for different fuel temperatures (a) Unheated VO, (b)
58o C, (c) 78 oC, (d) 99 oC, (e) 121 oC
52
(a)
(b)
(c)
(d)
(e)
Figure 3.6. VO Flame Images at ALR 4.0 for different fuel temperatures (a) Unheated VO, (b)
58o C, (c) 78oC, (d) 99 oC, (e) 121 oC
53
(a
75
0
10
20
30
40
Tf= 58oC
78 oC
99 oC
o
121 C
CO (ppm)
60
50
75
60
45
45
30
30
ALR 2.4
15
0
15
0
10
0
30
10
20
30
0
50
40
50
180
160
160
140
140
120
NOx (ppm)
40
Axial Distance (cm)
(b)
180
20
120
o
Tf= 58 C
78 oC
99 oC
121oC
100
100
80
80
60
60
ALR 2.4
40
40
20
20
0
0
10
20
30
40
0
50
Axial Distance (cm)
Figure 3.7. Axial Profiles for CO and NOx for different fuel temperatures at different ALR’s [(a)
and (b)] ALR 2.4
54
75
0
(c)
10
20
30
40
50
75
o
Tf= 58 C
o
78 C
o
99 C
o
121 C
CO (ppm)
60
45
45
30
30
ALR 3.5
15
0
0
10
180
0
15
20
30
40
0
50
Axial Distance (cm)
(d)
10
20
30
40
Tf= 58 oC
o
78 C
o
99 C
121 oC
160
140
NOx (ppm)
60
50
180
160
140
120
120
100
100
80
80
60
60
40
40
ALR 3.5
20
0
20
0
10
20
30
40
0
50
Axial Distance (cm)
Figure 3.7. Axial Profiles for CO and NOx for different fuel temperatures at different ALR’s [(c)
and (d)] ALR 3.5
55
(e)
75
0
10
20
30
Unheated VO
Tf= 58 oC
78 oC
99 oC
o
121 C
60
CO (ppm)
40
60
45
45
30
30
15
0
15
ALR 4.0
0
10
180
0
20
30
40
0
50
40
50
180
Axial Distance (cm)
(f)
10
20
30
160
160
Unheated VO
Tf= 58 oC
78 oC
99 oC
121oC
140
120
NOx (ppm)
50
75
140
120
100
100
80
80
60
60
ALR 4.0
40
40
20
20
0
0
10
20
30
40
0
50
Axial Distance (cm)
Figure 3.7. Axial Profiles for CO and NOx for different fuel temperatures at different ALR’s [(e)
and (f)] ALR 4.0
56
(a)
-4
20
-3
-2
-1
0
2
3
4
Tf= 58oC
78 oC
o
99 C
o
121 C
18
16
CO (ppm)
1
18
16
14
14
12
12
10
10
8
8
6
6
ALR 2.4
4
4
2
2
0
-4
-3
-2
-4
180
-3
-1
0
1
2
3
4
0
Radial Distance (cm)
(b)
-2
-1
0
1
2
3
4
180
160
160
140
140
o
120
NOx (ppm)
20
120
Tf= 58 C
78 oC
o
99 C
121oC
100
100
80
80
60
60
40
40
ALR 2.4
20
0
-4
20
-3
-2
-1
0
1
2
3
4
0
Radial Distance (cm)
Figure 3.8. Radial Profiles for CO and NOx for different fuel temperatures at different
ALR’s [(a) and (b)] ALR 2.4.
57
(c)
-4
20
-3
-2
-1
0
1
2
3
4
18
20
18
o
Tf= 58 C
o
78 C
o
99 C
o
121 C
16
14
16
14
CO (ppm)
ALR 3.5
12
12
10
10
8
8
6
6
4
4
2
2
0
-4
-3
-2
-1
0
1
2
3
4
0
Radial Distance (cm)
(d)
-4
180
-3
-2
-1
0
1
2
3
4
160
180
160
o
Tf= 58 C
78 oC
99 oC
121 oC
140
NOx (ppm)
120
ALR 3.5
140
120
100
100
80
80
60
60
40
40
20
20
0
-4
-3
-2
-1
0
1
2
3
4
0
Radial Distance (cm)
Figure 3.8. Radial Profiles for CO and NOx for different fuel temperatures at different ALR’s
[(a) and (b)] ALR 3.5.
58
(e)
-4
20
-3
-2
-1
0
1
18
CO (ppm)
ALR 4.0
4
20
18
16
14
12
12
10
10
8
8
6
6
4
4
2
2
0
-4
-3
-2
(f)
-4
180
-1
0
1
2
3
4
0
Radial Distance (cm)
-3
-2
-1
0
1
2
3
4
180
160
160
140
140
Unheated VO
o
Tf= 58 C
o
78 C
o
99 C
o
121 C
120
NOx (ppm)
3
Unheated VO
Tf= 58 oC
78oC
99 oC
121 oC
16
14
2
100
120
100
80
80
ALR 4.0
60
60
40
40
20
20
0
-4
-3
-2
-1
0
1
2
3
4
0
Radial Distance (cm)
Figure 3.8 Radial Profiles for CO and NOx for different fuel temperatures at different ALR’s
[(a) and (b)] ALR 4.0.
59
(a)
-4
20
-3
-2
-1
0
1
18
3
4
20
18
ALR 2.4
ALR 3.5
ALR 4.0
16
CO (ppm)
2
16
14
14
12
12
10
10
8
8
6
6
4
4
Tf= 58oC
2
2
0
-4
-3
-2
0
1
2
3
4
2
3
4
-3
-2
-1
0
1
160
180
160
140
140
ALR 2.4
ALR 3.5
ALR 4.0
120
NOX (ppm)
0
Radial Distance (cm)
(b)
-4
180
-1
120
100
100
80
80
60
60
o
Tf= 58 C
40
40
20
20
0
-4
-3
-2
-1
0
1
2
3
4
0
Radial Distance (cm)
Figure 3.9. Radial Profiles for CO and NOx for ALR 2.4, ALR 3.5 and ALR 4.0 at different fuel
inlet temperatures. [(a) and (b)] Tf = 58oC.
60
(c)
-4
20
-3
-2
-1
0
1
2
3
4
18
18
ALR 2.4
ALR 3.5
ALR 4.0
16
CO (ppm)
14
16
14
12
12
10
10
8
8
6
6
4
o
4
Tf= 78 C
2
2
0
-4
-3
-2
-4
180
-1
0
1
2
3
4
2
3
4
0
Radial Distance (cm)
(d)
-3
-2
-1
0
1
160
180
160
140
140
ALR 2.4
ALR 3.5
ALR 4.0
120
NOx (ppm)
20
120
100
100
80
80
60
60
Tf= 78oC
40
40
20
20
0
-4
-3
-2
-1
0
1
2
3
4
0
Radial Distance (cm)
Figure 3.9. Radial Profiles for CO and NOx for ALR 2.4, ALR 3.5 and ALR 4.0 at different fuel
inlet temperatures. [(c) and (d)] Tf = 78oC.
61
(e)
-4
20
-3
-2
-1
0
1
2
3
4
18
18
ALR 2.4
ALR 3.5
ALR 4.0
16
o
CO (ppm)
14
20
Tf= 99 C
16
14
12
12
10
10
8
8
6
6
4
4
2
2
0
-4
-3
-2
0
1
2
3
4
0
Radial Distance (cm)
(f)
-4
180
-1
-3
-2
-1
0
1
2
3
4
160
160
140
140
ALR 2.4
ALR 3.5
ALR 4.0
120
NOx (ppm)
180
120
100
100
Tf= 99oC
80
80
60
60
40
40
20
20
0
-4
-3
-2
-1
0
1
2
3
4
0
Radial Distance (cm)
Figure 3.9. Radial Profiles for CO and NOx for ALR 2.4, ALR 3.5 and ALR 4.0 at different fuel
inlet temperatures. [(e) and (f)] Tf = 99oC.
62
(g)
-4
20
-3
-2
-1
0
1
2
4
18
16
16
Tf= 121 C
12
14
ALR 2.4
ALR 3.5
ALR 4.0
o
12
10
10
8
8
6
6
4
4
2
2
0
-4
-3
-2
-4
180
-1
0
1
2
3
4
0
Radial Distance (cm)
(h
-3
-2
-1
0
1
2
3
4
180
160
160
140
140
120
NOx (ppm)
20
18
14
CO (ppm)
3
120
o
100
Tf= 121 C
100
ALR 2.4
ALR 3.5
ALR 4.0
80
80
60
60
40
40
20
20
0
-4
-3
-2
-1
0
1
2
3
4
0
Radial Distance (cm)
Figure 3.9. Radial Profiles for CO and NOx for ALR 2.4, ALR 3.5 and ALR 4.0 at different fuel
inlet temperatures. [(g) and (h)] Tf = 121oC
63
CHAPTER 4
CHARACTERISTICS OF PREHEATED NON-EVAPORATING BIO-OIL SPRAYS
Background
In chapter 2, we have investigated biodiesel fuels and diesel-VO blends as potential gas
turbine fuels to generate power. The fuel properties were characterized by gas chromatographmass spectrometer, thermogravimetric analysis, and density, kinematic viscosity, surface tension
and water content measurements. The combustion performance (in terms of carbon monoxide or
CO and nitric oxides or NOx emissions) of different fuels was compared experimentally in an
atmospheric pressure burner using an air-blast (AB) injector and swirling primary air around it
[Panchasara, 2009]. Results show that the atomization and fuel-air mixing processes have a
major impact on combustion emissions for all fuels. For example, an increase in the atomizing
airflow rate decreased CO emissions by a factor of up to 5 and NOx emissions by a factor of up
to 10. Most importantly, the 70-30 diesel-VO fuel blend was equivalent to biodiesel fuels, in
terms of NOx and CO emissions from the combustor.
One of the key parameters affecting atomization is the kinematic viscosity of the fuel
(Lefebvre, 1989; Lasheras & Hopfinger, 2000; Faeth, et.al, 1995; Babinsky & Sojka, 2002).
Figure 4.1 shows that the kinematic viscosity of VO is higher than that of diesel by more than an
order of magnitude which has been shown in chapter 2. The diesel kinematic viscosity at 25 0C
matches the viscosity of 90-10 diesel-VO blend at 40 0C, 80-20 blend at 50 0C and 70-30 blend
64
at 60°C. At 25 0C, the biodiesel viscosity is nearly the same as that of the 80-20 diesel-VO blend.
For a given temperature, the kinematic viscosity of 70-30 diesel-VO blend is similar to that of
biodiesel and about twice that of the diesel. Thus, emissions from 70-20 diesel-VO blend were
comparable to those from biodiesel (Panchasara, et.al. 2009).
An alternative to biodiesel and diesel-VO blends is direct combustion of source oils,
which can result in significant economic and environmental benefits. It will however require
preheating of the oil to reduce its kinematic viscosity so that the pressure drop in the fuel supply
line is reasonable, atomization occurs with fine droplets, and diffusion mode of combustion is
avoided to reduce pollutants of PM, NOx, CO, etc. In Chapter 3, we have shown that increasing
VO inlet temperature decreased both CO and NOx concentrations in the flame (Faeth, 1987). For
a given air to liquid mass ratio (ALR), the CO emissions decreased by 1 to 2 ppm for each 20 C
rise in the fuel temperature. The decrease in NOx emissions was much greater, especially at
higher fuel temperatures whereby a 20 C rise in fuel temperature decreased NOx emissions by up
to 20 ppm.
Spray characteristics such as drop diameter, drop size distribution, cone angle and
penetration determine fuel-air mixing in the combustor, and hence, pollutant formation, life,
durability, and efficiency of the gas turbine engine. Spray is produced by break-up of liquid fuel
into small droplets through the atomization process. Small droplets produce larger surface area,
thereby reducing the liquid vaporization time and improving fuel-air mixing. As stated by
Lefebvre, efficient combustion requires optimal droplet size distribution within the spray, for a
range of operating conditions to include droplets both large enough to penetrate into the
combustion chamber and small enough to pre-vaporize within the short residence time in the
flame region.
65
Past studies have presented the structure and break-up properties of sprays
(Faeth,et.al.,1995;Babinsky & Sojka,2002), different modeling techniques of drop size
distributions (Simmons, et.al, 2009) and processes of mixing, transport and combustion in sprays
(Faeth,et.al., 1995). Figure 4.2 depicts the geometry, instabilities, wavelength and the break up
process in a spray of liquid jet injected into a high velocity annular coaxial gas stream (Bari,
et.al. 2002). Small jet diameters and low Reynolds numbers demonstrate a spray with Rayleigh
instability. At higher Reynolds numbers, the jet is subjected to waviness caused by aerodynamic
effects, a regime called non-axisymmetric Rayleigh break-up. At even higher Reynolds numbers,
droplets are stripped off by wind stress at the liquid/gas interface and atomization results from
short wavelength shear instability (Lasheras & Hopfinger,2000).Related research on preheated
oils has focused mainly on injector performance and emissions characteristics of diesel engines
operated on cottonseed and palm oils (Bari, et.al.,2002;Nwafor,2003). These studies showed that
the reduction in kinematic viscosity resulting from fuel preheating improves the combustion and
emissions performance of the engine.
The objective of the current study is to examine the spray characteristics of preheated
soybean oil atomized by an AB injector. Such spray measurements of preheated VO are nonexistent at this time. The non-reacting spray is created at ambient conditions of temperature and
pressure. First a laser sheet visualization system is used to obtain the qualitative spray images.
Next, a two component Phase Doppler Particle Analyzer (PDPA) system is used to obtain
detailed quantitative measurements of droplet diameter and gas phase velocities. Details of the
experimental set-up, results and discussions, and conclusions are presented in the following
sections.
66
4.2. Experimental set-up
Figure 4.3 shows a schematic of the experimental set-up consisting of fuel (VO) and air
supply systems, fuel heater, AB injector, and spray collection system. The VO from the reservoir
is pumped by a high pressure peristaltic metering pump (Cole Parmer Model EW-07523-70)
with an uncertainty of ± 0.25% of the reading. The pressure oscillations introduced by the pump
are eliminated by a pulse dampener introduced in the flow path to achieve a steady flow. Next,
an electric inline liquid heater (CRES-ILB-12/24) is used to preheat the VO upstream of the AB
injector. The heater uses a Proportional Derivative Controller (PID) unit to control the VO
temperature within accuracy of ± 0.5 C, measured at the heater outlet by a K-type thermocouple.
Heated VO enters the injector through an insulated 5 cm long tube of 3 mm ID to minimize the
heat loss. The VO temperature at the injector inlet is measured using a K-type thermocouple. Air
for atomization is supplied by an air compressor and is measured using a mass flow meter with
0-100 liters per minute range.
The spray collection system includes a funnel, filters, collector, and an exhaust fan. The
VO in the spray is collected by the funnel and removed from the test area by the suction fan
located downstream. The resulting flow passes through a filter element where the VO is
recovered prior to exhausting out nearly oil-free air to the atmosphere. Figure 4 shows the
detailed view of the commercially available AB injector (Delavan Model 30609-2). In this
injector, atomization occurs by the shear interaction between gas and liquid phases. The liquid
supply tube of 0.3 mm inside diameter is surrounded by a swirling stream of atomizing air to
break up the liquid jet into fine droplets.
A laser sheet visualization system was used for qualitative visualization of the spray. The
laser sheet was created by a 200 mW diode-pumped solid state laser producing 532 nm green
67
laser beam. A combination of cylindrical and spherical lenses from Edmund Optics was used to
form a laser sheet at the mid-plane of the spray. Uncoated cylindrical lens of 12.5mm diameter x
12.5mm focal length and an uncoated Plano convex lens of 25 mm diameter x 750 mm focal
length used created a laser sheet of about 1mm thick Spray images were taken by a digital
camera with exposure times of 40µs and 100µs was mounted on a fixed tripod to ensure
consistent field of view for different operating conditions.
Quantitative drop size measurements were obtained using 2D Phase Doppler Particle
Analyzer (PDPA). PDPA is a point sampling device based on the light scattering
interferometery. The laser beams from the transmitter probe intersect to form a sample
measurement volume. Principally when a particle or drop passes through the beam intersection
region, the scattered light forms a fringe pattern. Since the droplet is moving, the fringe pattern
sweeps past the receiver aperture at the Doppler difference frequency, which is directly
proportional to the drop velocity. The spatial frequency of the fringe pattern is inversely
proportional to the drop diameter. The phase shift between the Doppler burst signals from
different detectors is proportional to the diameter of the spherical shaped droplet. There is no
calibration required for the PDPA method since the drop size and the droplet velocity are
dependant only on laser wavelength and optical configuration.
The schematic of the 2D PDPA system is as shown in Figure 4.5. The laser beam from a
2-W water cooled argon-ion laser is separated into a pair of 514.5 nm green beams and a pair of
488 nm blue beams using a beam separator assembly. One beam of each pair is shifted by a 40
MHz Bragg cell. Next, each beam is focused onto a fiber optic cable to deliver the beams to a
250 mm focal length PDPA transmitter. The PDPA receiver is set at an angle of 135 degrees
from the transmitter to collect the refracted light intensities from the spray. The detected signal is
68
acquired by a data acquisition system, and analyzed using the TSI Flow Sizer software to obtain
mean and root mean square (RMS) velocities, Sauter mean diameter (SMD), and drop size
distribution data.
Experiments were conducted for fixed VO flow rate of 12 liters per minute (lpm). The
atomizing air flow rate through the injector was varied to obtain ALR of 2.0 and 4.0.
Measurements were obtained by moving the AB injector using a three-way traversing system,
while the PDPA system was stationary. The injector was traversed to acquire radial profiles at
axial planes between Y = 10 mm and 100 mm, in 0.5 mm intervals. The center of the spray
pertained to the peak location of the axial velocity profile. Data rates of up to 55 kHz were
obtained towards the centre of the spray. The data rate and mean axial velocity both decreased to
nearly zero as the detection volume reached the outer edge of the spray.
4.3. Results and Discussion
4.3.1. Spray Images
The qualitative spray images for VO at 40 C, 70 C and 100 C are shown in Figure 4.6 for
ALR = 2.0 and in Figure 7 for ALR = 4.0. For each case, images were obtained for 100µs (top)
and 40µs (bottom) exposure times, to capture steady and transient features of the spray. Figure
4.6 shows a decreasing trend of spray angle with increasing VO temperature. This result can be
explained by an increasing number of larger droplets reaching farther away from the center of
the spray for lower VO temperatures. Because of the high momentum, large droplets also tend to
move farther downstream in the spray. The bottom images at a smaller exposure time reveal
similar characteristics, but they also reveal the waviness at the edge of the spray. The
observations of Figure 4.6 are replicated in Figure 7 for ALR = 4.0. An increase in ALR has
69
similar effect as an increase in VO inlet temperature, i.e., the spray becomes narrower, shorter,
and denser with an increase in the VO inlet temperature and/or ALR.
4.3.2. SMD Contour Plots
Point-wise measurements acquired at different axial and radial locations were used to
construct contour plots for different operating conditions. Figure 4.8 shows the contour plots of
SMD for T = 40 C, 70 C and 100 C at ALR 2.0, and T = 70 C at ALR = 4.0. For the lowest VO
inlet temperature of 40 C, the contour plot in Figure 4.8 (a) can be divided into three regions: the
smallest droplets of 5 to 15 µm reside in the center core region, mid-size droplets up to 40 µm
are found in the shear layer region, both near and farther away from the injector exit. The largest
droplets with diameter exceeding 40 µm are observed at the outer edge of the spray, both in the
near and far downstream locations.
Evidently, the high momentum of large droplets is
responsible for carrying them away from the injector exit. At this low VO inlet temperature, the
maximum SMD is about 60 µm, and the SMD range is correspondingly wider.
Figure 4.8 (b) shows the contour plot of SMD for VO at T = 70 C. Results show similar
trends; smallest droplets in the core region and largest droplets at the edge of the spray. The
maximum SMD value for this case is however around 40 µm, which is significantly lower than
that of 60 µm for T = 40C. Figure 4.8 (c) shows the same trend in the SMD for VO at T = 100
C. The maximum SMD for this case is still around 42 µm. Results show that an increase in VO
temperature improves atomization by decreasing the SMD, although benefits diminish at higher
VO temperatures. The effect of ALR on SMD is observed in Figure 4.8(d) showing SMD
contour plot for T = 70 C and ALR = 4.0. For this high ALR, the maximum SMD is around 34
µm and the spray cone angle is smaller than that for lower ALR. Results show that increases in
both the VO temperature and ALR improve atomization. Since increasing VO temperature
70
requires thermal energy input and increasing ALR increases the pressure drop penalty, a proper
balance between the two parameters is important to achieve the optimum injector performance.
Pressure drop measurements for the AB injector used in this study are given by Simmons et al
2009.
4.3.3. Axial Velocity Contour Plots
Figure 4.9 shows the contour plots of axial velocity at aforementioned VO inlet
temperatures and ALR values of 2.0 and 4.0. Figure 4.9 (a) for T = 40 C and ALR = 2.0 shows
that the axial velocity peaks in the center of the spray, with velocity magnitude diminishing
towards the edge of the spray. Moreover, the axial velocity decreases in the axial direction
indicating the loss of momentum (or kinetic energy) of the droplets. For ALR = 2.0, contour
plots in Figure 4.9(a)-(c) show that the peak axial velocity increases with increasing VO inlet
temperature, evidently because of the heating of the atomizing air by VO. For T = 70 C, an
increase in ALR from 2.0 to 4.0 changes the peak axial velocity near the injector exit from 40
m/s to about 95 m/s. At the higher ALR, the radial extent of the flow field decreases indicating a
narrow spray.
Figure 4.10 shows the contour plots of RMS axial velocity for conditions pertaining to
those in Figures 4.8 and 4.9. Figures 4.10(a)-(c) show that an increase in VO temperature
increases the RMS axial velocity in the near field. For example, the maximum RMS axial
velocity is about 7 m/s, 14 m/s and 15 m/s, respectively, for VO inlet temperature of 40 C, 70 C,
and 100 C. Clearly, the higher axial velocity resulting from heating of the atomizing air by the
VO is also producing higher turbulence labels, which would tend to reduce the droplet diameter
as observed in Figure 4.8. These results show that reduction in kinematic viscosity associated
with heated VO, increase in mean axial velocity of the atomizing air because of the heat transfer
71
from the heated VO, and increase in turbulent fluctuations, possibly due to high mean velocity,
are complimentary effects in reducing the SMD of heated VO. Figure 4.10 (d) shows the contour
of RMS axial velocity for T = 70 C and ALR = 4.0. In this case, the RMS axial velocity is
significantly higher, with the peak value reaching up to 32 m/s. Again, this increase in RMS
axial velocity is a major contributing factor in reducing the SMD at the higher ALR as shown in
Figure 4.8(d).
4.3.4. Radial Profiles of SMD
Figure 4.11(a) shows that the VO inlet temperature and/or ALR have a minor effect on
SMD in the center region of the spray in the near field, i.e., Y = 20 mm. For all cases, the
minimum SMD at the center of the spray is nearly the same, i.e., 8 µm. Away from the center,
the radial profile is steeper at the lower temperature, which signifies a spray with a wider range
of droplet diameters. The VO inlet temperature and ALR have the greatest effect in the outer
region of the spray. The effects of VO temperature and ALR are more obvious at downstream
locations of z = 50 and 80 mm, as shown in Figures 4.11(b and c). Increase in ALR has the most
dramatic effect; the radial profile for ALR = 4.0 is much flatter compared to that for ALR = 2.0.
Increase in VO inlet temperature results in smaller droplets in the outer regions of the spray.
Thus, the minimum SMD in the spray is independent of the VO temperature or ALR. However,
increasing VO temperature or ALR decreases the radial spread of the spray as well as the SMD
range.
Results in Figure 4.11 are rearranged in Figure 4.12 to illustrate how VO inlet
temperature, and hence, fuel kinematic viscosity affects SMD at different locations in the spray.
Results are plotted in terms of SMD versus kinematic viscosity for the three test cases with ALR
= 2.0. Figure 4.12(a) shows that near the injector exit, Y = 15 mm, for all radial locations, the
72
SMD increases with an increase in the fuel kinematic viscosity. However, the SMD increase is
minor near the center (r = 0 and 5 mm) and much greater away from the center (r = 8 and 12
mm). Similar results are obtained at Y = 20 mm and Y = 30 mm as shown in Figures 4.12(b and
c).
4.3.5. Droplet Diameter Distributions
Figure 4.13 (a, b and c) show the droplet diameter distributions in the spray for VO at
40°C, 70°C and 100°C at ALR = 2.0 and Figure 4.13 (d) shows the distribution profile for VO at
70°C and ALR = 4.0. The distribution profiles are taken at the edge of the spray at Y = 20 mm, r
= 10 mm and Y = 80 mm, r = 25 Figure 4.13(b and c) show the same trend of droplet diameter
distribution profiles for VO at 70°C and 100°C and ALR = 2.0. For a given ALR and given axial
location in the spray, higher VO inlet temperature shows a lower range of droplet diameters with
a narrower distribution profile. At Y = 20 mm, the largest droplet diameter is about 120 µm for
VO at 70°C whereas for VO at 100°C the largest droplet diameter is about 80 µm (see Figures
4.13 b-c). At Y = 80 mm, the largest droplet diameter is about 225 µm for VO at 70°C whereas
for VO at 100°C the largest droplet diameter is about 130 µm. The larger percentage of smaller
droplets at higher VO inlet temperatures result in the decrease in the SMD as observed in Figure
4.8. Figure 13(d) shows the droplet distribution profile for VO at 70 C and ALR 4.0. Profiles
show a higher percentage of larger droplets in the near field, Y = 20 mm, compared to those Y =
80mm.
As shown in Figure 4.13 (a), for VO at 40°C, the droplet diameter varies widely at Y =
20mm and Y = 80 mm. The count of larger droplets increases at the downstream location of Y =
80 mm. For example, the largest droplet diameter in the near field (Y = 20 mm) is about 150 µm
whereas at Y = 80 mm, the largest drop diameter has increased to about 250 µm. The wider drop
73
diameter distribution in the far field is attributed to the larger droplets of high momentum
migrating towards the periphery of the spray.
Figure 13(b and c) show the same trend of droplet distribution profiles for VO at 70°C
and 100°C for ALR = 2.0. For a given ALR and given axial location in the spray, higher VO
inlet temperature results in a lower range of droplet diameters with a narrower distribution
profile. At Y = 20 mm, the largest diameter is about 120 µm for VO at 70°C whereas for VO at
100°C the largest droplet diameter is of about 80 µm (see Figures 4.13 b-c). =At Y = 80 mm, the
largest droplet diameter is about 225 µm for VO at 70°C whereas for VO at 100°C the largest
droplet diameter is about 130 µm. The larger percentage of smaller droplets at higher VO inlet
temperatures result in the decrease in the SMD as observed in Figure 8. Figure 4.13(d) shows the
droplet distribution profile for VO at 70°C and ALR = 4.0. Profiles show a higher percentage of
larger droplets in the near field, Y = 20 mm, compared to those at Y = 80 mm. The largest drop
diameter for this case is about 180 µm and 80 µm, respectively, at Y = 20 mm and 80 mm.
4.4. Conclusions
In this chapter, measurements of SMD, axial velocity, axial RMS velocity and droplet
distribution are presented to explain the effect of operating conditions in a non-evaporating spray
of VO created by an AB injector. Laser sheet visualization system was used to acquire the spray
images.
•
For ALR = 2.5 and VO inlet temperature of 40 to 100°C, the SMD of droplets in the spray
ranged from 35 to 60 µm. The SMD decreased down to 30 µm for a higher ALR of 4.0.
•
The SMD of the VO spray decreased as the VO inlet temperature increased, indicating
improved atomization.
•
Because of the high momentum, the larger droplets migrated towards the edge of the spray
74
•
The number of larger droplets increased for the lower VO temperature, which led to an
increase in the SMD.
•
The spray angle decreased with an increase in VO inlet temperatures and ALR, thereby,
improving the spray quality in terms of decreased droplet diameter.
•
The mean axial velocity was the highest at the center of the spray and it decreased towards
the edge of the spray and at downstream locations.
•
The RMS axial velocity increased with an increase in VO inlet temperature and/or ALR.
Experiments presented in the next chapter will focus on drop size measurements in VO
spray flames to delineate the effects of heat release rate on fuel vaporization. The spray
characteristics are expected to be much different because of the thermal feedback from the flame.
75
70
20
40
60
Diesel
Vegetable Oil
60
60
2
Kinematic Viscosity mm /s
80
70
50
50
40
40
30
30
20
20
10
10
0
20
40
60
0
80
o
Temperature C
Figure 4.1
Kinematic Viscosity of Diesel and Vegetable Oil.
76
Figure 4.2 Schematic representation of the liquid break-up indicating geometry and
different lengths.
77
X
Y
Figure 4.3 Schematic diagram of the experimental setup
78
Orifice disc
Swirler vanes
Fuel supply Tube
Fuel inlet
Figure 4.4 Air-Blast Injector Details
79
X
Y
Figure 4.5 Schematic of PDPA system
80
(a)
(b)
Figure 4.6 Laser sheet spray images for VO at 40 C, 70 C and 100 C at ALR 2.0. (a)
Spray Images at Exposure time 100 ms. (b) Spray Images at Exposure time 40 ms.
81
(a)
(b)
Figure 4.7 Laser sheet spray images for VO at 40 C, 70 C and 100 C at ALR 4.0. (a)
Spray Images at Exposure time 100 ms. (b) Spray Images at Exposure time 40 ms.
82
Axial location (mm)
-30
100
-20
-10
0
10
20
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
-30
-20
-10
0
10
20
o
T = 40 C,
ALR = 2.0
SMD (microns)
100
60
55
50
45
40
35
30
25
20
15
10
5
10
Radial location (mm)
Axial locaiton (mm)
-30
100
-20
-10
0
10
20
100
90
T = 70 oC,
ALR = 2.0
90 SMD (microns)
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
-30
-20
-10
0
10
20
60
55
50
45
40
35
30
25
20
15
10
5
10
Radial location (mm)
Figure 4.8. Contours of Sauter Mean Diamter for (a) T = 40 C, ALR = 2.0, (b) T = 70 C, ALR
2.0.
83
100
-24 -16
-8
0
8
16
24
90
100
90 T = 100 C ,
Axial location (mm)
ALR = 2.0
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
-24 -16
-8
0
8
16
24
SMD (microns)
60
55
50
45
40
35
30
25
20
15
10
5
10
Radial location (mm)
100
-30
-20
-10
0
10
20
T = 70 oC,
ALR = 4.0
90 SMD (microns)
90
Axial locaiton (mm)
100
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
-30
-20
-10
0
10
20
60
55
50
45
40
35
30
25
20
15
10
5
10
Radial location (mm)
Figure 4.8. Contours of Sauter Mean Diamter for (c) T = 100 C, ALR = 2.0, (d) T = 70 C, ALR
4.0.
84
(a)
-30
100
-20
-10
0
10
20
100
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
T = 40 C ,
ALR = 2.0
Axial Location (mm)
Axial Velocity (m/s)
10
-30
-20
-10
0
10
20
26
24
22
20
18
16
14
12
10
8
6
4
2
10
Radial Location (mm)
(b)
Axial Location (mm)
-30
100
-20
-10
0
10
20
100
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
-30
-20
-10
0
10
20
T = 70 C ,
ALR = 2.0
Axial Velocity (m/s)
38
35
32
29
26
23
20
17
14
11
8
5
2
10
Radial Location (mm)
Figure 4.9. Contours of Axial Velocity for (a) T = 40 C, ALR = 2.0, (b) T = 70 C, ALR 2.0.
85
(c)
-20
100
-10
0
10
20
100
90
90
80
80
T = 100 C ,
ALR = 2.0
Axial Location (mm)
Axial Velocity (m/s)
.0
70
70
60
60
50
50
40
40
30
30
20
20
10
-20
-10
0
10
20
41
38
35
32
29
26
23
20
17
14
11
8
5
2
10
Radial Location (mm)
(d)
100
90
-30 -20 -10
0
10
20
100
90 T = 70 C ,
.0
Axial Location (mm)
ALR = 4.0
80
80 Axial Velocity (m/s)
70
70
60
60
50
50
40
40
30
30
20
20
10
-30 -20 -10
0
10
20
93
85
77
69
61
53
45
37
29
21
13
5
10
Radial Location (mm)
Figure 4.9.Contours of Axial Velocity for (c) T = 100 C, ALR = 2.0 and (d) T = 70 C,
ALR = 4.0.
86
(a)
-30
100
-20
-10
0
10
20
100
90
90
80
80
T = 40 C ,
ALR = 2 .0
Axial Locaiton (mm)
Axial RMS Velocity
70
.0
70
60
60
50
50
40
40
30
30
20
20
10
-30
-20
-10
0
10
20
7
6.5
6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
10
Radial Locaiton (mm)
(b)
-30 -20 -10
100
90
0
10
20
100
90 T = 70 C ,
.0
Axial Location (mm)
ALR = 2 .0
80
80 Axial RMS Velocity
70
70
60
60
50
50
40
40
30
30
20
20
10
-30 -20 -10
0
10
20
10
13
12
11
10
9
8
7
6
5
4
3
2
1
Radial Location (mm)
Figure 4.10.Contours of Axial RMS Velocity for (a) T = 40 C, ALR = 2.0, (b) T = 70 C, ALR
2.0.
87
(c)
-24 -16 -8
Axial Location (mm)
100
0
8
16 24
100
90
T = 100 C ,
90 ALR = 2.0
80
80
Axial RMS Velocity
70
70
.0
60
60
50
50
40
40
30
30
20
20
10
-24 -16 -8
0
8
16 24
10
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
Radial Location (mm)
(d)
100
Axial Location (mm)
90
-30 -20 -10
0
10
20
100 T = 70 C ,
ALR = 4 .0
90
.0
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
-30 -20 -10
0
10
20
10
Axial RMS Velocity
32
30
28
26
24
22
20
18
16
14
12
10
8
6
4
2
Radial Location (mm)
Figure 4.10.Contours of Axial RMS Velocity for (c) T = 100 C, ALR = 2.0 and (d)
T = 70 C, ALR = 4.0
88
-40
70
-30
-20
-10
0
10
20
30
40
70
T = 40 C
T = 70
T = 100 C
T = 70 C, ALR=4.0
60
Y = 20 mm
60
SMD (microns)
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Radial Distance (mm)
Figure 4.11.Profiles of Sauter Mean Diameter for (a) Y =20 mm
89
-40
70
-30
-20
-10
0
10
20
30
40
70
Y = 50 mm
60
50
SMD (microns)
60
T = 40 C
T = 70 C
T = 100 C
T = 70 C, ALR=4.0
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
Radial Distance (mm)
Figure 4.11.Profiles of Sauter Mean Diameter for (b) Y =50 mm
90
0
40
-40
70
-30
-20
-10
0
10
20
30
40
70
Y = 80 mm
60
50
SMD (microns)
60
T = 40 C
T = 70 C
T = 100 C
T = 70 C_ALR_4
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Radial Distance (mm)
Figure 4.11.Profiles of Sauter Mean Diameter for (c) Y =80 mm
91
40
0
5
10
35
15
20
25
30
35
40
35
Y = 15 mm
SMD (microns)
30
30
25
25
r = 0 mm
r = 5 mm
r = 8 mm
r = 12 mm
20
20
15
15
10
10
5
5
0
0
5
10
15
20
25
30
0
35
Viscosity (mm2/s)
Figure 4.12. Profiles of Sauter Mean Diameter versus Fuel
Viscosity for (a) Y =15 mm.
92
40
0
5
SMD (microns)
35
10
15
20
25
30
35
40
35
Y= 20 mm
30
30
25
25
r = 0 mm
r = 5 mm
r = 8 mm
r = 12 mm
20
15
20
15
10
10
5
5
0
0
5
10
15
20
25
30
0
35
Viscosity (mm2/s)
Figure 4.12.Profiles of Sauter Mean Diameter versus Fuel
Viscosity for (b) Y = 20 mm
93
40
0
5
10
15
20
25
Y = 30 mm
35
SMD (microns)
30
30
35
40
r = 0 mm
r = 5 mm
r = 8 mm
r = 12 mm
35
30
25
25
20
20
15
15
10
10
5
5
0
0
5
10
15
20
25
30
0
35
2
Viscosity (mm /s)
Figure 4.12. Profiles of Sauter Mean Diameter versus Fuel
Viscosity for (c) Y = 30 mm
94
0
10
50
100
150
200
T = 40 C ,
ALR = 2.0
3
250
10
3
Diameter Count
Y= 20mm, r= 10mm
Y = 80mm, r = 25mm
10 2
10 2
10
1
10
1
10
0
10
0
0
50
100
150
200
250
200
250
Diameter (microns)
0
50
100
150
T = 100 C ,
ALR = 2.0
103
103
Diameter Count
Y = 20 mm, r =10 mm
Y = 80 mm, r =25 mm
102
102
101
101
100
100
0
50
100
150
200
250
Diameter (microns)
Figure 4.13.Profiles of Diameter Distribution for (a) T = 40 C, ALR = 2.0, (b) T = 100 C, ALR
2.0
95
0
50
100
150
200
T = 70 C ,
ALR = 2.0
10 3
250
103
Diameter Count
Y =20 mm, r=10 mm
Y =80 mm, r=25 mm
10
2
10
10 1
10
2
101
0
10
0
50
100
150
200
250
200
250
0
Diameter (microns)
0
Diameter Count
10
50
100
150
T = 70 C ,
ALR = 4.0
3
10 3
Y = 20mm , r = 10 mm
Y = 80mm , r = 25 mm
10
2
10
2
10 1
10 1
10 0
10 0
0
50
100
150
200
250
Diameter (microns)
Figure 4.13.Profiles of Diameter Distribution for (c) T = 70 C, ALR = 2.0, (b) T = 70 C, ALR
4.0
96
CHAPTER 5
CHARACTERISTICS OF PREHEATED BIO-OIL SPRAYS
Background
Global atmosphere pollution has become a serious problem for today. The emissions
from the combustion of fossil fuels contribute a notable part to this pollution. Environmental
care together with the limited stock and growing prices of fossil fuels has given alternative fuels
the potential to supplant a significant portion of fuel for combustion applications such as gas
turbine engines and IC engines. Given a wide spread of different biofuels available for
combustion applications, the present study concentrates on atomization spray characteristics of
vegetable oils. Vegetable oils have energy density, cetane number, heat of vaporization and
stoichiometric air/fuel ratio comparable to diesel fuel. Different techniques have been employed
so far to improve on the physical properties of bio-oils and thus opening the doors to clean
combustion. The high kinematic viscosity has an adverse effect on the combustion of vegetable
oils, posing problems in the associated fuel supply line and injector system, as discussed in
chapters 2 and 3. Some well-known techniques to deal with high kinematic viscosity levels of
neat vegetable oils include dilution, pyrolysis, micro-emulsion and trans-esterification. These
techniques however, require additional energy input to improve the physical properties of the
fuel. Preheating of the fuel is also one of the ways that reduces the kinematic viscosity to
improve the atomization. Preheating is employed in the present study to investigate the spray
characteristics in a non evaporating spray as well as flame spray.
97
The atomization and subsequent propagation of the fuel droplets, their vaporization and
combustion are the most important processes concerning the formation of pollutants with the use
of liquid fuels. For example in diesel engines, gas turbine engines, and oil burners, the
combustion rate of fuel is controlled by effective vaporization of the fuel. The liquid fuel
atomization rate has a strong influence on vaporization rates because the total surface area of the
fuel is increased greatly by the atomization process. The fundamental mechanisms of
atomization have been under extensive experimental and theoretical study for more than a
century (Liu & Reitz, 1993). Still, one of the major thrusts in worldwide combustion research has
been to gain insight into the physics of liquid fuel combustion in the primary zone of the
combustor (Kneer, et.al.1993).
Traditionally, the fundamental basis for the atomization process relies on either injecting
the fuel under relatively high pressure into a relatively slow moving gas or subjecting the fuel to
a high velocity air blast or a combination of two atomization mechanisms. The fuel injector
assumes an important role in the combustor by providing some degree of fuel/air mixing close to
the atomizer. Almost all combustor performance characteristics are strongly governed by the
quality of the spray produced by the fuel atomizer. Calculation of the evaporation and reaction
rates in the combustor involves the evaluation of parameters such as mean spray angle, range of
drop sizes in the spray, and drop trajectory. It is a well known fact that the achievement of
efficient fuel atomization and rapid evaporation in addition to mixing with combustion air can
have a significant effect on such parameters as emissions, exhaust gas temperature profiles, and
pattern factor.
There are numerous studies on liquid breakup in the literature. As the relative velocity
between the drop and gas increases, the drop break up regimes such as bag break up, shear
98
breakup and catastrophic break up are encountered. The jet break up as summarized by Liu &
Reitz (1993) comprises of four liquid break up regimes based on breakup drop size and unbroken
jet length. Rayleigh breakup regime; first-wind induced break up regime; second break up
regime and the atomization break up regimes which are encountered as the jet velocities
increases. At lower jet velocities, the growth of the small disturbances on the liquid surface due
to the interaction with the surrounding gas is believed to be the dominant reason for the liquid
break up. As the jet velocity is increased and the aerodynamic effect becomes more prominent,
the breakup mechanisms become increasingly complex.
Air blast atomization is an attractive strategy for liquid fuel breakup in gas turbines.
Principally, the air blast atomizer functions by employing the kinetic energy of a flowing
airstream to break the fuel jet or sheet into ligaments and then drops. Air blast atomizers offer
great advantages over pressure atomizers, since a finer spray can be produced with a lower fuel
pumping pressure. Air blast atomization maximizes the interaction between the air and liquid
flows by, taking advantage of high velocity airflow to produce fine drops in a well distributed
spray. The atomizing air also serves as emissions reduction strategy improving the fuel air
mixing in the combustor, to reduce soot particulates, CO and NOx emissions.
The performance of a given spray combustion system depends not only on the fuel
droplet size distribution but also on the spray spatial distribution and the interaction of the
droplets with the gas turbulence that involves a physical mechanism that is yet to be well
understood. For this reason spatially and temporally resolved information such as mean droplet
size, droplet size distribution, mean drop velocity rms velocities needs to be studied to
understand most favorable spray conditions for optimal combustion performance.
99
Physical properties such as kinematic viscosity, surface tension and volatility are the key
parameters that affect the process of fuel atomization and evaporation. The liquid kinematic
viscosity affects not only the drop size distribution of the spray but also the fuel injector pressure
drop. An increase in kinematic viscosity lowers the Reynolds number, hindering the
development of any natural instabilities in the fuel jet or sheet, which help to further disintegrate
the drops. These combined effects delay any further disintegration thus increasing the droplet
sizes in the spray. Many alternative fuels are expected to have high kinematic viscosity which
makes them difficult to atomize well and thus affecting the combustion efficiency.
A comprehensive study on turbulent diffusion flames using intrusive probing techniques
was made by Onuma and Ogasawara, (1975). They suggested that the spray flame structure is
similar to that of a gaseous diffusion flame in turbulent flow. Chigier,(1974) used a non intrusive
detection technique to measure the Sauter Mean diameter (SMD), drop velocity, and number
density of air assisted spray and spray flames. A series of experimental and numerical studies of
air assisted sprays and spray flames have been made by McDonell, & Samuelsen, (1991). Their
observations concluded that the presence of fuel drops and reactions alters the structure of the
gas-phase turbulence and that local clustering of drops exists for both non-reacting and reacting
cases. A large portion of the experimental research in liquid fuel combustion is focused on
pressure atomization mainly in the diesel engine applications. Relatively few studies have been
reported on air blast atomization and their potential optimum strategies in alternative fuel
combustion. Moreover, very little attention has been given to the evaporation characteristics of
the air blast atomized sprays of alternative fuels. Detailed studies on the characteristics of spray
flames are necessary to mitigate environment problems and enhance the performance and
efficiency of liquid bio fuel combustion systems.
100
The present work seeks to experimentally investigate the spray characteristics of the fuel
droplets in a non-evaporating as well as flame spray conditions using the Phase Doppler Particle
Analyzer. An air blast atomizer is selected for the present investigation to generate the spray.
The photographic image and the key structural features of a flame from a typical air blast
atomizer are shown in figure 5.1. Larger drops are distributed mostly on the outer edge of the
spray in region ‘C’and the smaller drops are located in the centre region ‘A’. These larger drops
on the edge are affected by the presence of the flame. Due to slower evaporation rate of the large
drops, a blue lean reaction sheath is formed inside the spray boundary as seen in the
photographic image. The main reaction zone is a mixture of fuel vapor and air that is sufficient
for combustion. In the central core of the spray, as seen in region ‘A’the fuel vapor concentration
is too rich to allow the chemical reactions to take place. Above the spray flame, the smaller
droplets produce tiny blue streak flames as well as the larger drops that burn incompletely
produces a yellow orange streak flame flying in all directions (regions ‘B’ and ‘C’). Mostly on
the outer periphery of the spray flame, single drop burning with self sustained envelope flame is
observed. The swirling air flow improves mixing and creates a homogenous combustible mixture
and hence stabilizes the liquid fuel flame. This is shown in figure 5.1 by the region ‘D’ which is
a methane air mixture flame. The swirling combustion air makes the liquid sheet unstable and
hence helps any further disintegration in to smaller droplets that follow the flow. Figure 5.2
shows the droplet vaporization in spray flames. The spray consists of initial cool and hot zones.
Within the cool zone, heat transfer is restricted to radiation from the flame front to the droplet
surface. In the hot zone, heat transfer takes place both by radiation from the flame front and by
turbulent convection. The droplet diameter reduces by the d2 law:
-d(d2)/dt = λ
101
Where λ is the evaporation constant for forced convection.
The objective of the present work is to investigate the effects of combustion on spray and
spray flames of bio-oil for which little experimental data are available. Experiments are
conducted using unheated VO as well as heated VO at 100°C. Measurements are obtained for
both in open flame and an enclosed flame to simulate realistic gas turbine conditions. The mean
axial and RMS velocities, SMD and drop size distribution data are acquired. The primary focus
is placed on liquid fuel spray characteristics, and their effects on emissions. It can be envisioned
that smaller droplets in the spray would lead to premixed combustion and hence lower emissions.
The inferences from this study would aid in designing future liquid fuel combustors. Hereafter
the details of the experimental set-up, results and discussion, and conclusions are presented in
this chapter.
5.2. Experimental Set-up and Procedure
The experimental set-up is shown schematically in Fig. 5.3. It consists of the combustion,
fuel injection, and flare systems as discussed below. Phase Doppler Particle Analyzer (PDPA)
mounted on a 3-way traversing system was used to acquire quantitative spray measurements in
the cold flow as well as flame. In addition, emissions data were acquired by a continuous
sampling probe and an infrared camera was used to record the combustor wall temperature.
5.2.1 Combustion System
Combustion system was housed within a test chamber of dimensions 63.5 cm by 63.5 cm
by 1.2192 m. The primary air enters the combustion system through a plenum filled with marbles
to breakdown the large vortical structures. The air passes through a swirler into the mixing
section, where the gaseous fuel is supplied during the startup. The reactants or combustion-air
102
enter the combustor through a swirler section shown schematically in Fig. 5.3.The swirler is used
to enhance fuel-air mixing and it also helps to stabilize the flame. The swirler has six vanes
positioned at 280 to the horizontal to produce swirl number of about 1.5. The bulk axial inlet
velocity of the primary air is 1.9 to 2.1 m/s, which resulted in Reynolds number varying from
5960 to 6750. The combustor is enclosed within a 15 cm inside diameter, 46 cm long pentagonal
enclosure, with two sides of quartz glass and remaining three sides of metal plates.
Table 5.1
Properties and Characteristics of the Insulating Material [www.zircarzirconia.com]
Nominal Composition, Wt%
Buster Blanket
Al2O3
97
SiO2
3
Organics
<3
Trace Elements
<0.5
Melting Point, ºC (ºF)
2038 (3700)
Maximum Use Temperature, ºC (ºF)
1600 (2912)
W/mK (BTU/hr ft2 oF/inch) at 1600oC (2912oF)
0.476 (3.3)
Bulk Density, g/cc
<4
In order to have optical access and to be able to measure the spray data, a flat sided
pentagonal enclosure was designed and fabricated in house. The photographic view of the
enclosure with out and with insulation and the schematic of the top view of the pentagonal
enclosure are shown in figure 5.4. Transparent quartz glass windows of 8.0 cm by 46.0 cm by
3mm thick, grounded and polished to optical quality were used for the enclosure to allow optical
103
access through the enclosure. The enclosure was insulated with 4 layers of insulation to
minimize the heat loss from the flame. Alumina Mat type Buster Blanket insulation (1” standard
thickness) from Zircar with properties listed in Table 5.1 was used to insulate the enclosure.
Buster blanket is a flexible, Hi-Alpha fiber insulation material. It is needled into a durable
blanket with the addition of organic fiber reinforcement with temperature up to 1600oC. These
polycrystalline fiber blankets offer higher temperature capability, less shrinkage and greater
chemical resistance than standard alumina-silica blankets. The insulation helped retaining heat
inside the combustion area and hence improves flame stability. For optical access a small
window of 10.16 cm by 10.16 cm was cut on the insulation on two glass windows at the bottom
of the enclosure as observed in the photographic image of figure 5.4.
Air for combustion and atomization was supplied by an air compressor. The air passed
through a pressure regulator, and a dehumidifier and water traps to remove the moisture. Then,
the air was split into combustion air supply and atomizing air supply lines. The combustion air
flow rate was measured by a laminar flow element (LFE) with reported calibration error of ±5
lpm. The pressure drop across the LFE was measured by a differential pressure transducer. An
absolute pressure transducer was used to measure air pressure in the LFE. The atomizing airflow
rate was measured by calibrated mass flow controller from Sierra (Model 810S-M., Mass-Track
Flow Controller, 15-100 SLPM), with measurement uncertainty of ±0.5 lpm.
5.2.2 Fuel Injection System
The fuel injector system runs through the plenum and the mixing chamber, as shown in
figure 5.3. An O-ring within a sleeve is located at the bottom of the plenum to prevent air
leakage. A commercial air-blast atomizer (Delavan Siphon type SNA 0.20 nozzle) with its
details shown schematically and photographically in Fig. 5.5 was used for the experiments. The
104
injector creates a swirling flow of atomizing air to breakdown the fuel jet as the fluids exit the
injector. The inside diameter of the fuel supply tube is about 0.4 mm.
The liquid fuel supplied by a peristaltic pump passes through a pulse dampener, a fuel
filter and an electric fuel heater (Infinity Fluids Corporations, CRES-ILB-12/24 inline
water/liquid heater) as shown in figure 5.6. The heater uses a Proportional Derivative Controller
(PID) control unit to control the fuel temperature within accuracy of ± 0.5 oC, measured at the
heater outlet by a K-type thermocouple. Heated fuel enters fuel injector through 5 cm long tube
of 5 mm ID to minimize the heat loss. The fuel temperature at the injector inlet was measured by
a K- Type thermocouple.
5.2.3 Flare system
To burn the fuel droplets before they enter the exhaust system was a major challenge
while conducting cold flow measurements. Thus a flare system that could successfully combust
the liquid fuel upstream of the exhaust duct was designed in house. Several different approaches
were attempted to create a flare system to burn most of the fuel droplets upstream of the exhaust
duct while also avoiding bellowing of the spray in the operating area. Figure 5.8 shows the
photographic view of the final flare system used with open cold spray. Spray measurements in
the cold flow were acquired with and without swirling flow of combustion air. The flare system
consists of 3 diffusion torches with co flowing methane and air streams as shown in figure 5.8.
The flare system was located at about 60.96 cm downstream of the open cold spray. The methane
flow rate through the flare system was kept to about 40 SLPM while the air flow rate was about
300 SLPM. As seen from the photographic image, the open cold spray is encompassed in the
long flame from the flare creating a combustible mixture with most of the fuel droplets burning
upstream of the exhaust duct as shown in figure 5.8. Sufficient air was entrained to burn the fuel
105
droplets using the co-flow air. The flares ensured that the fuel spray mixture passed through a
sufficiently long flame so that enough residence time and high flame temperature would ensure
proper droplet burning of the spray upstream of the exhaust duct.
5.2.4 Emissions Measurement System
The product gas was sampled continuously by a quartz probe (OD = 7.0 mm) attached to
a three-way manual traversing system. The upstream tip of the probe was tapered to 1 mm ID to
quench reactions inside the probe. The probe was traversed in the axial direction at the center of
the combustor and in the radial direction at the combustor exit plane. The analyzer consist of a
built-in sample pump which draws the flue gases through the stainless steel probe, pre-cooler,
condensate trap and filter, flow meter, internal secondary filter, Teflon liquid blocker and the
sample collecting sensors as the basic built in functions. Water traps mounted upstream of the
emission analyzer prevent the sensors from the contamination. The dry sample was sent through
the electrochemical analyzers to measure the concentrations of CO and NOx in ppm. As shown
in figure 5.9, a NOVA gas analyzer was used to measure the concentrations of CO and NOx in
the exhaust gas sample. The emission concentrations are measured using electrochemical
sensors. Electrochemical ‘fuel cell’ type sensors produce a small electrical output proportional to
the gas being detected. The output signal is amplified and displayed on LCD digital panel
meters. The emission analyzer measures the concentrations of CO in the range of 0 to 2000 ppm
and NOx in the range of 0 to 1000 ppm. The analyzer also measured oxygen and carbon dioxide
concentrations, which were used to cross-check the equivalence ratio computed from the
measured fuel and air flow rates. The uncorrected emissions data on dry basis are reported with
uncertainty of +/-2 ppm.
106
5.2.5 Phase Doppler Particle Analyzer (PDPA) Set-up
A 2D Phase Doppler Particle Analyzer (PDPA) was used for flow velocity and drop size
measurements. PDPA is a point sampling device based on the light scattering interferometery
principle. The laser beams from the transmitter probe intersect to form a sample measurement
volume. Principally when a particle or drop passes through the beam intersection region, the
scattered light forms a fringe pattern. Since the droplet is moving, the fringe pattern sweeps past
the receiver aperture at the Doppler difference frequency, which is directly proportional to the
drop velocity. The droplet size is measured by the phase shift of the light encoded in the spatial
variation of fringes reaching two detectors after travelling paths of different lengths through the
droplets. The phase shift is measured by the two detectors, each focused at a spatially distinct
portion of the receiver lens. The spatial frequency of the fringe pattern is inversely proportional
to the drop diameter. The phase shift between the Doppler burst signals from different detectors
is proportional to the diameter of the spherical shaped droplet. There is no calibration required
for the PDPA method since the drop size and the droplet velocity are dependant only on laser
wavelength and optical configuration.
The schematic of the 2D PDPA system is shown in Figure 5.10. The laser beam from a 2W water cooled argon-ion laser is separated into a pair of 514.5 nm green beams and a pair of
488 nm blue beams using a beam separator assembly. One beam of each pair is shifted by a 40
MHz Bragg cell. Next, each beam is focused onto a fiber optic cable to deliver the beams to a
750 mm focal length PDPA transmitter. In order to be able to take flame spray measurements
using PDPA, the lenses were replaced to the focal length of 750 mm (29 inches) for both
transmitter and receiver probes so that the optical probes are kept away from the flame.
Precisions Achromatic Doublet lenses from Newport Optics were used. The transmitter lens, of
107
50.8 mm diameter, 750 mm focal length (Model No. PAC086) and the receiver lens of 76.2 mm
diameter, 750 mm focal length (Model No. PAC046) are mounted on the respective probes. The
PDPA receiver is set at an angle of 144 degrees from the transmitter to collect the refracted light
intensities from the spray. The detected signal is acquired by a data acquisition system, and
analyzed using the TSI Flow Sizer software to obtain mean and root mean square (RMS)
velocities, Sauter mean diameter (SMD), and drop size distribution data. The photographic view
of the PDPA system integrated with the combustion system is shown in figure 5.13.
5.2.6 Traversing System
Measurements were obtained by moving the PDPA system using a three-way traversing
system, while the combustor was kept stationary. The PDPA system was traversed to acquire
radial profiles at axial planes between Y = 5 mm and 75 mm, in 5mm intervals for the cold spray
as well as from Y = 5 mm and 40 mm, in 5 mm intervals for the flame spray measurements. The
3-way traversing system (Model No. MT10-0100-M02-31), from Velmex Inc., was used to
mount the PDPA system. The schematic of the system is shown in figure 5.11. The plan view of
the PDPA traversing mechanism in radial and axial direction is shown in figure 5.12. Each (x, y
and z) axis was traversed by the single shaft stepper motors Model NO. PK296-03AA-A6-3/8.
The Y-axis was traversed to measure the radial velocity and the Z-axis measured the axial
velocity. The traversing system was bolt mounted on a carbon steel table of 8.0 cm by 8.0 cm
and 82 cm high. The transmitter and the receiver probes were mounted on the rail attached to the
Z axis of the traversing system. To obtain a good signal quality, the angle between the
transmitter and receiver was maintained at 144°. To obtain this angle, both the probes were set at
18° with respective to the optical rail attached to the Z axis of the traversing system carrying the
transmitter and receiver. To simplify the measurements in radial directions, we aligned the
108
transmitter direction with respect to one of the traversing system axis, i.e. the X axis. This
demands the optical rail to be set at precisely the same angle with respect to X axis. This
arrangement of setting the optical rail at 18° facilitates to take radial location measurements
along Y axis. The center of the spray pertained to the peak location of the axial velocity profile.
Data rates of up to 30-40 kHz were obtained towards the centre of the spray. The data rate and
mean axial velocity both decreased to nearly zero as the detection volume reached the outer edge
of the spray.
5.2.7 Infrared (IR) Imaging
An infra red camera was used to measure the exterior surface temperature of the insulated
enclosure as shown in the photographic image in figure 5.14. The camera used was form Mikron
Infrared, Inc, Model no. 7200v. The IR imaging camera (MikroScan 7200 V) is a non-contact,
high sensitivity infrared radiometer. It measures the infrared radiation emitted by the target
surface and converts this radiation into a two-dimensional image related to the temperature
distribution at the target surface. The IR camera senses thermal energy that is emitted from the
target object. Through use of the software (MikroSpec), temperature variations over the area
included in the field of view can be displayed. The total radiant emission of a blackbody is given
by Stefan Boltzman equation
W = aT4.
The temperature of the blackbody can be obtained directly from the radiant energy of the
blackbody by this equation. For normal objects, the right side of above equation multiplied by
the emissivity. The data processing software (MikroSpec) allows the use r
to
view
the
temperature data of one or more points at selected locations anywhere within the field of view.
The camera was mounted on a regular tripod. Two different temperature ranges were used to
109
take IR images at different location on the enclosure. The ranges used were 0 to 500°C and 0 to
2000°C. The default emissivity value of 0.85 was used.
5.2.8 Test Conditions
Experiments were conducted for fixed VO flow rate of 12 liters per minute (lpm). The
atomizing air flow rate through the injector was kept constant at about 25 LPM which
corresponds to an ALR of 2.5. The combustion air flow rate was kept to about 125 LPM to
maintain a total air flow rate of 150 LPM throughout the experiments. As a first step the cold
open spray experiments were done with and without the swirling air to study the effect of
swirling air on the spray. Then the open flame spray data were taken which required swirling air
flow with small amounts of methane fuel to stabilize the flame. Experiments were conducted
with unheated and heated (to 100°C) VO. The PDPA system was traversed radially and axially
to acquire flow and drop size measurements at different locations in the open cold spray as well
as open flame. Next measurements were acquired in spray flame within an insulated enclosure.
Again, the VO was supplied either at room temperature or at 100°C.
The experiment was started by supplying gaseous methane and then, igniting the
methane-air reactant mixture in the combustor. Next, the liquid fuel flow rate was gradually
increased in steps of 2 mlpm to attain the desired value of 12 mlpm, while the methane flow rate
was slowly decreased to zero. In this study, the volume flow rate of total air (combustion +
atomizing) was constant at 150 standard lpm. Experiments were conducted for fixed volume
flow rate of fuel at 12 mlpm. The time required for the fuel to flow through the system and reach
the atomizer was about 30 minutes. Another 0-40 minutes were required before the fuel
temperature reached the set value.
110
Once a stable flame was achieved, the liquid fuel flow rate was gradually increased while
the methane flow rate was gradually reduced. Initially the methane fuel was supplied at 18 lpm
for about 1.5 hours and then gradually reduced to about 3.88 lpm. This procedure took about 2.5
hours before a stable VO flame at 12 mlpm was achieved without any condensation in the optical
measurement area. The equivalence ratio (φ) was maintained constant at 0.89-0.9 throughout the
experiment. The equivalence ratio in case of open flame was about 1.2 which did not account for
the ambient air entrainment. The emissions measurements were taken for the spray flame along
with the spray measurements using a 2 way velmex traversing system shown in figure 5.9 (b).
5.3 Results and Discussion
Experiments were conducted to measure the spray characteristics to demonstrate the (i)
effects of swirling air flow on open cold spray, (ii) effects of flame on open spray, and (iii)
effects of enclosure on spray flame and emissions. The effect of fuel inlet temperature is also
studied for unheated and preheated VO at 100°C. The following sections will discuss important
observations from the current study. Results are shown in the form of contour plots and profiles
of axial velocity, RMS velocity, SMD, and droplet diameter distribution. For each case, the
contour plots will be discussed first followed by the profiles plots.
5.3.1 Effect of Swirling Air Flow on Open Cold Spray
(a) Axial Velocity Contours (Mean and RMS)
Figures 5.15 and 5.16 shows the contour plots of axial velocity for cold spray with and
without swirl, respectively. The plots were constructed using measurements taken for different
axial and radial locations in the spray. The axial location was varied from 5 mm to 75 mm from
the injector exit in steps of 5 mm interval. Transverse measurements were taken in steps of 1 mm
111
across the spray. As mentioned earlier, the experimental system was kept stationary while the
PDPA system was traversed radially and axially.
In general, for the contour plots shown in both the cases, the mean axial velocity peaks in
the centre of the spray, with the velocity magnitude diminishing towards the edge of the spray.
The peak axial velocities are seen to be slightly higher in the near field axial locations from 5
mm to 20 mm in the cold spray with swirl as compared to the cold spray without swirl. Farther
downstream in the mid field (Y = 20 mm to 40 mm) and far field (Y = 45 mm to 75 mm)
locations in the spray, the peak axial velocities for cold spray without swirl are higher as
compared to the cold spray with swirl. For both cases as the flow field diverges axially
downstream, the velocity decreases with the loss of kinetic energy of the droplets. For the cold
spray without swirl, the axial velocity in the midfield and far field locations is about 12 m/s and
8 m/s higher as compared to cold spray with swirl. Figure 5.16 reveals a well-pronounced flow
recirculation zone near the outer edge of the spray between the radial locations of 10 mm and 40
mm .Note that the near field region of the spray, the values of axial velocity are negative. The
presence of swirl widens the spread of the spray and increases the velocity near the injector exit
while drastically reducing the velocity magnitude farther downstream compared to cold spray
without swirl. Near the injector exit, the effect of swirl is to create high axial velocity owing to
the increased airflow rate. However, farther downstream, the swirling air flow introduces
additional turbulent mixing to widen the radial extent of the jet spray with sufficient reduction in
axial velocity as compared to cold spray without swirl.
Figure 5.17 shows the mean axial velocity contour plot for cold spray with swirling air
flow for preheated VO at 100°C. As expected, the axial velocities show an increase in
magnitudes compared to the unheated case. This increase in the axial velocity results from the
112
heating of the atomizing air by heated VO. For preheated VO, the peak velocities at any given
axial location in the spray are higher by about 10-15 m/s as compared to the unheated VO case in
figure 5.8. With increase in fuel inlet temperature, the radial extent of the spray also decreases as
compared to the unheated case. Higher axial momentum decreases the radial diffusion of the jet
spray, resulting in a narrower spray in case of preheated fuel.
Figure 5.18, 5.19 and 5.20 shows the contour plots for RMS axial velocity for no swirl
flow, swirl flow and cold spray with swirl flow for VO at 100°C, respectively. As shown in
figure 5.15, for cold spray without swirl flow, the peak RMS axial velocity value is observed to
be of 18 m/s while for cold spray with swirl the peak RMS velocity in figure 5.19 is around 22
m/s. Both the plots in figures 5.17 and 5.19 show a central dip and peaks on either side of the jet
due indicating shear interactions between the jet and surrounding gaseous phase in the near
injector exit locations. The peak in the RMS axial velocity for cold spray with no swirl flow is
observed at X = 5 mm for Y = 5 mm to 10 mm while for cold spray with swirl flow the RMS
peaks at X = 10 mm for Y = 5 mm to 20 mm. Farther downstream for both the cases without and
with swirl flow, the RMS velocities show more uniform distribution compared to the near
injector plane locations. The RMS velocities trail off trend on the outside edge of the spray. The
higher magnitude of RMS velocity in cold spray with swirl flow indicates improved inherent
turbulent mixing due to the surrounding swirling air flow as compared to no swirl flow in figure
5.17. Figure 5.21 shows the RMS contour for cold spray with swirl for VO at 100°C. Higher fuel
inlet temperature shows a slight increase in RMS axial velocity compared to the unheated cold
spray. The increase in the RMS axial velocity results from the heated fuel raising the temperature
of the atomizing air to produce higher turbulence levels leading into smaller diameter droplets.
The peak RMS value is about 25 m/s.
113
(b) SMD Contours
The effective method of expressing the quality of the atomization is the mean diameter.
Probably, the most common form of mean droplet size is the Sauter Mean Diameter (SMD),
which has the physical interpretation as the diameter of the drop having the same volume/surface
area ratio as the entire spray. Figure 5.21 and 5.22 shows the contour plots constructed from
point-wise measurements acquired at different axial and radial locations.
Figure 5.21 and 5.22 shows respectively the contour plots of SMD for open cold spray
with and without co-flow swirling air. As seen from the plot the spray consists of the droplets
with wide range of SMD values of 20 microns to 80 microns. The plot shows the distribution of
the droplets in the various regions of the spray in the mid field (Y = 20 mm to 40 mm) and far
field (Y = 45 mm to 75 mm) locations. The larger droplets of 55 microns and above diameter
reside in the centre core region near the injector exit (Y= 5 mm to 15 mm) as well as at the outer
edge (X = ± 20 mm) of the spray for Y = 50 mm to 75 mm. The presence of larger droplets in
centre of the spray can be attributed due to higher axial momentum and weaker mixing in the jet
core region compared to the shear layer region of the spray, i.e. X = ± 5 to15 mm.
Some larger droplets are observed towards the spray edge due to their higher momentum.
At radial locations of X = ±10 to 25 mm and Y = 30 mm to 75 mm. As observer from the plot,
the smaller and intermediate size droplets reside in the shear layer (X = ± 5 -15 mm) both near
and far away from the injector exit. Droplets that escape from the centre liquid core undergo
further breakup which leads to formation of smaller droplets. However, farther downstream,
more uniform and homogenous distribution of the droplet size is observed in the radial direction.
For all the axial locations of the spray, the SMD values were observed to range from 20 microns
to 45 microns.
114
Figure 5.22, shows the SMD contour for cold spray with swirling air flow and unheated
VO. The SMD ranges from 20 to 60 microns. The radial spread in this case is seen to be about 30
mm as compared to 15 mm in case of no swirling air as shown in figure 5.19. As observed from
the plot the larger droplets are located near the jet centre, while the droplets diameter decreases
away from the jet centre. Droplet diameter in the near field location of the spray, i.e. Y = 5 mm
to 15 mm, ranges from 50 to 60 microns. For downstream spray locations the droplet diameter
gradually decreases and farther downstream of the spray at Y = 60 mm to 75 mm the SMD
decreases to about 30 to 35 microns. The droplet diameter in the shear layer, i.e. X = 10 to 20
mm, is in the range of 20 to 35 microns. It can be seen from figures 5.21 and 5.22; that the SMD
is smaller for spray with swirling air compared to spray with no swirling air. The swirling air
promotes secondary droplet breakup mechanism resulting into smaller SMD in the spray. Hence,
the droplets escaping from the central jet core undergo further disintegration into smaller
droplets ranging from 20 microns to 35 microns, due to the co flowing swirling air. This process
takes place for X = 15 mm to 45 mm for almost all the axial locations in the spray. The jet
spreads wider in case of swirling air compared to without swirling air. Thus swirling air flow
widens the spray and it also decreases the droplet diameter because of the secondary
disintegration process.
Figure 5.23 shows the SMD contour plot for cold spray with swirling air for VO
preheated to 100°C. Results show similar general trends as discussed above for figures 5.21 and
5.22. Larger droplets in the central core region and smaller droplets in the shear layer region near
and far field locations in the spray. The maximum SMD for this case however is around 40 µm
which is significantly lower than the maximum SMD of 60 µm for the unheated VO cold spray.
The difference in the results is attributed to the improved atomization due to increase in VO
115
temperature. Increase in VO temperature, improves the thermo physical properties of the VO,
e.g. reducing the kinematic viscosity and surface tension which make it easier to break-up the
droplets producing finer spray. Increasing the VO temperature however requires thermal energy
of about 0.53 % of the VO energy content (LHV = 37000 kJ/kg) to achieve the benefits of
improved spray characteristics.
(c)Transverse Profiles of Mean and RMS Axial Velocity
Figure 5.24 (a- z) shows transverse profiles of mean axial velocity for cold spray with
and without swirling air and unheated VO. The trend showing the effect of swirling air on cold
spray is explained by figure 5.24 (a) at one axial location and for the rest of the axial locations
the same trends follow. Figure 5.24 (a) shows a dome shaped velocity profile for Y = 5 mm, for
cold spray with and without swirl air flow. The mean axial velocity is highest at the center and it
decreases towards the periphery of the spray. The peak axial velocity is similar for both the
cases. But the general trend also observed in other profiles of figure 5.24 is that in the near field
region, the cold spray with swirling air flow shows marginally higher axial velocity, which
decreases at farther downstream locations. The peak axial velocity near the injector exit region of
the spray with swirling air is about ~5-8 m/s higher compared to non swirling air spray. From
figures 5.24 (a, b, c and d), we can observe a well pronounced recirculation zone in case of
swirling air flow, characterized by negative velocities on either side of the spray centre. The
recirculation flow improves mixing of the fuel and air to further improve atomization. Swirling
air flow is also the cause of the further disintegration the larger droplets. The swirling air flow
tends to increase the radial spread of the spray compared to that for the non swirling air flow.
The recirculation zone can be identified from the dotted line in the plot and it can be seen at X =
10mm to 40mm for Y = 5 mm to 20 mm.
116
Further downstream, the peak axial velocity decreases for each case with non swirl air
flow resulting in higher velocities compared to the swirl air flow case. With increased axial
distance, (Y = 30 mm to 75 mm), peak axial velocities show an increase for the non swirl flow
compared to the swirl flow. This effect is attributed to the intense mixing and radial spread of the
spray with swirling airflow. While the non-swirling air spray maintains the high axial momentum
because of poor mixing with the ambient air.
Figure 5.25 shows the effect of fuel temperature on cold spray. The figure 5.25 (a) shows
the axial velocity at Y = 5 mm, which shows a slight increase in peak axial velocity for VO at
100°C mostly in the region near the injector exit in the spray. The peak velocities of about 45
m/s, 34 m/s and 33 m/s are observed at respectively Y =5 mm, 10 mm and 15 mm in the near
field region of the spray. A similar trend of mean axial velocity profile is observed for both the
cases at all axial locations in the spray. Only a slight increase in mean axial velocity with heated
VO is observed in the near and far field regions of the spray. The effect of fuel inlet temperature
is more pronounced to reduce the droplet diameter which can be seen from the SMD profile
plots.
For all of the aforementioned test conditions, RMS axial velocities are shown in figure
5.24 (a-z). The ratio of RMS over the mean axial velocity represents the turbulence intensity.
Figure 5.24 (b), shows the RMS axial velocity at Y = 5 mm for cold spray with and without
swirl. In general the profiles for both the cases show a dip in the centre of the spray, which is
also observed at Y = 10 mm, 15 mm and 20 mm. This result is attributed to the intense turbulent
fluctuations in the shear region of the spray where the respective gradients reach the highest
values to dominate the turbulence generation rate. Further downstream the turbulence level
begins to decay owing to the diffusion and reduction in kinetic energy of the droplets. In
117
particular, turbulent kinetic energy attenuation is significant in the shear layer because of the
additional dissipation introduced by the presence of the droplets. As seen in the figure 5.24 (a),
the magnitude of the RMS axial velocity values for the swirl air flow as well as non swirl flow is
almost overlapping. This result is consistent with the higher axial mean velocities as discussed
previously in the near field regions of the spray. A local maximum is seen at radial location X =
5 mm, indicating that the faster and smaller drops have higher RMS axial velocity values. For Y
= 30 mm to 75 mm, the dip in the RMS axial velocity is no longer observed and the profiles are
more uniform indicating shear layer merging with the centerline at these downstream locations in
the spray . The RMS axial velocity values for Y = 30 mm, 35 mm , 40 mm are about 10 m/s, 8
m/s and 7 m/s for both the cases. Further downstream at Y = 60 mm, 70 mm and 75 mm, the
RMS axial velocity of about 3 to 5 m/s are similar for swirl air flow and non swirl air flow
respectively.
(d)Transverse Profiles of SMD
Figure 5.26 shows the radial profiles of SMD at different axial locations in the spray to
demonstrate the effect of swirl air flow on the cold spray. Figure 5.26 (a) shows the SMD profile
for cold spray with and without swirl air flow at of Y = 5 mm. In general, the SMD values are
seen to be higher at the centerline with a diminishing trend of smaller droplets towards the edge
of the spray. For the near field location in the spray i.e. Y = 5mm, SMD values at the centre is
around 75 µm. Moving away from the centre, the droplet diameter decreases. A slight difference
is observed in the SMD values for cold spray with and without swirling air, the latter showing
smaller droplets. The profile for cold spray with swirl is tapered radially at X = 20 mm, since the
swirling co flow air tends to spread the spray and further disintegrate the droplets to form smaller
drops. The radial extent of the spray is wider for swirl air flow case compared to the without
118
swirl case, i.e. 55 mm for swirl case and 30 mm for non swirl spray. SMD values at Y = 5 mm
range from 25 µm to 72µm for cold spray with swirl air flow, while that of cold flow without
swirl air flow ranges from 40 µm to 80 µm. Similar trend is observed for other near field
locations in the spray i.e. at Y = 10 mm, 15 mm and 20 mm. The maximum SMD is nearly the
same at Y = 10 mm, 15 mm and 20 mm with and without swirl air flow. However, the magnitude
of the SMD decreases in the axial direction.
The noticeable difference in the droplet diameter is observed from Y = 30 mm to 75 mm
in both the cases. The co flow swirling spray decreases the SMD by about 25 µm as compared to
the non swirling spray. For co flow swirling spray there is a decrease of about 25 microns in
SMD for Y = 30 mm to 75 mm. In the same region the SMD for the non swirling spray decreases
from 57 microns to 45 microns. Thus the swirl air flow in the cold spray decreases the droplet
diameter which would improve atomization and combustion.
Figure 5.27 shows the effect of fuel inlet temperature on radial profiles for SMD in the
spray. With the preheated VO at 100°C, the maximum SMD decreases to 40 microns form 70
microns at Y = 5 mm. The range of SMD at this axial location is about 17 microns as compared
to the wider range of 38 microns in case of unheated VO spray. The radial extent with the
smaller droplets is greater in case of preheated VO spray compared to the unheated VO spray.
Increase in fuel inlet temperature produces a spray with droplet diameter ranging from 18
microns to 40 microns; whereas the droplet diameter range in the unheated fuel spray is from 26
microns to 70 microns. Evidently the effect of fuel inlet temperature is seen at almost all axial
locations in the spray ranging from near field to far field locations .The higher fuel inlet
temperature decreases the fuel kinematic viscosity and surface tension, which aids in improving
the atomization characteristics to produce finer droplets. Increased fuel inlet temperature
119
decreases the SMD and its range as well as the radial spread of the spray as seen from the mean
and RMS axial velocity profiles.
(e) Droplet Diameter Distribution Profiles
To better understand the spray structure, figures 5.28 compares the droplet size
distribution profiles at the centre of the injector exit. Figure 5.28 (a) shows the droplet size
distribution profiles at the Y = 5mm, and X = 0 mm. The SMD is 64 µm and 76 µm, respectively
for cold spray with swirling air and without swirling air. As seen from the plot, cold flow with
swirling air shows a higher percentage of smaller diameter droplets in the range of < 50 µm as
compared to the non swirling spray. The largest droplet diameter is about 200 µm for both
swirling and non swirling spray, although very few larger diameter droplets are observed in case
of the swirling spray. For swirling spray, the SMD distribution shows most of the droplets with
diameter of about 100 µm or less while for non swirling spray the distribution spread out with a
higher percentage of larger diameter droplets. The greater number of smaller diameter resulted in
the smaller SMD in case of swirling spray as compared to non swirling spray.
Figure 5.28 (b) shows the distribution profiles at Y = 40 mm, X = 0 mm which is
considered here as the mid field location in the spray. Close to 90% of the droplets in case of
swirling spray are smaller than 100 µm. The largest diameter in non swirling spray is about 150
µm. A significant difference in the distribution count is seen for droplets of diameter < 35µm,
with swirling flow showing higher percentage compared to the non swirling flow. The presence
of these larger droplets in case of non swirling spray is the cause of higher SMD of about 46 µm
compared to that of 36 µm in case of swirling spray. Comparatively, the near field region of the
spray shows greater number of larger droplets than at the axial locations away from the injector
exit.
120
Figure 5.28 (c) shows the distribution profile for the swirling and non swirling spray at Y
= 60 mm, X = 0 mm, i.e. at a far field location in the spray. The profiles for both the cases are
very similar, still showing greater number of larger droplets in case of non swirling spray. The
highest droplet diameter in case of non swirling spray is 150 µm as compared to that of 80 µm
for swirling flow. The presence of these larger droplets dominates the SMD in case of non
swirling spray showing higher overall SMD’s compared to the cold spray with swirling air flow.
The SMD’s are 36 µm and 44 µm for swirling spray and non swirling spray, respectively. As
explained earlier, the droplet distribution is narrower with greater number of smaller drops at far
field regions of the spray than that observed at the near field locations, which is consistent with
the of higher SMD in the near the injector exit at the centre of the spray and smaller SMD father
downstream. Close to the injector exit the spray is dense and due to the presence of various
droplet sizes the distribution becomes rather wider displaying discrete maximum that that father
downstream centre of the spray.
5.3.2 Effect of Flame on Open Spray
(a) Velocity Contours (Mean and RMS)
Figure 5.29 and 5.30 shows respectively the mean axial velocity contour for spray flame
of unheated and preheated VO. The measurements were acquired for Y = 5 mm to 40 mm in the
flame spray. The measurements in the flame spray downstream of this axial location were
exceedingly difficult because of the high intensity flame radiation and extremely low data rate.
For the PDPA system, the signal to noise ratio was too small to obtain any accurate information
at low data rates. In general the mean axial velocity exhibits a peak at the centre and decays
gradually in radial direction, similar to the cold spray results discussed above. However, for the
spray flame the peak velocities are higher because of the increased temperature resulting from
121
heat release. Unlike, cold spray with swirl flow, the negative velocities for the droplets is not
measured in flame spray, since they evaporate completely
The mean axial velocity increases drastically in the flame spray compared to the cold
spray. The difference of about 35-40 m/s is observed in the peak mean axial velocity for both the
cases. The rapidly moving smaller drops in the mid spray region are attributed to the smaller
drag drop forces and volumetric expansion of the hot gases. Slow moving larger drops in the
outer edge of the spray result from high momentum of such drops. Similar trend is observed for
the flame spray for VO at 100°C as well. The peak mean axial velocities show a marginal
difference for unheated and heated VO. Results show that fuel preheating increases mean axial
velocity leading to improved atomization. Similar improvement is also observed in the
combusting sprays compared to the non-burning cold sprays.
(b) SMD Contours
The SMD contour plot for open flame with unheated VO is presented in figure 5.33. The
larger droplets are seen close to the injector exit between Y = 5 and 10 mm, and also on the outer
edge at farther downstream locations. The magnitude of the SMD ranges from about 6 µm to 37
µm. The maximum SMD of 37 µm in case of flame spray is significantly smaller compared to
that of 60 µm for cold spray as seen in figure 5.22. The decrease in the SMD is seen in the near
field i.e. Y = 5 mm to 10 mm. The decrease in the droplet size can be attributed to the droplets
passing through the flame zone. The high temperatures in the flame tend to reduce the droplet
diameter throughout the spray including the core region near the jet centre where larger droplet
sizes are observed in the cold spray. The SMD range in case of flame is narrower than the cold
spray indicating high evaporation rate for the former case .Farther downstream of the injector
exit, SMD decreases rapidly indicating faster evaporation of droplets in the flame. Large droplets
122
do not evaporate completely and thus they are still able to migrate towards the outer edge of the
spray.
Figure 5.34 shows the SMD contour of an open flame for preheated VO at 100°C. In this
case the maximum SMD decreases to about 30 µm compared to 37 µm for unheated VO. The
flame tends to reduce the radial jet spread and produces a narrower range of SMD indicating
droplet evaporation associated with high flame and fuel temperatures.
(c) Transverse Profiles of Mean and RMS Axial Velocity
Figure 5.35 shows the mean axial velocity profiles in the cold spray and spray flame. The
measurements were taken for axial locations of 5 mm to 40 mm in the flame, because of the low
data rate at Y > 40 mm, indicating nearly complete fuel vaporization at these locations. In
general, the mean axial velocity peaks at the centre and decreases towards the edge of the spray,
similar to the trends in the cold spray discussed above. Figure 5.35 (a) compares the mean axial
velocity for cold spray and flame spray at Y = 5 mm. The flame spray shows peak velocity of 70
m/s as compared to that of 45 m/s in the cold spray. Thermal expansion in the flame results in
droplets with higher mean axial velocity. At Y = 30 mm, the peak axial velocity decreases to
about 52 m/s as compared to peak axial velocity of 14 m/s for cold spray. The decrease in the
axial velocity is attributed to the spray extending over a wider region at downstream locations.
Figure 5.35 (h) shows that at Y = 40 mm, the peak axial velocity in the flame spray decreased to
about 38 m/s compared to that of 10 m/s in the cold spray .
Figure 5.36 shows the effect of flame on spray for VO at 100°C. The observed trend for
the axial velocities is similar to as discussed above. The mean axial velocity at Y = 5 mm, peaks
to 75 m/s in the flame spray as compared to peak value of 48 m/s in the cold spray. At Y = 10
mm and Y = 15 mm, the profiles show a dip in the jet centre while peak velocity on either side of
123
centerline at X = 10 mm. Farther downstream in the spray at Y = 30 mm and Y = 40 mm, the
flame spray velocity peaks at 58 m/s and 50 m/s compared to 20 m/s and 15 m/s in case of cold
spray.
Figure 5.35 shows that the RMS axial velocity is observed in the flame spray as well, as
discussed above for cold spray. The flame spray profiles show a dip in the centre and peaks on
either side of the centerline which represent the flame locations at X = 15 to 28 mm, unlike the
cold spray having a decaying trend with peak at the centre. This is attributed due to the presence
of higher turbulent fluctuations in case of the flame spray as compared to that of the cold spray.
The magnitude of RMS axial velocity ranges from 10 m/s to 22 m/s which is consistent to the
mean axial velocity discussed in the previous sections. The maximum RMS axial velocity values
of about 22m/s, 18 m/s and 16 m/s are observed at Y = 5 mm, 10 mm and 15 mm, respectively.
For the flame spray, the RMS axial velocity is higher than those of the cold spray on the outer
edge of the spray while they are almost similar in magnitude in the center of the spray.
Figure 5.36 shows the similar trend for flame spray for preheated VO at 100°C. At the
near field locations, the profiles peak nearly at same values for both flame spray and cold spray
at Y = 5 mm, 10 mm and 15 mm. The flame sprays profiles show a slight asymmetry at the
centre which can be attributed due to the asymmetry in the flame. The magnitude in the RMS
axial velocity values varies from 5 m/s to 20 m/s for the axial locations of Y = 5 mm to 40 mm in
the flame spray. As discussed above the RMS values are smaller in case of cold spray compared
to that of flame.
(d) Transverse Profiles of SMD
Transverse profiles of SMD for flame spray and cold spray are presented in figure 5.37.
Figure 5.37 (a), shows the effect of flame on spray at axial location of Y = 5 mm. For the flame
124
spray the profiles of SMD show smaller droplets in the centre and slightly larger droplets on each
side, and gradual decrease in droplet diameter in the transverse direction. This trend is same for
near field axial locations of Y = 5 mm, 10 mm, 15 mm and 20 mm. The maximum SMD for the
near field location in the flame spray is observed to be about35 µm at Y = 5 mm, 25 µm at 10
mm and 15 µm at 15 mm. Farther downstream, a noticeable change in the SMD trend is
observed; smaller droplets are at the centre while the larger droplets on the outer edge. This
result is attributed to the fact that the droplets at these locations are passing through the flame
zone. The size difference farther away from the centerline is observed to be larger because at the
outer edge of the spray the smaller droplets evaporate faster compared to the large droplets and
the larger droplets with high inertia penetrate out of the flame. The escaped larger droplets are
believed to burn in diffusion mode. Comparatively the SMD range for the burning spray is about
35 µm smaller than that of the cold spray for near field locations. The presence of flame
increases the evaporation rate of the droplets hence producing smaller droplets. The radial spread
for the flame spray is narrower compared to cold spray indicating that the droplets in case of cold
spray possess higher initial momentum hence penetrating farther and wider.
Figure 5.38 shows the radial profiles for SMD for fuel inlet temperature of 100°C for
flame and cold spray. Similar trend is observed for all the cases in figure 5.37 as discussed
above. The effect of higher fuel inlet temperature is to decrease the fuel kinematic viscosity and
surface tension, hence improving atomization characteristics by decreasing the mean droplet
size. For cold spray the maximum SMD is 45 µm compared to 30 µm for the flame spray at Y =
5 mm. The SMD values decreases farther downstream in the spray compared to the near field
locations. And this trend decreases at farther downstream locations in the spray. Evidently the
higher fuel inlet temperature helped burning of the droplets much faster compared to unheated
125
case discussed in figure 5.37. The minimum SMD measured for the flame spray is reduced to
about 10 µm compared to 30 µm in the cold spray. The minimum for unheated flame spray case
was about 20 µm. Also, the axial locations in the flame spray were measured to only about 40
mm, since most of the droplets were evaporated by the time they reached farther downstream.
Hence the data rate was significantly low to measure any signal from the droplets above this
axial locations in the spray.
(e) Droplet Diameter Distribution Profiles
Figure 5.39 shows the droplet diameter distribution profiles for flame spray and cold
spray at near and far field locations of Y = 5 mm, 10 mm, 35 mm and 40 mm and X = 0 mm
.Figure 5.39 (a) shows the distribution profile for near the injector exit in the centre of the spray
at Y = 5 mm. Results show a significant difference in the droplet diameter distribution for flame
and cold spray. The flame spray is observed to have most of the droplets of ≤50 µm while large
droplets with > 100 µm are found in the cold spray. The largest droplet in the flame spray is
about 100 µm where as for the cold spray it is about 200 µm. For the flame spray the droplets
tend to start vaporizing due to the high flame temperature zones, impact of phase interactions and
combustion producing smaller droplets and narrower distribution compared to the wider spread
distribution as observed in the cold spray. The SMD for flame spray and cold spray is
respectively 26 µm and 64 µm, indicating rapid evaporation of the droplets in the flame
compared to the cold spray.
Figure 5.39 (b, and c) show respectively the droplet diameter distribution profiles for Y =
10 mm and 35 mm and X = 0 mm respectively. At Y = 10 mm, X = 0 mm, the flame spray has
most of the droplets of diameter <30 µm while the cold spray has majority of droplets 130 µm
diameter. The largest diameter for flame spray is 75 µ m while that of cold spray is 175 µm. The
126
SMD is 19 µm and 50 µm, respectively, for flame spray and cold spray. The higher count of
larger droplets dominates the higher SMD for cold spray. The larger numbers of smaller droplets
in case of flame spray result in a decrease in SMD. At Y = 35 mm as shown in figure 5.39 (c),
the largest droplet diameter in cold spray is 125 µm where as that in the flame spray is 60 µm.
The respective SMD for these cases is 17 µm and 38 µm for flame spray and cold spray. Again
there is a higher count of smaller droplets in case of flame spray resulting in the overall decrease
in the SMD. As observed the distribution gets narrower with increase in axial distance, due to the
droplets getting evaporated rapidly and combusted in the flame locations and leading to smaller
number of larger droplets. The distribution profile is narrower in for far field locations in the
spray for both the cases compared to the locations near the nozzle exit. Moreover the number
density of the droplet further downstream decreases in the flame spray due to the droplet
evaporation and burning.
Figure 5.40 shows the droplet diameter distribution profiles for the flame spray for
preheated VO at 100°C. The largest droplet diameter for cold spray is observed to be 140 µm
where as for flame spray it is 90 µm. The increase in fuel inlet temperature produces a narrow
distribution profile with lower range of droplet diameter and hence with overall smaller SMD.
The SMD’s for both the cases is 28 µm and 37 µm for flame spray and cold spray respectively.
At Y = 10 mm, X = 0 mm the largest SMD for flame spray is of about 75 µm as well as for cold
spray is of about 115 µm respectively. This decreases the global SMD to 22 µm and 33 µm for
flame spray and cold spray. The increase in fuel inlet temperature aids to further reduce the SMD
as compared to the unheated case. As we move axially farther downstream, the droplets
distribution becomes narrower and the number density also reduces especially significantly for
the flame spray attributed to the fact that the droplets get vaporized in the flame zone. Also, the
127
preheating of the oil produces the finer droplets to begin with since it improves the thermo
physical properties of the fuel itself, and hence the finer droplets get vaporized and consumed
faster.
Farther downstream at Y = 35 mm, the largest droplet diameter is of about 100 µm for
cold spray and 40 µm for flame spray producing the SMD of 31 µm and 13 µm respectively.
Similar to the previous discussions the far field region in the spray has narrower distribution with
smaller droplet number density compared to that of the centre of the spray. This is consistent to
that of the SMD profiles with dense centre liquid jet near the injector exit showing higher SMD
values and decreasing trend as we move axially downstream with a wider and more dispersed
spray. Also the fact that the droplets in case of flame spray get consumed faster and get
evaporated rapidly, mostly the smaller ones compared to the larger drops and hence globally
being responsible for smaller SMD values.
5.3.3 Effect of Enclosure on Spray Flame and Emissions
Experiments were done to study the effect enclosure on flame spray characteristics and
combustion emissions. The enclosure was insulated to provide thermal feedback to the VO flame
and maintain a stable VO flame during all experimental conditions. It required about 2 hours to
preheat the enclosure with a stable VO flame before the methane flow rate was reduced to about
3.8 SLPM which corresponds to φ = 0.89. This mode of operation resulted in no condensation of
the droplets on the glass window of the enclosure. Practically in the continuous combustion
operations of gas turbine applications the enclosed flames are used, and hence the present work
also incorporates enclosure effect on the spray characteristics and emissions. At steady state
insulation around the enclosure, retained sufficient heat within the system to raise the VO
temperature at the injector exit to 90-100°C
128
(a) Transverse Profiles of Mean and RMS Axial Velocity
Figure 5.41 shows the mean axial velocity profiles for enclosed flames of VO preheated
to 100°C and 150 C. As seen in figure 5.41 (a), at Y = 5 mm, mean axial velocity shows a peak
at the centre and gradual decrease towards the edge of the spray. The peak mean axial velocity
for both cases is nearly the same. For both cases, the axial velocity peaks at 70 m/s, and both the
profiles almost overlap each other, with slight difference towards the edge of the spray. At Y =
10 mm and 15 mm the profiles are again observed to be overlapping, although at higher fuel inlet
temperature the spread of the spray is observed to be narrower as compared to VO at 100°C. The
transverse spray for VO at 100°C is extended widely from -40 mm to 30 mm while for VO at
150°C, the spread narrows down to a -15 mm 20 mm for Y = 5, 10 ,and 15 mm axially. Farther
downstream, the peak axial velocity decreased down to 60 m/s for VO at 150°C and 50 m/s for
VO at 100°C. Unlike velocities, a significant difference is observed in the droplet sizes as shown
in the SMD profiles for the enclosed flame. In enclosure, the co flow swirling air is the
combustion air for the flame without any ambient entrainment. In open flame there is no heat
feedback but there is a lot of ambient entrainment, while in enclosure heat feedback is there due
to insulation.
Figure 5.41 (b) shows the RMS axial velocity profiles for enclosed flame. For VO at
100°C at Y = 5 mm, the peak RMS axial velocity is 22 m/s while for VO at 150°C the peak RMS
axial velocity is 20 m/s. For VO at 100°C, the profiles show a local minima at the jet centre line
while double peaks at flame locations i.e. X = -30 mm to 15 mm, and Y = 10 mm to 35 mm. The
magnitude of the RMS axial velocity range from about 2 m/s to 22 m/s for Y = 5 mm to Y = 35
mm. Slight asymmetry in the RMS axial velocity profiles is observed at far field locations in the
spray i.e. Y = 25 mm to 35 mm. Comparatively for VO at 150°C, the RMS axial velocity profiles
129
looks more symmetrical with local minima dip in the centerline as well as double peaks on the
flame locations. This result is attributed to the fact that the jet centre is subjected to the small
turbulent fluctuations with lower RMS values compared to that at the edge of the spray. Near the
injector exit i.e. at Y = 5 mm to 15 mm, the peak RMS axial velocity for both the cases overlaps
with no significant difference in the velocity values. At locations farther downstream, the peak
RMS axial velocities for 150ºC are higher at X = 20 mm, indicating higher velocity fluctuations
compared to VO at 100°C.
(b) Transverse Profiles of SMD
Figure 5.42 shows the transverse profiles of SMD for enclosed VO flame at 100°c and
150°C. Figure 5.42 (a) shows the SMD profile at Y = 5 mm. The profiles show peak SMD value
at the centre of the spray. The maximum SMD for VO at 100°C is 34µm, while for VO at 150°C
is 32 µm. At Y = 15 mm and 20 mm, the range of SMD values decreases. The profiles no longer
show a well-defined central peak as compared to near the injector exit. SMD values range from
10 µm to 25 µm for VO at 100°C while for VO at 150°C the SMD ranges from 18 µm to 28 µm.
Farther downstream, at Y = 25 mm to 35 mm, the larger droplets are observed to have moved
towards the periphery of the spray causing higher SMDs at the outer edge of the spray as seen in
figurer 5.44 (e and f).The profiles show a central depression as compared to central peak
observed in the near injector locations (Y = 5 and 10 mm). For VO at 100°C, the SMD ranges
from 16 µm to 25 µm, which is larger compared to 18 µm to 22 µm for VO at 150°C. For VO at
150°C, the transverse distribution of SMD is narrower compared to VO at 100°C. This result is
attributed to the higher fuel inlet temperature and improved vaporization of the droplets thus
reducing the SMD at downstream locations in the flame. At Y = 35 mm, the SMD profile for VO
at 100°C by 5 µm compared to that for VO at 150°C. For VO at 100°C, the SMD values show a
130
decrease of about 12 µm with axial distance while that for VO at 150°C, the SMD decreases by
about 15 µm.
(c) Droplet Diameter Distribution Profiles
Figure 5.43 shows the droplet size distribution profiles for enclosed VO flame at 100 °C
and 150 °C respectively. At Y = 5 mm, X = 0 mm, the largest diameter for VO 100 °C is 150 µm
while that for VO at 150 °C is 60 µm. For VO at 150°C, a narrow distribution profile is observed
with greater number of smaller droplets. 90 % of the droplet sizes are in the range of < 50 µm.
The presence of few large drops highly influences the SMD in case of lower fuel inlet
temperature, i.e. VO at 100 °C. The SMD for VO at 150°C is 28 µm while for VO at 100°C is 32
µm. At Y = 10 mm, X = 0 mm, the distribution profile is almost similar for the drops < 30 µm .A
minor difference in droplet diameter distribution is observed in the droplets > 50 µm. The lower
fuel inlet temperature i.e. VO at 100°C, has more large droplets compared to that for VO at 150
°C. The SMD for VO at 100°C is 27 µm and for VO at 150°C is 25 µm. Largest diameter for VO
at 150 °C is of about 60 µm and for VO at 100 °C is of about 80 µm respectively.
At Y = 35 mm, X = 0 mm, VO at 150 °C shows higher count of smaller droplets for
drops < 11 µm compared to that of VO T 100 °C. The largest droplet for VO 150 °C is 70 µm
and for VO 100 °C 58 µm respectively. For drops > 40 µm, the distribution in case of VO 150 °
C is higher than that of VO at 100 ° C. The respective SMDs for both the cases are 20 µm and 18
µm which are influenced by the presence of few large droplets in the case of VO at 150 °C. At Y
= 35 mm, X = 0 mm, the largest diameter for VO at 150 oC is 60 µm where as for VO 100 oC is
55 µm. The respective SMDs for both the cases are 16 µm and 20 µm, which is attributed to the
fact of larger percentage of smaller droplets for higher fuel inlet temperature i.e. VO at 150 oC
compared to that of lower fuel inlet temperatures. As the axial distance increases the droplet
131
distribution is observed to be narrower and also more number of smaller droplets is observed for
both the cases. Comparatively the higher fuel inlet temperature shows improved atomization
with higher percentage of smaller drops compared to that of lower fuel inlet temperature.
(d) Effect of Enclosed Spray on Combustion Emissions
The emissions of CO and NOx were measured for three different axial locations in the
flame. The radial profiles are plotted for axial locations at Y = 43.5 cm, 41.0 cm and 38.0 cm.
The combustor enclosure is of a regular pentagonal shape of about 15 cm measured diagonally.
Figure 5.44 shows the CO profile for VO at 100°C and 150°C at Y = 43.5 cm, 41.0 cm and 38.0
cm at the combustor exit plane. At 100°C, the CO emissions are seen to be higher on negative
transverse location of the flame as compared to positive transverse location as seen in the plot.
This is likely due to the emissions probe extending across the combustor exit plane, which results
in heating of the probe due to exposure to the flame at the exit plane. This causes further
dissociation of the CO2, resulting in higher CO emissions as compared to the positive transverse
locations where the reactions get cooled down rapidly resulting in lower CO emissions. The CO
emissions for VO at150°C range from 2 to 5 ppm while those of VO at 100°C range from 4 to 12
ppm. The CO emissions at the combustor exit plane indicate that sufficient flow mixing has
taken place, forming a homogenous product gas mixture. The CO emissions show a minor
decreasing trend in axial direction but mostly they are in similar range. The decrease in
emissions for VO at 150°C is about 8 ppm compared to VO at 100°C, which is attributed to the
smaller droplet diameter for flame with higher fuel inlet temperature.
Figure 5.45 shows the NOx emissions profile at the aforementioned axial locations in the
flame. A significant reduction in NOx emissions is observed for VO at 150°C compared to VO at
100°C. The NOx emissions for VO at 150°C are constant at about 25 ppm, where as for VO at
132
100°C, the NOx emissions are significantly higher, in the range of 120 ppm to 160 ppm. The
NOx emissions for different fuel temperatures are shown in chapter 3, which show a similar
trend of emissions for a given ALR i.e. for VO at 99°C, the NOx emissions range from 130 ppm
to 155 ppm. An increase in fuel inlet temperature decreases the kinematic viscosity of the fuel
hence improving the atomization process, having significant effect on emissions. Spray droplet
characteristics cannot be directly compared to the emissions, since the emissions data were taken
at the axial locations farther downstream near the combustor exit plane. Most of the droplets
were consumed and vaporized in the flame and the PDPA system could not detect the droplets,
because of the low data rate after Y = 40 mm. near the injector exit, the CO emissions could not
be measured because it was not possible to quench the reactions consuming CO. The reactions
near the injector exit are supposed to be incomplete, due to continuous combustion and the
emissions probing within the flame should give detectable CO emissions. However in our case
the probe was exposed to a high temperature environment, which allowed the reactions to
proceed further within the sampling probe. Since we did not use a water-cooled sampling probe,
the measurement of CO emissions was not possible within the flame.
(e) Enclosure Exterior Surface Temperature Distributions
Figure 5.46 shows the infra red camera image of the enclosure. The exterior surface
temperature data is inferred from the thermal radiation captured by the infra red image as shown
in figure 5.45. Image reveals the transverse profiles of temperature across the enclosure surface
at 5 different axial locations. The surface temperature profile shows lower temperature for points
1 and 2 and higher for the points 3 and 4 and again lower for point 5. This decrease can be
observed as point 5 is exposed to exit of the combustor close to the exhaust ambient air while for
points 1 and 2 the heat loss is seen to occur due to the uninsulated glass area. Point 4 has the
133
higher average temperature of about 132°C which indicated that most of the heat is retained in
the flame zone at that axial location. The average temperature as seen from the figure varies from
a minimum of 113°C to 127°C. The minimum temperature observed on each axial location
would in fact be higher than the indicated because of variation in view factor, since the enclosure
is not a flat surface facing normal to the viewing direction. The principle on which IR camera
works is based on the radiosity readings, and hence the view factor, emissivity of the object in
the field of view would affect the temperature readings. The IR image provided with an idea on
how the exterior surface temperature varied. The red contour region in the figure shows the
highest temperature which is for the steel bars used to support the glass in the enclosure. Again a
significant heat loss is there due to the surface exposed to the ambient air and also the small
window kept uninsulated for optical access. However, the insulation helped retaining heat with
minimum heat loss through the surface and thus allowing minimum supply of methane to
achieve a stable VO flame throughout the experimental duration. The temperature profile at five
different axial locations is shown in figure 5.47.
5.4. Conclusions
The present study investigates the spray characteristics of heated and unheated vegetable
oil. The measurements of mean axial velocity, RMS axial velocity, SMD, and droplet diameter
distributions are presented to explain the effect of swirling air on a non-evaporating spray, effect
of flame on spray, and effect of enclosure on spray.
•
The mean axial velocity for the swirling air flow is observed to be higher than the non
swirling air flow. The axial velocity further decrease in the farther downstream locations
in the spray. The swirling spray shows higher axial velocity in the near field region of the
spray; whereas for far field regions the non swirling spray shows the axial velocity
134
profiles with higher magnitudes. Similar trend is observed for RMS axial velocity
profiles; higher RMS axial velocity near the injector exit compared to farther downstream
locations in the spray. And higher RMS axial velocity in the centre of the spray while
lower RMS axial velocity at the edge of the spray.
•
Swirling air flow in non evaporating spray lowers the SMD. The reduction is
significantly in the far field locations of the spray i.e. Y = 25 mm and higher. In the near
field location of the spray i.e. Y = 5 mm to 20 mm, the swirling air flow tends to increase
the radial extent of the spray hence further disintegrating the droplets. The effect of
swirling air flow is more prominent near the injector exit while the spray flattens out
farther downstream, overall improving the atomization compared to the non swirling air
flow.
•
Higher flame temperatures tend to accelerate the droplets leading into increase in mean
axial velocity at all the measured axial locations compared to the cold spray. The mean
axial velocity peaked at the centre while they showed a decreasing trend at the outer
edges of the spray. RMS axial velocity increases for flame spray compared to cold spray.
The droplet diameter is observed to be smaller in case of flame spray compared to the
cold spray. SMD significantly reduces with axial distance. A decrease of about 40-50 µm
is observed in the SMD for the flame spray compared to cold spray with swirl air flow.
This result is attributed to the improved vaporization because of high flame temperatures
versus cold spray.
•
The effect of fuel inlet temperature improves the fuel kinematic viscosity and surface
tension hence improving the atomization leading to reduction in SMD. This result is
observed in both cold spray and also for the flame spray with heated and unheated VO.
135
Increase in mean and RMS axial velocity is observed for higher VO temperature
compared to unheated VO. The greater number of larger droplets in case of unheated VO
dominates the higher droplet diameter compared to VO at 100°C. Higher fuel inlet
temperatures aids to improve droplet vaporization and hence combustion reactions thus
improving the overall spray characteristics .This further improves chemical reactions thus
reducing the emissions due to the thermal feedback from the flame delineating the effect
of heat release rate on the fuel vaporization.
•
The effect of enclosed flame was seen on two different fuel inlet temperatures with
higher fuel inlet temperature producing finer droplets lowering the SMD, increasing the
mean and RMS axial velocity. The insulated enclosure helped to provide heat feedback
which enabled the PDPA measurements avoiding any fuel condensation in the
measurement area. The confinement of the flame due to enclosure reduced the radial
spread compared to open flame conditions. There was no significant difference in the
mean axial velocities and SMD values.
•
Based on this study we conclude that the smaller drop size distribution results in lower
emissions of CO and NOx. Larger droplets tend to burn in diffusion mode compared to
finer drops that lead to premix mode of combustion. Larger droplets that are burning in
diffusion mode results into higher local temperatures and higher flame temperatures
result in increase in NOx and the more likely chances of fuel pyrolysis, fuel coking
problems and fuel decomposition resulting in higher CO emissions.
•
It can be seen from the present investigation, that the flame environment enhances the
reduction of droplet size and smaller droplet size distribution, significantly altering the
136
mixing and entrainment due to high flame temperature, as well as faster fuel vaporization
as a secondary effect of high temperatures consuming larger droplets.
•
A significant change is not observed in the SMD and axial velocity for open and enclosed
flame because of the addition of large amount of methane to sustain a stable VO flame in
open conditions which results, in significant heat feedback in the near field region of the
flame spray, thus neutralizing the influence of the insulated enclosure, effect that is
observed in other cases.
137
Figure 5.1 Flame structure of an air blast atomizer
138
Flame
Fuel Vapor Spray Boundary
Air
Droplet Boundary Layer
Air Supply
Air Supply
Atomizer
Fuel
Figure 5.2 Vaporization of a typical droplet in an idealized spray flame
139
Figure 5.3 Schematic of the combustor experimental set-up
140
(a)
(b)
Pentagonal
Enclosure
Stable VO
Flame
Insulation
8 cm
Φ =15 cm
46 cm
Stainless Steel key
stock bars to support
the glass plates
Figure 5.4 Photographic view of the enclosure (a) without and (b) with insulation and schematic
of the top view and vertical view of the pentagonal enclosure
141
Figure 5.5 Airblast Injector Details
142
Fuel
Heater
Thermocouple
Liquid
Fuel
Fuel
Pump
Combustor
Pulse
Dampener
Fuel
Reservoir
Thermal
Logger
Figure 5.6 Schematic of Liquid fuel supply system.
Pressure
Regulator
Control
Valve
LFE
Air
Water Traps / Filters
MFC
Methane
Needle
Valve
MFM
Check
Valve
Figure 5.7 Schematic of gaseous fuel air supply system.
143
Diffusion
flame
Exhaust Duct
Cold Spray
Diffusion flame torches
Laser Beams passing
through the spray
Non-evaporating VO Spray
Figure 5.8 Photographic representation of the Flare system
144
Figure 5.9 (a) Emissions Analyzer; (b) Emissions Measurement Traversing system
145
Figure 5.10 Schematic of the PDPA system
146
Axial
velocity
Radial
velocity
Figure 5.11 Experimental set-up of a PDPA system mounted on a 3-way traversing system
147
Direction of
motion of
the probe
volume
Traversing
rail
18°
18°
Y
Z
Pentagonal
Enclosure
X
Origin of the
coordinate system
at the injector exit
Probe volume
Combustion
chamber
Figure 5.12 Plan view of the PDPA traversing mechanism in radial and axial coordinates
148
Receive
r Probe
Traversing
system
Non
evaporating
VO Spray
Transmitte
r Probe
Combust
or
Figure 5.13 Photographic view of the PDPA system integrated with the combustor assembly
149
Figure 5.14 Photographic Image of the MikroScan 7200v Infrared Camera
150
Mean axial velocity (m/s)
No swirl flow
75
50
45
40
35
30
25
20
15
10
5
-0.5
Axial Location (mm)
60
45
30
15
-30
-15
0
15
30
Transverse location (mm)
Figure 5.15 Mean axial velocity contour for cold spray without swirl
151
Mean axial velocity (m/s)
Swirl flow
75
50
45
40
35
30
25
20
15
10
5
0
-0.5
Axial Locaiton (mm)
60
45
30
15
-30
-15
0
15
30
Transverse location (mm)
Figure 5.16 Mean axial velocity contour for cold spray with swirl
152
Mean axial velocity (m/s)
Swirl flow
o
Tf=100 C
75
50
45
40
35
30
25
20
15
10
5
-0.5
Axial location (mm)
60
45
30
15
-30
-15
0
15
30
Transverse location (mm)
Figure 5.17 Mean axial velocity contour for cold spray with swirl at Tf = 100°C
153
RMS axial Velocity (m/s)
No swirl flow
75
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Axial location (mm)
60
45
30
15
-30
-15
0
15
30
Transverse location (mm)
Figure 5.18 RMS axial velocity contour for cold spray without swirl
154
RMS axial velocity (m/s)
Swirl flow
75
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Axial location (mm)
60
45
30
15
-30
-15
0
15
30
Transverse location (mm)
Figure 5.19 RMS axial velocity contour for cold spray with swirl
155
RMS axial velocity (m/s)
Swirl flow
Tf=100 oC
75
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Axial location (mm)
60
45
30
15
-30
-15
0
15
30
Transverse location (mm)
Figure 5.20 RMS axial velocity contour for cold spray with swirl at Tf = 100°C
156
SMD (microns)
No swirl flow
75
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
Axial location (mm)
60
45
30
15
-30
-15
0
15
30
Transverse location (mm)
Figure 5.21 SMD contour for cold spray without swirling air
157
SMD (microns)
Swirl flow
75
75
70
60
55
50
45
40
35
30
25
20
15
10
5
Axial locations (mm)
60
45
30
15
-30
-15
0
15
30
Transverse location (mm)
Figure 5.22 SMD contour for cold spray with swirling air
158
SMD (microns)
Swirl lfow
Tf=100oC
75
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
Axial location (mm)
60
45
30
15
-30
-15
0
15
30
Transverse location (mm)
Figure 5.23 SMD contour for cold spray with swirling air at Tf = 100°C
159
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
30
40
30
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (a and b) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 5 mm
160
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (c and d) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 10 mm
161
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (e and f) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 15 mm
162
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (g and h) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 20 mm
163
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (i and j) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 25 mm
164
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (k and l) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 30 mm
165
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
30
40
30
Transverse location (mm)
-10
0
10
20
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (m and n) Transverse profiles of mean axial velocity and RMS axial velocity for
cold spray with and without swirl at Y = 35 mm
166
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (o and p) Transverse profiles of mean axial velocity and RMS axial velocity for
cold spray with and without swirl at Y = 40 mm
167
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (q and r) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 45 mm
168
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (s and t) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 50 mm
169
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (u and v) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 60 mm
170
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (w and x) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 70 mm.
171
-40
60
-30
-20
-10
0
10
20
30
40
60
Mean axial velocity (m/s)
Swirling air
No swirling air
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Swirling air
No swirling air
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.24 (y and z) Transverse profiles of mean axial velocity and RMS axial velocity for cold
spray with and without swirl at Y = 75 mm.
172
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
Swirling air, Tf= 100oC
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
30
40
30
Transverse location (mm)
-40
30
-30
-20
-10
0
20
Swirling air, unheated VO
o
Swirling air, Tf= 100 C
25
RMS axial velocity (m/s)
10
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (a and b) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 5 mm
173
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
o
Swirling air, Tf= 100 C
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
20
30
Swirling air, unheated VO
Swirling air, Tf= 100oC
25
RMS axial velocity (m/s)
10
40
30
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (c and d) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 10 mm
174
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
o
Swirling air, Tf= 100 C
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
20
30
Swirling air, unheated VO
Swirling air, Tf= 100oC
25
RMS axial velocity (m/s)
10
40
30
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (e and f) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 15 mm
175
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
o
Swirling air, Tf= 100 C
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
20
30
Swirling air, unheated VO
o
Swirling air, Tf= 100 C
25
RMS axial velocity (m/s)
10
40
30
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (g and h) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 20 mm
176
-40
60
-30
-20
-10
0
10
30
40
60
Swirling air; unheated VO
Swirling air, Tf= 100oC
50
Mean axial velocity (m/s)
20
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
20
30
Swirling air, unheated VO
Swirling air, Tf= 100oC
25
RMS axial velocity (m/s)
10
40
30
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (i and j) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 25 mm
177
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
o
Swirling air, Tf= 100 C
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
20
30
Swirling air, unheated VO
Swirling air, Tf= 100oC
25
RMS axial velocity (m/s)
10
40
30
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (k and l) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 30 mm
178
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
Swirling air, Tf= 100 oC
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
20
30
Swirling air, unheated VO
Swirling air, Tf= 100oC
25
RMS axial velocity (m/s)
10
40
30
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (m and n) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100 C Y = 35 mm
179
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
o
Swirling air, Tf= 100 C
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
30
40
30
Transverse location (mm)
-10
0
20
Swirling air, unheated VO
o
Swirling air, Tf= 100 C
25
RMS axial velocity (m/s)
10
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (o and p) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 40 mm
180
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
o
Swirling air, Tf= 100 C
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
20
30
Swirling air, unheated VO
Swirling air, Tf= 100 oC
25
RMS axial velocity (m/s)
10
40
30
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (q and r) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 50 mm
181
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
o
Swirling air, Tf= 100 C
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
20
30
Swirling air, unheated VO
Swirling air, Tf= 100oC
25
RMS axial velocity (m/s)
10
40
30
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (s and t) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 60 mm
182
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
o
Swirling air, Tf= 100 C
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
20
30
Swirling air, unheated VO
Swirling air, Tf= 100oC
25
RMS axial velocity (m/s)
10
40
30
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (u and v) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 70 mm
183
-40
60
-30
-20
-10
0
10
30
Swirling air; unheated VO
o
Swirling air, Tf= 100 C
50
Mean axial velocity (m/s)
20
40
60
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-40
30
-30
-20
-10
0
20
30
Swirling air, unheated VO
Swirling air, Tf= 100oC
25
RMS axial velocity (m/s)
10
40
30
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.25 (w and x) Transverse profiles of mean axial velocity and RMS axial velocity for
swirling air cold spray; unheated and VO at Tf = 100°C Y = 75 mm
184
-40
100
-30
-20
-10
0
10
30
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-40
100
-30
-20
-10
0
10
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.26 (a and b) Transverse profiles of SMD for cold spray with and without swirl at Y = 5
mm and 10 mm
185
-40
100
-30
-20
-10
0
10
30
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-40
100
-30
-20
-10
0
10
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.26 (c and d) Transverse profiles of SMD for cold spray with and without swirl at Y =
15 mm and 20 mm
186
-40
100
-30
-20
-10
0
10
30
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-40
100
-30
-20
-10
0
10
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.26 (e and f) Transverse profiles of SMD for cold spray with and without swirl at Y = 25
mm and 30 mm
187
-40
100
-30
-20
-10
0
10
30
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-40
100
-30
-20
-10
0
10
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.26 (g and h) Transverse profiles of SMD for cold spray with and without swirl at Y =
35 mm and 40 mm
188
-40
100
-30
-20
-10
0
10
30
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-40
100
-30
-20
-10
0
10
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.26 (i and j) Transverse profiles of SMD for cold spray with and without swirl at Y = 45
mm and 50 mm
189
-40
100
-30
-20
-10
0
10
30
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-40
100
-30
-20
-10
0
10
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.26 (k and l) Transverse profiles of SMD for cold spray with and without swirl at Y = 60
mm and 70 mm
190
-40
100
-30
-20
-10
0
10
30
Swirling air
No swirling air
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.26 (m) Transverse profiles of SMD for cold spray with and without swirl at Y = 75mm
191
-40
100
-30
-20
-10
10
20
30
Swirling air, unheated VO
o
Swirling air, Tf= 100 C
90
Sauter mean diameter (μm)
0
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
100
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-10
10
20
Swirling air, unheated VO
o
Swirling air, Tf= 100 C
90
Sauter mean diameter (μm)
0
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.27 (a and b) Transverse profiles of SMD for swirling air cold spray; unheated and VO
at Tf = 100 C Y = 5 mm and Y = 10 mm
192
-40
100
-30
-20
-10
10
20
30
Swirling air, unheated VO
o
Swirling air, Tf= 100 C
90
Sauter mean diameter (μm)
0
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-40
100
-30
-20
-10
10
20
Swirling air, unheated VO
o
Swirling air, Tf= 100 C
90
Sauter mean diameter (μm)
0
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.27 (c and d) Transverse profiles of SMD for swirling air cold spray; unheated and VO
at Tf = 100 C Y = 15 mm and Y = 20 mm
193
-40
100
-30
-20
-10
10
20
30
Swirling air, unheated VO
Swirling air, Tf= 100 oC
90
Sauter mean diameter (μm)
0
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-40
100
-30
-20
-10
10
20
Swirling air, unheated VO
o
Swirling air, Tf= 100 C
90
Sauter mean diameter (μm)
0
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.27 (e and f) Transverse profiles of SMD for swirling air cold spray; unheated and VO at
Tf = 100 C Y = 25 mm and Y = 30 mm
194
-40
100
-30
-20
-10
10
20
30
Swirling air, unheated VO
Swirling air, Tf= 100oC
90
Sauter mean diameter (μm)
0
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-40
100
-30
-20
-10
10
20
Swirling air, unheated VO
Swirling air, Tf= 100 oC
90
Sauter mean diameter (μm)
0
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.27 (g and h) Transverse profiles of SMD for swirling air cold spray; unheated and VO
at Tf = 100 C Y = 35 mm and Y = 40 mm
195
-40
100
-30
-20
-10
10
20
30
Swirling air, unheated VO
Swirling air, Tf= 100 oC
90
Sauter mean diameter (μm)
0
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-40
100
-30
-20
-10
10
20
Swirling air, unheated VO
Swirling air, Tf= 100 oC
90
Sauter mean diameter (μm)
0
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.27 (i and j) Transverse profiles of SMD for swirling air cold spray; unheated and VO at
Tf = 100 C Y = 50 mm and Y = 60 mm
196
-40
100
-30
-20
-10
10
20
30
Swirling air, unheated VO
o
Swirling air, Tf= 100 C
90
Sauter mean diameter (μm)
0
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
100
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-10
10
20
Swirling air, unheated VO
Swirling air, Tf= 100 oC
90
Sauter mean diameter (μm)
0
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.27 (k and l) Transverse profiles of SMD for swirling air cold spray; unheated and VO at
Tf = 100 C Y = 70 mm and Y = 75 mm
197
0
D iameter Count
10
25
50
75
100
3
125
150
175
200
225
Swirling air, SMD = 64 μm
No swirling air, SMD = 76 μm
10 2
250
10
3
102
10
1
10
1
10
0
10
0
0
25
50
75
0
25
50
75
100
125
150
175
200
225
250
200
225
250
Diameter (microns)
100
150
175
Swirling air, SMD = 50 μm
No swirling air, SMD = 52 μm
103
D iameter Count
125
103
102
102
101
101
100
100
0
25
50
75
100
125
150
175
200
225
250
Diameter (microns)
Figure 5.28 (a and b) Droplet Distribution Profile for cold spray with and without swirl at Y = 5
mm, X = 0 mm and Y = 10 mm, X = 0 mm
198
0
25
50
75
100
150
175
200
225
Swirling air, SMD = 37 μm
No swirling air, SMD = 46 μm
103
D iameter Count
125
250
103
10
2
10
2
10
1
10
1
10
0
10
0
0
25
50
75
0
25
50
75
100
125
150
175
200
225
250
200
225
250
Diameter (microns)
100
150
175
Swirling air, SMD = 36 μm
No swirling air, SMD = 44 μm
103
D iameter Count
125
102
103
102
10
1
10
1
10
0
10
0
0
25
50
75
100
125
150
175
200
225
250
Diameter (microns)
Figure 5.28 (c and d) Droplet Distribution Profile for cold spray with and without swirl at Y =
40 mm, X = 0 mm and Y = 60 mm, X = 0 mm
199
Axial location (mm)
Mean axial velocity (m/s)
Flame
40
30
20
10
-30
-20
-10
0
10
20
Transverse location (mm)
30
65
60
55
50
45
40
35
30
25
20
15
10
5
Figure 5.29 Mean axial velocity contour for flame spray
200
Axial location (mm)
Mean axial velocity(m/s)
Flame
o
Tf=100 C
40
30
20
10
-30
-20
-10
0
10
20
Transverse location (mm)
30
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
Figure 5.30 Mean axial velocity contour for flame spray for VO at 100°C
201
Axial location (mm)
RMS axial location (m/s)
Flame
40
30
20
10
-30
-20
-10
0
10
20
Transverse location (mm)
30
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Figure 5.31 RMS axial velocity contour for flame spray
202
Axial location (mm)
RMS axial velocity (m/s)
Flame
Tf=100oC
40
30
20
10
-30
-20
-10
0
10
20
Transverse location (mm)
30
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Figure 5.32 RMS axial velocity contour for flame spray for VO at 100°C
203
Axial location (mm)
SMD (microns)
Flame
40
30
20
10
-30
-20
-10
0
10
20
Transverse location (mm)
Figure 5.33 SMD contour for flame spray
204
30
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
Axial location (mm)
SMD (microns)
Flameo
Tf=100 C
40
30
20
10
-30
-20
-10
0
10
20
30
Transverse location (mm)
Figure 5.34 SMD contour for flame spray for VO at 100°C
205
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Mean axial velocity (m/s)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
30
40
30
Transverse location (mm)
-10
0
10
20
RMS axial velocity (m/s)
Flame
Swirl spray
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.35 (a and b) Transverse profiles of axial mean velocity and RMS axial velocity for
flame spray and cold spray at Y = 5 mm
206
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Mean Axial Velocity (m/s)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
30
40
30
Transverse location (mm)
-10
0
10
20
RMS axial velocity (m/s)
Flame
Swirl spray
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.35 (c and d) Transverse profiles of axial mean velocity and RMS axial velocity for
flame spray and cold spray at Y = 10 mm
207
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Mean axial velocity (m/s)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Flame
Swirl spray
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.35 (e and f) Transverse profiles of axial mean velocity and RMS axial velocity for
flame spray and cold spray at Y = 15 mm
208
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Mean axial velocity (m/s)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Flame
Swirl spray
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.35 (g and h) Transverse profiles of axial mean velocity and RMS axial velocity for
flame spray and cold spray at Y = 20 mm
209
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Mean axial velocity (m/s)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
30
40
30
Transverse location (mm)
-10
0
10
20
RMS axial velocity (m/s)
Flame
Swirl spray
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.35 (i and j) Transverse profiles of axial mean velocity and RMS axial velocity for flame
spray and cold spray at Y = 25 mm
210
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Mean axial velocity (m/s)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
30
40
30
Transverse location (mm)
-10
0
10
20
RMS axial velocity (m/s)
Flame
Swirl spray
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.35 (k and l) Transverse profiles of axial mean velocity and RMS axial velocity for
flame spray and cold spray at Y = 30 mm
211
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Mean axial velocity (m/s)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Flame
Swirl spray
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.35 (m and n) Transverse profiles of axial mean velocity and RMS axial velocity for
flame spray and cold spray at Y = 35 mm
212
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Mean axial velocity (m/s)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
-10
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
-10
40
Transverse location (mm)
-10
0
10
20
30
40
30
RMS axial velocity (m/s)
Flame
Swirl spray
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.35 (o and p) Transverse profiles of axial mean velocity and RMS axial velocity for
flame spray and cold spray at Y = 40 mm
213
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
Swirl spray, Tf=100 oC
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
0
40
30
40
30
Transverse location (mm)
-10
0
10
20
o
RMS axial velcoity (m/s)
Flame, Tf=100 C
o
Swirl spray, Tf=100 C
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.36 (a and b) Transverse profiles of mean axial velocity and RMS axial velocity for
flame spray and cold spray for Tf = 100°C at Y = 5 mm.
214
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
o
Swirl spray, Tf=100
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
-10
0
10
20
30
40
30
o
RMS axial velcoity (m/s)
Flame, Tf=100 C
Swirl spray, Tf=100 o
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.36 (c and d) Transverse profiles of mean axial velocity and RMS axial velocity for
flame spray and cold spray for Tf = 100°C at Y = 10 mm.
215
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
o
Swirl spray, Tf=100
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
-10
0
10
20
30
40
30
o
RMS axial velcoity (m/s)
Flame, Tf=100 C
Swirl spray, Tf=100o
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.36 (e and f) Transverse profiles of mean axial velocity and RMS axial velocity for
flame spray and cold spray for Tf = 100°C at Y = 15 mm.
216
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
o
Swirl spray, Tf=100
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
-10
0
10
20
30
40
30
RMS axial velcoity (m/s)
Flame, Tf=100 oC
o
Swirl spray, Tf=100
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.36 (g and h) Transverse profiles of mean axial velocity and RMS axial velocity for
flame spray and cold spray for Tf = 100°C at Y = 20 mm.
217
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
Swirl spray, Tf=100 o
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
-10
0
10
20
30
40
30
o
RMS axial velcoity (m/s)
Flame, Tf=100 C
o
Swirl spray, Tf=100
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.36(i and j) Transverse profiles of mean axial velocity and RMS axial velocity for flame
spray and cold spray for Tf = 100°C at Y = 25 mm.
218
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
Swirl spray, Tf=100 o
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
-10
0
10
20
30
40
30
o
RMS axial velcoity (m/s)
Flame, Tf=100 C
o
Swirl spray, Tf=100
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.36 (k and l) Transverse profiles of mean axial velocity and RMS axial velocity for
flame spray and cold spray for Tf = 100°C C at Y = 30 mm.
219
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
Cold spray, Tf=100oC
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
-10
0
10
20
30
40
30
RMS axial velcoity (m/s)
Flame, Tf=100oC
o
Cold spray, Tf=100 C
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.36 (m and n) Transverse profiles of mean axial velocity and RMS axial velocity for
flame spray and cold spray for Tf = 100°C at Y = 35 mm.
220
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
Cold spray, Tf=100 oC
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
-10
0
10
20
30
40
30
o
RMS axial velcoity (m/s)
Flame, Tf=100 C
o
Cold spray, Tf=100 C
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.36 (o and p) Transverse profiles of mean axial velocity and RMS axial velocity for
flame spray and cold spray for Tf = 100°C C at Y = 40 mm.
221
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
100
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-10
0
10
Flame
Swirl spray
90
Sauter mean diameter (μm)
20
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.37 (a and b) Transverse profiles of SMD for flame spray and cold spray at Y = 5 mm
and 10 mm.
222
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
100
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-10
0
10
Flame
Swirl spray
90
Sauter mean diameter (μm)
20
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.37 (c and d) Transverse profiles of SMD for flame spray and cold spray at Y = 15 mm
and 20 mm.
223
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
100
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-10
0
10
Flame
Swirl spray
90
Sauter mean diameter (μm)
20
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.37 (e and f) Transverse profiles of SMD for flame spray and cold spray at Y = 25 mm
and 30 mm
224
-40
100
-30
-20
-10
0
10
30
Flame
Swirl spray
90
Sauter mean diameter (μm)
20
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
100
-30
-20
-10
0
10
20
30
0
40
30
40
100
Transverse location (mm)
-10
0
10
Flame
Swirl spray
90
Sauter mean diameter (μm)
20
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.37 (g and h) Transverse profiles of SMD for flame spray and cold spray at Y = 35 mm
and 40 mm
225
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
Swirl spray, Tf=100o
S auter mean diameter (μm)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
100
-30
-20
-10
0
20
30
0
40
20
30
40
100
10
Transverse location (mm)
-10
0
Flame, Tf=100oC
Swirl spray, Tf=100o
90
S auter mean diameter (μm)
10
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.38(a and b) Transverse profiles of SMD for flame spray and cold spray at Y = 5 mm
and 10 mm for Tf = 100°C.
226
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
Swirl spray, Tf=100oC
S auter mean diameter (μm)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
100
-30
-20
-10
0
20
30
0
40
20
30
40
100
10
Transverse location (mm)
-10
0
10
o
Flame, Tf=100 C
o
Swirl spray, Tf=100 C
Sauter mean diameter (μm)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.38 (c and d) Transverse profiles of SMD for flame spray and cold spray at Y = 15 mm
and 20 mm for Tf = 100°C.
227
-40
100
-30
-20
-10
0
20
30
Flame, Tf=100oC
Swirl spray, Tf=100oC
90
Sauter mean diameter (μm)
10
40
100
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
100
-30
-20
-10
0
20
30
0
40
20
30
40
100
10
Transverse location (mm)
-10
0
10
o
Flame, Tf=100 C
o
Swirl spray, Tf=100 C
Sauter mean diameter (μm)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.38 (e and f) Transverse profiles of SMD for flame spray and cold spray at Y = 25 mm
and 30 mm for Tf = 100°C.
228
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Flame, Tf=100 C
o
Swirl spray, Tf=100 C
Sauter mean diameter (μm)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
0
40
30
Transverse location (mm)
-30
-20
-10
0
10
20
30
40
90
90
Flame, Tf=100oC
Swirl spray, Tf=100oC
Sauter mean diameter (μm)
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
-30
-20
-10
0
10
20
30
40
Transverse location (mm)
Figure 5.38 (g and h) Transverse profiles of SMD for flame spray and cold spray at Y = 35 mm
and 40 mm for Tf = 100°C.
229
0
25
50
75
100
125
D iameter Count
103
175
200
225
Flame, SMD = 26μm
Cold spray ,SMD = 64μm
102
250
103
102
10
1
10
1
10
0
10
0
0
25
50
75
0
25
50
75
100
125
150
175
Diameter (microns)
100
125
103
D iameter Count
150
150
175
200
225
250
200
225
250
Flame, SMD = 19 μm
Cold spray, SMD = 51 μm
103
102
102
101
101
10
0
10
0
25
50
75
100
125
150
175
200
225
0
250
Diameter (microns)
Figure 5.39 (a and b) Droplet Distribution Profile for flame spray and cold spray at Y = 5 mm, X
= 0 mm and Y = 10 mm, X = 0 mm
230
0
25
50
75
100
125
D iameter Count
103
150
175
200
225
Flame, SMD = 17 μm
Cold spray, SMD = 38 μm
250
103
102
102
101
101
100
100
0
25
50
75
0
25
50
75
100
125
150
175
200
225
250
200
225
250
D iameter Count
Diameter (microns)
10
3
10
100
125
150
175
10
3
2
10
2
10
1
10
1
10
0
10
0
Flame, SMD = 19μm
Cold spray, SMD = 37μm
0
25
50
75
100
125
150
175
200
225
250
Diameter (microns)
Figure 5.39 (c and d) Droplet Distribution Profile for flame spray and cold spray at Y = 35 mm,
X = 0 mm and Y = 40 mm, X = 0 mm
231
D iameter Count
0
10
3
10
25
50
75
100
125
150
175
200
225
250
Flame, Tf= 100oC, SMD = 28μm
Cold Spray,Tf= 100oC, SMD = 37μm
10
3
2
10
2
10
1
10
1
10
0
10
0
0
25
50
75
0
25
50
75
100
125
150
175
200
225
250
200
225
250
Diameter (microns)
D iameter Count
10
100
125
150
175
o
Flame,Tf= 100 C, SMD = 22μm
Cold Spray,Tf= 100oC, SMD = 33μm
3
10
102
3
102
10
1
10
1
10
0
10
0
0
25
50
75
100
125
150
175
200
225
250
Diameter (microns)
Figure 5.40 (a and b) Droplet Distribution Profile for flame spray and cold spray for Tf = 100°C
at Y = 5 mm, X = 0 mm and Y = 10 mm, X = 0 mm
232
0
25
50
75
100
150
175
200
225
250
o
Flame,Tf= 100 C, SMD = 13μm
o
Cold spray,Tf= 100 C, SMD = 30μm
103
D iameter Count
125
103
102
102
10
1
10
1
10
0
10
0
0
25
50
75
0
25
50
75
100
125
150
175
200
225
250
200
225
250
Diameter (microns)
100
150
175
o
Flame,Tf= 100 C, SMD = 14μm
Cold spray,Tf= 100oC, SMD =29μm
103
D iameter Count
125
103
102
102
10
1
10
1
10
0
10
0
0
25
50
75
100
125
150
175
200
225
250
Diameter (microns)
Figure 5.40 (c and d) Droplet Distribution Profile for flame spray and cold spray for Tf = 100°C
at Y = 35 mm, X = 0 mm and Y = 40 mm, X = 0 mm.
233
-40
100
-30
-20
-10
10
20
30
40
100
Enclosed flame, Tf = 100oC
o
Enclosed flame, Tf = 150 C
90
Mean axial velocity (m/s)
0
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
0
40
30
Transverse location (mm)
-10
0
10
20
30
40
30
o
RMS axial velocity(m/s)
Enclosed flame, Tf = 100 C
o
Enclosed flame, Tf = 150 C
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.41 (a and b) Transverse profiles of mean axial velocity and RMS axial velocity for
enclosed flame for Tf = 100°C and Tf = 150°C at Y = 5 mm
234
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Enclosed flame, Tf = 100 C
o
Enclosed flame, Tf = 150 C
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
0
40
30
Transverse location (mm)
-10
0
10
20
30
40
30
o
RMS axial velocity(m/s)
Enclosed flame, Tf = 100 C
Enclosed flame, Tf = 150 oC
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.41 (c and d) Transverse profiles of mean axial velocity and RMS axial velocity for
enclosed flame for Tf = 100°C and Tf = 150°C at Y = 10 mm
235
-40
100
-30
-20
-10
10
20
30
40
100
Enclosed flame, Tf = 100oC
Enclosed flame, Tf = 150oC
90
Mean axial velocity (m/s)
0
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
0
40
30
Transverse location (mm)
-10
0
10
20
30
40
30
o
RMS axial velocity(m/s)
Enclosed flame, Tf = 100 C
o
Enclosed flame, Tf = 150 C
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.41 (e and f) Transverse profiles of mean axial velocity and RMS axial velocity for
enclosed flame for Tf = 100°C and Tf = 150°C at Y = 15 mm
236
-40
100
-30
-20
-10
10
20
30
40
100
Enclosed flame, Tf = 100oC
o
Enclosed flame, Tf = 150 C
90
Mean axial velocity (m/s)
0
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
0
40
30
Transverse location (mm)
-10
0
10
20
30
40
30
o
RMS axial velocity(m/s)
Enclosed flame, Tf = 100 C
Enclosed flame, Tf = 150 oC
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.41 (g and h) Transverse profiles of mean axial velocity and RMS axial velocity for
enclosed flame for Tf = 100°C and Tf = 150°C at Y = 20 mm
237
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Enclosed flame, Tf = 100 C
o
Enclosed flame, Tf = 150 C
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
30
40
30
Transverse location (mm)
-40
30
-30
-20
-10
0
10
20
o
RMS axial velocity(m/s)
Enclosed flame, Tf = 100 C
o
Enclosed flame, Tf = 150 C
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.41 (i and j) Transverse profiles of mean axial velocity and RMS axial velocity for
enclosed flame for Tf = 100°C and Tf = 150°C at Y = 25 mm
238
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Enclosed flame, Tf = 100 C
o
Enclosed flame, Tf = 150 C
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
0
40
30
Transverse location (mm)
-10
0
10
20
30
40
30
o
RMS axial velocity(m/s)
Enclosed flame, Tf = 100 C
Enclosed flame, Tf = 150oC
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.41 (k and l) Transverse profiles of mean axial velocity and RMS axial velocity for
enclosed flame for Tf = 100°C and Tf = 150°C at Y = 30 mm
239
-40
100
-30
-20
-10
0
10
20
30
40
100
o
Enclosed flame, Tf = 100 C
o
Enclosed flame, Tf = 150 C
Mean axial velocity (m/s)
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
30
-30
-20
-10
0
10
20
30
0
40
30
40
30
Transverse location (mm)
-10
0
10
20
o
RMS axial velocity(m/s)
Enclosed flame, Tf = 100 C
Enclosed flame, Tf = 150oC
25
25
20
20
15
15
10
10
5
5
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.41 (m and n) Transverse profiles of mean axial velocity and RMS axial velocity for
enclosed flame for Tf = 100°C and Tf = 150°C at Y = 35 mm
240
-40
50
-30
-20
-10
0
10
20
30
40
50
Sauter mean diameter (microns)
Enclosed flame, Tf= 100 oC
o
Enclosed Flame, Tf = 150 C
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
50
-30
-20
-10
0
10
20
30
0
40
30
40
50
Transverse location (mm)
-10
0
10
20
o
Sauter mean diameter (microns)
Enclosed flame, Tf= 100 C
o
Enclosed Flame, Tf = 150 C
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.42 (a and b) Transverse profiles of SMD for enclosed flame for Tf = 100°C and Tf =
150°C at Y = 5 mm and Y = 10 mm.
241
-40
50
-30
-20
-10
0
10
20
30
40
50
o
Sauter mean diameter (microns)
Enclosed flame, Tf= 100 C
o
Enclosed Flame, Tf = 150 C
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
50
-30
-20
-10
0
10
20
30
0
40
30
40
50
Transverse location (mm)
-10
0
10
20
Sauter mean diameter (microns)
Enclosed flame, Tf= 100oC
o
Enclosed Flame, Tf = 150 C
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.42(c and d) Transverse profiles of SMD for enclosed flame for Tf = 100°C and Tf =
150°C at Y = 15 mm and Y = 20 mm.
242
-40
50
-30
-20
-10
0
10
20
30
40
50
o
Sauter mean diameter (microns)
Enclosed flame, Tf= 100 C
o
Enclosed Flame, Tf = 150 C
40
40
30
30
20
20
10
10
0
-40
-30
-20
-40
50
-30
-20
-10
0
10
20
30
0
40
30
40
50
Transverse location (mm)
-10
0
10
20
Sauter mean diameter (microns)
Enclosed flame, Tf= 100oC
o
Enclosed Flame, Tf = 150 C
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.42 (e and f) Transverse profiles of SMD for enclosed flame for Tf = 100°C and Tf =
150°C at Y = 25 mm and Y = 30 mm.
243
-40
50
-30
-20
-10
0
10
20
30
40
50
o
Sauter mean diameter (microns)
Enclosed flame, Tf= 100 C
o
Enclosed Flame, Tf = 150 C
40
40
30
30
20
20
10
10
0
-40
-30
-20
-10
0
10
20
30
0
40
Transverse location (mm)
Figure 5.42 (g) Transverse profiles of SMD for enclosed flame for Tf = 100°C and Tf = 150°C at
Y = 35mm.
244
0
25
50
75
100
D iameter Count
103
125
150
175
200
225
o
Tf= 100 C, SMD = 32μm
Tf= 150oC, SMD = 28μm
102
250
103
102
10
1
10
1
10
0
10
0
0
25
50
75
100
125
150
175
200
225
250
Diameter (microns)
25
D iameter Count
10
50
75
100
3
125
150
175
200
225
Tf= 100 C, SMD = 27 μm
Tf= 150oC, SMD = 26μm
o
250
10
3
102
102
101
101
10
0
10
25
50
75
100
125
150
175
200
225
0
250
Diameter (microns)
Figure 5.43(a and b) Droplet Distribution Profile for Enclose Flame for Tf = 100 °C and Tf =
150 °C at Y = 5 mm, X = 0 mm and Y = 10 mm, X = 0 mm
245
25
50
75
100
D iameter Count
103
125
150
175
200
225
Tf= 100 C, SMD = 18 μm
o
Tf= 150 C, SMD = 19μm
o
102
250
103
102
10
1
10
1
10
0
10
0
25
50
75
50
75
100
125
150
175
200
225
250
200
225
250
Diameter (microns)
25
D iameter Count
10
100
3
125
150
175
Tf= 100 C, SMD = 20 μm
o
Tf= 150 C, SMD = 17 μm
o
102
10
3
102
10
1
10
1
10
0
10
0
25
50
75
100
125
150
175
200
225
250
Diameter (microns)
Figure 5.43 (c and d) Droplet Distribution Profile for Enclosed Flame for Tf = 100 °C and Tf =
150 °C at Y = 30 mm, X = 0 mm and Y = 35 mm, X = 0 mm
246
-6
-4
-2
0
14
4
6
14
o
Y = 43.5 cm, Tf = 100 C
Y = 41.0 cm, Tf = 100oC
o
Y = 38.0 cm, Tf = 100 C
o
Y = 43.5 cm, Tf = 150 C
o
Y = 41.0 cm, Tf = 150 C
Y = 38.0 cm, Tf = 150oC
12
10
CO (ppm)
2
12
10
8
8
6
6
4
4
2
2
0
-6
-4
-2
0
2
4
6
0
Transverse location (cm)
Figure 44. Transverse profiles of CO for enclosed VO flame at 100°C and 150°C
247
NOx (ppm)
-6
200
-4
-2
0
2
4
6
200
175
175
150
150
125
125
o
Y = 43.5 cm, Tf = 100 C
Y = 41.0 cm, Tf = 100oC
Y = 38.0 cm, Tf = 100oC
Y = 43.5 cm, Tf = 150oC
o
Y = 41.0 cm, Tf = 150 C
o
Y = 38.0 cm, Tf = 150 C
100
75
100
75
50
50
25
25
-6
-4
-2
0
2
4
6
Transverse distance (cm)
Figure 45. Transverse profiles of NOx for enclosed VO flame at 100°C and 150°C
248
0.00 C
Axial location (mm)
Min C
Max C
Avg C
Range C
Line - 1
166.00
82.3
146.5
113.3
64.2
Line - 2
222.00
95.3
149.8
119.0
54.5
Line - 3
280.00
89.2
156.5
125.7
67.3
Line - 4
336.00
90.1
180.2
131.1
90.1
Line - 5
388.00
80.4
153.9
127.0
73.5
Figure 5.46.IR Image of the Enclosure surface temperature
249
-200
200
-150
-100
-50
0
100
150
Line 1 (Y = 166 mm)
Line 2 (Y = 222mm)
Line 3 (Y = 280 mm)
Line 4 (Y = 336 mm)
Line 5 (Y = 388 mm)
180
Exterior surface temperature ( o C)
50
160
200
200
180
160
140
140
120
120
100
100
80
80
60
60
-200
-150
-100
-50
0
50
100
150
200
Position on diametral projeciton (mm)
Figure 5.47.Profile plot for Exterior surface temperature at different axial location on the
enclosure
250
251
CHAPTER 6
CONCLUSIONS
Present research work involved detailed analysis of various parameters associated with
combustion of viscous bio-oils. From the perspective of improving the combustion performance,
detailed analysis into effect of composition, thermo physical properties, experimental operating
conditions of liquid fuel preheating and altering the mixing dynamics of the atomizer (through
ALR) was carried out using various investigative tools such as GC-MS, TGA, PDPA emissions
analyzer, etc. Fuels used in this study included diesel, two types of biodiesel, diesel-VO blends,
preheated VO. The important inferences from this investigation are listed below:
™ Comparative study of liquid fuel composition and their thermo physical properties shows
that the blending of high viscosity fuel with low viscosity fuel would be the simple viable
option to improve the fuel spray characteristics required to effectively atomize the fuel to
reduce emissions in liquid fuel combustion.
™ Physical properties such as kinematic viscosity and volatility of the bio-oils were
improved by blending them with a low viscosity and high volatility fuel such as diesel.
For a given ALR, the CO emissions decreased by 1 to 2 ppm and NOx emissions reduced
by 20 ppm with increase in fuel inlet temperature with increase in ALR.
™ The emissions also decreased significantly; by a factor of 2 to 3 for CO and 6 to 15 for
NOx for a given fuel inlet temperature. Wetting of the combustor wall by unheated fuel
251
droplets was observed visually in the flame images at all test conditions, although its
extent decreased significantly with an increase in the fuel inlet temperature and/or ALR.
™ Increase in ALR decreases the SMD, delineating the importance of spray mixing
dynamics in an atomizer.
™ Lower drop diameter improved the spray quality by decreasing the spray angle observed
visually.
™ Owing to higher momentum, larger droplets migrate towards the edge of the spray
indicating that a secondary mechanism for droplet break-up would be beneficial to obtain
fine spray with good combustion performance. Swirling air flow facilitates such
secondary break-up mechanism.
™ Effect of swirling air flow on a non-evaporating spray is to lower the SMD, increase the
mean axial velocity, and widen the radial extent of the spray.
™ At any given axial location, spray with swirling air flow spray showed finer atomization
because of secondary break up that was absent in the spray with non swirling air flow.
™ Mean axial velocities were higher at the centre of the spray and it decreased towards the
edge and farther downstream in the spray. These effects were more noticeable at higher
ALR, higher fuel inlet temperatures and in flame spray conditions.
™ At any axial location, the droplet diameter was smaller for the burning spray compared to
non-burning spray. The axial velocities were much higher in the burning spray compared
to the non burning spray.
™ Increase in fuel inlet temperature is observed to improve the atomization which results in
reduction in SMD. This effect is observed in both cold spray and flame spray.
252
™ At lower fuel inlet temperature, larger droplets dominate the SMD value. Higher fuel
inlet temperatures enhance vaporization and combustion reactions, thus reducing the drop
diameter in the flame, which helps in reducing emissions.
™ Compared with conventional pressure atomizers "dual-orifice" type, airblast atomizer
requires lower fuel supply pressures, to produce a fine spray. Moreover, because the fuel
drops can mix with the atomizing air at the injector exit primary zone, the ensuing
combustion process is characterized by very low soot formation and a blue flame of low
luminosity, resulting in relatively cool liner walls and a minimum of exhaust smoke. A
further asset of the airblast atomizer is that it provides a constant fuel distribution over
the entire range of fuel flows.
™ Overall the results from this research showed that by tailoring the fluid mechanics
associated with the atomization process, the on flame structure, combustion emissions
and spray characteristics for different fuels can be altered significantly. Thus, the
combustion performance clearly does not only depend on the fuel physical and chemical
properties but also on the flame characteristics that are determined by the flow processes.
6.2. Recommendations
The following recommendations are made based on the results from this study:
™ Improvement in clamping of insulation material in the near field spray region where the
optical windows for PDPA measurements are present would improve near field heat
feedback, and completely eliminate the methane requirement to sustain the flame only on
liquid VO.
™ A water cooled sampling probe for emissions measurements would freeze the chemical
reactions downstream of the point of sampling, so that true NOx and CO readings would
253
be acquired to establish the link between spray characteristics (like droplet diameter,
droplet diameter distribution, droplet velocities) and emissions.
™ The present research investigates the spray characteristics of an AB atomizer
experimentally; a computational model of an air blast atomizer can be developed to study
the spray behavior. The computational study for the cold spray can be validated based on
the experimental cold spray results.
™ Required modifications in the model can implemented to compute the flame spray
characteristics using the spray combustion model to simulate the actual conditions of the
spray in the flame, which further can be validated using the flame spray experimental
results presented in this study.
™ The performance of AB injector gets really affected for high viscosity fuels and there is a
need to condition the thermo physical properties of liquid fuel to improve the spray
characteristics of AB injector.
™ Flow blurring atomizers as discussed/introduced in chapter 1 can be explored that are
least vulnerable for handling viscous bio-oils.
254
REFERENCES
Agarwal, A. K., (2007), Biofuels, Alcohol and Biodiesel, Application as Fuels for Internal
Combustion Engines, Prog. Energy Combust. Sci., 33, pp. 233–271.
Bolszo, C., McDonell, V., and Samuelsen, S., (2007), Impact of Biodiesel on Fuel Preparation
and Emissions for a Liquid Fired Gas Turbine Engine, ASME Paper GT2007-27652.
Bari, S., Lim, T.H., and Yu, C.W., (2007), Effects of Preheating Crude Palm oil (CPO) on
Injection System, Performance and Emission of a Diesel Engine, Biomass and
Bioenergy, 27, pp.339-351.
Balasubramanyam, M.S., Chen, C.P., and Trinh, H.P., (2006), Numerical Modeling of
Turbulence Effects within an Evaporating Droplet in Atomizing Sprays, NASA Report
20060047741.
Chiang, H-W., Ching, I-Che, and Li, H-L, (2007), Performance Testing of Microturbine
Generator System Fueled by Biodiesel, ASME Paper GT2007-28075.
Chisti, Y., (2007), Biodiesel for Microalgae, Biotechnol. Adv., 25, pp. 294–306.
Cayli, G., and Kusefoglu, S., 2007, Increased yields in biodiesel production from cooking oils by
a two step process: Comparison with one step process using TGA, Fuel Process Technology,
Article in Press.
255
Chien-Pei Mao,Geng Wang, Norman Chiger,(1986), An Experimetnal Study of Air-Assist
Atomizer Spray Flames, Twenty-first Symposium (International) on Combustion/The
Combustion Institute, pp.665-673
Demirbas, A., (2005), Biodiesel production from vegetable oils via catalytic and non-catalytic
supercritical methanol transesterification methods, Progress in Energy and Combustion
Science, vol. 31, pp. 466-487.
Ejim, C.E., Fleck, B.A., and Amirfazli, A. (2007), Analytical Study for atomization of biodiesels
and their blends in a typical injector: Surface tension and viscosity effects, Fuel, vol.
86, pp. 1534-1544.
Faeth, G. M., Hsiang, L.P., and Wu, P.K., (1995), Structure and Break-up Properties of Spray,
International Journal of Multiphase Flow, 25, pp. 99-127.
Faeth, G. M., (1987), Mixing, Transport and Combustion in Sprays, Prog. Energy
Combustion.Sci, 13, pp. 293-345.
Gañán-Calvo, A. M., (2005) Enhanced Liquid Atomization: From flow-Focusing to Flow
Blurring, Applied Physics Letters, Vol. 86, pp. 2141-2143
Hadef, R., and Lnze, B., (2005), Measurements of Droplets Characteristics in a Swirl Stabilized
Spray Flame, Experimental Thermal and Fluid Science, 30, pp.117-130
256
Kneer, R., Willmann,M., Zeitler,R.,Wittig,S., and Collin, K.H., (1994), Phase Doppler Particle
Sizing Inside a Gas Turbine Combustor, International Journal of energetic Materials
and Chemical Propulsion, vol. 3,(1), pp 301-312
Lang, X., Dalai, A.K., Bakshi, N.N, Reaney, M.J., Hertz, P.B., (2001), Preparation and
characterization of bio-diesels from various bio-oils, Biosource Technology, vol. 80, pp.
53-62.
Lasheras, J. C., and Hopfinger, E. J.,(2000), Liquid jet instability and atomization in a coaxial
gas stream, Annual Review Fluid Mechanics , 32, pp. 275-308.
Lefebvre, A. H., (1999), Gas Turbine Combustion, 2nd ed., Taylor and Francis Publications.
Lefebvre, A.H., (1989), Atomization and Sprays, Hemisphere, New York.
Lin, S. P. and Reitz, R. D., (1998), Drop and Spray Formation from a Liquid Jet, Annual Reviews
of Fluid Mechanics, Vol. 30, pp. 85-105.
Liu, A.B., and Reitz, R.D., (1993), Mechanisms of Air-Assisted Liquid Atomization,
Atomization and Sprays, Vol. 3, pp. 55-75
McDonell, V. G. and Samuelsen, S., (1991), Gas and Drop Behavior in Reacting and NonReacting Airblast Atomizer Sprays, Journal of Propulsion and Power, Vol. 7, pp. 684691.
Mansour, A., (2005), Gas Turbine Fuel Injection Technology, ASME Paper GT-2005-68173.
257
Monyem, A., and Germen, J. H. V., (2001), The Effect of Biodiesel Oxidation on Engine
Performance and Emissions, Biomass Bioenergy, 20, pp. 317–325.
Moliere, M., Panarotto, E., Aboujaib, M., Bisseaud, J.M., Cambell, A., Citeno, J., Maire, P-A,
and Ducrest, L., (2007), Gas Turbine in Alternative Fuel Applications: Biodiesel Field
Test, ASME Paper 2007-27212.
Nakamura, S., McDonell, V., and Samuelsen, S., (2006), The Effect of Liquid-fuel Preparation
on Gas Turbine Emissions, ASME Paper GT-2006-90730.
Nwafor, O.M.I., The Effect of Elevated Fuel Inlet Temperature on Performance of Diesel Engine
Running on Neat Vegetable Oil at Constant Speed Conditions, Renewable Energy,
28,171-181.
Panchasara, H. V., Simmons, B. M., and Agrawal, A.K., (2009), Combustion Performance of
Biodiesel and Diesel-Vegetable Oil Blends in a Simulated Gas Turbine Burner, Journal
of Engineering for Gas Turbines and Power, 131, 031503:1-11.
Panchasara, H.V., Sequera, D.S., Schreiber,W., Agrawal,A.K., (2009), Emissions Reductions in
Diesel and Kerosene Flames Using a Novel Injector, Journal of Propulsion and Power
25, (4), pp. 984-987.
Razdan, M. K., (1998), Fuel/Air Preparation in the Design of Low Emissions Gas Turbine
Combustion Systems, Proceedings of Fourteenth NATO RTO Meeting on Gas Turbine
Combustion Emissions and Alternative Fuels, Paper No. 34.
258
Rizk, N.K., and Lefebvre, A.H., (1984), Spray Characteristics of Plain-Jet Airblast Atomizers,
Journal of Engineering for Gas Turbines and Power, 106, pp. 634-638.
Rakopoulos, C.D., Antonopoulos, K.A., Rakopoulos, D.C., Hountalas, D.T., and Giakoumis,
E.G., (2006), Comparative performance and emissions study of a direct injection diesel
engine using blends of diesel fuel with vegetable oils or bio-diesels of various origins,
Energy Conversion and Management, 47, pp. 3272–3287.
Sovani, S. D., Sojka, P. E., and Lefebvre, A. H., (2001), Effervescent Atomization, Progress in
Energy and Combustion Science, 27, pp. 483-521.
Sequera, D., Agrawal, A.K., Spears, S.K., and Daly, D.T., (2007), Combustion performance of
liquid-biofuels in a swirl stabilized burner, ASME Paper GT2007-2838.
Simmons, B.M., Panchasara, H.V., and Agrawal, A.K., (2008), Effect of Fuel Injection Concept
on Combustion Performance of Liquid Biofuels, Proceedings of the Meeting at the
Central States Combustion Institute.
Simmons, B. M., Panchasra, H. V.,Agrawal., A.K., (2009), “A Comparison of Air-blast and
Flow-blurring injectors using Phase Doppler Particle Analyzer Technique ASME Paper
2009-60239
Turns, S.R., (2000), An Introduction to Combustion- Concepts and Applications, 2nd ed.,
McGraw Hill.
Trinh, H.P., (2005), Numerical Simulation of Liquid Jet Atomization Including Turbulence
Effects, AIAA-2005-3973.
259
Tilman, D., et.al, (2009), Beneficial Biofuels – The Food, Energy, and Environment Trilemma,
Science, 325, (5938), pp. 270-271
Zuo, B., Black, D.L., and Crocker, D. S., (2002), Fuel Atomization and Drop Breakup Models
for Advanced Combustion CFD Codes, AIAA-2002-417.
National Institute of Standards and Technology, Standard Reference Database Number 69,
Retrieved on September 2007 from http://webbook.nist.gov/chemistry
Royal Society of Chemistry, Database of Chemical Structures and Property Predictions,
Retrieved on September 2007 from http://www.chemspider.com
Zhou, X., Luo, K.H., & Williams, J.J.R., (2001). Numerical Studies on Vortex Structures in the
Near Field of Oscillating Diffusion Flames. Heat and Mass Transfer, 37(2), 101-110.
Zhou, X., Luo, K.H., & Williams, J.J.R., (2002). Vortex dynamics in spatio-temporal
development of reacting plumes. Combustion and Flame, 129, 11-29.
260
APPENDIX –A
EQUIVALENCE RATIO CALCULATION
Commercially available Soybean oil (VO) form Sam’s Club was approximated as
C17.6376 H32.24 O2.
Assuming complete combustion (with no dissociation), the stoichiometric reaction of
VO is represented as:
x ⋅C
H
O + y ⋅ CH +
17.6376 32.24 2
4
(24.6976x + 2y ) ⋅ ⎡O
φ
⇒ (17.6376x + y ) ⋅ CO + (16.12x + 2y ) ⋅ H O +
2
2
⎢⎣ 2
+ 3.76 ⋅ N ⎤
2 ⎥⎦
(92.862976x + 7.52y ) ⋅ N
φ
2
⎡⎛ 1 − φ ⎞
⎤
+ ⎢⎜⎜
⎟ ⋅ (24.6976x + 2y )⎥ ⋅ O
⎟
⎣⎝ φ ⎠
⎦ 2
In the above reaction, the reactants molar flow rates are given by,
N
= x moles / min; N
= ymoles / min;
VO
methane
(24.6976 x + 2 y ) = (117.56 x + 9.52 y ) moles / min
N
= 4.76 ⋅
air
φ
φ
The conversion of molar flow rate to volumetric flow rate is given by,
Q (LPM) = {[N (moles/min) x Mol Wt. (gm/mol)] /density (kg/m3)} LPM
Q =
air
N
⎛ (117.56 x + 9.52 y ) ⎞
⎟ ⋅ 28.966
(moles / min ) ⋅ MW (gm/mol) ⎜⎝
φ
⎠
air
air
=
(P (R ⋅ T ))
(101325 (287 × 288))
261
⇒Q
Q
=
methane
N
methane
air
=
1
φ
⋅ (2777.8314x + 224.9486y )
(moles / min ) ⋅ MW
(gm/mol)
methane
=
(kg/m )
ρ
3
methane
y × 16.043
= 22.375 y LPM
0.717
N (moles / min ) ⋅ MW (gm/mol) x × 275.8912
VO
Q
= VO
=
= 0.29826 x LPM = 298.26 x ml/min
VO
3
925
⎛
ρ ⎜ kg/m ⎞⎟
VO ⎝
⎠
The conversion of volumetric flow rate to Heat release rate is given by (assuming 100%
combustion efficiency)
H (kW) = {[Q (slpm) x density (kg/m3)] x LHV (kJ/kg) / 60000} kW
Heat release rate for methane and glycerol based on their flow rates are given as follows
(Note that the combustion efficiency is assumed to be 100%).
Q (ml / min )
H (kW ) = VO
× ρ (kg / m 3 ) × LHV (kJ/kg)
VO
VO
60 × 1000 × 1000 VO
Q
VO
⇒ H (kW ) =
× 925 × 37000 = 0.5704 ⋅ Q (ml / min )
VO
VO
60 × 1000 × 1000
H
methane
⇒H
(kW ) =
methane
(slpm )
Q
methane
⎛⎜ kg / m3 ⎞⎟ × LHV
(kJ/kg)
×ρ
methane ⎝
methane
⎠
60 × 1000
(kW ) =
(slpm )
Q
methane
(slpm )
× 0.717 × 50016 = 0.5977 ⋅ Q
methane
60 × 1000
262
Fuel Properties
Low Heating Value
Density
Fuel
(LHV)
3
kg/m
kJ/kg
Air
1.226
Methane
0.717
37000
VO
925
50016
Air
Methane
VO flow
Methane
VO heat
Net heat
Mixture
flow
flow rate
rate
heat
release
release
Equivalence
rate
(slpm)
(mlpm)
release
150
20
12
11.954
6.8448
18.7988
2.085
150
3.8
12
2.27126
6.8448
9.11606
0.8997
263
ratio
APPENDIX B
EQUIVALENCE RATIO VALIDATION
Vegetable Oil Combustion
Assumptions: - Complete combustion (no dissociation).
- Sam’s club Soybean Oil is used as vegetable oil and is approximated as C17.6376
H32.24 O2.
The generalized lean (complete) combustion reaction for VO and air is represented by:
x ⋅C
H
O +
17.6376 32.24 2
(24.6976x ) ⋅ ⎡O
φ
⎢⎣ 2
⇒ (17.6376x ) ⋅ CO + (16.12x ) ⋅ H O +
2
2
(1)
+ 3.76 ⋅ N ⎤
2 ⎥⎦
(92.862976x ) ⋅ N
φ
2
⎡⎛ 1 − φ ⎞
⎤
+ ⎢⎜⎜
⎟⎟ ⋅ (24.6976x )⎥ ⋅ O
⎣⎝ φ ⎠
⎦ 2
The stoichiometric air/fuel ratio is calculated on mass basis of the reactant (air and VO)
flow rates.
m
N × MWair
24.6976 x × (1 + 3.76) × 28.966
⎛ A⎞
= air = air
=
= 12.3428
⎜ ⎟
x × 275.8912
⎝ F ⎠ stoich . m fuel N fuel × MW fuel
(2)
For sample calculation, the measured air flow rate was 150 l/min, which is the summation
of 125 l/min of primary air flow measured by the LFE and 25 l/min of atomizing air through
injectors, measured by the mass flow meter. The measured kerosene flow rate was 12 ml/min.
The air density was 1.2 kg/m3, based on the measured air flow temperature upstream the
combustor inlet section. The air density at standard conditions is 1.27 kg/m3, based on the ideal
gas law. The VO density was taken as 928 kg/m3 at the ambient conditions (NIST database).
Thus, the actual air/fuel ratio is found
264
found from Equation 2 as:
ρ air [kg / m 3 ] × A[lpm]
1.2 × 150
⎛ A⎞
=
=
= 16.1638
⎜ ⎟
3
⎝ F ⎠ actual ρ fuel [kg / m ] × F [lpm] 928 × (12 / 1000)
(3)
The equivalence ratio is found as:
φ=
( A / F ) stoich . 12.3428
=
= 0.7636
( A / F ) actual 16.1638
(4)
Alternate Calculation
The lean combustion of VO and air is represented by:
x ⋅C
H
O +
17.6376 32.24 2
(24.6976x ) ⋅ ⎡O
φ
⎢⎣ 2
⇒ (17.6376x ) ⋅ CO + (16.12x ) ⋅ H O +
2
2
+ 3.76 ⋅ N ⎤
2 ⎥⎦
(92.862976x ) ⋅ N
φ
(5)
⎡⎛ 1 − φ ⎞
⎤
+ ⎢⎜⎜
⎟⎟ ⋅ (24.6976x )⎥ ⋅ O
2 ⎣⎝ φ ⎠
⎦ 2
Percentage volume O2 ( X O2 ) in dry combustion products is given by,
X O2
⎤
⎡⎛ 1 − φ ⎞
⎢⎜ φ ⎟ ⋅ 24.6976 x ⎥
(1 − φ ) × 24.6976 x
⎠
⎦
⎣⎝
=
=
⎡⎛ 1 − φ ⎞
⎛ 92.862976 x ⎞⎤ 117.560576 − 7.06 x × φ
⎟⎥
⎟ ⋅ 24.6976 x + 17.6376 x + ⎜
⎢⎜
φ
⎝
⎠⎦
⎣⎝ φ ⎠
(6)
0For simplicity if we consider x = 1 as number of moles for fuel in reactants then
volumetric percentage of oxygen in dry combustion product is given by,
X O2 =
(1 − φ ) × 24.6976
117.560576 − 7.06 × φ
(7)
For the above operating conditions (Air Flow = 150 l/min, Fuel Flow = 12 ml/min), % O2
from the product gas measured by the gas analyzer on dry basis was 5.5 %
Then the equivalence ratio resulting from Equation 7 is:
φ=
24.6976 − 117.560576 X O2
24.6976 − 7.06 X O2
265
(8)
Substituting the measured value XO2 = 0.055 in Equation 8, the resulting equivalence ratio is Ø =
0.74999, which is the same as the one computed from the measured air and fuel flow rates
266
APPENDIX C
LOW HEATING VALUE CALCULAITON
The generalized lean (complete) combustion reaction for any fuel and air is represented by:
(1)
y z⎞
⎛
C H O + ⎜ x + − ⎟ ⋅ ⎡O + 3.76 ⋅ N ⎤
x y z ⎝
2 ⎥⎦
4 2 ⎠ ⎢⎣ 2
y
y z⎞
⎛
⇒ x ⋅ CO + ⋅ H O + ⋅⎜ x + − ⎟ × 3.76 ⋅ N
2 2 2
2
2 2⎠
⎝
The low heating value (LHV) for the fuel is generalized by:
y⎡
⎡
⎤
⎡
⎤
⎤
LHV = ⎢ ΔH ⎥
− x ⎢ ΔH ⎥ − ⎢ ΔH ⎥
f ⎦ Cx H yOz
f ⎦ CO2 2 ⎣
f ⎦ H 2O
⎣
⎣
(2)
For simplicity, a sample calculation for Lenoleic acid LHV is shown here and the
respective heat of formations ΔH are listed.
f
The chemical formula for Lenoleic acid is C H O
19 34 2
⎡
⎢⎣ ΔH f
⎤
= -513 kJ/mol
⎥⎦
C H O
19 34 2
⎡
⎤
= -393.5 kJ/mol
⎢⎣ ΔH f ⎥⎦
CO
2
⎡
⎢⎣ ΔH f
⎤
=-241.8kJ/mol
⎥⎦
H O
2
267
Substituting these values in equation 2 above,
⎡
⎤
34
LHV = ⎢[− 513]C H O − 19[− 393.5]CO − [− 241.8]H O ⎥ × 1000
2 ⎦
19 34 2
2 2
⎣
(3)
∴ LHV = 11075773 kJ/kmol
The molecular weight of Lenoleic acid is 294.3 kg/kmol
LHV in kJ/kg is given by:
LHV 11075773
=
= 37634kJ / kg
MW
294.3
The high heating value is calculated using the latent heat of vaporization of H2O = 285.8 kJ/mol
Substituting this value in equation 3,
HHV = 1182397 kJ/kmol
HHV in kJ/kg is given by:
HHV 1182397
=
= 40176.6kJ / kg
MW
294.3
Calculating the LHV’s for all the other components of BD-1 and BD-2 listed in Table 2.2 of
chapter 2, the global LHV of the BD-1 is given based on the % mass of each component as:
LHVBD −1 = 0.5 × LHVC19 H 34O2 + 0.358 × LHVC19 H 36O2 + 0.103 × LHVC17 H 34O2 + 0.034 × LHVC19 H 38O2
LHVBD −1 = 0.5 × 37634 + 0.358 × 37756 + 0.103 × 37284 + 0.034 × 38314
∴ LHVBD −1 = 38002.3kJ / kg
Similarly for BD-2:
LHVBD −1 = 0.203 × LHVC19 H 34O2 + 0.45 × LHVC19 H 36O2 + 0.193 × LHVC17 H 34O2 + 0.086 × LHVC19 H 38O2
+ 0.002 × LHVC15H 30O2 + 0.066 × LHVC17 H 32O2
LHV
= 0.203 × 37634 + 0.45 × 37756 + 0.193 × 37284 + 0.086 × 38314
BD − 1
+ 0.002 × 36526 + 0.066 × 37184
268
∴ LHVBD − 2 = 37659.7kJ / kg
The list of heat of formation for all the BD’s components is listed in table below.
Formula
Methyl Esters
Lenoleic Methyl Ester
Oleic Methyl Ester
Palmitic Methyl Ester
Steartic Methyl Ester
Tetradecanoic Methyl Ester
9-Hexadecenoic acid, Methyl Ester
C19H34O2
C19H36O2
C17H34O2
C19H38O2
C15H30O2
C17H32O2
269
MW
kg/kmol
294.3
296.3
270.2
298.3
242.2
268.2
ΔfHliquid
kJ/mol
-513.00
-643.25
-727.50
787.40
-684.30
-587.00
APPENDIX –D
RENOYLDS NUMBER CALCULATION
The general equation of Reynolds (Re) number for a fluid flowing in a pipe is:
Re =
ρ ×V × D
μ
H
(1)
where ρ is the fluid density in kg/m3, V is the fluid velocity in m/s, DH is the hydraulic diameter
of the annular cross section of the pipe in meters (m) and μ is the absolute viscosity in N ⋅ s / m 2 .
The hydraulic diameter for the annular cross sectional area of the pipe is given by:
DH = Do-Di, where, Do and Di are the outer and inner diameters of pipe
The flow velocity: V =
m&
, where m& is the mass flow rate in kg/s and A represents the
ρ×A
pipe area: A = π Do2 − Di 2 / 4 in m2. Substituting in Equation 1:
Re =
4 × m&
π ×D ×μ
H
(2)
Equation 2 was used to estimate the Reynolds number. The mass flow rate was measured
during the experiments, while the hydraulic diameter is used as the equivalent diameter of
annular cross-section of the combustion air tube is D = 0.01898 m .
The flow dynamic viscosity µ for air at standard atmospheric conditions is taken as
μ = 1.983 × 10 − 5 kg / m − s
270
The density of air is taken as ρ = 1.2kg / m 3
The total air flow rate for the experiments is taken as Q = 150 LPM
Calculating the flow inlet velocity of combustion air:
V =
Q
m&
=
=
ρ×A
A
150 × 10 − 3
2
2⎞
⎛
π × ⎜ ⎛⎜ 40.08 × 10 − 3 ⎞⎟ − ⎛⎜ 21.1 × 10 − 3 ⎞⎟ ⎟
⎜⎝
⎠
⎝
⎠ ⎟⎠
⎝
∴V = 2.7m / s
Substituting these values in equation 1 above, the Reynolds’s number is calculated as:
Re =
ρ ×V × D
∴ Re =
μ
H
1.2kg / m 3 × 2.7m / s × 0.01898m
1.983 × 10 −5 kg / m − s
∴ Re = 3150 for combustion air flow rate of Q = 150 LPM.
271
APPENDIX –E
UNCERTAINITY ANALYSIS
With the confidence interval (CI), one can estimate the likelihood that a calculated mean
lies within a true mean for a given population. A confidence interval of 95% implies that 95% of
all the samples would give an interval that includes true mean (µ), and only 5% of all the
samples would yield an erroneous interval.
The confidence interval is based on standard deviation and desired confidence level.
The standard deviation for a population is given as:
⎡ Σ ( xi − x ) ⎤
⎥
n
⎦
1/ 2
σ =⎢
⎣
(1)
where, x = population mean and n = population count. CI is specified by:
x ± zα / 2 ⋅ σ / n
(2)
where α based on the desired confidence level, 1-95% for 95% CI, which yields zα / 2 = 1.96.
Droplet Diameter Confidence Interval
Mean droplet diameter and the standard deviation was calculated from a sample size of
10000 droplets for the diameter distribution at each transverse and axial location in the spray.
A sample calculation for 95% CI is presented for non swirl air flow at Y = 5 mm and X = 0 mm
in the cold spray.
272
The mean diameter for this case is x = 35.17 microns. The standard deviation as calculated form
above equation 1 is σ =26.42 microns.
Substituting the values into equation 2 above yields:
95% CI =± 0.517 microns.
This means that there is a 95% confidence that the true population mean droplet diameter is
within the range of 35.17 ± 0.517 microns.
Instrument Uncertainties
Instruments
Reported
Uncertainty
LFE
±0.5 LPM
Peristaltic fuel pump
±0.25% of the
flow rate
Sierra Mass Flow
±0.5 LPM
controller for Atomizing
air
NOVA emissions
±2 ppm
analyzer
Fuel heater
±0.5 °C
273